1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
52 PP/01 Solaris build fix for Oracle's LDAP libraries.
53 Bugzilla 1109, patch from Stephen Usher.
55 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
57 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
60 TF/02 Fix a couple more cases where we did not log the error message
61 when unlink() failed. See also change 4.74-TF/03.
63 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
64 lock up or crash if it happened to be inside a call to libc when it
65 got a SIGUSR1 from exiwhat.
67 The SIGUSR1 handler appends the current process status to the process
68 log which is later printed by exiwhat. It used to use the general
69 purpose logging code to do this, but several functions it calls are
72 The new output code in the SIGUSR1 handler is specific to the process
73 log, and simple enough that it's easy to inspect for signal safety.
74 Removing some special cases also simplifies the general logging code.
75 Removing the spurious timestamps from the process log simplifies
78 TF/04 Improved ratelimit ACL condition.
80 The /noupdate option has been deprecated in favour of /readonly which
81 has clearer semantics. The /leaky, /strict, and /readonly update modes
82 are mutually exclusive. The update mode is no longer included in the
83 database key; it just determines when the database is updated. (This
84 means that when you upgrde Exim will forget old rate measurements.)
86 Exim now checks that the per_* options are used with an update mode that
87 makes sense for the current ACL. For example, when Exim is processing a
88 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
89 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
90 must specify per_mail/readonly. If you omit the update mode it defaults to
91 /leaky where that makes sense (as before) or /readonly where required.
93 The /noupdate option is now undocumented but still supported for
94 backwards compatibility. It is equivalent to /readonly except that in
95 ACLs where /readonly is required you may specify /leaky/noupdate or
96 /strict/noupdate which are treated the same as /readonly.
98 A useful new feature is the /count= option. This is a generalization
99 of the per_byte option, so that you can measure the throughput of other
100 aggregate values. For example, the per_byte option is now equivalent
101 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
103 The per_rcpt option has been generalized using the /count= mechanism
104 (though it's more complicated than the per_byte equivalence). When it is
105 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
106 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
107 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
108 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
109 Note that using per_rcpt with a non-readonly update mode in more than
110 one ACL will cause the recipients to be double-counted. (The per_mail
111 and per_byte options don't have this problem.)
113 The handling of very low rates has changed slightly. If the computed rate
114 is less than the event's count (usually one) then this event is the first
115 after a long gap. In this case the rate is set to the same as this event's
116 count, so that the first message of a spam run is counted properly.
118 The major new feature is a mechanism for counting the rate of unique
119 events. The new per_addr option counts the number of different
120 recipients that someone has sent messages to in the last time period. It
121 behaves like per_rcpt if all the recipient addresses are different, but
122 duplicate recipient addresses do not increase the measured rate. Like
123 the /count= option this is a general mechanism, so the per_addr option
124 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
125 example, measure the rate that a client uses different sender addresses
126 with the options per_mail/unique=$sender_address. There are further
127 details in the main documentation.
129 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
131 TF/06 Removed a few PCRE remnants.
133 TF/07 Automatically extract Exim's version number from tags in the git
134 repository when doing development or release builds.
136 PP/02 Raise smtp_cmd_buffer_size to 16kB.
137 Bugzilla 879. Patch from Paul Fisher.
139 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
140 Heavily based on revision 40f9a89a from Simon Arlott's tree.
143 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
145 PP/05 Variable $av_failed, true if the AV scanner deferred.
146 Bugzilla 1078. Patch from John Horne.
148 PP/06 Stop make process more reliably on build failure.
149 Bugzilla 1087. Patch from Heiko Schlittermann.
151 PP/07 Make maildir_use_size_file an _expandable_ boolean.
152 Bugzilla 1089. Patch from Heiko Schlittermann.
154 PP/08 Handle ${run} returning more data than OS pipe buffer size.
155 Bugzilla 1131. Patch from Holger Weiß.
157 PP/09 Handle IPv6 addresses with SPF.
158 Bugzilla 860. Patch from Wolfgang Breyha.
160 PP/10 GnuTLS: support TLS 1.2 & 1.1.
162 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
165 PP/11 match_* no longer expand right-hand-side by default.
166 New compile-time build option, EXPAND_LISTMATCH_RHS.
167 New expansion conditions, "inlist", "inlisti".
169 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
171 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
173 PP/14 fix log_write() format string regression from TF/03.
174 Bugzilla 1152. Patch from Dmitry Isaikin.
180 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
182 PP/02 Harmonised TLS library version reporting; only show if debugging.
183 Layout now matches that introduced for other libraries in 4.74 PP/03.
185 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
187 PP/04 New "dns_use_edns0" global option.
189 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
192 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
193 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
195 TK/01 Updated PolarSSL code to 0.14.2.
196 Bugzilla 1097. Patch from Andreas Metzler.
198 PP/07 Catch divide-by-zero in ${eval:...}.
201 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
204 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
205 format-string attack -- SECURITY: remote arbitrary code execution.
207 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
208 time unintentionally subject to list matching rules, letting the header
209 cause arbitrary Exim lookups (of items which can occur in lists, *not*
210 arbitrary string expansion). This allowed for information disclosure.
212 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
213 INT_MIN/-1 -- value coerced to INT_MAX.
219 NM/01 Workround for PCRE version dependency in version reporting
222 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
223 This fixes portability to compilers other than gcc, notably
224 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
226 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
227 makefiles for portability to HP-UX and POSIX correctness.
229 PP/01 Permit LOOKUP_foo enabling on the make command-line.
230 Also via indented variable definition in the Makefile.
231 (Debugging by Oliver Heesakkers).
233 PP/02 Restore caching of spamd results with expanded spamd_address.
234 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
236 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
237 Improves build reliability. Fix from: Frank Elsner
239 NM/02 Fix wide character breakage in the rfc2047 coding
240 Fixes bug 1064. Patch from Andrey N. Oktyabrski
242 NM/03 Allow underscore in dnslist lookups
243 Fixes bug 1026. Patch from Graeme Fowler
245 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
246 Code patches from Adam Ciarcinski of NetBSD.
248 NM/04 Fixed exiqgrep to cope with mailq missing size issue
251 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
252 is logged, to avoid truncation. Patch from John Horne.
254 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
255 Patch from Jakob Hirsch.
257 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
258 SQL string expansion failure details.
259 Patch from Andrey Oktyabrski.
261 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
262 Patch from Simon Arlott.
264 PP/09 New lookups functionality failed to compile on old gcc which rejects
265 extern declarations in function scope.
266 Patch from Oliver Fleischmann
268 PP/10 Use sig_atomic_t for flags set from signal handlers.
269 Check getgroups() return and improve debugging.
270 Fixed developed for diagnosis in bug 927 (which turned out to be
273 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
274 Patch from Mark Zealey.
276 PP/12 Bugzilla 1056: Improved spamd server selection.
277 Patch from Mark Zealey.
279 PP/13 Bugzilla 1086: Deal with maildir quota file races.
280 Based on patch from Heiko Schlittermann.
282 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
283 Patch from Uwe Doering, sign-off by Michael Haardt.
285 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
286 variable declaration deep within a block. Bug and patch from
289 PP/15 lookups-Makefile IRIX compatibilty coercion.
291 PP/16 Make DISABLE_DKIM build knob functional.
293 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
294 Patch by Simon Arlott
296 TF/03 Fix valgrind.h portability to C89 compilers that do not support
297 variable argument macros. Our copy now differs from upstream.
