1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
24 PP/01 Solaris build fix for Oracle's LDAP libraries.
25 Bugzilla 1109, patch from Stephen Usher.
27 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
29 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
32 TF/02 Fix a couple more cases where we did not log the error message
33 when unlink() failed. See also change 4.74-TF/03.
35 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
36 lock up or crash if it happened to be inside a call to libc when it
37 got a SIGUSR1 from exiwhat.
39 The SIGUSR1 handler appends the current process status to the process
40 log which is later printed by exiwhat. It used to use the general
41 purpose logging code to do this, but several functions it calls are
44 The new output code in the SIGUSR1 handler is specific to the process
45 log, and simple enough that it's easy to inspect for signal safety.
46 Removing some special cases also simplifies the general logging code.
47 Removing the spurious timestamps from the process log simplifies
50 TF/04 Improved ratelimit ACL condition.
52 The /noupdate option has been deprecated in favour of /readonly which
53 has clearer semantics. The /leaky, /strict, and /readonly update modes
54 are mutually exclusive. The update mode is no longer included in the
55 database key; it just determines when the database is updated. (This
56 means that when you upgrde Exim will forget old rate measurements.)
58 Exim now checks that the per_* options are used with an update mode that
59 makes sense for the current ACL. For example, when Exim is processing a
60 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
61 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
62 must specify per_mail/readonly. If you omit the update mode it defaults to
63 /leaky where that makes sense (as before) or /readonly where required.
65 The /noupdate option is now undocumented but still supported for
66 backwards compatibility. It is equivalent to /readonly except that in
67 ACLs where /readonly is required you may specify /leaky/noupdate or
68 /strict/noupdate which are treated the same as /readonly.
70 A useful new feature is the /count= option. This is a generalization
71 of the per_byte option, so that you can measure the throughput of other
72 aggregate values. For example, the per_byte option is now equivalent
73 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
75 The per_rcpt option has been generalized using the /count= mechanism
76 (though it's more complicated than the per_byte equivalence). When it is
77 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
78 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
79 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
80 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
81 Note that using per_rcpt with a non-readonly update mode in more than
82 one ACL will cause the recipients to be double-counted. (The per_mail
83 and per_byte options don't have this problem.)
85 The handling of very low rates has changed slightly. If the computed rate
86 is less than the event's count (usually one) then this event is the first
87 after a long gap. In this case the rate is set to the same as this event's
88 count, so that the first message of a spam run is counted properly.
90 The major new feature is a mechanism for counting the rate of unique
91 events. The new per_addr option counts the number of different
92 recipients that someone has sent messages to in the last time period. It
93 behaves like per_rcpt if all the recipient addresses are different, but
94 duplicate recipient addresses do not increase the measured rate. Like
95 the /count= option this is a general mechanism, so the per_addr option
96 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
97 example, measure the rate that a client uses different sender addresses
98 with the options per_mail/unique=$sender_address. There are further
99 details in the main documentation.
101 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
103 TF/06 Removed a few PCRE remnants.
105 TF/07 Automatically extract Exim's version number from tags in the git
106 repository when doing development or release builds.
108 PP/02 Raise smtp_cmd_buffer_size to 16kB.
109 Bugzilla 879. Patch from Paul Fisher.
111 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
112 Heavily based on revision 40f9a89a from Simon Arlott's tree.
115 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
117 PP/05 Variable $av_failed, true if the AV scanner deferred.
118 Bugzilla 1078. Patch from John Horne.
120 PP/06 Stop make process more reliably on build failure.
121 Bugzilla 1087. Patch from Heiko Schlittermann.
123 PP/07 Make maildir_use_size_file an _expandable_ boolean.
124 Bugzilla 1089. Patch from Heiko Schlittermann.
126 PP/08 Handle ${run} returning more data than OS pipe buffer size.
127 Bugzilla 1131. Patch from Holger Weiß.
129 PP/09 Handle IPv6 addresses with SPF.
130 Bugzilla 860. Patch from Wolfgang Breyha.
132 PP/10 GnuTLS: support TLS 1.2 & 1.1.
134 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
137 PP/11 match_* no longer expand right-hand-side by default.
138 New compile-time build option, EXPAND_LISTMATCH_RHS.
139 New expansion conditions, "inlist", "inlisti".
141 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
143 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
145 PP/14 fix log_write() format string regression from TF/03.
146 Bugzilla 1152. Patch from Dmitry Isaikin.
152 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
154 PP/02 Harmonised TLS library version reporting; only show if debugging.
155 Layout now matches that introduced for other libraries in 4.74 PP/03.
157 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
159 PP/04 New "dns_use_edns0" global option.
161 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
164 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
165 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
167 TK/01 Updated PolarSSL code to 0.14.2.
168 Bugzilla 1097. Patch from Andreas Metzler.
170 PP/07 Catch divide-by-zero in ${eval:...}.
173 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
176 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
177 format-string attack -- SECURITY: remote arbitrary code execution.
179 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
180 time unintentionally subject to list matching rules, letting the header
181 cause arbitrary Exim lookups (of items which can occur in lists, *not*
182 arbitrary string expansion). This allowed for information disclosure.
184 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
185 INT_MIN/-1 -- value coerced to INT_MAX.
191 NM/01 Workround for PCRE version dependency in version reporting
194 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
195 This fixes portability to compilers other than gcc, notably
196 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
198 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
199 makefiles for portability to HP-UX and POSIX correctness.
201 PP/01 Permit LOOKUP_foo enabling on the make command-line.
202 Also via indented variable definition in the Makefile.
203 (Debugging by Oliver Heesakkers).
205 PP/02 Restore caching of spamd results with expanded spamd_address.
206 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
208 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
209 Improves build reliability. Fix from: Frank Elsner
211 NM/02 Fix wide character breakage in the rfc2047 coding
212 Fixes bug 1064. Patch from Andrey N. Oktyabrski
214 NM/03 Allow underscore in dnslist lookups
215 Fixes bug 1026. Patch from Graeme Fowler
217 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
218 Code patches from Adam Ciarcinski of NetBSD.
220 NM/04 Fixed exiqgrep to cope with mailq missing size issue
223 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
224 is logged, to avoid truncation. Patch from John Horne.
226 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
227 Patch from Jakob Hirsch.
229 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
230 SQL string expansion failure details.
231 Patch from Andrey Oktyabrski.
233 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
234 Patch from Simon Arlott.
236 PP/09 New lookups functionality failed to compile on old gcc which rejects
237 extern declarations in function scope.
238 Patch from Oliver Fleischmann
240 PP/10 Use sig_atomic_t for flags set from signal handlers.
241 Check getgroups() return and improve debugging.
242 Fixed developed for diagnosis in bug 927 (which turned out to be
245 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
246 Patch from Mark Zealey.
248 PP/12 Bugzilla 1056: Improved spamd server selection.
249 Patch from Mark Zealey.
251 PP/13 Bugzilla 1086: Deal with maildir quota file races.
252 Based on patch from Heiko Schlittermann.
254 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
255 Patch from Uwe Doering, sign-off by Michael Haardt.
257 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
258 variable declaration deep within a block. Bug and patch from
261 PP/15 lookups-Makefile IRIX compatibilty coercion.
263 PP/16 Make DISABLE_DKIM build knob functional.
265 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
266 Patch by Simon Arlott
268 TF/03 Fix valgrind.h portability to C89 compilers that do not support
269 variable argument macros. Our copy now differs from upstream.
275 TF/01 Failure to get a lock on a hints database can have serious
276 consequences so log it to the panic log.
278 TF/02 Log LMTP confirmation messages in the same way as SMTP,
279 controlled using the smtp_confirmation log selector.
281 TF/03 Include the error message when we fail to unlink a spool file.
283 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
284 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
285 for maintaining out-of-tree patches for some time.
287 PP/01 Bugzilla 139: Documentation and portability issues.
288 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
289 Handle per-OS dynamic-module compilation flags.
291 PP/02 Let /dev/null have normal permissions.
292 The 4.73 fixes were a little too stringent and complained about the
293 permissions on /dev/null. Exempt it from some checks.
294 Reported by Andreas M. Kirchwitz.
296 PP/03 Report version information for many libraries, including
297 Exim version information for dynamically loaded libraries. Created
298 version.h, now support a version extension string for distributors
299 who patch heavily. Dynamic module ABI change.
301 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
302 privilege escalation vulnerability whereby the Exim run-time user
303 can cause root to append content of the attacker's choosing to
306 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
309 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
310 If dropping privileges for untrusted macros, we disabled normal logging
311 on the basis that it would fail; for the Exim run-time user, this is not
312 the case, and it resulted in successful deliveries going unlogged.
313 Fixed. Reported by Andreas Metzler.
319 PP/01 Date: & Message-Id: revert to normally being appended to a message,
320 only prepend for the Resent-* case. Fixes regression introduced in
321 Exim 4.70 by NM/22 for Bugzilla 607.
323 PP/02 Include check_rfc2047_length in configure.default because we're seeing
324 increasing numbers of administrators be bitten by this.
326 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
328 PP/03 Bugzilla 994: added openssl_options main configuration option.
330 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
332 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
334 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
336 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
337 without a peer certificate, leading to a segfault because of an
338 assumption that peers always have certificates. Be a little more
339 paranoid. Problem reported by Martin Tscholak.
341 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
342 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
343 NB: ClamAV planning to remove STREAM in "middle of 2010".
344 CL also introduces -bmalware, various -d+acl logging additions and
345 more caution in buffer sizes.
347 PP/09 Implemented reverse_ip expansion operator.
349 PP/10 Bugzilla 937: provide a "debug" ACL control.
351 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
353 PP/12 Bugzilla 973: Implement --version.
355 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
357 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
359 PP/15 Bugzilla 816: support multiple condition rules on Routers.
361 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
362 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
363 ignore trailing whitespace.
365 JJ/02 prevent non-panic DKIM error from being sent to paniclog
367 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
370 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
371 Notification from Dr Andrew Aitchison.
373 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
374 ExtendedDetectionInfo response format.
375 Notification from John Horne.
377 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
380 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
381 XSL and documented dependency on system catalogs, with examples of how
384 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
387 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
388 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
389 configuration file which is writeable by the Exim user or group.
391 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
392 of configuration files to cover files specified with the -C option if
393 they are going to be used with root privileges, not just the default
396 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
397 option (effectively making it always true).
399 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
400 files to be used while preserving root privileges.
402 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
403 that rogue child processes cannot use them.
405 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
406 run-time user, instead of root.
408 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
409 Exim run-time user without dropping privileges.
411 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
412 result string, instead of calling string_vformat() twice with the same
415 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
416 for other users. Others should always drop root privileges if they use
417 -C on the command line, even for a whitelisted configure file.
419 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
421 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
427 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
428 $data_path, and $header_path variables; fixed documentation bugs and
431 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
432 exipick to access non-standard spools, including the "frozen" queue
435 NM/01 Bugzilla 965: Support mysql stored procedures.
436 Patch from Alain Williams
438 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
440 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
441 Patch from Andreas Metzler
443 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
444 Patch from Kirill Miazine
446 NM/05 Bugzilla 671: Added umask to procmail example.
448 JJ/03 installed exipick 20100323.0, fixing doc bug
450 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
451 directory. Notification and patch from Dan Rosenberg.
453 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
455 TK/02 Improve log output when DKIM signing operation fails.
457 MH/01 Treat the transport option dkim_domain as a colon separated
458 list, not as a single string, and sign the message with each element,
459 omitting multiple occurences of the same signer.
461 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
462 Bugzilla 985, 986. Patch by Simon Arlott
464 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
465 Patch by Simon Arlott
467 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
468 MBX locking. Notification from Dan Rosenberg.
474 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
476 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
478 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
480 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
482 NM/04 Bugzilla 915: Flag broken perl installation during build.
488 TK/01 Added patch by Johannes Berg that expands the main option
489 "spamd_address" if it starts with a dollar sign.
491 TK/02 Write list of recipients to X-Envelope-Sender header when building
492 the mbox-format spool file for content scanning (suggested by Jakob
495 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
496 (http://www.dcc-servers.net/) support via dccifd. Activated by
497 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
499 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
500 by Mark Daniel Reidel <mr@df.eu>.
502 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
503 When building exim an external PCRE library is now needed -
504 PCRE is a system library on the majority of modern systems.
505 See entry on PCRE_LIBS in EDITME file.
507 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
508 conversation. Added nologin parameter to request.
509 Patch contributed by Kirill Miazine.
511 TF/01 Do not log submission mode rewrites if they do not change the address.
513 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
515 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
516 log files in place. Contributed by Roberto Lima.
518 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
520 TF/03 Bugzilla 615: When checking the local_parts router precondition
521 after a local_part_suffix or local_part_prefix option, Exim now
522 does not use the address's named list lookup cache, since this
523 contains cached lookups for the whole local part.
525 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
526 Robert Millan. Documentation is in experimental-spec.txt.
528 TF/04 Bugzilla 668: Fix parallel build (make -j).
530 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
532 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
533 Patch provided by Jan Srzednicki.
535 TF/05 Leading white space used to be stripped from $spam_report which
536 wrecked the formatting. Now it is preserved.
538 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
539 that they are available at delivery time.
541 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
543 TF/08 TLS error reporting now respects the incoming_interface and
544 incoming_port log selectors.
546 TF/09 Produce a more useful error message if an SMTP transport's hosts
547 setting expands to an empty string.
549 NM/06 Bugzilla 744: EXPN did not work under TLS.
550 Patch provided by Phil Pennock.
552 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
553 Patch provided by Richard Godbee.
555 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
556 acl_smtp_notquit, added index entry.
558 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
559 Patch provided by Eugene Bujak.
561 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
562 accept(). Patch provided by Maxim Dounin.
564 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
565 Patch provided by Phil Pennock.
567 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
569 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
570 Patch provided by Brad "anomie" Jorsch.
572 NM/14 Bugzilla 622: Added timeout setting to transport filter.
573 Patch provided by Dean Brooks.
575 TK/05 Add native DKIM support (does not depend on external libraries).
577 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
578 Patch provided by Graeme Fowler.
580 NM/16 Bugzilla 851: Documentation example syntax fix.
582 NM/17 Changed NOTICE file to remove references to embedded PCRE.
584 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
587 NM/19 Bugzilla 745: TLS version reporting.
588 Patch provided by Phil Pennock.
590 NM/20 Bugzilla 167: bool: condition support.
591 Patch provided by Phil Pennock.
593 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
594 clients. Patch provided by Phil Pennock.
596 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
597 Patch provided by Brad "anomie" Jorsch.
599 NM/23 Bugzilla 687: Fix misparses in eximstats.
600 Patch provided by Heiko Schlittermann.
602 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
603 Patch provided by Heiko Schlittermann.
605 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
606 plus update to original patch.
608 NM/26 Bugzilla 799: Documentation correction for ratelimit.
610 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
611 Patch provided by David Brownlee.
613 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
615 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
617 NM/30 Bugzilla 888: TLS documentation bugfixes.
619 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
621 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
622 Unlike the original bugzilla I have changed all shell scripts in src tree.
624 NM/33 Bugzilla 898: Transport filter timeout fix.
625 Patch by Todd Rinaldo.
627 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
628 Patch by Serge Demonchaux.
630 NM/35 Bugzilla 39: Base64 decode bug fixes.
631 Patch by Jakob Hirsch.
633 NM/36 Bugzilla 909: Correct connect() call in dcc code.
635 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
637 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
639 NM/39 Bugzilla 911: Fixed MakeLinks build script.
645 TK/01 Add preliminary DKIM support. Currently requires a forked version of
646 ALT-N's libdkim that I have put here:
647 http://duncanthrax.net/exim-experimental/
649 Note to Michael Haardt: I had to rename some vars in sieve.c. They
650 were called 'true' and it seems that C99 defines that as a reserved
651 keyword to be used with 'bool' variable types. That means you could
652 not include C99-style headers which use bools without triggering
653 build errors in sieve.c.
655 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
656 as mailq or other aliases. Changed the --help handling significantly
657 to do whats expected. exim_usage() emits usage/help information.
659 SC/01 Added the -bylocaldomain option to eximstats.
661 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
663 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
665 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
671 PH/01 Another patch from the Sieve maintainer.
673 PH/02 When an IPv6 address is converted to a string for single-key lookup
674 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
675 dots are used instead of colons so that keys in lsearch files need not
676 contain colons. This was done some time before quoting was made available
677 in lsearch files. However, iplsearch files do require colons in IPv6 keys
678 (notated using the quote facility) so as to distinguish them from IPv4
679 keys. This meant that lookups for IP addresses in host lists did not work
680 for iplsearch lookups.
682 This has been fixed by arranging for IPv6 addresses to be expressed with
683 colons if the lookup type is iplsearch. This is not incompatible, because
684 previously such lookups could never work.
686 The situation is now rather anomolous, since one *can* have colons in
687 ordinary lsearch keys. However, making the change in all cases is
688 incompatible and would probably break a number of configurations.
690 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
693 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
694 conversion specification without a maximum field width, thereby enabling
695 a rogue spamd server to cause a buffer overflow. While nobody in their
696 right mind would setup Exim to query an untrusted spamd server, an
697 attacker that gains access to a server running spamd could potentially
698 exploit this vulnerability to run arbitrary code as the Exim user.
700 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
701 $primary_hostname instead of what libspf2 thinks the hosts name is.
703 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
704 a directory entry by the name of the lookup key. Previously, if a
705 symlink pointed to a non-existing file or a file in a directory that
706 Exim lacked permissions to read, a lookup for a key matching that
707 symlink would fail. Now it is enough that a matching directory entry
708 exists, symlink or not. (Bugzilla 503.)
710 PH/03 The body_linecount and body_zerocount variables are now exported in the
713 PH/04 Added the $dnslist_matched variable.
715 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
716 This means they are set thereafter only if the connection becomes
719 PH/06 Added the client_condition to authenticators so that some can be skipped
720 by clients under certain conditions.
722 PH/07 The error message for a badly-placed control=no_multiline_responses left
723 "_responses" off the end of the name.
725 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
727 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
728 (without spaces) instead of just copying the configuration text.
730 PH/10 Added the /noupdate option to the ratelimit ACL condition.
732 PH/11 Added $max_received_linelength.
734 PH/12 Added +ignore_defer and +include_defer to host lists.
736 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
737 way in which PCRE > 7.0 is built.
739 PH/14 Implemented queue_only_load_latch.
741 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
742 MAIL command. The effect was to mangle the value on 64-bit systems.
744 PH/16 Another patch from the Sieve maintainer.
746 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
748 PH/18 If a system quota error occurred while trying to create the file for
749 a maildir delivery, the message "Mailbox is full" was not appended to the
750 bounce if the delivery eventually timed out. Change 4.67/27 below applied
751 only to a quota excession during the actual writing of the file.
753 PH/19 It seems that peer DN values may contain newlines (and other non-printing
754 characters?) which causes problems in log lines. The DN values are now
755 passed through string_printing() before being added to log lines.
757 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
758 and InterBase are left for another time.)
760 PH/21 Added message_body_newlines option.
762 PH/22 Guard against possible overflow in moan_check_errorcopy().
764 PH/23 POSIX allows open() to be a macro; guard against that.
766 PH/24 If the recipient of an error message contained an @ in the local part
767 (suitably quoted, of course), incorrect values were put in $domain and
768 $local_part during the evaluation of errors_copy.
774 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
775 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
778 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
779 issue a MAIL command.
781 PH/02 In an ACL statement such as
783 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
785 if a client was not listed at all, or was listed with a value other than
786 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
787 the condition was not true (as it should be), so access was not denied.
788 The bug was that the ! inversion was incorrectly passed on to the second
789 item. This has been fixed.
791 PH/03 Added additional dnslists conditions == and =& which are different from
792 = and & when the dns lookup returns more than one IP address.
794 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
795 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
797 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
798 FSYNC, which compiles an option called disable_fsync that allows for
799 bypassing fsync(). The documentation is heavily laced with warnings.
801 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
803 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
804 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
805 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
806 including adding "make clean"; (3) Added -fPIC when compiling the test
807 dynamically loaded module, to get rid of a warning.
809 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
810 message fails, move_frozen_messages = true and ignore_bounce_errors_after
811 = 0s. The bug is otherwise harmless.
813 PH/07 There was a bug in the dovecot authenticator such that the value of
814 $auth1 could be overwritten, and so not correctly preserved, after a
815 successful authentication. This usually meant that the value preserved by
816 the server_setid option was incorrect.
818 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
820 PH/09 Installed PCRE release 7.0.
822 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
823 run for batched SMTP input. It is now run at the start of every message
824 in the batch. While fixing this I discovered that the process information
825 (output by running exiwhat) was not always getting set for -bs and -bS
826 input. This is fixed, and it now also says "batched" for BSMTP.
828 PH/11 Added control=no_pipelining.
830 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
831 patch, slightly modified), and move the expansion of helo_data till after
832 the connection is made in the smtp transport (so it can use these
835 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
837 PH/14 Added log_selector = +pid.
839 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
841 PH/16 Add ${if forany and ${if forall.
843 PH/17 Added dsn_from option to vary the From: line in DSNs.
845 PH/18 Flush SMTP output before performing a callout, unless control =
846 no_callout_flush is set.
848 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
849 was true (the default) a successful delivery failed to delete the retry
850 item, thus causing premature timeout of the address. The bug is now
853 PH/20 Added hosts_avoid_pipelining to the smtp transport.
855 PH/21 Long custom messages for fakedefer and fakereject are now split up
856 into multiline reponses in the same way that messages for "deny" and
857 other ACL rejections are.
859 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
860 with slight modification.
862 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
863 draft, changing the syntax and factoring some duplicate code".