303 TF/01 Failure to get a lock on a hints database can have serious
304 consequences so log it to the panic log.
306 TF/02 Log LMTP confirmation messages in the same way as SMTP,
307 controlled using the smtp_confirmation log selector.
309 TF/03 Include the error message when we fail to unlink a spool file.
311 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
312 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
313 for maintaining out-of-tree patches for some time.
315 PP/01 Bugzilla 139: Documentation and portability issues.
316 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
317 Handle per-OS dynamic-module compilation flags.
319 PP/02 Let /dev/null have normal permissions.
320 The 4.73 fixes were a little too stringent and complained about the
321 permissions on /dev/null. Exempt it from some checks.
322 Reported by Andreas M. Kirchwitz.
324 PP/03 Report version information for many libraries, including
325 Exim version information for dynamically loaded libraries. Created
326 version.h, now support a version extension string for distributors
327 who patch heavily. Dynamic module ABI change.
329 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
330 privilege escalation vulnerability whereby the Exim run-time user
331 can cause root to append content of the attacker's choosing to
334 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
337 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
338 If dropping privileges for untrusted macros, we disabled normal logging
339 on the basis that it would fail; for the Exim run-time user, this is not
340 the case, and it resulted in successful deliveries going unlogged.
341 Fixed. Reported by Andreas Metzler.
347 PP/01 Date: & Message-Id: revert to normally being appended to a message,
348 only prepend for the Resent-* case. Fixes regression introduced in
349 Exim 4.70 by NM/22 for Bugzilla 607.
351 PP/02 Include check_rfc2047_length in configure.default because we're seeing
352 increasing numbers of administrators be bitten by this.
354 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
356 PP/03 Bugzilla 994: added openssl_options main configuration option.
358 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
360 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
362 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
364 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
365 without a peer certificate, leading to a segfault because of an
366 assumption that peers always have certificates. Be a little more
367 paranoid. Problem reported by Martin Tscholak.
369 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
370 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
371 NB: ClamAV planning to remove STREAM in "middle of 2010".
372 CL also introduces -bmalware, various -d+acl logging additions and
373 more caution in buffer sizes.
375 PP/09 Implemented reverse_ip expansion operator.
377 PP/10 Bugzilla 937: provide a "debug" ACL control.
379 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
381 PP/12 Bugzilla 973: Implement --version.
383 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
385 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
387 PP/15 Bugzilla 816: support multiple condition rules on Routers.
389 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
390 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
391 ignore trailing whitespace.
393 JJ/02 prevent non-panic DKIM error from being sent to paniclog
395 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
398 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
399 Notification from Dr Andrew Aitchison.
401 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
402 ExtendedDetectionInfo response format.
403 Notification from John Horne.
405 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
408 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
409 XSL and documented dependency on system catalogs, with examples of how
412 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
415 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
416 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
417 configuration file which is writeable by the Exim user or group.
419 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
420 of configuration files to cover files specified with the -C option if
421 they are going to be used with root privileges, not just the default
424 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
425 option (effectively making it always true).
427 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
428 files to be used while preserving root privileges.
430 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
431 that rogue child processes cannot use them.
433 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
434 run-time user, instead of root.
436 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
437 Exim run-time user without dropping privileges.
439 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
440 result string, instead of calling string_vformat() twice with the same
443 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
444 for other users. Others should always drop root privileges if they use
445 -C on the command line, even for a whitelisted configure file.
447 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
449 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
455 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
456 $data_path, and $header_path variables; fixed documentation bugs and
459 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
460 exipick to access non-standard spools, including the "frozen" queue
463 NM/01 Bugzilla 965: Support mysql stored procedures.
464 Patch from Alain Williams
466 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
468 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
469 Patch from Andreas Metzler
471 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
472 Patch from Kirill Miazine
474 NM/05 Bugzilla 671: Added umask to procmail example.
476 JJ/03 installed exipick 20100323.0, fixing doc bug
478 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
479 directory. Notification and patch from Dan Rosenberg.
481 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
483 TK/02 Improve log output when DKIM signing operation fails.
485 MH/01 Treat the transport option dkim_domain as a colon separated
486 list, not as a single string, and sign the message with each element,
487 omitting multiple occurences of the same signer.
489 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
490 Bugzilla 985, 986. Patch by Simon Arlott
492 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
493 Patch by Simon Arlott
495 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
496 MBX locking. Notification from Dan Rosenberg.
502 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
504 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
506 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
508 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
510 NM/04 Bugzilla 915: Flag broken perl installation during build.
516 TK/01 Added patch by Johannes Berg that expands the main option
517 "spamd_address" if it starts with a dollar sign.
519 TK/02 Write list of recipients to X-Envelope-Sender header when building
520 the mbox-format spool file for content scanning (suggested by Jakob
523 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
524 (http://www.dcc-servers.net/) support via dccifd. Activated by
525 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
527 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
528 by Mark Daniel Reidel <mr@df.eu>.
530 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
531 When building exim an external PCRE library is now needed -
532 PCRE is a system library on the majority of modern systems.
533 See entry on PCRE_LIBS in EDITME file.
535 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
536 conversation. Added nologin parameter to request.
537 Patch contributed by Kirill Miazine.
539 TF/01 Do not log submission mode rewrites if they do not change the address.
541 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
543 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
544 log files in place. Contributed by Roberto Lima.
546 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
548 TF/03 Bugzilla 615: When checking the local_parts router precondition
549 after a local_part_suffix or local_part_prefix option, Exim now
550 does not use the address's named list lookup cache, since this
551 contains cached lookups for the whole local part.
553 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
554 Robert Millan. Documentation is in experimental-spec.txt.
556 TF/04 Bugzilla 668: Fix parallel build (make -j).
558 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
560 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
561 Patch provided by Jan Srzednicki.
563 TF/05 Leading white space used to be stripped from $spam_report which
564 wrecked the formatting. Now it is preserved.
566 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
567 that they are available at delivery time.
569 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
571 TF/08 TLS error reporting now respects the incoming_interface and
572 incoming_port log selectors.
574 TF/09 Produce a more useful error message if an SMTP transport's hosts
575 setting expands to an empty string.
577 NM/06 Bugzilla 744: EXPN did not work under TLS.
578 Patch provided by Phil Pennock.
580 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
581 Patch provided by Richard Godbee.
583 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
584 acl_smtp_notquit, added index entry.
586 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
587 Patch provided by Eugene Bujak.
589 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
590 accept(). Patch provided by Maxim Dounin.
592 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
593 Patch provided by Phil Pennock.
595 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
597 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
598 Patch provided by Brad "anomie" Jorsch.
600 NM/14 Bugzilla 622: Added timeout setting to transport filter.
601 Patch provided by Dean Brooks.
603 TK/05 Add native DKIM support (does not depend on external libraries).
605 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
606 Patch provided by Graeme Fowler.
608 NM/16 Bugzilla 851: Documentation example syntax fix.
610 NM/17 Changed NOTICE file to remove references to embedded PCRE.
612 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
615 NM/19 Bugzilla 745: TLS version reporting.
616 Patch provided by Phil Pennock.
618 NM/20 Bugzilla 167: bool: condition support.
619 Patch provided by Phil Pennock.
621 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
622 clients. Patch provided by Phil Pennock.
624 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
625 Patch provided by Brad "anomie" Jorsch.
627 NM/23 Bugzilla 687: Fix misparses in eximstats.
628 Patch provided by Heiko Schlittermann.
630 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
631 Patch provided by Heiko Schlittermann.
633 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
634 plus update to original patch.
636 NM/26 Bugzilla 799: Documentation correction for ratelimit.
638 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
639 Patch provided by David Brownlee.
641 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
643 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
645 NM/30 Bugzilla 888: TLS documentation bugfixes.
647 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
649 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
650 Unlike the original bugzilla I have changed all shell scripts in src tree.
652 NM/33 Bugzilla 898: Transport filter timeout fix.
653 Patch by Todd Rinaldo.
655 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
656 Patch by Serge Demonchaux.
658 NM/35 Bugzilla 39: Base64 decode bug fixes.
659 Patch by Jakob Hirsch.
661 NM/36 Bugzilla 909: Correct connect() call in dcc code.
663 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
665 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
667 NM/39 Bugzilla 911: Fixed MakeLinks build script.
673 TK/01 Add preliminary DKIM support. Currently requires a forked version of
674 ALT-N's libdkim that I have put here:
675 http://duncanthrax.net/exim-experimental/
677 Note to Michael Haardt: I had to rename some vars in sieve.c. They
678 were called 'true' and it seems that C99 defines that as a reserved
679 keyword to be used with 'bool' variable types. That means you could
680 not include C99-style headers which use bools without triggering
681 build errors in sieve.c.
683 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
684 as mailq or other aliases. Changed the --help handling significantly
685 to do whats expected. exim_usage() emits usage/help information.
687 SC/01 Added the -bylocaldomain option to eximstats.
689 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
691 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
693 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
699 PH/01 Another patch from the Sieve maintainer.
701 PH/02 When an IPv6 address is converted to a string for single-key lookup
702 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
703 dots are used instead of colons so that keys in lsearch files need not
704 contain colons. This was done some time before quoting was made available
705 in lsearch files. However, iplsearch files do require colons in IPv6 keys
706 (notated using the quote facility) so as to distinguish them from IPv4
707 keys. This meant that lookups for IP addresses in host lists did not work
708 for iplsearch lookups.
710 This has been fixed by arranging for IPv6 addresses to be expressed with
711 colons if the lookup type is iplsearch. This is not incompatible, because
712 previously such lookups could never work.
714 The situation is now rather anomolous, since one *can* have colons in
715 ordinary lsearch keys. However, making the change in all cases is
716 incompatible and would probably break a number of configurations.
718 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
721 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
722 conversion specification without a maximum field width, thereby enabling
723 a rogue spamd server to cause a buffer overflow. While nobody in their
724 right mind would setup Exim to query an untrusted spamd server, an
725 attacker that gains access to a server running spamd could potentially
726 exploit this vulnerability to run arbitrary code as the Exim user.
728 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
729 $primary_hostname instead of what libspf2 thinks the hosts name is.
731 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
732 a directory entry by the name of the lookup key. Previously, if a
733 symlink pointed to a non-existing file or a file in a directory that
734 Exim lacked permissions to read, a lookup for a key matching that
735 symlink would fail. Now it is enough that a matching directory entry
736 exists, symlink or not. (Bugzilla 503.)
738 PH/03 The body_linecount and body_zerocount variables are now exported in the
741 PH/04 Added the $dnslist_matched variable.
743 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
744 This means they are set thereafter only if the connection becomes
747 PH/06 Added the client_condition to authenticators so that some can be skipped
748 by clients under certain conditions.
750 PH/07 The error message for a badly-placed control=no_multiline_responses left
751 "_responses" off the end of the name.
753 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
755 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
756 (without spaces) instead of just copying the configuration text.
758 PH/10 Added the /noupdate option to the ratelimit ACL condition.
760 PH/11 Added $max_received_linelength.
762 PH/12 Added +ignore_defer and +include_defer to host lists.
764 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
765 way in which PCRE > 7.0 is built.
767 PH/14 Implemented queue_only_load_latch.
769 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
770 MAIL command. The effect was to mangle the value on 64-bit systems.
772 PH/16 Another patch from the Sieve maintainer.
774 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
776 PH/18 If a system quota error occurred while trying to create the file for
777 a maildir delivery, the message "Mailbox is full" was not appended to the
778 bounce if the delivery eventually timed out. Change 4.67/27 below applied
779 only to a quota excession during the actual writing of the file.
781 PH/19 It seems that peer DN values may contain newlines (and other non-printing
782 characters?) which causes problems in log lines. The DN values are now
783 passed through string_printing() before being added to log lines.
785 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
786 and InterBase are left for another time.)
788 PH/21 Added message_body_newlines option.
790 PH/22 Guard against possible overflow in moan_check_errorcopy().
792 PH/23 POSIX allows open() to be a macro; guard against that.
794 PH/24 If the recipient of an error message contained an @ in the local part
795 (suitably quoted, of course), incorrect values were put in $domain and
796 $local_part during the evaluation of errors_copy.
802 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
803 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
806 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
807 issue a MAIL command.
809 PH/02 In an ACL statement such as
811 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
813 if a client was not listed at all, or was listed with a value other than
814 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
815 the condition was not true (as it should be), so access was not denied.
816 The bug was that the ! inversion was incorrectly passed on to the second
817 item. This has been fixed.
819 PH/03 Added additional dnslists conditions == and =& which are different from
820 = and & when the dns lookup returns more than one IP address.
822 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
823 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
825 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
826 FSYNC, which compiles an option called disable_fsync that allows for
827 bypassing fsync(). The documentation is heavily laced with warnings.
829 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
831 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
832 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
833 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
834 including adding "make clean"; (3) Added -fPIC when compiling the test
835 dynamically loaded module, to get rid of a warning.
837 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
838 message fails, move_frozen_messages = true and ignore_bounce_errors_after
839 = 0s. The bug is otherwise harmless.
841 PH/07 There was a bug in the dovecot authenticator such that the value of
842 $auth1 could be overwritten, and so not correctly preserved, after a
843 successful authentication. This usually meant that the value preserved by
844 the server_setid option was incorrect.
846 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
848 PH/09 Installed PCRE release 7.0.
850 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
851 run for batched SMTP input. It is now run at the start of every message
852 in the batch. While fixing this I discovered that the process information
853 (output by running exiwhat) was not always getting set for -bs and -bS
854 input. This is fixed, and it now also says "batched" for BSMTP.
856 PH/11 Added control=no_pipelining.
858 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
859 patch, slightly modified), and move the expansion of helo_data till after
860 the connection is made in the smtp transport (so it can use these
863 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
865 PH/14 Added log_selector = +pid.
867 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
869 PH/16 Add ${if forany and ${if forall.
871 PH/17 Added dsn_from option to vary the From: line in DSNs.
873 PH/18 Flush SMTP output before performing a callout, unless control =
874 no_callout_flush is set.