865 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
866 for deliveries of the second and subsequent messages over the same SMTP
869 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
870 ${reduce, with only minor "tidies".
872 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
874 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
875 expansion side effects.
877 PH/27 When a message times out after an over-quota error from an Exim-imposed
878 quota, the bounce message says "mailbox is full". This message was not
879 being given when it was a system quota that was exceeded. It now should
882 MH/03 Made $recipients available in local_scan(). local_scan() already has
883 better access to the recipient list through recipients_list[], but
884 $recipients can be useful in postmaster-provided expansion strings.
886 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
887 in the case of a MAIL command with additional options following the
888 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
889 were accidentally chopped off.
891 PH/29 SMTP synchronization checks are implemented when a command is read -
892 there is a check that no more input is waiting when there shouldn't be
893 any. However, for some commands, a delay in an ACL can mean that it is
894 some time before the response is written. In this time, more input might
895 arrive, invalidly. So now there are extra checks after an ACL has run for
896 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
897 pipelining has not been advertised.
899 PH/30 MH's patch to allow iscntrl() characters to be list separators.
901 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
902 returned in the SMTP response when smtp_return_error_details was false.
905 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
906 instead of the C I/O that was originally supplied, because problems were
909 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
910 Exim which did not show up earlier: it was assuming that a call to
911 SSL_CTX_set_info_callback() might give an error value. In fact, there is
912 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
913 was a macro that became an assignment, so it seemed to work. This has
914 changed to a proper function call with a void return, hence the compile
915 error. Exim's code has been fixed.
917 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
920 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
922 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
924 PH/37 If a message is not accepted after it has had an id assigned (e.g.
925 because it turns out to be too big or there is a timeout) there is no
926 "Completed" line in the log. When some messages of this type were
927 selected by exigrep, they were listed as "not completed". Others were
928 picked up by some special patterns. I have improved the selection
929 criteria to be more general.
931 PH/38 The host_find_failed option in the manualroute router can now be set
932 to "ignore", to completely ignore a host whose IP address cannot be
933 found. If all hosts are ignored, the behaviour is controlled by the new
934 host_all_ignored option.
936 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
937 homing or because of multiple MX records with /mx) generated more than
938 one IP address, and the following item turned out to be the local host,
939 all the secondary addresses of the first item were incorrectly removed
940 from the list, along with the local host and any following hosts (which
941 is what is supposed to happen).
943 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
944 whoever called Exim into the -H file. In the case of the daemon it was
945 behaving confusingly. When first started, it used values for whoever
946 started the daemon, but after a SIGHUP it used the Exim user (because it
947 calls itself on a restart). I have changed the code so that it now always
950 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
951 message are rejected with the same error (e.g. no authentication or bad
952 sender address), and a DATA command is nevertheless sent (as can happen
953 with PIPELINING or a stupid MUA), the error message that was given to the
954 RCPT commands is included in the rejection of the DATA command. This is
955 intended to be helpful for MUAs that show only the final error to their
958 PH/42 Another patch from the Sieve maintainer.
960 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
961 Eximstats - Fixed some broken HTML links and added missing column headers
963 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
964 columns for Rejects, Temp Rejects, Ham, and Spam rows.
966 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
968 PH/43 Yet another patch from the Sieve maintainer.
970 PH/44 I found a way to check for a TCP/IP connection going away before sending
971 the response to the final '.' that terminates a message, but only in the
972 case where the client has not sent further data following the '.'
973 (unfortunately, this is allowed). However, in many cases there won't be
974 any further data because there won't be any more messages to send. A call
975 to select() can be used: if it shows that the input is "ready", there is
976 either input waiting, or the socket has been closed. An attempt to read
977 the next input character can distinguish the two cases. Previously, Exim
978 would have sent an OK response which the client would never have see.
979 This could lead to message repetition. This fix should cure that, at
980 least in a lot of common cases.
982 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
983 advertised in response to EHLO.
989 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
990 fixed by 4.65/MH/01 (is this a record?) are fixed:
992 (i) An empty string was always treated as zero by the numeric comparison
993 operators. This behaviour has been restored.
995 (ii) It is documented that the numeric comparison operators always treat
996 their arguments as decimal numbers. This was broken in that numbers
997 starting with 0 were being interpreted as octal.
999 While fixing these problems I realized that there was another issue that
1000 hadn't been noticed. Values of message_size_limit (both the global option
1001 and the transport option) were treated as octal if they started with 0.
1002 The documentation was vague. These values are now always treated as
1003 decimal, and I will make that clear in the documentation.
1009 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1010 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1013 MH/01 Don't check that the operands of numeric comparison operators are
1014 integers when their expansion is in "skipping" mode (fixes bug
1015 introduced by 4.64-PH/07).
1017 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1018 child addresses, Exim now panics and dies. Previously, because the count
1019 is held in a short int, deliveries were likely to be lost. As such a
1020 large number of recipients for a single message is ridiculous
1021 (performance will be very, very poor), I have chosen to impose a limit
1022 rather than extend the field.
1028 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1029 leftover -K file (the existence of which was triggered by #402).
1030 While we were at it, introduced process PID as part of the -K
1031 filename. This should rule out race conditions when creating
1034 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1035 processing considerably. Previous code took too long for large mails,
1036 triggering a timeout which in turn triggers #401.
1038 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1039 in the DK code in transports.c. sendfile() is not really portable,
1040 hence the _LINUX specificness.
1042 TF/01 In the add_headers option to the mail command in an Exim filter,
1043 there was a bug that Exim would claim a syntax error in any
1044 header after the first one which had an odd number of characters
1047 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1048 callout verification, Exim cached a "reject" for the entire domain. This
1049 is correct for most verifications, but it is not correct for a recipient
1050 verification with use_sender or use_postmaster set, because in that case
1051 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1052 case of MAIL FROM:<> rejection from other early rejections (e.g.
1053 rejection of HELO). When verifying a recipient using a non-null MAIL
1054 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1055 Whatever the result of the callout, the value of the domain cache is
1056 left unchanged (for any other kind of callout, getting as far as trying
1057 RCPT means that the domain itself is ok).
1059 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1062 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1063 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1064 session, the daemon ignores ECONNECT errors and logs others; it now
1065 ignores EPIPE as well.
1067 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1068 (quoted-printable decoding).
1070 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1071 later the small subsequent patch to fix an introduced bug.
1073 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1075 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1077 PH/08 An error is now given if message_size_limit is specified negative.
1079 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1080 to be given (somewhat) arbitrary names.
1082 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1085 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1086 miscellaneous code fixes
1088 PH/10 Added the log_reject_target ACL modifier to specify where to log
1091 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1092 hostname. This is wrong, because it relates to the incoming message (and
1093 probably the interface on which it is arriving) and not to the outgoing
1094 callout (which could be using a different interface). This has been
1095 changed to use the value of the helo_data option from the smtp transport
1096 instead - this is what is used when a message is actually being sent. If
1097 there is no remote transport (possible with a router that sets up host
1098 addresses), $smtp_active_hostname is used.
1100 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1101 tweaks were necessary in order to get it to work (see also 21 below):
1102 (a) The code assumed that strncpy() returns a negative number on buffer
1103 overflow, which isn't the case. Replaced with Exim's string_format()
1105 (b) There were several signed/unsigned issues. I just did the minimum
1106 hacking in of casts. There is scope for a larger refactoring.
1107 (c) The code used strcasecmp() which is not a standard C function.
1108 Replaced with Exim's strcmpic() function.
1109 (d) The code set only $1; it now sets $auth1 as well.
1110 (e) A simple test gave the error "authentication client didn't specify
1111 service in request". It would seem that Dovecot has changed its
1112 interface. Fortunately there's a specification; I followed it and
1113 changed what the client sends and it appears to be working now.
1115 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1118 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1119 address A is aliased to B and C, where B exists and C does not. Without
1120 -v the output is "A verified" because verification stops after a
1121 successful redirection if more than one address is generated. However,
1122 with -v the child addresses are also verified. Exim was outputting "A
1123 failed to verify" and then showing the successful verification for C,
1124 with its parentage. It now outputs "B failed to verify", showing B's
1125 parentage before showing the successful verification of C.
1127 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1128 look up a TXT record in a specific list after matching in a combined
1131 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1132 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1133 they consult the DNS. I had assumed they would set it the way they
1134 wanted; and indeed my experiments on Linux seem to show that in some
1135 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1136 To be on the safe side, however, I have now made the interface to
1137 host_find_byname() similar to host_find_bydns(), with an argument
1138 containing the DNS resolver options. The host_find_byname() function now
1139 sets these options at its start, just as host_find_bydns() does. The smtp
1140 transport options dns_qualify_single and dns_search_parents are passed to
1141 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1142 of host_find_byname() use the default settings of RES_DEFNAMES
1143 (qualify_single) but not RES_DNSRCH (search_parents).
1145 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1146 spool_read_header() do less string testing, by means of a preliminary
1147 switch on the second character of optional "-foo" lines. (This is
1148 overdue, caused by the large number of possibilities that now exist.
1149 Originally there were few.) While I was there, I also converted the
1150 str(n)cmp tests so they don't re-test the leading "-" and the first
1151 character, in the hope this might squeeze out yet more improvement.
1153 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1154 flag allowing group syntax was set by the header_syntax check but not
1155 turned off, possible causing trouble later; (2) The flag was not being
1156 set at all for the header_verify test, causing "group"-style headers to
1157 be rejected. I have now set it in this case, and also caused header_
1158 verify to ignore an empty address taken from a group. While doing this, I
1159 came across some other cases where the code for allowing group syntax
1160 while scanning a header line wasn't quite right (mostly, not resetting
1161 the flag correctly in the right place). These bugs could have caused
1162 trouble for malformed header lines. I hope it is now all correct.
1164 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1165 with the "reply" argument non-NULL. The code, however (which originally
1166 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1167 but it didn't always do it. This confused somebody who was copying the
1168 code for some other use. I have removed all the tests.
1170 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1171 feature that was used to support insecure browsers during the U.S. crypto
1172 embargo. It requires special client support, and Exim is probably the
1173 only MTA that supported it -- and would never use it because real RSA is
1174 always available. This code has been removed, because it had the bad
1175 effect of slowing Exim down by computing (never used) parameters for the
1176 RSA_EXPORT functionality.
1178 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1179 authenticator to fail if there's a tab character in the incoming data
1180 (there should never be unless someone is messing about, as it's supposed
1181 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1182 if the connection is using TLS or if the remote IP is the same as the
1183 local IP, and the "valid-client-cert option" if a client certificate has
1186 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1187 authenticators. This can be used for authorization after authentication
1188 succeeds. (In the case of plaintext, it servers for both authentication
1191 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1192 if any retry times were supplied.
1194 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1195 connection, where there is no client IP address to check. In this
1196 situation, the verify now always succeeds.
1198 PH/25 Applied John Jetmore's -Mset patch.
1200 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1202 PH/27 In a string expansion for a processed (not raw) header when multiple
1203 headers of the same name were present, leading whitespace was being
1204 removed from all of them, but trailing whitespace was being removed only
1205 from the last one. Now trailing whitespace is removed from each header
1206 before concatenation. Completely empty headers in a concatenation (as
1207 before) are ignored.
1209 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1210 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1212 PH/29 [Removed. This was a change that I later backed out, and forgot to
1213 correct the ChangeLog entry (that I had efficiently created) before
1214 committing the later change.]
1216 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1217 address errors (4xx response to RCPT) over the same connection as other
1218 messages routed to the same hosts. Such deliveries are always "forced",
1219 so retry times are not inspected. This resulted in far too many retries
1220 for the affected addresses. The effect occurred only when there were more
1221 hosts than the hosts_max_try setting in the smtp transport when it had
1222 the 4xx errors. Those hosts that it had tried were not added to the list
1223 of hosts for which the message was waiting, so if all were tried, there
1224 was no problem. Two fixes have been applied:
1226 (i) If there are any address or message errors in an SMTP delivery, none
1227 of the hosts (tried or untried) are now added to the list of hosts
1228 for which the message is waiting, so the message should not be a
1229 candidate for sending over the same connection that was used for a
1230 successful delivery of some other message. This seems entirely
1231 reasonable: after all the message is NOT "waiting for some host".
1232 This is so "obvious" that I'm not sure why it wasn't done
1233 previously. Hope I haven't missed anything, but it can't do any
1234 harm, as the worst effect is to miss an optimization.
1236 (ii) If, despite (i), such a delivery is accidentally attempted, the
1237 routing retry time is respected, so at least it doesn't keep
1238 hammering the server.
1240 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1241 in ${readsocket because some servers need this prod.
1243 PH/32 Added some extra debug output when updating a wait-xxx database.
1245 PH/33 The hint "could be header name not terminated by colon", which has been
1246 given for certain expansion errors for a long time, was not being given
1247 for the ${if def:h_colon_omitted{... case.
1249 PH/34 The spec says: "With one important exception, whenever a domain list is
1250 being scanned, $domain contains the subject domain." There was at least
1251 one case where this was not true.
1253 PH/35 The error "getsockname() failed: connection reset by peer" was being
1254 written to the panic log as well as the main log, but it isn't really
1255 panic-worthy as it just means the connection died rather early on. I have
1256 removed the panic log writing for the ECONNRESET error when getsockname()
1259 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1260 runs only) independently of the message's sender address. This meant
1261 that, if the 4xx error was in fact related to the sender, a different
1262 message to the same recipient with a different sender could confuse
1263 things. In particualar, this can happen when sending to a greylisting
1264 server, but other circumstances could also provoke similar problems.
1265 I have changed the default so that the retry time for these errors is now
1266 based a combination of the sender and recipient addresses. This change
1267 can be overridden by setting address_retry_include_sender=false in the
1270 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1271 remote server are returned as part of bounce messages. This was not
1272 happening for LMTP over a pipe (the lmtp transport), but now it is the
1273 same for both kinds of LMTP.
1275 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1276 in header lines that were in fact CNAMEs. This is no longer the case.
1278 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1279 and queue runs started by the daemon processed all messages. This has
1280 been fixed so that -R and -S can now usefully be given with -q<time>.
1282 PH/40 Import PCRE release 6.7 (fixes some bugs).
1284 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1286 PH/42 Give an error if -q is specified more than once.
1288 PH/43 Renamed the variables $interface_address and $interface_port as
1289 $received_ip_address and $received_port, to make it clear that these
1290 values apply to message reception, and not to the outgoing interface when
1291 a message is delivered. (The old names remain recognized, of course.)
1293 PH/44 There was no timeout on the connect() call when using a Unix domain
1294 socket in the ${readsocket expansion. There now is.
1296 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1297 be meaningful with "accept".
1299 SC/01 Eximstats V1.43
1300 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1302 SC/02 Eximstats V1.44
1303 Use a glob alias rather than an array ref in the generated
1304 parser. This improves both readability and performance.
1306 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1307 Collect SpamAssassin and rejection statistics.
1308 Don't display local sender or destination tables unless
1309 there is data to show.
1310 Added average volumes into the top table text output.
1312 SC/04 Eximstats V1.46
1313 Collect data on the number of addresses (recipients)
1314 as well as the number of messages.
1316 SC/05 Eximstats V1.47
1317 Added 'Message too big' to the list of mail rejection
1318 reasons (thanks to Marco Gaiarin).
1320 SC/06 Eximstats V1.48
1321 Mainlog lines which have GMT offsets and are too short to
1322 have a flag are now skipped.
1324 SC/07 Eximstats V1.49 (Alain Williams)
1325 Added the -emptyok flag.
1327 SC/08 Eximstats V1.50
1328 Fixes for obtaining the IP address from reject messages.
1330 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1331 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1332 whitesspace changes from 4.64-PH/27
1334 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1337 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1338 are found, allow negative numbers in numeric criteria)
1340 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1342 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1343 to match changes made in 4.64-PH/43
1345 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1347 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1348 "domain" is actually an IP address, and force a failure. This locks out
1349 those revolvers/nameservers that support "A-for-A" lookups, in
1350 contravention of the specifications.
1352 PH/48 When a host name was looked up from an IP address, and the subsequent
1353 forward lookup of the name timed out, the host name was left in
1354 $sender_host_name, contrary to the specification.
1356 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1357 restricted to single-key lookups, Exim was not diagnosing an error if
1358 * or *@ was used with a query-style lookup.
1360 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1362 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1363 long ago, but noone interested enough thought of it. Let's just say that
1364 the "1.1" means that there are some new functions that weren't there at
1365 some point in the past.
1367 PH/51 Error processing for expansion failure of helo_data from an smtp
1368 transport during callout processing was broken.
1370 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1371 tested/used via the -bh/-bhc/-bs options.
1373 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1374 bug, fixed in subsequent PCRE releases).
1376 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1377 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1379 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1385 SC/01 Use a glob alias rather than an array ref in eximstats generated
1386 parser. This improves both readability and performance.
1388 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1389 Don't display local sender or destination tables in eximstats unless
1390 there is data to show.
1391 Added average volumes into the eximstats top table text output.
1393 SC/03 Collect data on the number of addresses (recipients) as well
1394 as the number of messages in eximstats.
1396 TF/01 Correct an error in the documentation for the redirect router. Exim
1397 does (usually) call initgroups() when daemonizing.
1399 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1400 with consistent privilege compared to when running as a daemon.
1402 TF/03 Note in the spec that $authenticated_id is not set for local
1403 submissions from trusted users.
1405 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1406 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1408 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1409 by adding some example configuration directives to the default
1410 configuration file. A little bit of work is required to uncomment the
1411 directives and define how usernames and passwords are checked, but
1412 there is now a framework to start from.
1414 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1415 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1416 without this. I don't know how relevant this is to other LDAP libraries.
1418 PH/02 Add the verb name to the "unknown ACL verb" error.
1420 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1422 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1424 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1425 directory not expanded when it should be if an expanded home directory
1426 was set for the address (which is overridden by the transport).
1428 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1431 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1432 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1433 because it is too late at that time, and has no effect.
1435 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1436 security issue with \' (bugzilla #107). I could not use the
1437 PQescapeStringConn() function, because it needs a PGconn value as one of
1440 PH/08 When testing addresses using -bt, indicate those final addresses that
1441 are duplicates that would not cause an additional delivery. At least one
1442 person was confused, thinking that -bt output corresponded to deliveries.
1443 (Suppressing duplicates isn't a good idea as you lose the information
1444 about possibly different redirections that led to the duplicates.)
1446 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1447 systems where poll() doesn't work, in particular OS X.
1449 PH/10 Added more information to debugging output for retry time not reached.
1451 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1452 operations in malware.c.
1454 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1457 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1458 syslog_duplication set false, log lines that would normally be written
1459 both the the main log and to the reject log were not written to syslog at
1462 PH/14 In the default configuration, change the use of "message" in ACL warn
1463 statements to "add_header".
1465 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1466 not followed by a command (e.g. "seen endif").
1468 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1469 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1472 PH/17 Added extra conditions to the default value of delay_warning_condition
1476 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1477 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1478 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1481 The Auto-Submitted: and various List- headers are standardised, whereas I
1482 don't think Precedence: ever was.
1484 PH/18 Refactored debugging code in route_finduser() to show more information,
1485 in particular, the error code if getpwnam() issues one.
1487 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1488 This is apparently needed in addition to the PH/07 change above to avoid
1489 any possible encoding problems.
1491 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1492 but not after initializing Perl.
1494 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1495 output them only if debugging. By default they are written stderr,
1496 apparently, which is not desirable.
1498 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1501 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1504 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1506 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1507 authenticated or an ident call has been made. Suppress the default
1508 values for $authenticated_id and $authenticated_sender (but permit -oMai
1509 and -oMas) when testing with -bh.
1511 PH/24 Re-jigged the order of the tests in the default configuration so that the
1512 tests for valid domains and recipients precede the DNS black list and CSA
1513 tests, on the grounds that those ones are more expensive.
1515 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1516 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1517 This bug exists in every version of Exim that I still have, right back to
1520 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1521 However, an attempt to turn on case-sensitivity in a regex key by
1522 including (?-i) didn't work because the subject string was already
1523 lowercased, and the effects were non-intuitive. It turns out that a
1524 one-line patch can be used to allow (?-i) to work as expected.
1530 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1531 other effects) broke the use of negated acl sub-conditions.
1533 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1536 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1537 "Deny" causes Exim to reject the incoming connection with a 554 error.
1538 Unfortunately, if there is a major crisis, such as a disk failure,
1539 tcp-wrappers gives "deny", whereas what one would like would be some
1540 kind of temporary error. A kludge has been added to help with this.
1541 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1542 554 error is used if errno is still zero or contains ENOENT (which occurs
1543 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1546 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1548 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1549 errors. Otherwise a message that provokes a temporary error (when other
1550 messages do not) can cause a whole host to time out.
1552 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1553 addresses were routed directly to files or pipes from a redirect router.
1554 File deliveries just didn't batch; pipe deliveries might have suffered
1557 PH/06 A failure to get a lock for a hints database would erroneously always say
1558 "Failed to get write lock", even when it was really a read lock.
1560 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1561 of 0600. This has been changed to use the value of the lockfile_mode
1562 option (which defaults to 0600).
1564 PH/08 Applied small patch from the Sieve maintainer.
1566 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1567 folder from quota calculations, a direct delivery into this folder messed
1568 up the contents of the maildirsize file. This was because the regex was
1569 used only to exclude .Trash (or whatever) when the size of the mailbox
1570 was calculated. There was no check that a delivery was happening into an
1571 excluded directory. This bug has been fixed by ignoring all quota
1572 processing for deliveries into excluded directories.