876 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
877 was true (the default) a successful delivery failed to delete the retry
878 item, thus causing premature timeout of the address. The bug is now
881 PH/20 Added hosts_avoid_pipelining to the smtp transport.
883 PH/21 Long custom messages for fakedefer and fakereject are now split up
884 into multiline reponses in the same way that messages for "deny" and
885 other ACL rejections are.
887 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
888 with slight modification.
890 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
891 draft, changing the syntax and factoring some duplicate code".
893 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
894 for deliveries of the second and subsequent messages over the same SMTP
897 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
898 ${reduce, with only minor "tidies".
900 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
902 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
903 expansion side effects.
905 PH/27 When a message times out after an over-quota error from an Exim-imposed
906 quota, the bounce message says "mailbox is full". This message was not
907 being given when it was a system quota that was exceeded. It now should
910 MH/03 Made $recipients available in local_scan(). local_scan() already has
911 better access to the recipient list through recipients_list[], but
912 $recipients can be useful in postmaster-provided expansion strings.
914 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
915 in the case of a MAIL command with additional options following the
916 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
917 were accidentally chopped off.
919 PH/29 SMTP synchronization checks are implemented when a command is read -
920 there is a check that no more input is waiting when there shouldn't be
921 any. However, for some commands, a delay in an ACL can mean that it is
922 some time before the response is written. In this time, more input might
923 arrive, invalidly. So now there are extra checks after an ACL has run for
924 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
925 pipelining has not been advertised.
927 PH/30 MH's patch to allow iscntrl() characters to be list separators.
929 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
930 returned in the SMTP response when smtp_return_error_details was false.
933 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
934 instead of the C I/O that was originally supplied, because problems were
937 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
938 Exim which did not show up earlier: it was assuming that a call to
939 SSL_CTX_set_info_callback() might give an error value. In fact, there is
940 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
941 was a macro that became an assignment, so it seemed to work. This has
942 changed to a proper function call with a void return, hence the compile
943 error. Exim's code has been fixed.
945 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
948 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
950 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
952 PH/37 If a message is not accepted after it has had an id assigned (e.g.
953 because it turns out to be too big or there is a timeout) there is no
954 "Completed" line in the log. When some messages of this type were
955 selected by exigrep, they were listed as "not completed". Others were
956 picked up by some special patterns. I have improved the selection
957 criteria to be more general.
959 PH/38 The host_find_failed option in the manualroute router can now be set
960 to "ignore", to completely ignore a host whose IP address cannot be
961 found. If all hosts are ignored, the behaviour is controlled by the new
962 host_all_ignored option.
964 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
965 homing or because of multiple MX records with /mx) generated more than
966 one IP address, and the following item turned out to be the local host,
967 all the secondary addresses of the first item were incorrectly removed
968 from the list, along with the local host and any following hosts (which
969 is what is supposed to happen).
971 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
972 whoever called Exim into the -H file. In the case of the daemon it was
973 behaving confusingly. When first started, it used values for whoever
974 started the daemon, but after a SIGHUP it used the Exim user (because it
975 calls itself on a restart). I have changed the code so that it now always
978 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
979 message are rejected with the same error (e.g. no authentication or bad
980 sender address), and a DATA command is nevertheless sent (as can happen
981 with PIPELINING or a stupid MUA), the error message that was given to the
982 RCPT commands is included in the rejection of the DATA command. This is
983 intended to be helpful for MUAs that show only the final error to their
986 PH/42 Another patch from the Sieve maintainer.
988 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
989 Eximstats - Fixed some broken HTML links and added missing column headers
991 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
992 columns for Rejects, Temp Rejects, Ham, and Spam rows.
994 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
996 PH/43 Yet another patch from the Sieve maintainer.
998 PH/44 I found a way to check for a TCP/IP connection going away before sending
999 the response to the final '.' that terminates a message, but only in the
1000 case where the client has not sent further data following the '.'
1001 (unfortunately, this is allowed). However, in many cases there won't be
1002 any further data because there won't be any more messages to send. A call
1003 to select() can be used: if it shows that the input is "ready", there is
1004 either input waiting, or the socket has been closed. An attempt to read
1005 the next input character can distinguish the two cases. Previously, Exim
1006 would have sent an OK response which the client would never have see.
1007 This could lead to message repetition. This fix should cure that, at
1008 least in a lot of common cases.
1010 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1011 advertised in response to EHLO.
1017 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1018 fixed by 4.65/MH/01 (is this a record?) are fixed:
1020 (i) An empty string was always treated as zero by the numeric comparison
1021 operators. This behaviour has been restored.
1023 (ii) It is documented that the numeric comparison operators always treat
1024 their arguments as decimal numbers. This was broken in that numbers
1025 starting with 0 were being interpreted as octal.
1027 While fixing these problems I realized that there was another issue that
1028 hadn't been noticed. Values of message_size_limit (both the global option
1029 and the transport option) were treated as octal if they started with 0.
1030 The documentation was vague. These values are now always treated as
1031 decimal, and I will make that clear in the documentation.
1037 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1038 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1041 MH/01 Don't check that the operands of numeric comparison operators are
1042 integers when their expansion is in "skipping" mode (fixes bug
1043 introduced by 4.64-PH/07).
1045 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1046 child addresses, Exim now panics and dies. Previously, because the count
1047 is held in a short int, deliveries were likely to be lost. As such a
1048 large number of recipients for a single message is ridiculous
1049 (performance will be very, very poor), I have chosen to impose a limit
1050 rather than extend the field.
1056 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1057 leftover -K file (the existence of which was triggered by #402).
1058 While we were at it, introduced process PID as part of the -K
1059 filename. This should rule out race conditions when creating
1062 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1063 processing considerably. Previous code took too long for large mails,
1064 triggering a timeout which in turn triggers #401.
1066 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1067 in the DK code in transports.c. sendfile() is not really portable,
1068 hence the _LINUX specificness.
1070 TF/01 In the add_headers option to the mail command in an Exim filter,
1071 there was a bug that Exim would claim a syntax error in any
1072 header after the first one which had an odd number of characters
1075 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1076 callout verification, Exim cached a "reject" for the entire domain. This
1077 is correct for most verifications, but it is not correct for a recipient
1078 verification with use_sender or use_postmaster set, because in that case
1079 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1080 case of MAIL FROM:<> rejection from other early rejections (e.g.
1081 rejection of HELO). When verifying a recipient using a non-null MAIL
1082 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1083 Whatever the result of the callout, the value of the domain cache is
1084 left unchanged (for any other kind of callout, getting as far as trying
1085 RCPT means that the domain itself is ok).
1087 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1090 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1091 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1092 session, the daemon ignores ECONNECT errors and logs others; it now
1093 ignores EPIPE as well.
1095 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1096 (quoted-printable decoding).
1098 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1099 later the small subsequent patch to fix an introduced bug.
1101 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1103 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1105 PH/08 An error is now given if message_size_limit is specified negative.
1107 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1108 to be given (somewhat) arbitrary names.
1110 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1113 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1114 miscellaneous code fixes
1116 PH/10 Added the log_reject_target ACL modifier to specify where to log
1119 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1120 hostname. This is wrong, because it relates to the incoming message (and
1121 probably the interface on which it is arriving) and not to the outgoing
1122 callout (which could be using a different interface). This has been
1123 changed to use the value of the helo_data option from the smtp transport
1124 instead - this is what is used when a message is actually being sent. If
1125 there is no remote transport (possible with a router that sets up host
1126 addresses), $smtp_active_hostname is used.
1128 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1129 tweaks were necessary in order to get it to work (see also 21 below):
1130 (a) The code assumed that strncpy() returns a negative number on buffer
1131 overflow, which isn't the case. Replaced with Exim's string_format()
1133 (b) There were several signed/unsigned issues. I just did the minimum
1134 hacking in of casts. There is scope for a larger refactoring.
1135 (c) The code used strcasecmp() which is not a standard C function.
1136 Replaced with Exim's strcmpic() function.
1137 (d) The code set only $1; it now sets $auth1 as well.
1138 (e) A simple test gave the error "authentication client didn't specify
1139 service in request". It would seem that Dovecot has changed its
1140 interface. Fortunately there's a specification; I followed it and
1141 changed what the client sends and it appears to be working now.
1143 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1146 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1147 address A is aliased to B and C, where B exists and C does not. Without
1148 -v the output is "A verified" because verification stops after a
1149 successful redirection if more than one address is generated. However,
1150 with -v the child addresses are also verified. Exim was outputting "A
1151 failed to verify" and then showing the successful verification for C,
1152 with its parentage. It now outputs "B failed to verify", showing B's
1153 parentage before showing the successful verification of C.
1155 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1156 look up a TXT record in a specific list after matching in a combined
1159 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1160 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1161 they consult the DNS. I had assumed they would set it the way they
1162 wanted; and indeed my experiments on Linux seem to show that in some
1163 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1164 To be on the safe side, however, I have now made the interface to
1165 host_find_byname() similar to host_find_bydns(), with an argument
1166 containing the DNS resolver options. The host_find_byname() function now
1167 sets these options at its start, just as host_find_bydns() does. The smtp
1168 transport options dns_qualify_single and dns_search_parents are passed to
1169 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1170 of host_find_byname() use the default settings of RES_DEFNAMES
1171 (qualify_single) but not RES_DNSRCH (search_parents).
1173 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1174 spool_read_header() do less string testing, by means of a preliminary
1175 switch on the second character of optional "-foo" lines. (This is
1176 overdue, caused by the large number of possibilities that now exist.
1177 Originally there were few.) While I was there, I also converted the
1178 str(n)cmp tests so they don't re-test the leading "-" and the first
1179 character, in the hope this might squeeze out yet more improvement.
1181 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1182 flag allowing group syntax was set by the header_syntax check but not
1183 turned off, possible causing trouble later; (2) The flag was not being
1184 set at all for the header_verify test, causing "group"-style headers to
1185 be rejected. I have now set it in this case, and also caused header_
1186 verify to ignore an empty address taken from a group. While doing this, I
1187 came across some other cases where the code for allowing group syntax
1188 while scanning a header line wasn't quite right (mostly, not resetting
1189 the flag correctly in the right place). These bugs could have caused
1190 trouble for malformed header lines. I hope it is now all correct.
1192 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1193 with the "reply" argument non-NULL. The code, however (which originally
1194 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1195 but it didn't always do it. This confused somebody who was copying the
1196 code for some other use. I have removed all the tests.
1198 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1199 feature that was used to support insecure browsers during the U.S. crypto
1200 embargo. It requires special client support, and Exim is probably the
1201 only MTA that supported it -- and would never use it because real RSA is
1202 always available. This code has been removed, because it had the bad
1203 effect of slowing Exim down by computing (never used) parameters for the
1204 RSA_EXPORT functionality.
1206 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1207 authenticator to fail if there's a tab character in the incoming data
1208 (there should never be unless someone is messing about, as it's supposed
1209 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1210 if the connection is using TLS or if the remote IP is the same as the
1211 local IP, and the "valid-client-cert option" if a client certificate has
1214 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1215 authenticators. This can be used for authorization after authentication
1216 succeeds. (In the case of plaintext, it servers for both authentication
1219 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1220 if any retry times were supplied.
1222 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1223 connection, where there is no client IP address to check. In this
1224 situation, the verify now always succeeds.
1226 PH/25 Applied John Jetmore's -Mset patch.
1228 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1230 PH/27 In a string expansion for a processed (not raw) header when multiple
1231 headers of the same name were present, leading whitespace was being
1232 removed from all of them, but trailing whitespace was being removed only
1233 from the last one. Now trailing whitespace is removed from each header
1234 before concatenation. Completely empty headers in a concatenation (as
1235 before) are ignored.
1237 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1238 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1240 PH/29 [Removed. This was a change that I later backed out, and forgot to
1241 correct the ChangeLog entry (that I had efficiently created) before
1242 committing the later change.]
1244 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1245 address errors (4xx response to RCPT) over the same connection as other
1246 messages routed to the same hosts. Such deliveries are always "forced",
1247 so retry times are not inspected. This resulted in far too many retries
1248 for the affected addresses. The effect occurred only when there were more
1249 hosts than the hosts_max_try setting in the smtp transport when it had
1250 the 4xx errors. Those hosts that it had tried were not added to the list
1251 of hosts for which the message was waiting, so if all were tried, there
1252 was no problem. Two fixes have been applied:
1254 (i) If there are any address or message errors in an SMTP delivery, none
1255 of the hosts (tried or untried) are now added to the list of hosts
1256 for which the message is waiting, so the message should not be a
1257 candidate for sending over the same connection that was used for a
1258 successful delivery of some other message. This seems entirely
1259 reasonable: after all the message is NOT "waiting for some host".
1260 This is so "obvious" that I'm not sure why it wasn't done
1261 previously. Hope I haven't missed anything, but it can't do any
1262 harm, as the worst effect is to miss an optimization.
1264 (ii) If, despite (i), such a delivery is accidentally attempted, the
1265 routing retry time is respected, so at least it doesn't keep
1266 hammering the server.
1268 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1269 in ${readsocket because some servers need this prod.
1271 PH/32 Added some extra debug output when updating a wait-xxx database.
1273 PH/33 The hint "could be header name not terminated by colon", which has been
1274 given for certain expansion errors for a long time, was not being given
1275 for the ${if def:h_colon_omitted{... case.
1277 PH/34 The spec says: "With one important exception, whenever a domain list is
1278 being scanned, $domain contains the subject domain." There was at least
1279 one case where this was not true.
1281 PH/35 The error "getsockname() failed: connection reset by peer" was being
1282 written to the panic log as well as the main log, but it isn't really
1283 panic-worthy as it just means the connection died rather early on. I have
1284 removed the panic log writing for the ECONNRESET error when getsockname()
1287 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1288 runs only) independently of the message's sender address. This meant
1289 that, if the 4xx error was in fact related to the sender, a different
1290 message to the same recipient with a different sender could confuse
1291 things. In particualar, this can happen when sending to a greylisting
1292 server, but other circumstances could also provoke similar problems.