1574 PH/10 Added the maildirfolder_create_regex option to appendfile.
1580 PH/01 The code for finding all the local interface addresses on a FreeBSD
1581 system running IPv6 was broken. This may well have applied to all BSD
1582 systems, as well as to others that have similar system calls. The broken
1583 code found IPv4 interfaces correctly, but gave incorrect values for the
1584 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1585 that it would not match correctly against @[] and not recognize the IPv6
1588 PH/02 The ipliteral router was not recognizing addresses of the form user@
1589 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1591 PH/03 Added disable_ipv6.
1593 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1594 decoded form, because it is most often used to construct To: headers
1595 lines in autoreplies, and the decoded form may well be syntactically
1596 invalid. However, $reply_address has leading white space removed, and all
1597 newlines turned into spaces so that the autoreply transport does not
1600 PH/05 If group was specified without a user on a router, and no group or user
1601 was specified on a transport, the group from the router was ignored.
1603 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1604 for visible compile-time settings that can be used to change these
1605 numbers, for those that want even more. Backwards compatibility with old
1606 spool files has been maintained. However, going back to a previous Exim
1607 release will lost any variables that are in spool files.
1609 PH/07 Two small changes when running in the test harness: increase delay when
1610 passing a TCP/IP connection to a new process, in case the original
1611 process has to generate a bounce, and remove special handling of
1612 127.0.0.2 (sic), which is no longer necessary.
1614 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1615 be the same on different OS.
1617 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1620 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1621 whether --show-vars was specified or not
1623 JJ/02 exipick: Added support for new ACL variable spool format introduced
1626 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1627 syntactically invalid From: or Reply-to: line, and a filter used this to
1628 generate an autoreply, and therefore failed to obtain an address for the
1629 autoreply, Exim could try to deliver to a non-existent relative file
1630 name, causing unrelated and misleading errors. What now happens is that
1631 it logs this as a hard delivery error, but does not attempt to create a
1634 PH/11 The exinext utility has a -C option for testing purposes, but although
1635 the given file was scanned by exinext itself; it wasn't being passed on
1636 when Exim was called.
1638 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1639 an end-of-file indication when reading a command response.
1641 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1642 compiled. In many other places in Exim, IPv6 addresses are always
1643 recognized, so I have changed this. It also means that IPv4 domain
1644 literals of the form [IPV4:n.n.n.n] are now always recognized.
1646 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1647 used if the router is not running as root, for example, when verifying at
1648 ACL time, or when using -bh. The debugging output from this situation was
1649 non-existent - all you got was a failure to exec. I have made two
1652 (a) Failures to set uid/gid, the current directory, or a process leader
1653 in a subprocess such as that created by queryprogram now generate
1654 suitable debugging ouput when -d is set.
1656 (b) The queryprogram router detects when it is not running as root,
1657 outputs suitable debugging information if -d is set, and then runs
1658 the subprocess without attempting to change uid/gid.
1660 PH/15 Minor change to Makefile for building test_host (undocumented testing
1663 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1664 additional section of a DNS packet that returns MX or SRV records.
1665 Instead, it always explicitly searches for A/AAAA records. This avoids
1666 major problems that occur when a DNS server includes only records of one
1667 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1668 fixed another bug: if SRV records were looked up and the corresponding
1669 address records were *not* found in the additional section, the port
1670 values from the SRV records were lost.
1672 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1673 using the correct key (the original address) when searching the retry
1674 rules in order to find which one to use for generating the retry hint.
1676 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1677 adding the default one. Similarly, if it contains a Reply-To: header, the
1678 errors_reply_to option, if set, is not used.
1680 PH/19 When calculating a retry time, Exim used to measure the "time since
1681 failure" by looking at the "first failed" field in the retry record. Now
1682 it does not use this if it is later than than the arrival time of the
1683 message. Instead it uses the arrival time. This makes for better
1684 behaviour in cases where some deliveries succeed, thus re-setting the
1685 "first failed" field. An example is a quota failure for a huge message
1686 when small messages continue to be delivered. Without this change, the
1687 "time since failure" will always be short, possible causing more frequent
1688 delivery attempts for the huge message than are intended.
1689 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1691 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1692 $1, $2, $3) because the numerical variables can be reset during some
1693 expansion items (e.g. "match"), thereby losing the authentication data.
1695 PH/21 Make -bV show the size of off_t variables so that the test suite can
1696 decide whether to run tests for quotas > 2G.
1698 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1699 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1700 filecount value is greater than 2G or if a quota value is greater than 2G
1701 on a system where the size of off_t is not greater than 4, a panic error
1704 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1705 never match. The debug and -bh output now contains an explicit error
1706 message indicating a malformed IPv4 address or mask.
1708 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1709 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1710 PH/23 above applies.
1712 PH/25 Do not write to syslog when running in the test harness. The only
1713 occasion when this arises is a failure to open the main or panic logs
1714 (for which there is an explicit test).
1716 PH/26 Added the /no_tell option to "control=freeze".
1718 PH/27 If a host name lookup failed very early in a connection, for example, if
1719 the IP address matched host_lookup and the reverse lookup yielded a name
1720 that did not have a forward lookup, an error message of the form "no IP
1721 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1722 could be logged. Now it outputs the IP address instead of "NULL".
1724 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1725 allows the sender and the authenticated sender to be set when
1726 submitting a message from within Exim. Since child_open_exim() is
1727 documented for local_scan(), the new function should be too.
1729 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1730 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1731 results in an empty string is now treated as unset.
1733 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1735 PH/31 Added sender_verify_fail logging option.
1737 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1738 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1739 tidied the source and removed it altogether.
1741 PH/33 When a queue run was abandoned because the load average was too high, a
1742 log line was always written; now it is written only if the queue_run log
1743 selector is set. In addition, the log line for abandonment now contains
1744 information about the queue run such as the pid. This is always present
1745 in "start" and "stop" lines but was omitted from the "abandon" line.
1747 PH/34 Omit spaces between a header name and the colon in the error message that
1748 is given when verify = headers_syntax fails (if there are lots of them,
1749 the message gets confusing).
1751 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1752 names, as there are now some PTR records that contain slashes. This check
1753 is only to protect against broken name servers that fall over on strange
1754 characters, so the fact that it applies to all lookups doesn't matter.
1756 PH/36 Now that the new test suite is complete, we can remove some of the
1757 special code in Exim that was needed for the old test suite. For example,
1758 sorting DNS records because real resolvers return them in an arbitrary
1759 order. The new test suite's fake resolver always returns records in the
1762 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1763 bounces) except when queue_only is set, to avoid logging races between
1764 the different processes.
1766 PH/38 Panic-die if .include specifies a non-absolute path.
1768 PH/39 A tweak to the "H" retry rule from its user.
1770 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1771 a label. They prevented compilation on older perls.
1773 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1774 a warning to be raised on newish perls.
1776 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1777 on queue. Changes to match documented behaviour of showing count of
1778 messages matching specified criteria.
1780 PH/40 Changed the default ident timeout from 30s to 5s.
1782 PH/41 Added support for the use of login_cap features, on those BSD systems
1783 that have them, for controlling the resources used by pipe deliveries.
1785 PH/42 The content-scanning code uses fopen() to create files in which to put
1786 message data. Previously it was not paying any attention to the mode of
1787 the files. Exim runs with umask(0) because the rest of the code creates
1788 files with open(), and sets the required mode explicitly. Thus, these
1789 files were ending up world-writeable. This was not a big issue, because,
1790 being within the spool directory, they were not world-accessible. I have
1791 created a function called modefopen, which takes an additional mode
1792 argument. It sets umask(777), creates the file, chmods it to the required
1793 mode, then resets the umask. All the relevant calls to fopen() in the
1794 content scanning code have been changed to use this function.
1796 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1797 to 24 hours. This avoids potential overflow problems when processing G
1798 and H retry rules. I suspect nobody ever tinkers with this value.
1800 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1802 PH/45 When the plaintext authenticator is running as a client, the server's
1803 challenges are checked to ensure they are valid base64 strings. By
1804 default, the authentication attempt is cancelled if an invalid string is
1805 received. Setting client_ignore_invalid_base64 true ignores these errors.
1806 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1807 they are received. Thus, the responses can be made to depend on the
1808 challenges. If an invalid string is ignored, an empty string is placed in
1811 PH/46 Messages that are created by the autoreply transport now contains a
1812 References: header, in accordance with RFCs 2822 and 3834.
1814 PH/47 Added authenticated_sender_force to the smtp transport.
1816 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1818 PH/49 Installed latest patch from the Sieve maintainer.
1820 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1821 was also set, the appendfile transport was unnecessarily scanning a
1822 directory of message files (e.g. for maildir delivery) to find the count
1823 of files (along with the size), even though it did not need this
1824 information. It now does the scan only if it needs to find either the
1825 size of the count of files.
1827 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1829 PH/52 Two bugs concerned with error handling when the smtp transport is
1832 (i) Exim was not creating retry information for temporary errors given
1833 for individual recipients after the DATA command when the smtp transport
1834 was used in LMTP mode. This meant that they could be retried too
1835 frequently, and not timed out correctly.
1837 (ii) Exim was setting the flag that allows error details to be returned
1838 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1839 recipients that were returned after the DATA command.
1841 PH/53 This is related to PH/52, but is more general: for any failing address,
1842 when detailed error information was permitted to be returned to the
1843 sender, but the error was temporary, then after the final timeout, only
1844 "retry timeout exceeded" was returned. Now it returns the full error as
1845 well as "retry timeout exceeded".
1847 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1848 do this, and (what is worse) MTAs that accept it.
1850 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1851 will now be deprecated.
1853 PH/56 New os.c-cygwin from the Cygwin maintainer.
1855 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1856 formats (previously only available in exim formats via -bpr, -bpru,
1857 and -bpra. Now also available in native and exiqgrep formats)
1859 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1860 with very large, slow to parse queues
1862 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1864 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1866 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1867 responses to authentication challenges, though it was showing the
1868 challenges; (ii) I've removed the CR characters from the debug output for
1871 PH/58 Allow for the insertion of a newline as well as a space when a string
1872 is turned into more than one encoded-word during RFC 2047 encoding. The
1873 Sieve code now uses this.
1875 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1876 data_4xx, lost_connection, tls_required.
1878 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1879 message was being sent as an SMTP response.
1881 PH/61 Add -l and -k options to exicyclog.
1883 PH/62 When verifying, if an address was redirected to one new address, so that
1884 verification continued, and the new address failed or deferred after
1885 having set something in $address_data, the value of $address_data was not
1886 passed back to the ACL. This was different to the case when no
1887 redirection occurred. The value is now passed back in both cases.
1889 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1890 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1891 use login_cap.h, so on its own it isn't the distinguishing feature. The
1892 new name refers directly to the setclassresources() function.
1894 PH/65 Added configuration files for NetBSD3.
1896 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1898 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1899 is preferred over IPv4.
1901 PH/68 The bounce_return_message and bounce_return_body options were not being
1902 honoured for bounces generated during the reception of non-SMTP messages.
1903 In particular, this applied to messages rejected by the ACL. This bug has
1904 been fixed. However, if bounce_return_message is true and bounce_return_
1905 body is false, the headers that are returned for a non-SMTP message
1906 include only those that have been read before the error was detected.
1907 (In the case of an ACL rejection, they have all been read.)
1909 PH/69 The HTML version of the specification is now built in a directory called
1910 spec_html instead of spec.html, because the latter looks like a path with
1911 a MIME-type, and this confuses some software.
1913 PH/70 Catch two compiler warnings in sieve.c.
1915 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1916 function verify_get_ident() calls ip_connect() to connect a socket, but
1917 if the "connect()" function timed out, ip_connect() used to close the
1918 socket. However, verify_get_ident() also closes the socket later, and in
1919 between Exim writes to the log, which may get opened at this point. When
1920 the socket was closed in ip_connect(), the log could get the same file
1921 descriptor number as the socket. This naturally causes chaos. The fix is
1922 not to close the socket in ip_connect(); the socket should be closed by
1923 the function that creates it. There was only one place in the code where
1924 this was missing, in the iplookup router, which I don't think anybody now
1925 uses, but I've fixed it anyway.
1927 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1928 well as to direct DNS lookups. Otherwise the handling of names in host
1929 lists is inconsistent and therefore confusing.
1935 PH/01 Two changes to the default runtime configuration:
1937 (1) Move the checks for relay_from_hosts and authenticated clients from
1938 after to before the (commented out) DNS black list checks.
1940 (2) Add control=submission to the relay_from_hosts and authenticated
1941 clients checks, on the grounds that messages accepted by these
1942 statements are most likely to be submissions.
1944 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1946 (1) Generate an error if the third argument for the ${prvs expansion is
1949 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1952 (3) Reset the variables that are obtained from the first argument of
1953 ${prvscheck and used in the second argument before leaving the code,
1954 because their memory is reclaimed, so using them afterwards may do
1957 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1958 one (it's much easier than Tom thought :-).
1960 (5) Because of (4), we can now allow for the use of $prvscheck_result
1961 inside the third argument.
1963 PH/03 For some reason, the default setting of PATH when running a command from
1964 a pipe transport was just "/usr/bin". I have changed it to
1967 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1968 anything to be listed in the output from -bV.
1970 PH/05 When a filter generated an autoreply, the entire To: header line was
1971 quoted in the delivery log line, like this:
1973 => >A.N.Other <ano@some.domain> <original@ddress> ...
1975 This has been changed so that it extracts the operative address. There
1976 may be more than one such address. If so, they are comma-separated, like
1979 => >ano@some.domain,ona@other.domain <original@ddress> ...
1981 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1982 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1983 not being looked up in the rDNS to get a host name, Exim was showing the
1984 IP address twice in Received: lines, even though the IP addresses were
1985 identical. For example:
1987 Received: from [1.2.3.4] (helo=[1.2.3.4])
1989 However, if the real host name was known, it was omitting the HELO data
1990 if it matched the actual IP address. This has been tidied up so that it
1991 doesn't show the same IP address twice.
1993 PH/07 When both +timestamp and +memory debugging was on, the value given by
1994 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1995 called by the debug printing, thereby overwriting the timestamp buffer.
1996 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1998 PH/08 When the original message was included in an autoreply transport, it
1999 always said "this is a copy of the message, including all the headers",
2000 even if body_only or headers_only was set. It now gives an appropriate
2003 PH/09 Applied a patch from the Sieve maintainer which:
2005 o fixes some comments
2006 o adds the (disabled) notify extension core
2007 o adds some debug output for the result of if/elsif tests
2008 o points to the current vacation draft in the documentation
2009 and documents the missing references header update
2013 o fixes a bug in processing the envelope test (when testing
2014 multiple envelope elements, the last element determinted the
2017 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2018 Electronic Mail") by including:
2020 Auto-submitted: auto-generated
2022 in the messages that it generates (bounce messages and others, such as
2023 warnings). In the case of bounce messages for non-SMTP mesages, there was
2024 also a typo: it was using "Auto_submitted" (underscore instead of
2025 hyphen). Since every message generated by Exim is necessarily in response
2026 to another message, thes have all been changed to:
2028 Auto-Submitted: auto-replied
2030 in accordance with these statements in the RFC:
2032 The auto-replied keyword:
2034 - SHOULD be used on messages sent in direct response to another
2035 message by an automatic process,
2037 - MUST NOT be used on manually-generated messages,
2039 - MAY be used on Delivery Status Notifications (DSNs) and Message
2040 Disposition Notifications (MDNs),
2042 - MUST NOT be used on messages generated by automatic or periodic
2043 processes, except for messages which are automatic responses to
2046 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2047 to the default Received: header definition.
2049 PH/12 Added log selector acl_warn_skipped (default on).
2051 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2052 variables because (a) they are in the wrong storage pool and (b) even if
2053 they were copied, it wouldn't work properly because of the caching.
2055 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2056 checking when decoding. Apparently there are clients that generate
2057 overlong encoded strings. Why am I not surprised?
2059 PH/15 If the first argument of "${if match_address" was not empty, but did not
2060 contain an "@" character, Exim crashed. Now it writes a panic log message
2061 and treats the condition as false.
2063 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2065 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2066 extension "envelope-auth". The code is finished and in agreement with
2067 other implementations, but there is no documentation so far and in fact,
2068 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2069 not changing the active code.
2071 Print executed "if" and "elsif" statements when debugging is used. This
2072 helps a great deal to understand what a filter does.
2074 Document more things not specified clearly in RFC3028. I had all this
2075 sorted out, when out of a sudden new issues came to my mind. Oops."
2077 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2080 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2081 canonical form (as documented). However, after a host name lookup from
2082 the IP address, check_host() was doing a simple string comparison with
2083 addresses acquired from the DNS when checking that the found name did
2084 have the original IP as one of its addresses. Since any found IPv6
2085 addresses are likely to be in abbreviated form, the comparison could
2086 fail. Luckily, there already exists a function for doing the comparison
2087 by converting both addresses to binary, so now that is used instead of
2088 the text comparison.
2090 PH/20 There was another similar case to PH/19, when a complete host name was
2091 given in a host list; looking up its IP address could give an abbreviated
2092 form, whereas the current host's name might or might not be abbreviated.
2093 The same fix has been applied.
2099 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2100 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2103 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2104 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2106 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2108 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2109 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2110 Also, for "Auto-Submitted", treat anything other than "no" as
2111 non-personal, in accordance with RFC 3834. (Previously it treated
2112 anything starting "auto-" as non-personal.)
2114 TF/01 The control=submission/name=... option had a problem with syntax
2115 errors if the name included a slash character. The /name= option
2116 now slurps the rest of the string, so it can include any characters
2117 but it must come last in the list of options (after /sender_retain
2120 PH/05 Some modifications to the interface to the fake nameserver for the new
2128 TK/01 Added the "success_on_redirect" address verification option. See
2129 NewStuff for rationale and an example.
2131 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2133 PH/02 Patch to exigrep to allow it to work on syslog lines.
2135 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2136 fread() to skip over the body file's header line, because in Cygwin the
2137 header line is locked and is inaccessible.
2139 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2140 co-exist for some time) to make it clear that it is the Exim ID that is
2141 referenced, not the Message-ID: header line.
2143 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2144 string_format() function, because snprintf() does not exist on all
2147 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2148 ${stat: expansion item.
2150 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2151 protocol synchronization error", to keep the pedants happy.
2153 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2154 well as for IRIX systems, when gcc is being used. See the host.c source
2157 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2159 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2162 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2163 transport and to the smtp transport in LMTP mode.
2165 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2167 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2168 run a filter in a subprocess. This could lead to confusion in subsequent
2169 lookups in the parent process. There should also be a search_tidyup() at
2170 the end of the subprocess.
2172 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2173 only if the host matched helo_try_verify_hosts, which caused the
2174 verification to occur when the EHLO/HELO command was issued. The ACL just
2175 tested the remembered result. Now, if a previous verification attempt has
2176 not happened, "verify = helo" does it there and then.
2178 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2180 TK/03 Fix log output including CR from clamd.
2182 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2183 exist provoked a memory error which could cause a segfault.
2185 PH/15 Installed PCRE 6.2
2187 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2189 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2190 of the problem. Specifically, suggested +O2 rather than +O1 for the
2193 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2195 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2196 record, the port was not correctly logged when the outgoing_port log
2197 selector was set (it logged the transort's default port).
2199 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2200 fallback_hosts, and "hosts" in the smtp transport.
2202 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2203 host errors such as "Connection refused".
2205 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2206 authentication with radiusclient 0.4.9:
2208 - Error returned from rc_read_config was caught wrongly
2209 - Username/password not passed on to radius server due to wrong length.
2211 The presumption is that some radiusclient API changes for 4.51/PH/17
2212 were not taken care of correctly. The code is still untested by me (my
2213 Linux distribution still has 0.3.2 of radiusclient), but it was
2214 contributed by a Radius user.
2216 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2217 expanding the "port" option of the smtp transport.
2219 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2220 while reading a MIME header. Thanks to Tom Hughes for a patch.
2222 PH/24 Include config.h inside local_scan.h so that configuration settings are
2225 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2226 that in an ACL for RCPT (for example), you can examine exactly what was
2229 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2230 commands, but it was not correctly comparing the address with the actual
2231 client host address. Thus, it would show the EHLO address in Received:
2232 header lines when this was not necessary.
2234 PH/27 Added the % operator to ${eval:}.
2236 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2237 it should be ignoring failures (because with -C, for example, it has lost
2238 privilege). It wasn't ignoring creation failures other than "already
2241 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2244 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2245 because an input line was too long, either on its own, or by virtue of
2246 too many continuations, the temporary file was not being removed, and the
2247 return code was incorrect.
2249 PH/31 Missing "BOOL" in function definition in filtertest.c.
2251 PH/32 Applied Sieve patches from the maintainer.
2253 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2255 PH/33 Added "verify = not_blind".
2257 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2258 Local/Makefile (with some defaults set). These are used in built scripts
2259 such as exicyclog, but they have never been used in the exim_install
2260 script (though there are many overriding facilities there). I have
2261 arranged that the exim_install script now takes note of these two
2264 PH/35 Installed configuration files for Dragonfly.
2266 PH/36 When a locally submitted message by a trusted user did not contain a
2267 From: header, and the sender address was obtained from -f or from an SMTP
2268 MAIL command, and the trusted user did not use -F to supply a sender
2269 name, $originator_name was incorrectly used when constructing a From:
2270 header. Furthermore, $originator_name was used for submission mode
2271 messages from external hosts without From: headers in a similar way,
2272 which is clearly wrong.
2274 PH/37 Added control=suppress_local_fixups.
2276 PH/38 When log_selector = +received_sender was set, and the addition of the
2277 sender made the log line's construction buffer exactly full, or one byte
2278 less than full, an overflow happened when the terminating "\n" was
2281 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2282 when the result of a list match is failure because a DNS lookup failed.
2284 PH/40 RM_COMMAND is now used in the building process.
2286 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2287 the "build-* directories that it finds.