1293 I have changed the default so that the retry time for these errors is now
1294 based a combination of the sender and recipient addresses. This change
1295 can be overridden by setting address_retry_include_sender=false in the
1298 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1299 remote server are returned as part of bounce messages. This was not
1300 happening for LMTP over a pipe (the lmtp transport), but now it is the
1301 same for both kinds of LMTP.
1303 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1304 in header lines that were in fact CNAMEs. This is no longer the case.
1306 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1307 and queue runs started by the daemon processed all messages. This has
1308 been fixed so that -R and -S can now usefully be given with -q<time>.
1310 PH/40 Import PCRE release 6.7 (fixes some bugs).
1312 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1314 PH/42 Give an error if -q is specified more than once.
1316 PH/43 Renamed the variables $interface_address and $interface_port as
1317 $received_ip_address and $received_port, to make it clear that these
1318 values apply to message reception, and not to the outgoing interface when
1319 a message is delivered. (The old names remain recognized, of course.)
1321 PH/44 There was no timeout on the connect() call when using a Unix domain
1322 socket in the ${readsocket expansion. There now is.
1324 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1325 be meaningful with "accept".
1327 SC/01 Eximstats V1.43
1328 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1330 SC/02 Eximstats V1.44
1331 Use a glob alias rather than an array ref in the generated
1332 parser. This improves both readability and performance.
1334 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1335 Collect SpamAssassin and rejection statistics.
1336 Don't display local sender or destination tables unless
1337 there is data to show.
1338 Added average volumes into the top table text output.
1340 SC/04 Eximstats V1.46
1341 Collect data on the number of addresses (recipients)
1342 as well as the number of messages.
1344 SC/05 Eximstats V1.47
1345 Added 'Message too big' to the list of mail rejection
1346 reasons (thanks to Marco Gaiarin).
1348 SC/06 Eximstats V1.48
1349 Mainlog lines which have GMT offsets and are too short to
1350 have a flag are now skipped.
1352 SC/07 Eximstats V1.49 (Alain Williams)
1353 Added the -emptyok flag.
1355 SC/08 Eximstats V1.50
1356 Fixes for obtaining the IP address from reject messages.
1358 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1359 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1360 whitesspace changes from 4.64-PH/27
1362 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1365 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1366 are found, allow negative numbers in numeric criteria)
1368 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1370 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1371 to match changes made in 4.64-PH/43
1373 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1375 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1376 "domain" is actually an IP address, and force a failure. This locks out
1377 those revolvers/nameservers that support "A-for-A" lookups, in
1378 contravention of the specifications.
1380 PH/48 When a host name was looked up from an IP address, and the subsequent
1381 forward lookup of the name timed out, the host name was left in
1382 $sender_host_name, contrary to the specification.
1384 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1385 restricted to single-key lookups, Exim was not diagnosing an error if
1386 * or *@ was used with a query-style lookup.
1388 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1390 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1391 long ago, but noone interested enough thought of it. Let's just say that
1392 the "1.1" means that there are some new functions that weren't there at
1393 some point in the past.
1395 PH/51 Error processing for expansion failure of helo_data from an smtp
1396 transport during callout processing was broken.
1398 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1399 tested/used via the -bh/-bhc/-bs options.
1401 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1402 bug, fixed in subsequent PCRE releases).
1404 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1405 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1407 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1413 SC/01 Use a glob alias rather than an array ref in eximstats generated
1414 parser. This improves both readability and performance.
1416 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1417 Don't display local sender or destination tables in eximstats unless
1418 there is data to show.
1419 Added average volumes into the eximstats top table text output.
1421 SC/03 Collect data on the number of addresses (recipients) as well
1422 as the number of messages in eximstats.
1424 TF/01 Correct an error in the documentation for the redirect router. Exim
1425 does (usually) call initgroups() when daemonizing.
1427 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1428 with consistent privilege compared to when running as a daemon.
1430 TF/03 Note in the spec that $authenticated_id is not set for local
1431 submissions from trusted users.
1433 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1434 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1436 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1437 by adding some example configuration directives to the default
1438 configuration file. A little bit of work is required to uncomment the
1439 directives and define how usernames and passwords are checked, but
1440 there is now a framework to start from.
1442 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1443 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1444 without this. I don't know how relevant this is to other LDAP libraries.
1446 PH/02 Add the verb name to the "unknown ACL verb" error.
1448 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1450 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1452 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1453 directory not expanded when it should be if an expanded home directory
1454 was set for the address (which is overridden by the transport).
1456 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1459 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1460 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1461 because it is too late at that time, and has no effect.
1463 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1464 security issue with \' (bugzilla #107). I could not use the
1465 PQescapeStringConn() function, because it needs a PGconn value as one of
1468 PH/08 When testing addresses using -bt, indicate those final addresses that
1469 are duplicates that would not cause an additional delivery. At least one
1470 person was confused, thinking that -bt output corresponded to deliveries.
1471 (Suppressing duplicates isn't a good idea as you lose the information
1472 about possibly different redirections that led to the duplicates.)
1474 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1475 systems where poll() doesn't work, in particular OS X.
1477 PH/10 Added more information to debugging output for retry time not reached.
1479 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1480 operations in malware.c.
1482 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1485 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1486 syslog_duplication set false, log lines that would normally be written
1487 both the the main log and to the reject log were not written to syslog at
1490 PH/14 In the default configuration, change the use of "message" in ACL warn
1491 statements to "add_header".
1493 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1494 not followed by a command (e.g. "seen endif").
1496 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1497 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1500 PH/17 Added extra conditions to the default value of delay_warning_condition
1504 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1505 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1506 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1509 The Auto-Submitted: and various List- headers are standardised, whereas I
1510 don't think Precedence: ever was.
1512 PH/18 Refactored debugging code in route_finduser() to show more information,
1513 in particular, the error code if getpwnam() issues one.
1515 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1516 This is apparently needed in addition to the PH/07 change above to avoid
1517 any possible encoding problems.
1519 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1520 but not after initializing Perl.
1522 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1523 output them only if debugging. By default they are written stderr,
1524 apparently, which is not desirable.
1526 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1529 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1532 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1534 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1535 authenticated or an ident call has been made. Suppress the default
1536 values for $authenticated_id and $authenticated_sender (but permit -oMai
1537 and -oMas) when testing with -bh.
1539 PH/24 Re-jigged the order of the tests in the default configuration so that the
1540 tests for valid domains and recipients precede the DNS black list and CSA
1541 tests, on the grounds that those ones are more expensive.
1543 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1544 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1545 This bug exists in every version of Exim that I still have, right back to
1548 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1549 However, an attempt to turn on case-sensitivity in a regex key by
1550 including (?-i) didn't work because the subject string was already
1551 lowercased, and the effects were non-intuitive. It turns out that a
1552 one-line patch can be used to allow (?-i) to work as expected.
1558 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1559 other effects) broke the use of negated acl sub-conditions.
1561 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1564 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1565 "Deny" causes Exim to reject the incoming connection with a 554 error.
1566 Unfortunately, if there is a major crisis, such as a disk failure,
1567 tcp-wrappers gives "deny", whereas what one would like would be some
1568 kind of temporary error. A kludge has been added to help with this.