2289 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2290 address in a domain literal was a prefix of an interface address.
2292 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2293 when verifying a sender address, unless rewrite_headers is false.
2295 PH/44 Wrote a long comment about why errors_to addresses are verified as
2296 recipients, not senders.
2298 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2299 the ratelimit ACL was added.
2301 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2303 PH/46 Added extra information about PostgreSQL errors to the error string.
2305 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2306 avoiding the need to install special zones in a real server. This is
2307 backwards compatible; if it can't find the fake resolver, it drops back.
2308 Thus, both old and new test suites can be run.
2310 TF/02 Added util/ratelimit.pl
2312 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2313 clock is set back in time.
2315 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2316 Candler <B.Candler@pobox.com>.
2318 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2319 OK for addresses that are the result of redirections.
2321 PH/48 A number of further additions for the benefit of the new test suite,
2322 including a fake gethostbyname() that interfaces to the fake DNS resolver
2325 TF/06 The fix for widen_domains has also been applied to qualify_single and
2326 search_parents which are the other dnslookup options that can cause
2329 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2332 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2334 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2335 service" if it encountered an MX record with an empty target hostname.
2336 The message is now "an MX or SRV record indicated no SMTP service".
2338 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2339 if the DNS of the sending site is misconfigured. This is quite a
2340 common situation. This change restores the behaviour of treating a
2341 helo verification defer as a failure.
2343 PH/51 If self=fail was set on a router, the bounce message did not include the
2344 actual error message.
2350 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2352 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2353 command itself ended in error, the underlying message about the transport
2354 filter timeout was being overwritten with the pipe command error. Now the
2355 underlying error message should be appended to the second error message.
2357 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2359 PH/02 Reduce the amount of output that "make" produces by default. Full output
2360 can still be requested.
2362 PH/03 The warning log line about a condition test deferring for a "warn" verb
2363 was being output only once per connection, rather than after each
2364 occurrence (because it was using the same function as for successful
2365 "warn" verbs). This seems wrong, so I have changed it.
2367 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2368 it should not have, which might have caused a crash in the right
2369 circumstances, but probably never did.
2371 PH/04 Installed a modified version of Tony Finch's patch to make submission
2372 mode fix the return path as well as the Sender: header line, and to
2373 add a /name= option so that you can make the user's friendly name appear
2376 TF/03 Added the control = fakedefer ACL modifier.
2378 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2379 Mark Lowes for thorough testing.
2381 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2383 TK/03 Merged latest SRS patch from Miles Wilton.
2385 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2386 with the definition in sysexits.h (which is #included earlier).
2387 Fortunately, Exim does not actually use EX_OK. The code used to try to
2388 preserve the sysexits.h value, by assumimg that macro definitions were
2389 scanned for macro replacements. I have been disabused of this notion,
2390 so now the code just undefines EX_OK before #including unistd.h.
2392 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2393 in the smtp transport. When a block could not be written in a single
2394 write() function, the timeout was being re-applied to each part-write.
2395 This seems wrong - if the receiver was accepting one byte at a time it
2396 would take for ever. The timeout is now adjusted when this happens. It
2397 doesn't have to be particularly precise.
2399 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2400 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2402 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2403 without a domain if the check to <postmaster@domain> fails.
2405 SC/01 Eximstats: added -xls and the ability to specify output files
2406 (patch written by Frank Heydlauf).
2408 SC/02 Eximstats: use FileHandles for outputing results.
2410 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2412 SC/04 Eximstats: fixed display of large numbers with -nvr option
2414 SC/05 Eximstats: fixed merging of reports with empty tables.
2416 SC/06 Eximstats: added the -include_original_destination flag
2418 SC/07 Eximstats: removed tabs and trailing whitespace.
2420 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2422 TK/06 MBOX spool code: Add real "From " MBOX separator line
2423 so the .eml file is really in mbox format (even though
2424 most programs do not really care). Patch from Alex Miller.
2426 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2427 The latter is generated from $received_to and is only set if the
2428 message has one envelope recipient. SA can use these headers,
2429 obviously out-of-the-box. Patch from Alex Miller.
2431 PH/08 The ${def test on a variable was returning false if the variable's
2432 value was "0", contrary to what the specification has always said!
2433 The result should be true unless the variable is empty.
2435 PH/09 The syntax error of a character other than { following "${if
2436 def:variable_name" (after optional whitespace) was not being diagnosed.
2437 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2438 accidental colon was present, for example, could give incorrect results.
2440 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2441 result is used (not including appendfile, where other changes are about
2444 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2445 This involved changing a lot of size variables from int to off_t. It
2446 should work with maildirs and everything.
2448 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2449 spamd dying while we are connected to it.
2451 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2454 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2455 function for BDB 4.3.
2457 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2459 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2460 the sources for PCRE have been re-arranged and more files are now
2463 PH/15 The code I had for printing potentially long long variables in PH/11
2464 above was not the best (it lost precision). The length of off_t variables
2465 is now inspected at build time, and an appropriate printing format (%ld
2466 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2467 to be "long long int" or "long int". This is needed for the internal
2468 formatting function string_vformat().
2470 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2471 the configuration file to be ":syslog", then the script "guesses" where
2472 the logs files are, rather than using the compiled in default. In our
2473 case the guess is not the same as the compiled default, so the script
2474 suddenly stopped working when I started to use syslog. The patch checks
2475 to see if log_file_path is "". If so, it attempts to read it from exim
2476 with no configuration file to get the compiled in version, before it
2477 falls back to the previous guessing code."
2479 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2480 implementing BATV in an Exim configuration. See NewStuff for the gory
2483 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2484 Makefile that are specific to HP-UX.
2486 PH/18 If the "use_postmaster" option was set for a recipient callout together
2487 with the "random" option, the postmaster address was used as the MAIL
2488 FROM address for the random test, but not for the subsequent recipient
2489 test. It is now used for both.
2491 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2492 patch removes a few documentation additions to RFC 3028, because the
2493 latest draft now contains them. It adds the new en;ascii-case comparator
2494 and a new error check for 8bit text in MIME parts. Comparator and
2495 require names are now matched exactly. I enabled the subaddress
2496 extension, but it is not well tested yet (read: it works for me)."
2498 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2499 rework some of the code of TK/09 above to avoid the hardwired use of
2500 "%lld" and "long long". Replaced the call to snprintf() with a call to
2503 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2504 records point to non-existent hosts", "retry timeout exceeded", and
2505 "retry time not reached for any host after a long failure period".
2507 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2508 experimental DomainKeys support:
2510 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2511 (2) On an error such as an illegally used "control", the wrong name for
2512 the control was given.
2514 These problems did NOT occur unless DomainKeys support was compiled.
2516 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2518 PH/24 Added ${if match_ip condition.
2520 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2521 databases so that it will be absolutely obvious if a crash occurs in the
2522 DB library. This is a regular occurrence (often caused by mis-matched
2525 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2526 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2527 picked up on a user's system that detects such things. There doesn't seem
2528 to be a gcc warning option for this - only an attribute that has to be
2529 put on the function's prototype. It seems that in Fedora Core 4 they have
2530 set this on a number of new functions. No doubt there will be more in due
2533 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2534 specify a transport. However, if an address that was verified by such a
2535 router was the subject of a callout, Exim crashed because it tried to
2536 read the rcpt_include_affixes from the non-existent transport. Now it
2537 just assumes that the setting of that option is false. This bug was
2538 introduced by 4.51/PH/31.
2540 PH/28 Changed -d+all to exclude +memory, because that information is very
2541 rarely of interest, but it makes the output a lot bigger. People tend to
2542 do -d+all out of habit.
2544 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2545 code in os-type was giving problems when libc.so lives in lib64, like on
2548 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2549 aren't the modern standard, and it seems that some systems' include files
2550 don't always have them. Exim was already checking for some of the newer
2551 ones like T_AAAA, and defining it itself. I've added checks for all the
2552 record types that Exim uses.
2554 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2555 not automatically generating a new one, as it is supposed to. This
2556 prevented TLS from working. If the file did exist, but contained invalid
2557 data, a new version was generated, as expected. It was only the case of a
2558 non-existent file that was broken.
2560 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2561 with a change in libdomainkeys > 0.64.
2563 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2564 from DNS. If the selector record carries the flag, it now has
2565 precedence over the domain-wide flag.
2567 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2569 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2570 the use of a port name that isn't defined in /etc/services) occurred, the
2571 message was deferred as in a normal delivery, and thus remained on the
2572 spool, instead of being failed because of the mua_wrapper setting. This
2573 is now fixed, and I tidied up some of the mua_wrapper messages at the
2576 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2577 the messages in a hash of arrays rather than using individual hashes.
2578 This is a bit cleaner and results in dramatic memory savings, albeit
2579 at a slight CPU cost.
2581 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2582 as requested by Marc Sherman.
2584 SC/10 Eximstats: added histograms for user specified patterns as requested
2587 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2589 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2590 fopen() in the content-scanning modules that did not already have it.
2596 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2597 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2599 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2601 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2603 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2604 to test Sieve filters that use "vacation".
2606 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2607 that changes the way the GnuTLS parameters are stored in the cache file.
2608 The new format can be generated externally. For backward compatibility,
2609 if the data in the cache doesn't make sense, Exim assumes it has read an
2610 old-format file, and it generates new data and writes a new file. This
2611 means that you can't go back to an older release without removing the
2614 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2615 work if there are any delivery delays because "one_time" forces the
2616 parent to be marked "delivered", so its unseen clone is never tried
2617 again. For this reason, Exim now forbids the simultaneous setting of
2620 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2621 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2622 entry if you want to know the details. The fix, however, neglected to
2623 consider the case when local delivery batching is involved. The test for
2624 "previously delivered" was not happening when checking to see if an
2625 address could be batched with a previous (undelivered) one; under
2626 certain circumstances this could lead to multiple deliveries to the same
2629 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2630 in its include files, and this causes problems building Exim.
2632 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2633 header_syntax) but Exim was just ignoring anything given after a slash.
2634 In particular, this caused confusion with an attempt to use "verify =
2635 reverse_host_lookup/defer_ok". An error is now given when options are
2636 supplied for verify items that do not have them. (Maybe reverse_host_
2637 lookup should have a defer_ok option, but that's a different point.)
2639 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2640 defined by RFC 821) to 2048, because there were problems with some AUTH
2641 commands, and RFC 1869 says the size should be increased for extended
2642 SMTP commands that take arguments.
2644 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2647 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2648 "unknown" error; now it says that the functionality isn't in the binary.
2650 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2651 an address' error message when a string expansion fails (syntax or
2652 whatever). Otherwise the password may appear in the log. Following change
2653 PH/42 below, there is no longer a chance of it appearing in a bounce
2656 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2658 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2659 of its addresses was ever tried. (Bugzilla bug #2.)
2661 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2662 the result incorrectly in the debug output. (It correctly added a newline
2663 to what was transported.)
2665 TF/01 Added $received_time.
2667 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2668 commented out examples of how to interface to a virus scanner and to
2669 SpamAssassin. Also added commented examples of av_scanner and
2670 spamd_address settings.
2672 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2673 and controls are allowed in which ACLs. There were a couple of minor
2674 errors. Some of the entries in the conditions table (which is a table of
2675 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2676 negation of where the condition IS allowed.
2678 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2680 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2681 header file does not have a version number, so I've had to invent a new
2682 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2683 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2684 radiusclient), but it was contributed by a Radius user.
2686 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2687 files or directories, for OpenSSL.
2689 PH/19 When an Exim process that is running as root has to create an Exim log
2690 file, it does so in a subprocess that runs as exim:exim so as to get the
2691 ownership right at creation (otherwise, other Exim processes might see
2692 the file with the wrong ownership). There was no test for failure of this
2693 fork() call, which would lead to the process getting stuck as it waited
2694 for a non-existent subprocess. Forks do occasionally fail when resources
2695 run out. I reviewed all the other calls to fork(); they all seem to check
2698 PH/20 When checking for unexpected SMTP input at connect time (before writing
2699 the banner), Exim was not dealing correctly with a non-positive return
2700 from the read() function. If the client had disconnected by this time,
2701 the result was a log entry for a synchronization error with an empty
2702 string after "input=" when read() returned zero. If read() returned -1
2703 (an event I could not check), uninitialized data bytes were printed.
2704 There were reports of junk text (parts of files, etc) appearing after
2707 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2709 PH/22 Added support for macro redefinition, and (re)definition in between
2710 driver and ACL definitions.
2712 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2713 forgetting to use the resulting value; it was using the unexpanded value.
2715 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2716 hadn't been configured. The fix is from Juergen Kreileder, who
2717 understands it better than I do:
2719 "Here's what I see happening with three configured cyrus_sasl
2720 authenticators configured (plain, login, cram-md5):
2722 On startup auth_cyrus_sasl_init() gets called for each of these.
2723 This means three calls to sasl_listmech() without a specified mech_list.
2724 => SASL tests which mechs of all available mechs actually work
2725 => three warnings about OTP not working
2726 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2728 With the patch, sasl_listmech() also gets called three times. But now
2729 SASL's mech_list option is set to the server_mech specified in the the
2730 authenticator. Or in other words, the answer from sasl_listmech()
2731 gets limited to just the mech you're testing for (which is different
2733 => the return list contains just 'plain' or 'login', 'cram-md5' or
2734 nothing depending on the value of ob->server_mech.
2736 I've just tested the patch: Authentication still works fine,
2737 unavailable mechs specified in the exim configuration are still
2738 caught, and the auth.log warnings about OTP are gone."
2740 PH/25 When debugging is enabled, the contents of the command line are added
2741 to the debugging output, even when log_selector=+arguments is not
2744 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2745 answer is "GNU", and only if the return is "GNU/something" is the answer
2748 PH/27 $acl_verify_message is now set immediately after the failure of a
2749 verification in an ACL, and so is available in subsequent modifiers. In
2750 particular, the message can be preserved by coding like this:
2752 warn !verify = sender
2753 set acl_m0 = $acl_verify_message
2755 Previously, $acl_verify_message was set only while expanding "message"
2756 and "log_message" when a very denied access.
2758 PH/28 Modified OS/os.c-Linux with
2760 -#ifndef OS_LOAD_AVERAGE
2761 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2763 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2764 nomenclature these days.)
2766 PH/29 Installed patch from the Sieve maintainer that adds the options
2767 sieve_useraddress and sieve_subaddress to the redirect router.
2769 PH/30 In these circumstances:
2770 . Two addresses routed to the same list of hosts;
2771 . First host does not offer TLS;
2772 . First host accepts first address;
2773 . First host gives temporary error to second address;
2774 . Second host offers TLS and a TLS session is established;
2775 . Second host accepts second address.
2776 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2777 and peerdn, if requested) that were in fact used only for the second
2780 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2781 attention to any local part prefix or suffix that was matched by the
2782 router that accepted the address. It now behaves in the same way as it
2783 does for delivery: the affixes are removed from the local part unless
2784 rcpt_include_affixes is set on the transport.
2786 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2787 timeout during the DATA phase of an incoming message.
2789 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2790 applied a patch sanctioned by the Sieve maintainer.
2792 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2793 the uid or gid is negative. A case of a negative gid caused this to be
2794 noticed. The fix allows for either to be negative.
2796 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2797 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2800 PH/36 The change PH/12 above was broken. Fixed it.
2802 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2803 the grounds that routing the same address twice would always produce the
2804 same answer. This might have been true once, but it is certainly no
2805 longer true now. Routing a child address may depend on the previous
2806 routing that produced that child. Some complicated redirection strategies
2807 went wrong when messages had multiple recipients, and made Exim's
2808 behaviour dependent on the order in which the addresses were given.
2810 I have moved the duplicate checking until after the routing is complete.
2811 Exim scans the addresses that are assigned to local and remote
2812 transports, and removes any duplicates. This means that more work will be
2813 done, as duplicates will always all be routed, but duplicates are
2814 presumably rare, so I don't expect this is of any significance.
2816 For deliveries to pipes, files, and autoreplies, the duplicate checking
2817 still happens during the routing process, since they are not going to be
2820 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2821 It corrects a timeout issue with spamd. This is Ian's comment: "The
2822 background is that sometimes spamd either never reads data from a
2823 connection it has accepted, or it never writes response data. The exiscan
2824 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2825 blindly assumes that writes won't block so it may never time out."
2827 PH/39 Allow G after quota size as well as K and M.
2829 PH/40 The value set for $authenticated_id in an authenticator may not contain
2830 binary zeroes or newlines because the value is written to log lines and
2831 to spool files. There was no check on this. Now the value is run through
2832 the string_printing() function so that such characters are converted to
2833 printable escape sequences.
2835 PH/41 $message_linecount is a new variable that contains the total number of
2836 lines in the message. Compare $body_linecount, which is the count for the
2839 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2840 bounce and delay warning messages, except in certain special cases, which
2843 (a) An SMTP error message from a remote host;
2844 (b) A message specified in a :fail: redirection;
2845 (c) A message specified in a "fail" command in a system filter;
2846 (d) A message specified in a FAIL return from the queryprogram router;
2847 (e) A message specified by the cannot_route_message router option.
2849 In these cases only, Exim does include the error details in bounce and
2850 warning messages. There are also a few cases where bland messages such
2851 as "unrouteable address" or "local delivery error" are given.
2853 PH/43 $value is now also set for the "else" part of a ${run expansion.
2855 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2856 being worked on, but at least Exim now implements the latest version to
2859 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2860 process to complete was treated as a delivery failure, a timeout while
2861 writing the message to the pipe was logged, but erroneously treated as a
2862 successful delivery. Such timeouts include transport filter timeouts. For
2863 consistency with the overall process timeout, these timeouts are now
2864 treated as errors, giving rise to delivery failures by default. However,
2865 there is now a new Boolean option for the pipe transport called
2866 timeout_defer, which, if set TRUE, converts the failures into defers for
2867 both kinds of timeout. A transport filter timeout is now identified in
2870 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2871 systems where "make" and "gmake" are different, calling "gmake" at top
2872 level broke things. I've arranged for the value of $(MAKE) to be passed
2873 from the Makefile to this script so that it can call the same version of
2877 A note about Exim versions 4.44 and 4.50
2878 ----------------------------------------
2880 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2881 changes of various kinds. As a consequence, a big documentation update was
2882 needed. This delayed the release for rather longer than seemed good, especially
2883 in the light of a couple of (minor) security issues. Therefore, the changes
2884 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2885 release. So 4.44 and 4.50 are in effect two different branches that both start
2888 I have left the 4.50 change log unchanged; it contains all the changes since
2889 4.43. The change log for 4.44 is below; many of its items are identical to
2890 those for 4.50. This seems to be the most sensible way to preserve the
2891 historical information.
2897 1. Minor wording change to the doc/README.SIEVE file.
2899 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2900 computation of the current number of files was incorrect.
2902 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2903 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2906 4. Give more explanation in the error message when the command for a transport
2907 filter fails to execute.
2909 5. There are several places where Exim runs a non-Exim command in a
2910 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2911 was being done only for the command run by the queryprogram router. It is
2912 now done for all such subprocesses. The other cases are: ${run, transport
2913 filters, and the commands run by the lmtp and pipe transports.
2915 6. Added CONFIGURE_GROUP build-time option.
2917 7. Some older OS have a limit of 256 on the maximum number of file
2918 descriptors. Exim was using setrlimit() to set 1000 as a large value
2919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2920 systems. I've change it so that if it can't get 1000, it tries for 256.
2922 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2923 was an oversight, and furthermore, ever since the addition of extra
2924 controls (e.g. 4.43/32), the checks on when to allow different forms of
2925 "control" were broken. There should now be diagnostics for all cases when a
2926 control that does not make sense is encountered.
2928 9. Added the /retain_sender option to "control=submission".
2930 10. $recipients is now available in the predata ACL (oversight).
2932 11. Tidy the search cache before the fork to do a delivery from a message
2933 received from the command line. Otherwise the child will trigger a lookup
2934 failure and thereby defer the delivery if it tries to use (for example) a
2935 cached ldap connection that the parent has called unbind on.
2937 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2938 of $address_data from the recipient verification was clobbered by the
2939 sender verification.
2941 13. The value of address_data from a sender verification is now available in
2942 $sender_address_data in subsequent conditions in the ACL statement.
2944 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2946 15. Added a new option "connect=<time>" to callout options, to set a different
2949 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2950 was its contents. (It was OK if the option was not defined at all.)
2952 17. A "Completed" log line is now written for messages that are removed from
2953 the spool by the -Mrm option.
2955 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2956 information about exactly what failed.
2958 19. Added -dd to debug only the daemon process.
2960 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2961 handles timeouts, both on the server side and network timeouts. Renamed the
2962 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2964 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2965 It is now set to "smtps".
2967 22. $host_address is now set to the target address during the checking of
2968 ignore_target_hosts.
2970 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2971 being passed; this would have caused $sender_host_name to have been used if
2972 matching the list had actually called for a host name (not very likely,
2973 since this list is usually IP addresses). A host name is now passed as
2976 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2977 code that specifies a non-restarting handler (typically sigaction() in
2978 modern systems) in an attempt to fix a rare and obscure crash bug.
2980 25. Narrowed the window for a race in the daemon that could cause it to ignore
2981 SIGCHLD signals. This is not a major problem, because they are used only to
2982 wake it up if nothing else does.
2984 26. A malformed maildirsize file could cause Exim to calculate negative values
2985 for the mailbox size or file count. Odd effects could occur as a result.
2986 The maildirsize information is now recalculated if the size or filecount
2989 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2990 support for a long time. Removed HAVE_SYS_VFS_H.
2992 28. Installed the latest version of exipick from John Jetmore.
2994 29. In an address list, if the pattern was not a regular expression, an empty
2995 subject address (from a bounce message) matched only if the pattern was an
2996 empty string. Non-empty patterns were not even tested. This was the wrong
2997 because it is perfectly reasonable to use an empty address as part of a
2998 database query. An empty address is now tested by patterns that are
2999 lookups. However, all the other forms of pattern expect the subject to
3000 contain a local part and a domain, and therefore, for them, an empty
3001 address still always fails if the pattern is not itself empty.
3003 30. Exim went into a mad DNS loop when attempting to do a callout where the
3004 host was specified on an smtp transport, and looking it up yielded more
3005 than one IP address.
3007 31. Re-factored the code for checking spool and log partition space into a
3008 function that finds that data and another that does the check. The former
3009 is then used to implement four new variables: $spool_space, $log_space,
3010 $spool_inodes, and $log_inodes.
3012 32. The RFC2047 encoding function was originally intended for short strings
3013 such as real names; it was not keeping to the 75-character limit for
3014 encoded words that the RFC imposes. It now respects the limit, and
3015 generates multiple encoded words if necessary. To be on the safe side, I
3016 have increased the buffer size for the ${rfc2047: expansion operator from
3019 33. It is now permitted to omit both strings after an "if" condition; if the
3020 condition is true, the result is "true". As before, when the second string
3021 is omitted, a false condition yields an empty string. This makes it less
3022 cumbersome to write custom ACL and router conditions.
3024 34. Failure to deliver a bounce message always caused it to be frozen, even if
3025 there was an errors_to setting on the router. The errors_to setting is now
3028 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3029 canonical form (fully expanded) before being placed in
3030 $sender_host_address.
3032 36. The table in the code that translates DNS record types into text (T_A to
3033 "A" for instance) was missing entries for NS and CNAME. It is just possible
3034 that this could have caused confusion if both these types were looked up
3035 for the same domain, because the text type is used as part of Exim's
3036 per-process caching. But the chance of anyone hitting this buglet seems
3039 37. The dnsdb lookup has been extended in a number of ways.
3041 (1) There is a new type, "zns", which walks up the domain tree until it
3042 finds some nameserver records. It should be used with care.
3044 (2) There is a new type, "mxh", which is like "mx" except that it returns
3045 just the host names, not the priorities.
3047 (3) It is now possible to give a list of domains (or IP addresses) to be
3048 looked up. The behaviour when one of the lookups defers can be
3049 controlled by a keyword.
3051 (4) It is now possible to specify the separator character for use when
3052 multiple records are returned.
3054 38. The dnslists ACL condition has been extended: it is now possible to supply
3055 a list of IP addresses and/or domains to be looked up in a particular DNS
3058 39. Added log_selector=+queue_time_overall.
3060 40. When running the queue in the test harness, wait just a tad after forking a
3061 delivery process, to get repeatability of debugging output.
3063 41. Include certificate and key file names in error message when GnuTLS fails
3064 to set them up, because the GnuTLS error message doesn't include the name
3065 of the failing file when there is a problem reading it.
3067 42. Allow both -bf and -bF in the same test run.
3069 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3071 44. The "Exiscan patch" is now merged into the mainline Exim source.
3073 45. Sometimes the final signoff response after QUIT could fail to get
3074 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3075 < 0 before doing a fflush(). This bug looks as though it goes back to the
3076 introduction of TLS in release 3.20, but "sometimes" must have been rare
3077 because the tests only now provoked it.
3079 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3080 (this can affect the format of dates).
3082 47. exim_tidydb, when checking for the continued existence of a message for
3083 which it has found a message-specific retry record, was not finding
3084 messages that were in split spool directories. Consequently, it was
3085 deleting retry records that should have stayed in existence.
3087 48. Steve fixed some bugs in eximstats.
3089 49. The SPA authentication driver was not abandoning authentication and moving
3090 on to the next authenticator when an expansion was forced to fail,
3091 contradicting the general specification for all authenticators. Instead it
3092 was generating a temporary error. It now behaves as specified.
3094 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3095 (the order specifies the preference for clients). The order is now AES256,
3096 AES128, 3DES, ARCFOUR128.
3098 51. Small patch to Sieve code - explicitly set From: when generating an
3101 52. Exim crashed if a remote delivery caused a very long error message to be
3102 recorded - for instance if somebody sent an entire SpamAssassin report back
3103 as a large number of 550 error lines. This bug was coincidentally fixed by
3104 increasing the size of one of Exim's internal buffers (big_buffer) that
3105 happened as part of the Exiscan merge. However, to be on the safe side, I
3106 have made the code more robust (and fixed the comments that describe what
3109 53. Now that there can be additional text after "Completed" in log lines (if
3110 the queue_time_overall log selector is set), a one-byte patch to exigrep
3111 was needed to allow it to recognize "Completed" as not the last thing in
3114 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3115 patch that reportedly fixes this has been added. I am not expert enough to
3116 create a test for it. This is what the patch creator wrote:
3118 "I found a little strange behaviour of ldap code when working with
3119 Windows 2003 AD Domain, where users was placed in more than one
3120 Organization Units. When I tried to give exim partial DN, the exit code
3121 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3122 But simultaneously result of request was absolutely normal ldap result,
3123 so I produce this patch..."
3125 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3126 REFERENCE, so I have modified the code to exclude the patch when that macro
3129 55. Some experimental protocols are using DNS PTR records for new purposes. The
3130 keys for these records are domain names, not reversed IP addresses. The
3131 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3132 leaves it alone. Component reversal etc. now happens only for IP addresses.
3135 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3137 57. Double the size of the debug message buffer (to 2048) so that more of very
3138 long debug lines gets shown.
3140 58. The exicyclog utility now does better if the number of log files to keep
3141 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3143 59. Two changes related to the smtp_active_hostname option:
3145 (1) $smtp_active_hostname is now available as a variable.
3146 (2) The default for smtp_banner uses $smtp_active_hostname instead
3147 of $primary_hostname.
3149 60. The host_aton() function is supposed to be passed a string that is known
3150 to be a valid IP address. However, in the case of IPv6 addresses, it was
3151 not checking this. This is a hostage to fortune. Exim now panics and dies
3152 if the condition is not met. A case was found where this could be provoked
3153 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3154 components; fortuitously, this particular loophole had already been fixed
3155 by change 4.50/55 above.
3157 If there are any other similar loopholes, the new check in host_aton()
3158 itself should stop them being exploited. The report I received stated that
3159 data on the command line could provoke the exploit when Exim was running as
3160 exim, but did not say which command line option was involved. All I could
3161 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3162 running as the user.
3165 61. There was a buffer overflow vulnerability in the SPA authentication code
3166 (which came originally from the Samba project). I have added a test to the
3167 spa_base64_to_bits() function which I hope fixes it.
3170 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3171 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3173 63. The daemon start-up calls getloadavg() while still root for those OS that
3174 need the first call to be done as root, but it missed one case: when
3175 deliver_queue_load_max is set with deliver_drop_privilege. This is
3176 necessary for the benefit of the queue runner, because there is no re-exec
3177 when deliver_drop_privilege is set.
3179 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3180 This has been fixed.
3182 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3183 was in use, was not putting the data itself into the right store pool;
3184 consequently, it could be overwritten for a subsequent message in the same
3185 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3188 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3190 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3191 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3192 the function were treating the return as a boolean value, which happened to
3193 work because 0=false and not-0=true, but is not correct code.
3195 68. The host_aton() function was not handling scoped IPv6 addresses (those
3196 with, for example, "%eth0" on the end) correctly.
3198 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3199 negated items (that is, ~something) in unsigned ints. Some compilers
3200 apparently mutter when there is no cast.
3202 70. If an address verification called from an ACL failed, and did not produce a
3203 user-specific message (i.e. there was only a "system" message), nothing was
3204 put in $acl_verify_message. In this situation, it now puts the system
3207 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3208 session; change 4.31/43 added the unwanted input to the log line - except
3209 that it did not do this in the start of session case. It now does.
3211 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3212 This is wrong and can cause the other end to generate a synchronization
3213 error if it is another Exim or anything else that does the synchronization
3214 check. A QUIT command is no longer sent after a timeout.
3216 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3217 during host lookups.
3219 74. The defer_ok option of callout verification was not working if it was used
3220 when verifying addresses in header lines, that is, for this case:
3222 verify = header_sender/callout=defer_ok
3224 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3225 those file descriptors could be used for SMTP connections. If anything
3226 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3227 could be sent to the SMTP client, causing chaos. The daemon now opens
3228 stdin, stdout, and stderr to /dev/null when it puts itself into the
3231 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3232 log by default. The user can override this with suitable Perl magic.
3234 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3235 the log message when discard triggers, was not working for the DATA ACL or
3236 for the non-SMTP ACL.
3238 78. Error message wording change in sieve.c.
3240 79. If smtp_accept_max_per_host was set, the number of connections could be
3241 restricted to fewer than expected, because the daemon was trying to set up
3242 a new connection before checking whether the processes handling previous
3243 connections had finished. The check for completed processes is now done
3244 earlier. On busy systems, this bug wouldn't be noticed because something
3245 else would have woken the daemon, and it would have reaped the completed
3248 80. If a message was submitted locally by a user whose login name contained one
3249 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3250 It caused a spool format error. I have fixed the spool reading code. A
3251 related problem was that the "from" clause in the Received: line became
3252 illegal because of the space(s). It is now covered by ${quote_local_part.
3254 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3257 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3259 83. Patch from Sieve maintainer to fix unterminated string problem in
3260 "vacation" handling.
3262 84. Some minor changes to the Linux configuration files to help with other
3263 OS variants using glibc.
3265 85. One more patch for Sieve to update vacation handling to latest spec.
3268 ----------------------------------------------------
3269 See the note above about the 4.44 and 4.50 releases.
3270 ----------------------------------------------------
3276 1. Change 4.43/35 introduced a bug that caused file counts to be
3277 incorrectly computed when quota_filecount was set in an appendfile
3280 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3281 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3284 3. Give more explanation in the error message when the command for a transport
3285 filter fails to execute.
3287 4. There are several places where Exim runs a non-Exim command in a
3288 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3289 was being done only for the command run by the queryprogram router. It is
3290 now done for all such subprocesses. The other cases are: ${run, transport
3291 filters, and the commands run by the lmtp and pipe transports.
3293 5. Some older OS have a limit of 256 on the maximum number of file
3294 descriptors. Exim was using setrlimit() to set 1000 as a large value
3295 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3296 systems. I've change it so that if it can't get 1000, it tries for 256.
3298 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3299 was an oversight, and furthermore, ever since the addition of extra
3300 controls (e.g. 4.43/32), the checks on when to allow different forms of
3301 "control" were broken. There should now be diagnostics for all cases when a
3302 control that does not make sense is encountered.
3304 7. $recipients is now available in the predata ACL (oversight).
3306 8. Tidy the search cache before the fork to do a delivery from a message
3307 received from the command line. Otherwise the child will trigger a lookup
3308 failure and thereby defer the delivery if it tries to use (for example) a
3309 cached ldap connection that the parent has called unbind on.
3311 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3312 of $address_data from the recipient verification was clobbered by the
3313 sender verification.
3315 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3316 was its contents. (It was OK if the option was not defined at all.)
3318 11. A "Completed" log line is now written for messages that are removed from
3319 the spool by the -Mrm option.
3321 12. $host_address is now set to the target address during the checking of
3322 ignore_target_hosts.
3324 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3325 being passed; this would have caused $sender_host_name to have been used if
3326 matching the list had actually called for a host name (not very likely,
3327 since this list is usually IP addresses). A host name is now passed as
3330 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3331 code that specifies a non-restarting handler (typically sigaction() in
3332 modern systems) in an attempt to fix a rare and obscure crash bug.
3334 15. Narrowed the window for a race in the daemon that could cause it to ignore
3335 SIGCHLD signals. This is not a major problem, because they are used only to
3336 wake it up if nothing else does.
3338 16. A malformed maildirsize file could cause Exim to calculate negative values
3339 for the mailbox size or file count. Odd effects could occur as a result.
3340 The maildirsize information is now recalculated if the size or filecount
3343 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3344 support for a long time. Removed HAVE_SYS_VFS_H.
3346 18. Updated exipick to current release from John Jetmore.
3348 19. Allow an empty sender to be matched against a lookup in an address list.
3349 Previously the only cases considered were a regular expression, or an
3352 20. Exim went into a mad DNS lookup loop when doing a callout where the
3353 host was specified on the transport, if the DNS lookup yielded more than
3356 21. The RFC2047 encoding function was originally intended for short strings
3357 such as real names; it was not keeping to the 75-character limit for
3358 encoded words that the RFC imposes. It now respects the limit, and
3359 generates multiple encoded words if necessary. To be on the safe side, I
3360 have increased the buffer size for the ${rfc2047: expansion operator from
3363 22. Failure to deliver a bounce message always caused it to be frozen, even if
3364 there was an errors_to setting on the router. The errors_to setting is now
3367 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3368 canonical form (fully expanded) before being placed in
3369 $sender_host_address.
3371 24. Updated eximstats to version 1.33
3373 25. Include certificate and key file names in error message when GnuTLS fails
3374 to set them up, because the GnuTLS error message doesn't include the name
3375 of the failing file when there is a problem reading it.
3377 26. Expand error message when OpenSSL has problems setting up cert/key files.
3380 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3381 (this can affect the format of dates).
3383 28. exim_tidydb, when checking for the continued existence of a message for
3384 which it has found a message-specific retry record, was not finding
3385 messages that were in split spool directories. Consequently, it was
3386 deleting retry records that should have stayed in existence.
3388 29. eximstats updated to version 1.35
3389 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3390 1.35 - bugfix such that pie charts by volume are generated correctly
3392 30. The SPA authentication driver was not abandoning authentication and moving
3393 on to the next authenticator when an expansion was forced to fail,
3394 contradicting the general specification for all authenticators. Instead it
3395 was generating a temporary error. It now behaves as specified.
3397 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3398 (the order specifies the preference for clients). The order is now AES256,
3399 AES128, 3DES, ARCFOUR128.
3401 31. Small patch to Sieve code - explicitly set From: when generating an
3404 32. Exim crashed if a remote delivery caused a very long error message to be
3405 recorded - for instance if somebody sent an entire SpamAssassin report back
3406 as a large number of 550 error lines. This bug was coincidentally fixed by
3407 increasing the size of one of Exim's internal buffers (big_buffer) that
3408 happened as part of the Exiscan merge. However, to be on the safe side, I
3409 have made the code more robust (and fixed the comments that describe what
3412 33. Some experimental protocols are using DNS PTR records for new purposes. The
3413 keys for these records are domain names, not reversed IP addresses. The
3414 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3415 leaves it alone. Component reversal etc. now happens only for IP addresses.
3418 34. The host_aton() function is supposed to be passed a string that is known
3419 to be a valid IP address. However, in the case of IPv6 addresses, it was
3420 not checking this. This is a hostage to fortune. Exim now panics and dies
3421 if the condition is not met. A case was found where this could be provoked
3422 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3423 components; fortuitously, this particular loophole had already been fixed
3424 by change 4.50/55 or 4.44/33 above.
3426 If there are any other similar loopholes, the new check in host_aton()
3427 itself should stop them being exploited. The report I received stated that
3428 data on the command line could provoke the exploit when Exim was running as
3429 exim, but did not say which command line option was involved. All I could
3430 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3431 running as the user.
3434 35. There was a buffer overflow vulnerability in the SPA authentication code
3435 (which came originally from the Samba project). I have added a test to the
3436 spa_base64_to_bits() function which I hope fixes it.
3439 36. The daemon start-up calls getloadavg() while still root for those OS that
3440 need the first call to be done as root, but it missed one case: when
3441 deliver_queue_load_max is set with deliver_drop_privilege. This is
3442 necessary for the benefit of the queue runner, because there is no re-exec
3443 when deliver_drop_privilege is set.
3445 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3446 was in use, was not putting the data itself into the right store pool;
3447 consequently, it could be overwritten for a subsequent message in the same
3448 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3451 38. Sometimes the final signoff response after QUIT could fail to get
3452 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3453 < 0 before doing a fflush(). This bug looks as though it goes back to the
3454 introduction of TLS in release 3.20, but "sometimes" must have been rare
3455 because the tests only now provoked it.
3461 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3462 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3463 Now it returns an integer. A number of calls were still expecting a T/F
3464 return. Fortuitously, in all cases, the tests worked in OK situations,
3465 which is the norm. However, things would have gone wrong on any write
3466 failures on the smtp file descriptor. This function is used when sending
3467 messages over SMTP and also when doing verify callouts.
3469 2. When Exim is called to do synchronous delivery of a locally submitted
3470 message (the -odf or -odi options), it no longer closes stderr before doing
3473 3. Implemented the mua_wrapper option.
3475 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3477 5. Implemented the functions header_remove(), header_testname(),
3478 header_add_at_position(), and receive_remove_recipient(), and exported them
3481 6. If an ACL "warn" statement specified the addition of headers, Exim already
3482 inserted X-ACL-Warn: at the start if there was no header name. However, it
3483 was not making this test for the second and subsequent header lines if
3484 there were newlines in the string. This meant that an invalid header could
3485 be inserted if Exim was badly configured.
3487 7. Allow an ACL "warn" statement to add header lines at the start or after all
3488 the Received: headers, as well as at the end.
3490 8. Added the rcpt_4xx retry error code.
3492 9. Added postmaster_mailfrom=xxx to callout verification option.
3494 10. Added mailfrom=xxxx to the callout verification option, for verify=
3497 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3498 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3500 12. Inserted some casts to stop certain compilers warning when using pointer
3501 differences as field lengths or precisions in printf-type calls (mostly
3502 affecting debugging statements).
3504 13. Added optional readline() support for -be (dynamically loaded).
3506 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3507 same clock tick as a message's arrival, so that its received time was the
3508 same as the "first fail" time on the retry record, and that message
3509 remained on the queue past the ultimate address timeout, every queue runner
3510 would try a delivery (because it was past the ultimate address timeout) but
3511 after another failure, the ultimate address timeout, which should have then
3512 bounced the address, did not kick in. This was a "< instead of <=" error;
3513 in most cases the first failure would have been in the next clock tick
3514 after the received time, and all would be well.
3516 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3517 being recognized when the domain list was tested by the match_domain
3518 condition in an expansion string.
3520 16. Added the ${str2b64: operator.
3522 17. Exim was always calling setrlimit() to set a large limit for the number of
3523 processes, without checking whether the existing limit was already
3524 adequate. (It did check for the limit on file descriptors.) Furthermore,
3525 errors from getrlimit() and setrlimit() were being ignored. Now they are
3526 logged to the main and panic logs, but Exim does carry on, to try to do its
3527 job under whatever limits there are.
3529 18. Imported PCRE 5.0.
3531 19. Trivial typo in log message " temporarily refused connection" (the leading
3534 20. If the log selector return_path_on_delivery was set and an address was
3535 redirected to /dev/null, the delivery process crashed because it assumed
3536 that a return path would always be set for a "successful" delivery. In this
3537 case, the whole delivery is bypassed as an optimization, and therefore no
3540 21. Internal re-arrangement: the function for sending a challenge and reading
3541 a response while authentication was assuming a zero-terminated challenge
3542 string. It's now changed to take a pointer and a length, to allow for
3543 binary data in such strings.
3545 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3547 23. Exim was not respecting finduser_retries when seeking the login of the
3548 uid under which it was called; it was always trying 10 times. (The default
3549 setting of finduser_retries is zero.) Also, it was sleeping after the final
3550 failure, which is pointless.
3552 24. Implemented tls_on_connect_ports.
3554 25. Implemented acl_smtp_predata.
3556 26. If the domain in control=submission is set empty, Exim assumes that the
3557 authenticated id is a complete email address when it generates From: or
3558 Sender: header lines.
3560 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3561 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3562 chown and chgrp in /bin and hostname in /usr/bin.
3564 28. Exim was keeping the "process log" file open after each use, just as it
3565 does for the main log. This opens the possibility of it remaining open for
3566 long periods when the USR1 signal hits a daemon. Occasional processlog
3567 errors were reported, that could have been caused by this. Anyway, it seems
3568 much more sensible not to leave this file open at all, so that is what now
3571 29. The long-running daemon process does not normally write to the log once it
3572 has entered its main loop, and it closes the log before doing so. This is
3573 so that log files can straightforwardly be renamed and moved. However,
3574 there are a couple of unusual error situations where the daemon does write
3575 log entries, and I had neglected to close the log afterwards.
3577 30. The text of an SMTP error response that was received during a remote
3578 delivery was being truncated at 512 bytes. This is too short for some of
3579 the long messages that one sometimes sees. I've increased the limit to
3582 31. It is now possible to make retry rules that apply only when a message has a
3583 specific sender, in particular, an empty sender.
3585 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3586 it possible to be selective about when SMTP synchronization is enforced.
3588 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3590 32. Implemented hosts_connection_nolog.
3592 33. Added an ACL for QUIT.
3594 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3597 35. Added mailbox_size and mailbox_filecount to appendfile.
3599 36. Added control = no_multiline_responses to ACLs.
3601 37. There was a bug in the logic of the code that waits for the clock to tick
3602 in the case where the clock went backwards by a substantial amount such
3603 that the microsecond fraction of "now" was more than the microsecond
3604 fraction of "then" (but the whole seconds number was less).
3606 38. Added support for the libradius Radius client library this is found on
3607 FreeBSD (previously only the radiusclient library was supported).
3613 1. When certain lookups returned multiple values in the form name=value, the
3614 quoting of the values was not always being done properly. Specifically:
3615 (a) If the value started with a double quote, but contained no whitespace,
3617 (b) If the value contained whitespace other than a space character (i.e.