1569 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1570 554 error is used if errno is still zero or contains ENOENT (which occurs
1571 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1574 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1576 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1577 errors. Otherwise a message that provokes a temporary error (when other
1578 messages do not) can cause a whole host to time out.
1580 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1581 addresses were routed directly to files or pipes from a redirect router.
1582 File deliveries just didn't batch; pipe deliveries might have suffered
1585 PH/06 A failure to get a lock for a hints database would erroneously always say
1586 "Failed to get write lock", even when it was really a read lock.
1588 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1589 of 0600. This has been changed to use the value of the lockfile_mode
1590 option (which defaults to 0600).
1592 PH/08 Applied small patch from the Sieve maintainer.
1594 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1595 folder from quota calculations, a direct delivery into this folder messed
1596 up the contents of the maildirsize file. This was because the regex was
1597 used only to exclude .Trash (or whatever) when the size of the mailbox
1598 was calculated. There was no check that a delivery was happening into an
1599 excluded directory. This bug has been fixed by ignoring all quota
1600 processing for deliveries into excluded directories.
1602 PH/10 Added the maildirfolder_create_regex option to appendfile.
1608 PH/01 The code for finding all the local interface addresses on a FreeBSD
1609 system running IPv6 was broken. This may well have applied to all BSD
1610 systems, as well as to others that have similar system calls. The broken
1611 code found IPv4 interfaces correctly, but gave incorrect values for the
1612 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1613 that it would not match correctly against @[] and not recognize the IPv6
1616 PH/02 The ipliteral router was not recognizing addresses of the form user@
1617 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1619 PH/03 Added disable_ipv6.
1621 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1622 decoded form, because it is most often used to construct To: headers
1623 lines in autoreplies, and the decoded form may well be syntactically
1624 invalid. However, $reply_address has leading white space removed, and all
1625 newlines turned into spaces so that the autoreply transport does not
1628 PH/05 If group was specified without a user on a router, and no group or user
1629 was specified on a transport, the group from the router was ignored.
1631 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1632 for visible compile-time settings that can be used to change these
1633 numbers, for those that want even more. Backwards compatibility with old
1634 spool files has been maintained. However, going back to a previous Exim
1635 release will lost any variables that are in spool files.
1637 PH/07 Two small changes when running in the test harness: increase delay when
1638 passing a TCP/IP connection to a new process, in case the original
1639 process has to generate a bounce, and remove special handling of
1640 127.0.0.2 (sic), which is no longer necessary.
1642 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1643 be the same on different OS.
1645 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1648 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1649 whether --show-vars was specified or not
1651 JJ/02 exipick: Added support for new ACL variable spool format introduced
1654 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1655 syntactically invalid From: or Reply-to: line, and a filter used this to
1656 generate an autoreply, and therefore failed to obtain an address for the
1657 autoreply, Exim could try to deliver to a non-existent relative file
1658 name, causing unrelated and misleading errors. What now happens is that
1659 it logs this as a hard delivery error, but does not attempt to create a
1662 PH/11 The exinext utility has a -C option for testing purposes, but although
1663 the given file was scanned by exinext itself; it wasn't being passed on
1664 when Exim was called.
1666 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1667 an end-of-file indication when reading a command response.
1669 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1670 compiled. In many other places in Exim, IPv6 addresses are always
1671 recognized, so I have changed this. It also means that IPv4 domain
1672 literals of the form [IPV4:n.n.n.n] are now always recognized.
1674 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1675 used if the router is not running as root, for example, when verifying at
1676 ACL time, or when using -bh. The debugging output from this situation was
1677 non-existent - all you got was a failure to exec. I have made two
1680 (a) Failures to set uid/gid, the current directory, or a process leader
1681 in a subprocess such as that created by queryprogram now generate
1682 suitable debugging ouput when -d is set.
1684 (b) The queryprogram router detects when it is not running as root,
1685 outputs suitable debugging information if -d is set, and then runs
1686 the subprocess without attempting to change uid/gid.
1688 PH/15 Minor change to Makefile for building test_host (undocumented testing
1691 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1692 additional section of a DNS packet that returns MX or SRV records.
1693 Instead, it always explicitly searches for A/AAAA records. This avoids
1694 major problems that occur when a DNS server includes only records of one
1695 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1696 fixed another bug: if SRV records were looked up and the corresponding
1697 address records were *not* found in the additional section, the port
1698 values from the SRV records were lost.
1700 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1701 using the correct key (the original address) when searching the retry
1702 rules in order to find which one to use for generating the retry hint.
1704 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1705 adding the default one. Similarly, if it contains a Reply-To: header, the
1706 errors_reply_to option, if set, is not used.
1708 PH/19 When calculating a retry time, Exim used to measure the "time since
1709 failure" by looking at the "first failed" field in the retry record. Now
1710 it does not use this if it is later than than the arrival time of the
1711 message. Instead it uses the arrival time. This makes for better
1712 behaviour in cases where some deliveries succeed, thus re-setting the
1713 "first failed" field. An example is a quota failure for a huge message
1714 when small messages continue to be delivered. Without this change, the
1715 "time since failure" will always be short, possible causing more frequent
1716 delivery attempts for the huge message than are intended.
1717 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1719 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1720 $1, $2, $3) because the numerical variables can be reset during some
1721 expansion items (e.g. "match"), thereby losing the authentication data.
1723 PH/21 Make -bV show the size of off_t variables so that the test suite can
1724 decide whether to run tests for quotas > 2G.
1726 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1727 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1728 filecount value is greater than 2G or if a quota value is greater than 2G
1729 on a system where the size of off_t is not greater than 4, a panic error
1732 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1733 never match. The debug and -bh output now contains an explicit error
1734 message indicating a malformed IPv4 address or mask.
1736 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1737 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1738 PH/23 above applies.
1740 PH/25 Do not write to syslog when running in the test harness. The only
1741 occasion when this arises is a failure to open the main or panic logs
1742 (for which there is an explicit test).
1744 PH/26 Added the /no_tell option to "control=freeze".
1746 PH/27 If a host name lookup failed very early in a connection, for example, if
1747 the IP address matched host_lookup and the reverse lookup yielded a name
1748 that did not have a forward lookup, an error message of the form "no IP
1749 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1750 could be logged. Now it outputs the IP address instead of "NULL".
1752 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1753 allows the sender and the authenticated sender to be set when
1754 submitting a message from within Exim. Since child_open_exim() is
1755 documented for local_scan(), the new function should be too.
1757 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1758 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1759 results in an empty string is now treated as unset.
1761 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1763 PH/31 Added sender_verify_fail logging option.
1765 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1766 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1767 tidied the source and removed it altogether.
1769 PH/33 When a queue run was abandoned because the load average was too high, a
1770 log line was always written; now it is written only if the queue_run log
1771 selector is set. In addition, the log line for abandonment now contains
1772 information about the queue run such as the pid. This is always present
1773 in "start" and "stop" lines but was omitted from the "abandon" line.
1775 PH/34 Omit spaces between a header name and the colon in the error message that
1776 is given when verify = headers_syntax fails (if there are lots of them,
1777 the message gets confusing).