3618 tabs or newlines or carriage returns) it was not quoted.
3619 This fix has been applied to the mysql and pgsql lookups by writing a
3620 separate quoting function and calling it from the lookup code. The fix
3621 should probably also be applied to nisplus, ibase and oracle lookups, but
3622 since I cannot test any of those, I have not disturbed their existing code.
3624 2. A hit in the callout cache for a specific address caused a log line with no
3625 reason for rejecting RCPT. Now it says "Previous (cached) callout
3626 verification failure".
3628 3. There was an off-by-one bug in the queryprogram router. An over-long
3629 return line was truncated at 256 instead of 255 characters, thereby
3630 overflowing its buffer with the terminating zero. As well as fixing this, I
3631 have increased the buffer size to 1024 (and made a note to document this).
3633 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3634 when Exim is waiting for an SMTP response from a remote server, Exim
3635 restarts its select() call on the socket, thereby resetting its timeout.
3636 This is not a problem when such interrupts are rare. Somebody set up a cron
3637 job to run exiwhat every 2 minutes, which is less than the normal select()
3638 timeout (5 or 10 minutes). This meant that the select() timeout never
3639 kicked in because it was always reset. I have fixed this by comparing the
3640 time when an interrupt arrives with the time at the start of the first call
3641 to select(). If more time than the timeout has elapsed, the interrupt is
3642 treated as a timeout.
3644 5. Some internal re-factoring in preparation for the addition of Sieve
3645 extensions (by MH). In particular, the "personal" test is moved to a
3646 separate function, and given an option for scanning Cc: and Bcc: (which is
3647 not set for Exim filters).
3649 6. When Exim created an email address using the login of the caller as the
3650 local part (e.g. when creating a From: or Sender: header line), it was not
3651 quoting the local part when it contained special characters such as @.
3653 7. Installed new OpenBSD configuration files.
3655 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3656 try to make them clearer.
3658 9. Callout options, other than the timeout value, were being ignored when
3659 verifying sender addresses in header lines. For example, when using
3661 verify = header_sender/callout=no_cache
3663 the cache was (incorrectly) being used.
3665 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3666 only the Cygwin environment.
3668 11. When return_path_on_delivery was set as a log selector, if different remote
3669 addresses in the same message used different return paths and parallel
3670 remote delivery occurred, the wrong values would sometimes be logged.
3671 (Whenever a remote delivery process finished, the return path value from
3672 the most recently started remote delivery process was logged.)
3674 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3675 lines when AUTH and/or TLS are in use. This is the "received protocol"
3676 field. Exim used to use "asmtp" for authenticated SMTP, without any
3677 indication (in the protocol name) for TLS use. Now it follows the RFC and
3678 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3679 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3680 names appear in log lines as well as in Received: header lines.
3682 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3683 extensions, and comparison tests, and to fix some bugs.
3685 14. Changes to the "personal" filter test:
3687 (1) The test was buggy in that it was just doing the equivalent of
3688 "contains" tests on header lines. For example, if a user's address was
3689 anne@some.where, the "personal" test would incorrectly be true for
3691 To: susanne@some.where
3693 This test is now done by extracting each address from the header in turn,
3694 and checking the entire address. Other tests that are part of "personal"
3695 are now done using regular expressions (for example, to check local parts
3696 of addresses in From: header lines).
3698 (2) The list of non-personal local parts in From: addresses has been
3699 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3700 taken from the Sieve specification recommendations.
3702 (3) If the message contains any header line starting with "List-" it is
3703 treated as non-personal.
3705 (4) The test for "circular" in the Subject: header line has been removed
3706 because it now seems ill-conceived.
3708 15. Minor typos in src/EDITME comments corrected.
3710 16. Installed latest exipick from John Jetmore.
3712 17. If headers_add on a router specified a text string that was too long for
3713 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3714 of string_sprintf() is now avoided.
3716 18. $message_body_size was not set (it was always zero) when running the DATA
3717 ACL and the local_scan() function.
3719 19. For the "mail" command in an Exim filter, no default was being set for
3720 the once_repeat time, causing a random time value to be used if "once" was
3721 specified. (If the value happened to be <= 0, no repeat happened.) The
3722 default is now 0s, meaning "never repeat". The "vacation" command was OK
3723 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3724 (I found it when inspecting the code).
3726 20. There is now an overall timeout for performing a callout verification. It
3727 defaults to 4 times the callout timeout, which applies to individual SMTP
3728 commands during the callout. The overall timeout applies when there is more
3729 than one host that can be tried. The timeout is checked before trying the
3730 next host. This prevents very long delays if there are a large number of
3731 hosts and all are timing out (e.g. when the network connections are timing
3732 out). The value of the overall timeout can be changed by specifying an
3733 additional sub-option for "callout", called "maxwait". For example:
3735 verify = sender/callout=5s,maxwait=20s
3737 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3738 to the end before writing, but this should make it even safer).
3740 22. Exim was forgetting that it had advertised PIPELINING for the second and
3741 subsequent messages on an SMTP connection. It was also not resetting its
3742 memory on STARTTLS and an internal HELO.
3744 23. When Exim logs an SMTP synchronization error within a session, it now
3745 records whether PIPELINING has been advertised or not.
3747 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3748 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3749 rather than long int.
3751 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3753 26. Added the never_mail option to autoreply.
3759 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3760 crash if the getsockname() call failed; this can happen if a connection is
3761 closed very soon after it is established. The problem was simply in the
3762 order in which certain operations were done, causing Exim to try to write
3763 to the SMTP stream before it had set up the file descriptor. The bug has
3764 been fixed by making things happen in the correct order.
3770 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3771 before the connection was closed, thus losing the rejection response.
3773 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3774 some early Solaris releases, but causes trouble in current releases where
3775 socklen_t is defined.
3777 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3780 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3783 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3784 that releasing the top of it at the end releases what was used for sub-
3785 expansions (unless the block got too big). However, discard this block if
3786 the first thing is a variable or header, so that we can use its block when
3787 it is dynamic (useful for very large $message_headers, for example).
3789 6. Lookups now cache *every* query, not just the most recent. A new, separate
3790 store pool is used for this. It can be recovered when all lookup caches are
3791 flushed. Lookups now release memory at the end of their result strings.
3792 This has involved some general refactoring of the lookup sources.
3794 7. Some code has been added to the store_xxx() functions to reduce the amount
3795 of flapping under certain conditions.
3797 8. log_incoming_interface used to affect only the <= reception log lines. Now
3798 it causes the local interface and port to be added to several more SMTP log
3799 lines, for example "SMTP connection from", and rejection lines.
3801 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3803 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3805 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3806 could be overwritten at the end of the current message (or the start of a
3807 new message if it was set in a HELO ACL). The value is now preserved for
3808 the duration of the SMTP connection.
3810 12. If a transport had a headers_rewrite setting, and a matching header line
3811 contained an unqualified address, that address was qualified, even if it
3812 did not match any rewriting rules. The underlying bug was that the values
3813 of the flags that permit the existence of unqualified sender and recipient
3814 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3815 non-local messages, and by -bnq for local messages) were not being
3816 preserved with the message after it was received.
3818 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3819 "next input=" as part of the text comprising the host identity instead of
3820 the correct text. The code was using the same buffer for two different
3821 strings. However, depending on which order the printing function evaluated
3822 its arguments, the bug did not always show up. Under Linux, for example, my
3823 test suite worked just fine.
3825 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3826 doesn't work with some older versions of Perl. It has been changed to "my",
3827 which in any case is probably the better facility to use.
3829 15. A really picky compiler found some instances of statements for creating
3830 error messages that either had too many or two few arguments for the format
3833 16. The size of the buffer for calls to the DNS resolver has been increased
3834 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3835 for addresses that have a lot of PTR records. This alleviates a problem; it
3836 does not fully solve it.
3838 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3839 buffer now truncates the list and logs the incident, which is the same
3840 action as happens when Exim is looking up a host name and its aliases.
3841 Previously in this situation something unpredictable would happen;
3842 sometimes it was "internal error: store_reset failed".
3844 18. If a server dropped the connection unexpectedly when an Exim client was
3845 using GnuTLS and trying to read a response, the client delivery process
3846 crashed while trying to generate an error log message.
3848 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3849 string, for example:
3851 warn message = H1: something\nH2: something
3853 the text was added as a single header line from Exim's point of view
3854 though it ended up OK in the delivered message. However, searching for the
3855 second and subsequent header lines using $h_h2: did not work. This has been
3856 fixed. Similarly, if a system filter added multiple headers in this way,
3857 the routers could not see them.
3859 20. Expanded the error message when iplsearch is called with an invalid key to
3860 suggest using net-iplsearch in a host list.
3862 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3863 ACLs are no longer actually imposed (and a message to that effect is
3866 22. If a "gecos" field in a passwd entry contained escaped characters, in
3867 particular, if it contained a \" sequence, Exim got it wrong when building
3868 a From: or a Sender: header from that name. A second bug also caused
3869 incorrect handling when an unquoted " was present following a character
3870 that needed quoting.
3872 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3873 was not being matched caselessly.
3875 24. Arranged for all hyphens in the exim.8 source to be escaped with
3878 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3879 properly. Recipient callout cache records were still being keyed to include
3880 the sender, even when use_sender was set false. This led to far more
3881 callouts that were necessary. The sender is no longer included in the key
3882 when use_sender is false.
3884 26. Added "control = submission" modifier to ACLs.
3886 27. Added the ${base62d: operator to decode base 62 numbers.
3888 28. dnsdb lookups can now access SRV records.
3890 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3891 the configuration file.
3893 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3894 (-v) mode. This makes the output for a verbose queue run more intelligible.
3896 31. Added a use_postmaster feature to recipient callouts.
3898 32. Added the $body_zerocount variable, containing the number of binary zero
3899 bytes in the message body.
3901 33. The time of last modification of the "new" subdirectory is now used as the
3902 "mailbox time last read" when there is a quota error for a maildir
3905 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3907 35. Added +ignore_unknown as a special item in host lists.
3909 36. Code for decoding IPv6 addresses in host lists is now included, even if
3910 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3911 address was recognized as an IP address, but was then not correctly decoded
3912 into binary, causing unexpected and incorrect effects when compared with
3919 1. Very minor rewording of debugging text in manualroute to say "list of
3920 hosts" instead of "hostlist".
3922 2. If verify=header_syntax was set, and a header line with an unqualified
3923 address (no domain) and a large number of spaces between the end of the
3924 name and the colon was received, the reception process suffered a buffer
3925 overflow, and (when I tested it) crashed. This was caused by some obsolete
3926 code that should have been removed. The fix is to remove it!
3928 3. When running in the test harness, delay a bit after writing a bounce
3929 message to get a bit more predictability in the log output.
3931 4. Added a call to search_tidyup() just before forking a reception process. In
3932 theory, someone could use a lookup in the expansion of smtp_accept_max_
3933 per_host which, without the tidyup, could leave open a database connection.
3935 5. Added the variables $recipient_data and $sender_data which get set from a
3936 lookup success in an ACL "recipients" or "senders" condition, or a router
3937 "senders" option, similar to $domain_data and $local_part_data.
3939 6. Moved the writing of debug_print from before to after the "senders" test
3942 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3943 problems for message scanning, either using a data ACL, or using
3944 local_scan() because the Received: header was not generated till after they
3945 were called (in order to set the time as the time of reception completion).
3946 I have revised the way this works. The header is now generated after the
3947 body is received, but before the ACL or local_scan() are called. After they
3948 are run, the timestamp in the header is updated.
3954 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3955 before starting a queue runner without re-exec. This happened only when
3956 deliver_drop_privilege was set or when the Exim user was set to root. The
3957 effect of the bug was that timeouts during subsequent deliveries caused
3958 crashes instead of being properly handled. The handler is now left at its
3959 default (and expected) setting.
3961 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3962 message, again when deliver_drop_privilege is set or Exim is run as root.
3963 The bug described in (1) was not present in this case, but the tidying up
3964 of the other signals was missing. I have made the two cases consistent.
3966 3. The ignore_target_hosts setting on a manualroute router was being ignored
3967 for hosts that were looked up using the /MX notation.
3969 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3972 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3973 operated on the sender address. After changing the $sender_address to <>
3974 for the sender address verify, Exim was re-instated it as the original
3975 (before rewriting) address, but remembering that it had rewritten it, so it
3976 wasn't rewriting it again. This bug also had the effect of breaking the
3977 sender address verification caching when the sender address was rewritten.
3979 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3980 This has been changed so that if the ip literal address matches
3981 ignore_target_hosts, the router declines.
3983 7. Added expansion conditions match_domain, match_address, and match_local_
3984 part (NOT match_host).
3986 8. The placeholder for the Received: header didn't have a length field set.
3988 9. Added code to Exim itself and to exim_lock to test for a specific race
3989 condition that could lead to file corruption when using MBX delivery. The
3990 issue is with the lockfile that is created in /tmp. If this file is removed
3991 after a process has opened it but before that process has acquired a lock,
3992 there is the potential for a second process to recreate the file and also
3993 acquire a lock. This could lead to two Exim processes writing to the file
3994 at the same time. The added code performs the same test as UW imapd; it
3995 checks after acquiring the lock that its file descriptor still refers to
3996 the same named file.
3998 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3999 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4002 11. Added the smtp_active_hostname option. If used, this will typically be made
4003 to depend on the incoming interface address. Because $interface_address is
4004 not set up until the daemon has forked a reception process, error responses
4005 that can happen earlier (such as "too many connections") no longer contain
4008 12. If an expansion in a condition on a "warn" statement fails because a lookup
4009 defers, the "warn" statement is abandoned, and the next ACL statement is
4010 processed. Previously this caused the whole ACL to be aborted.
4012 13. Added the iplsearch lookup type.
4014 14. Added ident_timeout as a log selector.
4016 15. Added tls_certificate_verified as a log selector.
4018 16. Added a global option tls_require_ciphers (compare the smtp transport
4019 option of the same name). This controls incoming TLS connections.
4021 17. I finally figured out how to make tls_require_ciphers do a similar thing
4022 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4023 before starting the TLS session.
4025 18. Tabs are now shown as \t in -bP output.
4027 19. If the log selector return_path_on_delivery was set, Exim crashed when
4028 bouncing a message because it had too many Received: header lines.
4030 20. If two routers both had headers_remove settings, and the first one included
4031 a superfluous trailing colon, the final name in the first list and the
4032 first name in the second list were incorrectly joined into one item (with a
4033 colon in the middle).
4039 1. Added -C and -D options to the exinext utility, mainly to make it easier
4040 to include in the automated testing, but these could be helpful when
4041 multiple configurations are in use.
4043 2. The exinext utility was not formatting the output nicely when there was
4044 an alternate port involved in the retry record key, nor when there was a
4045 message id as well (for retries that were specific to a specific message
4046 and a specific host). It was also confused by IPv6 addresses, because of
4047 the additional colons they contain. I have fixed the IPv4 problem, and
4048 patched it up to do a reasonable job for IPv6.
4050 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4051 delivery, the log line now contains "pipelined" if PIPELINING was used.
4053 4. An SMTP transport process used to panic and die if the bind() call to set
4054 an explicit outgoing interface failed. This has been changed; it is now
4055 treated in the same way as a connect() failure.
4057 5. A reference to $sender_host_name in the part of a conditional expansion
4058 that was being skipped was still causing a DNS lookup. This no longer
4061 6. The def: expansion condition was not recognizing references to header lines
4062 that used bh_ and bheader_.
4064 7. Added the _cache feature to named lists.
4066 8. The code for checking quota_filecount in the appendfile transport was
4067 allowing one more file than it should have been.
4069 9. For compatibility with Sendmail, the command line option
4077 and sets the incoming protocol and host name (for trusted callers). The
4078 host name and its colon can be omitted when only the protocol is to be set.
4079 Note the Exim already has two private options, -pd and -ps, that refer to
4080 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4081 "s", but I don't think that's a major issue.
4083 10. A number of refactoring changes to the code, none of which should affect
4086 (a) The number of logging options was getting close to filling up the
4087 32-bit word that was used as a bit map. I have split them into two classes:
4088 those that are passed in the argument to log_write(), and those that are
4089 only ever tested independently outside of that function. These are now in
4090 separate 32-bit words, so there is plenty of room for expansion again.
4091 There is no change in the user interface or the logging behaviour.
4093 (b) When building, for example, log lines, the code previously used a
4094 macro that called string_cat() twice, in order to add two strings. This is
4095 not really sufficiently general. Furthermore, there was one instance where
4096 it was actually wrong because one of the argument was used twice, and in
4097 one call a function was used. (As it happened, calling the function twice
4098 did not affect the overall behaviour.) The macro has been replaced by a
4099 function that can join an arbitrary number of extra strings onto a growing
4102 (c) The code for expansion conditions now uses a table and a binary chop
4103 instead of a serial search (which was left over from when there were very
4104 few conditions). Also, it now recognizes conditions like "pam" even when
4105 the relevant support is not compiled in: a suitably worded error message is
4106 given if an attempt is made to use such a condition.
4108 11. Added ${time_interval:xxxxx}.
4110 12. A bug was causing one of the ddress fields not to be passed back correctly
4111 from remote delivery subprocesses. The field in question was not being
4112 subsequently used, so this caused to problems in practice.
4114 13. Added new log selectors queue_time and deliver_time.
4116 14. Might have fixed a bug in maildirsizefile handling that threw up
4117 "unexpected character" debug warnings, and recalculated the data
4118 unnecessarily. In any case, I expanded the warning message to give more
4121 15. Added the message "Restricted characters in address" to the statements in
4122 the default ACL that block characters like @ and % in local parts.
4124 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4125 Three changes have been made:
4127 (a) There was a serious bug; a negative response to MAIL caused the whole
4128 recipient domain to be cached as invalid, thereby blocking all messages
4129 to all local parts at the same domain, from all senders. This bug has
4130 been fixed. The domain is no longer cached after a negative response to
4131 MAIL if the sender used is not empty.
4133 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4136 (c) A new callout option, "use_sender" has been added for people who want
4137 the modified behaviour.
4143 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4146 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4147 indeed breaks things for older releases.
4149 3. Added additional logging to the case where there is a problem reading data
4150 from a filter that is running in a subprocess using a pipe, in order to
4151 try to track down a specific problem.
4153 4. Testing facility fudge: when running in the test harness and attempting
4154 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4155 getting "No route to host". Convert this to a timeout.
4157 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4160 6. Some OS don't have socklen_t but use size_t instead. This affects the
4161 fifth argument of getsockopt() amongst other things. This is now
4162 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4163 can be set for individual OS. I have set it for SunOS5, OSF1, and
4164 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4165 some earlier ones do not.
4167 7. Change 4.30/15 was not doing the test caselessly.
4169 8. The standard form for an IPv6 address literal was being rejected by address
4170 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4171 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4172 this, as well as the form without the "IPv6" on the front (but only when
4173 address literals are enabled, of course).
4175 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4177 10. Exim crashed if a message with an empty sender address specified by -f
4178 encountered a router with an errors_to setting. This could be provoked only
4179 by a command such as
4183 where an empty string was supplied; "<>" did not hit this bug.
4185 11. Installed PCRE release 4.5.
4187 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4188 remained set. It is now erased.
4190 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4191 times from message ids (which are base 36 rather than the normal 62).
4193 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4194 were being counted as actual protocol errors, and logged if the log
4195 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4196 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4197 and DATA following a set of rejected RCPTs do not count as protocol errors.
4198 In other words, Exim assumes they were pipelined, though this may not
4199 actually be the case. Of course, in all cases the client gets an
4200 appropriate error code.
4202 15. If a lookup fails in an ACL condition, a message about the failure may
4203 be available; it is used if testing the ACL cannot continue, because most
4204 such messages specify what the cause of the deferral is. However, some
4205 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4206 that caused an old message to be retained and used if a later statement
4207 caused a defer, replacing the real cause of the deferral.
4209 16. If an IP address had so many PTR records that the DNS lookup buffer
4210 was not large enough to hold them, Exim could crash while trying to process
4211 the truncated data. It now detects and logs this case.
4213 17. Further to 4.21/58, another change has been made: if (and only if) the
4214 first line of a message (the first header line) ends with CRLF, a bare LF
4215 in a subsequent header line has a space inserted after it, so as not to
4216 terminate the header.
4218 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4219 unnecessarily, used atoi() instead of strtol(), and didn't check the
4220 termination when getting file sizes from file names by regex.
4222 19. Completely re-implemented the support for maildirsize files, in the light
4223 of a number of problems with the previous contributed implementation
4224 (4.30/29). In particular:
4226 . If the quota is zero, the maildirsize file is maintained, but no quota is
4229 . If the maildir directory does not exist, it is created before any attempt
4230 to write a maildirsize file.
4232 . The quota value in the file is just a cache; if the quota is changed in
4233 the transport, the new value overrides.
4235 . A regular expression is available for excluding directories from the
4238 20. The autoreply transport checks the characters in options that define the
4239 message's headers; it allows continued headers, but it was checking with
4240 isspace() after an embedded newline instead of explicitly looking for a
4243 21. If all the "regular" hosts to which an address was routed had passed their
4244 expiry times, and had not reached their retry times, the address was
4245 bounced, even if fallback hosts were defined. Now Exim should go on to try
4248 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4249 equivalent code in the SMTP transport. Some hosts send humungous responses
4250 to HELO/EHLO, more than 1024 it seems.
4252 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4253 gives compiler warnings in some environments. I've now done it "properly",
4256 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4257 (because of problems with the built-in one) was declared to return uschar *
4258 instead of char *, causing compiler failure.
4260 25. Fixed a file descriptor leak when processing alias/forward files.
4262 26. Fixed a minor format string issue in dbfn.c.