1779 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1780 names, as there are now some PTR records that contain slashes. This check
1781 is only to protect against broken name servers that fall over on strange
1782 characters, so the fact that it applies to all lookups doesn't matter.
1784 PH/36 Now that the new test suite is complete, we can remove some of the
1785 special code in Exim that was needed for the old test suite. For example,
1786 sorting DNS records because real resolvers return them in an arbitrary
1787 order. The new test suite's fake resolver always returns records in the
1790 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1791 bounces) except when queue_only is set, to avoid logging races between
1792 the different processes.
1794 PH/38 Panic-die if .include specifies a non-absolute path.
1796 PH/39 A tweak to the "H" retry rule from its user.
1798 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1799 a label. They prevented compilation on older perls.
1801 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1802 a warning to be raised on newish perls.
1804 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1805 on queue. Changes to match documented behaviour of showing count of
1806 messages matching specified criteria.
1808 PH/40 Changed the default ident timeout from 30s to 5s.
1810 PH/41 Added support for the use of login_cap features, on those BSD systems
1811 that have them, for controlling the resources used by pipe deliveries.
1813 PH/42 The content-scanning code uses fopen() to create files in which to put
1814 message data. Previously it was not paying any attention to the mode of
1815 the files. Exim runs with umask(0) because the rest of the code creates
1816 files with open(), and sets the required mode explicitly. Thus, these
1817 files were ending up world-writeable. This was not a big issue, because,
1818 being within the spool directory, they were not world-accessible. I have
1819 created a function called modefopen, which takes an additional mode
1820 argument. It sets umask(777), creates the file, chmods it to the required
1821 mode, then resets the umask. All the relevant calls to fopen() in the
1822 content scanning code have been changed to use this function.
1824 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1825 to 24 hours. This avoids potential overflow problems when processing G
1826 and H retry rules. I suspect nobody ever tinkers with this value.
1828 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1830 PH/45 When the plaintext authenticator is running as a client, the server's
1831 challenges are checked to ensure they are valid base64 strings. By
1832 default, the authentication attempt is cancelled if an invalid string is
1833 received. Setting client_ignore_invalid_base64 true ignores these errors.
1834 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1835 they are received. Thus, the responses can be made to depend on the
1836 challenges. If an invalid string is ignored, an empty string is placed in
1839 PH/46 Messages that are created by the autoreply transport now contains a
1840 References: header, in accordance with RFCs 2822 and 3834.
1842 PH/47 Added authenticated_sender_force to the smtp transport.
1844 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1846 PH/49 Installed latest patch from the Sieve maintainer.
1848 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1849 was also set, the appendfile transport was unnecessarily scanning a
1850 directory of message files (e.g. for maildir delivery) to find the count
1851 of files (along with the size), even though it did not need this
1852 information. It now does the scan only if it needs to find either the
1853 size of the count of files.
1855 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1857 PH/52 Two bugs concerned with error handling when the smtp transport is
1860 (i) Exim was not creating retry information for temporary errors given
1861 for individual recipients after the DATA command when the smtp transport
1862 was used in LMTP mode. This meant that they could be retried too
1863 frequently, and not timed out correctly.
1865 (ii) Exim was setting the flag that allows error details to be returned
1866 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1867 recipients that were returned after the DATA command.
1869 PH/53 This is related to PH/52, but is more general: for any failing address,
1870 when detailed error information was permitted to be returned to the
1871 sender, but the error was temporary, then after the final timeout, only
1872 "retry timeout exceeded" was returned. Now it returns the full error as
1873 well as "retry timeout exceeded".
1875 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1876 do this, and (what is worse) MTAs that accept it.
1878 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1879 will now be deprecated.
1881 PH/56 New os.c-cygwin from the Cygwin maintainer.
1883 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1884 formats (previously only available in exim formats via -bpr, -bpru,
1885 and -bpra. Now also available in native and exiqgrep formats)
1887 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1888 with very large, slow to parse queues
1890 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1892 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1894 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1895 responses to authentication challenges, though it was showing the
1896 challenges; (ii) I've removed the CR characters from the debug output for
1899 PH/58 Allow for the insertion of a newline as well as a space when a string
1900 is turned into more than one encoded-word during RFC 2047 encoding. The
1901 Sieve code now uses this.
1903 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1904 data_4xx, lost_connection, tls_required.
1906 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1907 message was being sent as an SMTP response.
1909 PH/61 Add -l and -k options to exicyclog.
1911 PH/62 When verifying, if an address was redirected to one new address, so that
1912 verification continued, and the new address failed or deferred after
1913 having set something in $address_data, the value of $address_data was not
1914 passed back to the ACL. This was different to the case when no
1915 redirection occurred. The value is now passed back in both cases.
1917 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1918 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1919 use login_cap.h, so on its own it isn't the distinguishing feature. The
1920 new name refers directly to the setclassresources() function.
1922 PH/65 Added configuration files for NetBSD3.
1924 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1926 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1927 is preferred over IPv4.
1929 PH/68 The bounce_return_message and bounce_return_body options were not being
1930 honoured for bounces generated during the reception of non-SMTP messages.
1931 In particular, this applied to messages rejected by the ACL. This bug has
1932 been fixed. However, if bounce_return_message is true and bounce_return_
1933 body is false, the headers that are returned for a non-SMTP message
1934 include only those that have been read before the error was detected.
1935 (In the case of an ACL rejection, they have all been read.)
1937 PH/69 The HTML version of the specification is now built in a directory called
1938 spec_html instead of spec.html, because the latter looks like a path with
1939 a MIME-type, and this confuses some software.
1941 PH/70 Catch two compiler warnings in sieve.c.
1943 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1944 function verify_get_ident() calls ip_connect() to connect a socket, but
1945 if the "connect()" function timed out, ip_connect() used to close the
1946 socket. However, verify_get_ident() also closes the socket later, and in
1947 between Exim writes to the log, which may get opened at this point. When
1948 the socket was closed in ip_connect(), the log could get the same file
1949 descriptor number as the socket. This naturally causes chaos. The fix is
1950 not to close the socket in ip_connect(); the socket should be closed by
1951 the function that creates it. There was only one place in the code where
1952 this was missing, in the iplookup router, which I don't think anybody now
1953 uses, but I've fixed it anyway.
1955 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1956 well as to direct DNS lookups. Otherwise the handling of names in host
1957 lists is inconsistent and therefore confusing.
1963 PH/01 Two changes to the default runtime configuration:
1965 (1) Move the checks for relay_from_hosts and authenticated clients from
1966 after to before the (commented out) DNS black list checks.
1968 (2) Add control=submission to the relay_from_hosts and authenticated
1969 clients checks, on the grounds that messages accepted by these
1970 statements are most likely to be submissions.
1972 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1974 (1) Generate an error if the third argument for the ${prvs expansion is
1977 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1980 (3) Reset the variables that are obtained from the first argument of
1981 ${prvscheck and used in the second argument before leaving the code,
1982 because their memory is reclaimed, so using them afterwards may do
1985 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1986 one (it's much easier than Tom thought :-).
1988 (5) Because of (4), we can now allow for the use of $prvscheck_result
1989 inside the third argument