4264 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4266 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4267 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4270 29. When a sender address is verified, it is cached, to save repeating the test
4271 when there is more than one recipient in a message. However, when the
4272 verification involves a callout, it is possible for different callout
4273 options to be set for different recipients. It is too complicated to keep
4274 track of this in the cache, so now Exim always runs a verification when a
4275 callout is required, relying on the callout cache for the optimization.
4276 The overhead is duplication of the address routing, but this should not be
4279 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4280 to be verified with callout=postmaster, and the main callout worked but the
4281 postmaster check failed, the verification correctly failed. However, if a
4282 subsequent RCPT command asked for sender verification *without* the
4283 postmaster check, incorrect caching caused this verification also to fail,
4286 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4287 it was not caching the DNS options (qualify_single, search_parents) that
4288 were used when the lookup failed. A subsequent lookup with different
4289 options therefore always gave the same answer, though there were cases
4290 where it should not have. (Example: a "domains = !$mx_any" option on a
4291 dnslookup router: the "domains" option is always processed without any
4292 widening, but the router might have qualify_single set.) Now Exim uses the
4293 cached value only when the same options are set.
4295 32. Added John Jetmore's "exipick" utility to the distribution.
4297 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4298 than a timeout (e.g. a certificate is required, and is not provided), an
4299 Exim server now closes the connection immediately. Previously it waited for
4300 the client to close - but if the client is SSL, it seems that they each
4301 wait for each other, leading to a delay before one of them times out.
4303 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4304 maintained 0.8.x compatibility because I don't think many are using it, and
4305 it is clearly obsolete.
4307 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4310 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4311 client certificate was expired. A simple patch fixes this, though I don't
4312 understand the full logic of why the verify callback is called multiple
4315 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4316 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4317 which causes problems with some clients (such as the Certicom SSL Plus
4318 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4319 disables the coutermeasure allowing Eudora to connect."
4321 38. Exim was not checking that a write() to a log file succeeded. This could
4322 lead to Bad Things if a log got too big, in particular if it hit a file
4323 size limit. Exim now panics and dies if it cannot write to a log file, just
4324 as it does if it cannot open a log file.
4326 39. Modified OS/Makefile-Linux so that it now contains
4328 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4330 The two -D definitions ensure that Exim is compiled with large file
4331 support, which makes it possible to handle log files that are bigger than
4334 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4335 instance) a domain was checked against a named list that involved a lookup,
4336 causing $domain_data to be set, then another domain was checked against the
4337 same list, then the first domain was re-checked, the value of $domain_data
4338 after the final check could be wrong. In particular, if the second check
4339 failed, it could be set empty. This bug probably also applied to
4342 41. The strip_trailing_dot option was not being applied to the address given
4343 with the -f command-line option.
4345 42. The code for reading a message's header from the spool was incrementing
4346 $received_count, but never initializing it. This meant that the value was
4347 incorrect (doubled) while delivering a message in the same process in which
4348 it was received. In the most common configuration of Exim, this never
4349 happens - a fresh exec is done - but it can happen when
4350 deliver_drop_privilege is set.
4352 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4353 it now includes up to 150 characters of the unexpected data in the log
4356 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4357 and building data strings. The size of both of these buffers was 10 000
4358 bytes - far larger than anybody would *ever* want, thought I. Needless to
4359 say, somebody hit the limit. I have increased the maximum line length to
4360 20 000 and the maximum data length of concatenated lines to 100 000. I have
4361 also fixed two bugs, because there was no checking on these buffers. Tsk,
4362 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4363 buffer is too small.
4365 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4366 lsearch lookups. Now it does.
4368 46. When parsing a route_list item in a manualroute router, a fixed-length
4369 buffer was used for the list of hosts. I made this 1024 bytes long,
4370 thinking that nobody would ever have a list of hosts that long. Wrong.
4371 Somebody had a whole pile of complicated expansion conditions, and the
4372 string was silently truncated, leading to an expansion error. It turns out
4373 that it is easier to change to an unlimited length (owing to other changes
4374 that have happened since this code was originally written) than to build
4375 structure for giving a limitation error. The length of the item that
4376 expands into the list of hosts is now unlimited.
4378 47. The lsearch lookup could not handle data where the length of text line was
4379 more than 4095 characters. Such lines were truncated, leading to shortened
4380 data being returned. It should now handle lines of any length.
4382 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4383 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4386 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4387 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4388 no longer happens in comment lines. A list of replacements is now placed
4389 at the head of all of the source files, except those whose only change is
4390 to replace PERL_COMMAND in the very first #! line.
4392 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4393 messages on the queue, with a bottom-up merge sort, using code contributed
4394 by Michael Haardt. This should make operations like -bp somewhat faster on
4395 large queues. It won't affect queue runners, except when queue_run_in_order
4398 51. Installed eximstats 1.31 in the distribution.
4400 52. Added support for SRV lookups to the dnslookup router.
4402 53. If an ACL referred to $message_body or $message_body_end, the value was not
4403 reset for any messages that followed in the same SMTP session.
4405 54. The store-handling optimization for building very long strings was not
4406 differentiating between the different store pools. I don't think this
4407 actually made any difference in practice, but I've tidied it.
4409 55. While running the routers to verify a sender address, $sender_address
4410 was still set to the sender address. This is wrong, because when routing to
4411 send a bounce to the sender, it would be empty. Therefore, I have changed
4412 it so that, while verifying a sender address, $sender_address is set to <>.
4413 (There is no change to what happens when verifying a recipient address.)
4415 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4416 target A or AAAA records (if not already returned) without resetting the
4417 qualify_single or search_parents options of the DNS resolver. These are
4418 inappropriate in this case because the targets of MX and SRV records must
4419 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4420 target that, when qualified, matched something in the local domain. These
4421 two options are now turned off when doing these lookups.
4423 57. It seems that at least some releases of Reiserfs (which does not have the
4424 concept of a fixed number of inodes) returns zero and not -1 for the
4425 number of available inodes. This interacted badly with check_spool_inodes,
4426 which assumed that -1 was the "no such thing" setting. What I have done is
4427 to check that the total number of inodes is greater than zero before doing
4428 the test of how many are available.
4430 58. When a "warn" ACL statement has a log_message modifier, the message is
4431 remembered, and not repeated. This is to avoid a lot of repetition when a
4432 message has many recipients that cause the same warning to be written.
4433 Howewer, Exim was preserving the list of already written lines for an
4434 entire SMTP session, which doesn't seem right. The memory is now reset if a
4435 new message is started.
4437 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4438 debugging output unless log_rewrite was also set.
4440 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4441 of (int)(handle) when we know that handle contains (void *)(-1).
4443 61. The Exim daemon panic-logs an error return when it closes the incoming
4444 connection. However "connection reset by peer" seems to be common, and
4445 isn't really an error worthy of noting specially, so that particular error
4448 62. When Exim is trying to find all the local interfaces, it used to panic and
4449 die if the ioctl to get the interface flags failed. However, it seems that
4450 on at least one OS (Solaris 9) it is possible to have an interface that is
4451 included in the list of interfaces, but for which you get a failure error
4452 for this call. This happens when the interface is not "plumbed" into a
4453 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4454 failure of the "get flags" call assumes that the interface is down.
4456 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4457 makes life easier for people who are doing arithmetic on fields extracted
4458 from dates, where you often get leading zeros that should not be
4459 interpreted as octal.
4461 64. Added qualify_domain to the redirect router, to override the global
4464 65. If a pathologically long header line contained very many addresses (the
4465 report of this problem mentioned 10 000) and each of them was rewritten,
4466 Exim could use up a very large amount of memory. (It kept on making new
4467 copies of the header line as it rewrote, and never released the old ones.)
4468 At the expense of a bit more processing, the header rewriting function has
4469 been changed so that it no longer eats memory in this way.
4471 66. The generation of the Received: header has been moved from the time that a
4472 message starts to be received, to the time that it finishes. The timestamp
4473 in the Received: header should now be very close to that of the <= log
4474 line. There are two side-effects of this change:
4476 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4477 logged header lines no longer include the local Received: line, because
4478 it has not yet been created. The same applies to a copy of the message
4479 that is returned to a non-SMTP sender when a message is rejected.
4481 (b) When a filter file is tested using -bf, no additional Received: header
4482 is added to the test message. After some thought, I decided that this
4485 This change does not affect the value of $received_for. It is still set
4486 after address rewriting, but before local_scan() is called.
4488 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4490 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4491 gave an unhelpful panic error message, and a defer error. I have managed to
4492 change this behaviour so that it now rejects any supplied certificate,
4493 which seems right, as the list of acceptable certificates is empty.
4495 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4496 gave an unhelpful defer error. I have not managed to make this reject any
4497 supplied certificates, but the error message it gives is "no certificate
4498 supplied", which is not helpful.
4500 70. exigrep's output now also includes lines that are not associated with any
4501 message, but which match the given pattern. Implemented by a patch from
4502 Martin Sluka, which also tidied up the Perl a bit.
4504 71. Recipient callout verification, like sender verification, was using <> in
4505 the MAIL FROM command. This isn't really the right thing, since the actual
4506 sender may affect whether the remote host accepts the recipient or not. I
4507 have changed it to use the actual sender in the callout; this means that
4508 the cache record is now keyed on a recipient/sender pair, not just the
4509 recipient address. There doesn't seem to be a real danger of callout loops,
4510 since a callout by the remote host to check the sender would use <>.
4511 [SEE ABOVE: changed after hitting problems.]
4513 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4514 temporary errors. However, in the case of such a code being given after
4515 the end of a data transmission (i.e. after ".") Exim was failing to write
4516 a retry record for the message. (Yes, there was some broken host that was
4517 actually sending 8xx at this point.)
4519 73. An unknown lookup type in a host list could cause Exim to panic-die when
4520 the list was checked. (An example that provoked this was putting <; in the
4521 middle of a list instead of at the start.) If this happened during a DATA
4522 ACL check, a -D file could be left lying around. This kind of configuration
4523 error no longer causes Exim to die; instead it causes a defer errror. The
4524 incident is still logged to the main and panic logs.
4526 74. Buglet left over from Exim 3 conversion. The message "too many messages
4527 in one connection" was written to the rejectlog but not the mainlog, except
4528 when address rewriting (yes!) was being logged.
4530 75. Added write_rejectlog option.
4532 76. When a system filter was run not as root (that is, when system_filter_user
4533 was set), the values of the $n variables were not being returned to the
4534 main process; thus, they were not subsequently available in the $sn
4537 77. Added +return_path_on_delivery log selector.
4539 78. A connection timeout was being treated differently from recipients deferred
4540 when testing hosts_max_try with a message that was older than the host's
4541 retry timeout. (The host should not be counted, thus allowing all hosts to
4542 be tried at least once before bouncing.) This may have been the cause of an
4543 occasionally reported bug whereby a message would remain on the queue
4544 longer than the retry timeout, but would be bounced if a delivery was
4545 forced. I say "may" because I never totally pinned down the problem;
4546 setting up timeout/retry tests is difficult. See also the next item.
4548 79. The ultimate address timeout was not being applied to errors that involved
4549 a combination of host plus message (for example, a timeout on a MAIL
4550 command). When an address resolved to a number of possible hosts, and they
4551 were not all tried for each delivery (e.g. because of hosts_max_try), a
4552 message could remain on the queue longer than the retry timeout.
4554 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4557 81. Fixed an obscure SMTP outgoing bug which required at least the following
4558 conditions: (a) there was another message waiting for the same server;
4559 (b) the server returned 5xx to all RCPT commands in the first message so
4560 that the message was not completed; (c) the server dropped the connection
4561 or gave a negative response to the RSET that Exim sends to abort the
4562 transaction. The observed case was a dropped connection after DATA that had
4563 been sent in pipelining mode. That is, the server had advertised PIPELINING
4564 but was not implementing it correctly. The effect of the bug was incorrect
4565 behaviour, such as trying another host, and this could lead to a crash.
4571 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4572 and daemon.c were passed as pointers to ints; they should have been
4573 pointers to socklen_t variables (which are typically unsigned ints).
4575 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4578 3. Fixed a really odd bug that affected only the testing scheme; patching a
4579 certain fixed string in the binary changed the value of another string that
4580 happened to be identical to the end of the original first string.
4582 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4583 name", it returns that address as the IP address. On some operating
4584 systems (e.g. Solaris), it also passes back the IP address string as the
4585 "host name". However, on others (e.g. Linux), it passes back an empty
4586 string. Exim wasn't checking for this, and was changing the host name to an
4587 empty string, assuming it had been canonicized.
4589 5. Although rare, it is permitted to have more than one PTR record for a given
4590 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4591 all the names associated with an address, because they do in Solaris.
4592 However, it seems that they do not in Linux for data that comes from the
4593 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4594 I found this out when I moved to a new Linux workstation and tried to run
4595 the Exim test suite.
4597 To get round this problem I have changed the code so that it now does its
4598 own call to the DNS to look up PTR records when searching for a host name.
4599 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4600 addresses that are only in /etc/hosts are still found.
4602 This behaviour is, however, controlled by an option called host_lookup_
4603 order, which defaults to "bydns:byaddr". If people want to use the other
4604 order, or indeed, just use one or the other means of lookup, they can
4605 specify it in this variable.
4607 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4608 some operating systems, this comes back from gethostbyaddr() as an empty
4609 string, and this is what Exim used to test for. However, it seems that in
4610 other systems, "." is yielded. Exim now tests for this case too.
4612 7. The values of check_spool_space and check_log_space are now held internally
4613 as a number of kilobytes instead of an absolute number of bytes. If a
4614 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4615 kilobyte. This means that much larger values can be stored.
4617 8. Exim monitor: an attempt to get the action menu when not actually pointing
4618 at a message produces an empty menu entitled "No message selected". This
4619 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4620 no entries in it ("Shell widget menu has zero width and/or height"). So I
4621 have added a single, blank menu entry in this case.
4623 9. Added ${quote_local_part.
4625 10. MIME decoding is now applied to the contents of Subject: header lines when
4628 11. Now that a reference to $sender_host_address automatically causes a reverse
4629 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4630 host lookup before query-style lookups in lists that might use this
4631 variable. This has therefore been abolished, and the "net-" prefix is no
4632 longer necessary for query-style lookups.
4634 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4635 have been a typo for LFLAGS, so it has been changed.
4637 13. The install script calls Exim with "-C /dev/null" in order to find the
4638 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4639 to be output. Howeve, since Exim outputs its version number before the
4640 error, it didn't break the script. It just looked ugly. I fixed this by
4641 always allowing "-C /dev/null" if the caller is root.
4643 14. Ignore overlarge ACL variable number when reading spool file - insurance
4644 against a later release with more variables having written the file.
4646 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4647 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4648 this, as well as the form without the "IPv6" on the front.
4650 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4651 OS/Makefile-Darwin file.
4653 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4654 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4656 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4657 variable. However, some OS use ints for inodes. I've added cast to long int
4658 to get rid of the compiler warning.
4660 19. I had forgotten to lock out "/../" in configuration file names when
4661 ALT_CONFIG_PREFIX was set.
4663 20. Routers used for verification do not need to specify transports. However,
4664 if such a router generated a host list, and callout was configured, Exim
4665 crashed, because it could not find a port number from the (non-existent)
4666 transport. It now assumes port 25 in this circumstance.
4668 21. Added the -t option to exigrep.
4670 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4671 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4672 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4673 them is set, LOOKUP_LSEARCH is forced.
4675 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4677 24. Added sender and host information to the "rejected by local_scan()" log
4678 line; previously there was no indication of these.
4680 25. Added .include_if_exists.
4682 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4683 Linux. It turns out that not all file systems support this. Apparently some
4684 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4685 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4686 not supported on the file descriptor, is now ignored when Exim is trying to
4687 sync a directory. This applies only to Linux.
4689 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4691 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4692 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4695 29. Added support for maildirsize files from supplied patch (modified a bit).
4697 30. The use of :fail: followed by an empty string could lead Exim to respond to
4698 sender verification failures with (e.g.):
4700 550 Verification failed for <xxx>
4701 550 Sender verify failed
4703 where the first response line was missing the '-' that indicates it is not
4704 the final line of the response.
4706 31. The loop for finding the name of the user that called Exim had a hardwired
4707 limit of 10; it now uses the value of finduser_retries, which is used for
4708 all other user lookups.
4710 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4713 33. Exim was neglecting to zero errno before one call of strtol() when
4714 expanding a string and expecting an integer value. On some systems this
4715 resulted in spurious "integer overflow" errors. Also, it was casting the
4716 result into an int without checking.
4718 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4719 did not work. The code looks as if it has *never* worked, though it appears
4720 to have been documented since at least releast 1.62. I have made it work.
4722 35. The "timeout_DNS" error in retry rules, also documented since at least
4723 1.62, also never worked. As it isn't clear exactly what this means, and
4724 clearly it isn't a major issue, I have abolished the feature by treating it
4725 as "timeout", and writing a warning to the main and panic logs.
4727 36. The display of retry rules for -brt wasn't always showing the error code
4730 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4731 timeout_connect_A, timeout_connect_MX.
4733 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4734 to the empty sender.
4736 39. The daemon was not analysing the content of -oX till after it had closed
4737 stderr and disconnected from the controlling terminal. This meant that any
4738 syntax errors were only noted on the panic log, and the return code from
4739 the command was 0. By re-arranging the code a little, I've made the
4740 decoding happen first, so such errors now appear on stderr, and the return
4741 code is 1. However, the actual setting up of the sockets still happens in
4742 the disconnected process, so errors there are still only recorded on the
4745 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4746 connections (as happens on some IP stacks) was logged at start up time as
4747 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4748 it from "IPv6 and IPv4", which means that two separate sockets are being
4751 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4752 says whether AF_INET or AF_INET6 was passed as an argument.
4754 42. Exiwhat output was messed up when time zones were included in log
4757 43. Exiwhat now gives more information about the daemon's listening ports,
4758 and whether -tls-on-connect was used.
4760 44. The "port" option of the smtp transport is now expanded.
4762 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4763 silently ignored. Now an error message is written to the main and panic
4766 46. There's a new ACL modifier called "logwrite" which writes to a log file
4767 as soon as it is encountered.
4769 47. Added $local_user_uid and $local_user_gid at routing time.
4771 48. Exim crashed when trying to verify a sender address that was being
4774 49. Exim was recognizing only a space character after ".include". It now also
4775 recognizes a tab character.
4777 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4778 extracting the relevant information from the specification. The man page no
4779 longer contains scrambled data for the -d option, and I've added a section
4780 at the front about calling Exim under different names.
4782 51. Added "extra_headers" argument to the "mail" command in filter files.
4784 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4787 53. Installed eximstats 1.29.
4789 54. Added transport_filter_timeout as a generic transport option.
4791 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4792 Cc: header lines. This was required by RFC 822, but it not required by RFC
4795 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4796 incoming messages that did not have them. Now it does so only if the
4797 message originates locally, that is, if there is no associated remote host
4798 address. When Resent- header lines are present, this applies to the Resent-
4799 lines rather than the non-Resent- lines.
4801 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4802 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4804 58. Messages for configuration errors now include the name of the main
4805 configuration file - useful now that there may be more than one file in a
4806 list (.included file names were always shown).
4808 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4809 for those rare installations that do not start the daemon as root or run it
4810 setuid root. I've cut out the call to initgroups() if the daemon is not
4813 60. The Exim user and group can now be bound into the binary as text strings
4814 that are looked up at the start of Exim's processing.
4816 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4818 62. Added $mailstore_basename variable.
4820 63. Installed patch to sieve.c from Michael Haardt.
4822 64. When Exim failed to open the panic log after failing to open the main log,
4823 the original message it was trying to log was written to stderr and debug
4824 output, but if they were not available (the usual case in production), it
4825 was lost. Now it is written to syslog before the two lines that record the
4826 failures to open the logs.
4828 65. Users' Exim filters run in subprocesses under the user's uid. It is
4829 possible for a "deliver" command or an alias in a "personal" command to
4830 provoke an address rewrite. If logging of address rewriting is configured,
4831 this fails because the process is not running as root or exim. There may be
4832 a better way of dealing with this, but for the moment (because 4.30 needs
4833 to be released), I have disabled address rewrite logging when running a
4834 filter in a non-root, non-exim process.
4840 1. The buildconfig auxiliary program wasn't quoting the value set for
4841 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4842 not defined. This bug was masked in 4.22 by the effect that was fixed in
4845 2. Some messages that were rejected after a message id was allocated were
4846 shown as "incomplete" by exigrep. It no longer does this for messages that
4847 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4849 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4850 have allow_domain_literals set, the ID did not get logged in the <= line.
4851 Domain literals are now always recognized in Message-ID: header lines.
4853 4. The first argument for a ${extract expansion item is the key name or field
4854 number. Leading and trailing spaces in this item were not being ignored,
4855 causing some misleading effects.
4857 5. When deliver_drop_privilege was set, single queue runner processes started
4858 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4859 same command in the process it spins off) were not dropping privilege.
4861 6. When the daemon running as "exim" started a queue runner, it always
4862 re-executed Exim in the spun-off process. This is a waste of effort when
4863 deliver_drop_privilege is set. The new process now just calls the
4864 queue-runner function directly.
4870 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4873 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4874 ignored. Though the use of -oP was forcing the writing of a pid file, it
4875 was always written to the default place.
4877 3. If the message "no IP address found for host xxxx" is generated during
4878 incoming verification, it is now followed by identification of the incoming
4879 connection (so you can more easily find what provoked it).
4881 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4883 5. Added some features to "harden" Exim a bit more against certain attacks:
4885 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4886 be put in Local/Makefile. This is like the never_users runtime option,
4887 but it cannot be overridden. The default setting is "root".
4889 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4890 prefix string with which any file named in a -C command line option
4893 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4894 is retained for -C and -D only if the caller of Exim is root. Without
4895 it, the exim user may also use -C and -D and retain privilege.
4897 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4898 command line option is disabled.
4900 6. Macro names set by the -D option must start with an upper case letter, just
4901 like macro names defined in the configuration file.
4903 7. Added "dereference=" facility to LDAP.
4905 8. Two instances of the typo "uknown" in the source files are fixed.
4907 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4908 the Configure-Makefile script screwed up while processing it.
4910 10. Incorporated PCRE 4.4.
4912 11. The SMTP synchronization check was not operating right at the start of an
4913 SMTP session. For example, it could not catch a HELO sent before the client
4914 waited for the greeting. There is now a check for outstanding input at the
4915 point when the greeting is written. Because of the duplex, asynchronous
4916 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4917 way, but not yet received, when the check is performed.
4919 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4920 on TCP/IP sockets, because this apparently causes some broken clients to
4923 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4924 unchanged) from the Cygwin maintainer.
4926 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4927 was supported instead of testing for mbx. Effectively a typo.
4929 15. The spa authenticator server code was not checking that the input it
4930 received was valid base64.
4932 16. The debug output line for the "set" modifier in ACLs was not showing the
4933 name of the variable that was being set.
4935 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4937 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4938 lookup on its own. Something else was needed to trigger the lookup. For
4939 example, a match in host_lookup or the need for a host name in a host list.
4940 Now, if $sender_host_name is referenced and the host name has not yet been
4941 looked up, a lookup is performed. If the lookup fails, the variable remains
4942 empty, and $host_lookup_failed is set to "1".
4944 19. Added "eqi" as a case-independent comparison operator.
4946 20. The saslauthd authentication condition could segfault if neither service
4947 nor realm was specified.
4949 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4950 error message that was logged was misleading, and incoming SMTP
4951 connections were dropped. The message is now more accurate, and temporary
4952 errors are given to SMTP connections.
4954 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4955 (except RSET) with a 421 error, until QUIT is received. However, it was
4956 failing to send a response to QUIT.
4958 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4959 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4960 ACL is now run after the helo_try_verify_hosts code.
4962 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4963 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4964 case-independent, so other case variants are also recognized). Apparently
4965 some systems use these upper case variants.
4967 25. If more than two messages were waiting for the same host, and a transport
4968 filter was specified for the transport, Exim sent two messages over the
4969 same TCP/IP connection, and then failed with "socket operation on non-
4970 socket" when it tried to send the third.
4972 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4974 27. The extern definition of crypt16() in expand.c was not being excluded when
4975 the OS had its own crypt16() function.
4977 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4978 as a preferred synonym for return_size_limit, both as an option and as an
4981 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4983 30. Changed the default configuration ACL to relax the local part checking rule
4984 for addresses that are not in any local domains. For these addresses,
4985 slashes and pipe symbols are allowed within local parts, but the sequence
4986 /../ is explicitly forbidden.
4988 31. SPA server authentication was not clearing the challenge buffer before
4991 32. log_message in a "warn" ACL statement was writing to the reject log as
4992 well as to the main log, which contradicts the documentation and doesn't
4993 seem right (because no rejection is happening). So I have stopped it.
4995 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4996 However, I am unable to do any testing of this.
4998 34. Fixed an infelicity in the appendfile transport. When checking directories
4999 for a mailbox, to see if any needed to be created, it was accidentally
5000 using path names with one or more superfluous leading slashes; tracing
5001 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5003 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5004 contents are added to the log line that is written for every discarded
5005 recipient. (Previously a log_message setting was ignored.)
5007 36. The ${quote: operator now quotes the string if it is empty.
5009 37. The install script runs exim in order to find its version number. If for
5010 some reason other than non-existence or emptiness, which it checks, it
5011 could not run './exim', it was installing it with an empty version number,
5012 i.e. as "exim-". This error state is now caught, and the installation is
5015 38. An argument was missing from the function that creates an error message
5016 when Exim fails to connect to the socket for saslauthd authentication.
5017 This could cause Exim to crash, or give a corrupted message.
5019 39. Added isip, isip4, and isip6 to ${if conditions.
5021 40. The ACL variables $acl_xx are now saved with the message, and can be
5022 accessed later in routers, transports, and filters.
5024 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5025 strings in the file are not string-expanded.
5027 42. If a MAIL command specified a SIZE value that was too large to fit into an
5028 int variable, the check against message_size_limit failed. Such values are
5029 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5030 day this will have to be increased, but I don't think I want to be around
5031 when emails are that large.
5038 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5039 iconv() is not standard in FreeBSD.
5041 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5042 IPv6 enabled. The observed symptom was a segmentation fault on return from
5043 the function os_common_find_running_interfaces() in src/os.c.
5045 3. In the check_special_case() function in daemon.c I had used "errno" as an
5046 argument name, which causes warnings on some systems. This was basically a
5047 typo, since it was named "eno" in the comments!
5049 4. The code that waits for the clock to tick (at a resolution of some fraction
5050 of a second) so as to ensure message-id uniqueness was always waiting for
5051 at least one whole tick, when it could have waited for less. [This is
5052 almost certainly not relevant at current processor speeds, where it is
5053 unlikely to ever wait at all. But we try to future-proof.]
5055 5. The function that sleeps for a time interval that includes fractions of a
5056 second contained a race. It did not block SIGALRM between setting the
5057 timer, and suspending (a couple of lines later). If the interval was short
5058 and the sigsuspend() was delayed until after it had expired, the suspension
5059 never ended. On busy systems this could lead to processes getting stuck for
5062 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5063 process, before it forks any delivery processes. The open lookup caching
5064 mechanism meant that the open file or database connection was passed into
5065 the delivery process. The problem was that delivery processes always tidy
5066 up cached lookup data. This could cause a problem for the next delivery
5067 process started by the queue runner, because the external queue runner
5068 process does not know about the closure. So the next delivery process
5069 still has data in the lookup cache. In the case of a file lookup, there was
5070 no problem because closing a file descriptor in a subprocess doesn't affect
5071 the parent. However, if the lookup was caching a connection to a database,
5072 the connection was closed, and the second delivery process was likely to
5073 see errors such as "PGSQL: query failed: server closed the connection
5074 unexpectedly". The problem has been fixed by closing all cached lookups
5075 in a queue runner before running a delivery process.
5077 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5078 seem to have the "const" qualifier which it has on other OS. I've
5081 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5082 *of the same type* (client or server) with the same public name that an
5083 error should be diagnosed.
5085 9. When Exim looked up a host name for an IP address, but failed to find the
5086 original IP address when looking up the host name (a safety check), it
5087 output the message "<ip address> does not match any IP for NULL", which was
5088 confusing, to say the least. The bug was that the host name should have
5089 appeared instead of "NULL".
5091 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5092 user that is built into the binary, and the -C or -D options is used, root
5093 privilege is dropped before the configuration file is read. In addition,
5094 logging is switched to stderr instead of the normal log files. If the
5095 configuration then re-defines the Exim user, the unprivileged environment
5096 is probably not what is expected, so Exim logs a panic warning message (but
5099 However, if deliver_drop_privilege is set, the unprivileged state may well
5100 be exactly what is intended, so the warning has been cut out in that case,
5101 and Exim is allowed to try to write to its normal log files.
5107 1. smtp_return_error_details was not giving details for temporary sender
5108 or receiver verification errors.
5110 2. Diagnose a configuration error if two authenticators have the same public
5113 3. Exim used not to create the message log file for a message until the first
5114 delivery attempt. This could be confusing when incoming messages were held
5115 for policy or load reasons. The message log file is now created at the time
5116 the message is received, and an initial "Received" line is written to it.
5118 4. The automatically generated man page for command line options had a minor
5119 bug that caused no ill effects; however, a more serious problem was that
5120 the procedure for building the man page automatically didn't always
5121 operate. Consequently, release 4.20 contains an out-of-date version. This
5122 shouldn't happen again.
5124 5. When building Exim with embedded Perl support, the script that builds the
5125 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5126 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5128 6. The freeze_tell option was not being used for messages that were frozen on
5129 arrival, either by an ACL or by local_scan().
5131 7. Added the smtp_incomplete_transaction log selector.
5133 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5134 was accepting AUTH without a new EHLO.
5136 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5137 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5140 10. Logging of TCP/IP connections (when configured) now happens in the main
5141 daemon process instead of the child process, so that the TCP/IP connection
5142 count is more accurate (but it can never be perfect).
5144 11. The use of "drop" in a nested ACL was not being handled correctly in the
5145 outer ACL. Now, if condition failure induced by the nested "drop" causes
5146 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5147 or "require"), the connection is dropped.
5149 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5150 that yield "discard" can now be used with an "accept" or a "discard" verb,
5151 but an error is generated for any others (because I can't see a useful way
5152 to define what should happen).
5154 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5155 are now processed in the same way as lines in the Exim configuration file.
5156 In particular, continuation lines are supported.
5158 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5160 15. Added -ti meaning -t -i.
5162 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5163 domains, and warn by logging if others are found.
5165 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5166 returned from GIFCONF when Exim is trying to find the list of interfaces on
5167 a host. The code in os.c has been modified to copy each ifreq to an aligned
5168 structure in all cases.
5170 Also, in some cases, the returned ifreq's were being copied to a 'struct
5171 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5172 means the last couple of bytes of an IPv6 address could be chopped if the
5173 ifreq contained only a normal sockaddr (14 bytes storage).
5175 18. Named domain lists were not supported in the hosts_treat_as_local option.
5176 An entry such as +xxxx was not recognized, and was treated as a literal
5179 19. Ensure that header lines added by a DATA ACL are included in the reject log
5180 if the ACL subsequently rejects the message.
5182 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5183 MD5 (which is deprecated).
5185 21. When testing a filter file using -bf, Exim was writing a message when it
5186 took the sender from a "From " line in the message, but it was not doing so
5187 when it took $return_path from a Return-Path: header line. It now does.
5189 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5190 with a valid header line field name (a series of printing characters
5191 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5193 23. Changed "disc" in the source to "disk" to conform to the documentation and
5194 the book and for uniformity.
5196 24. Ignore Sendmail's -Ooption=value command line item.
5198 25. When execve() failed while trying to run a command in a pipe transport,
5199 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5200 could be confused with a return value of 69 from the command itself. This
5201 has been changed to 127, the value the shell returns if it is asked to run
5202 a non-existent command. The wording for the related log line suggests a
5203 non-existent command as the problem.
5205 26. If received_header_text expands to an empty string, do not add a Received:
5206 header line to the message. (Well, it adds a token one on the spool, but
5207 marks it "old" so that it doesn't get used or transmitted.)
5209 27. Installed eximstats 1.28 (addition of -nt option).
5211 28. There was no check for failure on the call to getsockname() in the daemon
5212 code. This can fail if there is a shortage of resources on the system, with
5213 ENOMEM, for example. A temporary error is now given on failure.
5215 29. Contrary to the C standard, it seems that in some environments, the
5216 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5217 program. Exim now does this explicitly; it affects the formatting of
5218 timestamps using strftime().
5220 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5221 complaints. I've now initialized all the variables, to avoid this.
5223 32. Header lines added by a system filter were not being "seen" during
5224 transport-time rewrites.
5226 33. The info_callback() function passed to OpenSSL is set up with type void
5227 (*)(SSL *, int, int), as described somewhere. However, when calling the
5228 function (actually a macro) that sets it up, the type void(*)() is
5229 expected. I've put in a cast to prevent warnings from picky compilers.
5231 34. If a DNS black list lookup found a CNAME record, but there were no A
5232 records associated with the domain it pointed at, Exim crashed.
5234 35. If a DNS black list lookup returned more than one A record, Exim ignored
5235 all but the first. It now scans all returned addresses if a particular IP
5236 value is being sought. In this situation, the contents of the
5237 $dnslist_value variable are a list of all the addresses, separated by a
5240 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5241 to accept a host name and all its aliases if the forward lookup for any of
5242 them yielded the IP address of the incoming connection. Now it accepts only
5243 those names whose forward lookup yields the correct IP address. Any other
5244 names are discarded. This closes a loophole whereby a rogue DNS
5245 administrator could create reverse DNS records to break through a
5246 wildcarded host restriction in an ACL.
5248 37. If a user filter or a system filter that ran in a subprocess used any of
5249 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5250 the wrong values were passed to the pipe command ($thisaddress had the
5251 value of $0, $0 had the value of $1, etc). This bug was introduced by
5252 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5254 38. Improved the line breaking for long SMTP error messages from ACLs.
5255 Previously, if there was no break point between 40 and 75 characters, Exim
5256 left the rest of the message alone. Two changes have been made: (a) I've
5257 reduced the minimum length to 35 characters; (b) if it can't find a break
5258 point between 35 and 75 characters, it looks ahead and uses the first one
5259 that it finds. This may give the occasional overlong line, but at least the
5260 remaining text gets split now.
5262 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5263 file descriptors might be low, and that setting 1000 would always raise it.
5264 It turns out that in some environments, the limit is already over 1000 and
5265 that lowering it causes trouble. So now Exim takes care not to decrease it.
5267 40. When delivering a message, the value of $return_path is set to $sender_
5268 address at the start of routing (routers may change the value). By an
5269 oversight, this default was not being set up when an address was tested by
5270 -bt or -bv, which affected the outcome if any router or filter referred to
5273 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5274 log only when "message" or "log_message" are set. However, if one of the
5275 conditions was an address verification, or a call to a nested ACL, the
5276 messages generated by the underlying test were being passed through. This
5277 no longer happens. The underlying message is available in $acl_verify_
5278 message for both "message" and "log_message" expansions, so it can be
5279 passed through if needed.
5281 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5282 new expansion $bh_ to give the encoded byte string without charset
5283 translation. Translation happens only if iconv() is available; HAVE_ICONV
5284 indicates this at build time. HEADERS_CHARSET gives the charset to
5285 translate to; headers_charset can change it in the configuration, and
5286 "headers charset" can change it in an individual filter file.
5288 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5289 that creates RFC 2047 encoded "words" labels them as that charset instead
5290 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5291 expansion operator; (ii) when Exim creates a From: line for a local
5292 message; (iii) when a header line is rewritten to include a "phrase" part.
5294 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5295 buggy, causing it to skip the first lines of messages whose message ID
5296 ended in 'D'. This would not have bitten before Exim release 4.14, because
5297 message IDs were unlikely to end in 'D' before then. The effect was to have
5298 incorrect size information for certain domains.
5300 45. #include "config.h" was missing at the start of the crypt16.c module. This
5301 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5304 46. If there was a timeout during a "random" callout check, Exim treated it as
5305 a failure of the random address, and carried on sending RSET and the real
5306 address. If the delay was just some slowness somewhere, the response to the
5307 original RCPT would be taken as a response to RSET and so on, causing
5308 mayhem of various kinds.
5310 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5311 when I implemented it. It didn't allow for the fact that some option values
5312 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5313 the right test for positive values.
5315 48. Domain names in DNS records are case-independent. Exim always looks them up
5316 in lower case. Some resolvers return domain names in exactly the case they
5317 appear in the zone file, that is, they may contain uppercase letters. Not
5318 all resolvers do this - some return always lower case. Exim was treating a
5319 change of case by a resolver as a change of domain, similar to a widening
5320 of a domain abbreviation. This triggered its re-routing code and so it was
5321 trying to route what was effectively the same domain again. This normally
5322 caused routing to fail (because the router wouldn't handle the domain
5323 twice). Now Exim checks for this case specially, and just changes the
5324 casing of the domain that it ultimately uses when it transmits the message
5327 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5330 50. If a filter generated a file delivery with a non-absolute name (possible if
5331 no home directory exists for the router), the forbid_file option was not
5334 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5335 the existing equality matching.
5337 52. Exim was using ints instead of ino_t variables in some places where it was
5338 dealing with inode numbers.
5340 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5341 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5342 called TMPDIR, and if it finds it is different, it changes its value.
5344 54. The smtp_printf() function is now made available to local_scan() so
5345 additional output lines can be written before returning. There is also an
5346 smtp_fflush() function to enable the detection of a dropped connection.
5347 The variables smtp_input and smtp_batched_input are exported to
5350 55. Changed the default runtime configuration: the message "Unknown user"
5351 has been removed from the ACL, and instead placed on the localuser router,
5352 using the cannot_route_message feature. This means that any verification
5353 failures that generate their own messages won't get overridden. Similarly,
5354 the "Unrouteable address" message that was in the ACL for unverifiable
5355 relay addresses has also been removed.
5357 56. Added hosts_avoid_esmtp to the smtp transport.
5359 57. The exicyclog script was not checking for the esoteric option
5360 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5361 will work only if exicyclog is run under the appropriate euid.
5363 58. Following a discussion on the list, the rules by which Exim recognises line
5364 endings on incoming messages have been changed. The -dropcr and drop_cr
5365 options are now no-ops, retained only for backwards compatibility. The
5366 following line terminators are recognized: LF CRLF CR. However, special
5367 processing applies to CR:
5369 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5370 nor a local message in the state where . is a terminator.
5372 (ii) If a bare CR is encountered in a header line, an extra space is added
5373 after the line terminator so as not to end the header. The reasoning
5374 behind this is that bare CRs in header lines are most likely either
5375 to be mistakes, or people trying to play silly games.
5377 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5378 was being incorrectly given as 18 bytes larger than it should have been.
5379 This is a VOB (very old bug).
5381 60. This may never have affected anything current, but just in case it has:
5382 When the local host is found other than at the start of a list of hosts,
5383 the local host, those with the same MX, and any that follow, are discarded.
5384 When the list in question was part of a longer list of hosts, the following
5385 hosts (not currently being processed) were also being discarded. This no
5386 longer happens. I'm not sure if this situation could ever has previously
5389 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5392 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5393 header. This is something that is recommended in a new Internet Draft, and
5394 is something that is documented as being done by Sendmail. There are two
5395 possible values. For messages generated by the autoreply transport, Exim
5398 Auto-Submitted: auto-replied
5400 whereas for all other generated messages (e.g. bounces) it adds
5402 Auto-Submitted: auto-generated
5404 63. The "personal" condition in filters now includes a test for the
5405 Auto-Submitted: header. If it contains the string "auto-" the message it
5406 not considered personal.
5408 64. Added rcpt_include_affixes as a generic transport option.
5410 65. Added queue_only_override (default true).
5412 66. Added the syslog_duplication option.
5414 67. If what should have been the first header line of a message consisted of
5415 a space followed by a colon, Exim was mis-interpreting it as a header line.
5416 It isn't of course - it is syntactically invalid and should therefore be
5417 treated as the start of the message body. The misbehaviour could have
5418 caused a number of strange effects, including loss of data in subsequent
5419 header lines, and spool format errors.
5421 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5422 client host had authenticated. This control can now be exercised by an ACL
5423 for more flexibility.
5425 69. By default, callouts do not happen when testing with -bh. There is now a
5426 variant, -bhc, which does actually run the callout code, including
5427 consulting and updating the callout cache.
5429 70. Added support for saslauthd authentication, courtesy of Alexander
5432 71. If statvfs() failed on the spool or log directories while checking their
5433 size for availability, Exim confusingly gave the error "space shortage".
5434 Furthermore, in debugging mode it crashed with a floating point exception.
5435 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5436 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5437 serious problem, Exim now writes to the main and panic logs when this
5438 happens, with details of the failure. It then refuses to accept the
5439 incoming message, giving the message "spool directory problem" or "log
5440 directory problem" with a 421 code for SMTP messages.
5442 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5443 0, 1, and 2 exist, because some OS complain for execs without them (see
5444 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5445 descriptors. However, the code omitted to check that the open succeeded,
5446 causing mysterious errors if for some reason the permissions on /dev/null
5447 got screwed. Now Exim writes a message to the main and panic logs, and
5448 bombs out if it can't open /dev/null.
5450 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5451 interact so that it is all more flexible. It is supposed to remain
5452 backwards compatible. Also added extra_local_interfaces.
5454 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5455 to bomb out with an assertion failure - to the client this appears as a
5456 connection drop. This problem occurs in the part of the code that was taken
5457 from the Samba project. Fortunately, the assertion is in a very simple
5458 function, so I have fixed this by reproducing the function inline in the
5459 one place where it is called, and arranging for authentication to fail
5460 instead of killing the process with assert().
5462 75. The SPA client code was not working when the server requested OEM rather
5463 than Unicode encoding.
5465 76. Added code to make require_files with a specific uid setting more usable in
5466 the case where statting the file as root fails - usually a non-root-mounted
5467 NFS file system. When this happens and the failure is EACCES, Exim now
5468 forks a subprocess and does the per-uid checking as the relevant uid.
5470 77. Added process_log_path.
5472 78. If log_file_path was not explicitly set, a setting of check_log_space or
5473 check_log_inodes was ignored.
5475 79. If a space check for the spool or log partitions fails, the incident is now
5476 logged. Of course, in the latter case the data may get lost...
5478 80. Added the %p formatting code to string_format() so that it can be used to
5479 print addresses in debug_print(). Adjusted all the address printing in the
5480 debugging in store.c to use %p rather than %d.
5482 81. There was a concern that a line of code in smtp_in.c could overflow a
5483 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5484 initially expressed, the concern was not well-founded, because trailing
5485 spaces are removed early. However, if the trailing spaces were followed by
5486 a NULL, they did not get removed, so the overflow was possible. Two fixes
5489 (a) I re-wrote the offending code in a cleaner fashion.
5490 (b) If an incoming SMTP command contains a NULL character, it is rejected
5493 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5494 runs initgroups(), so that if the Exim user is in any additional groups,
5495 they will be used during message reception.
5501 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob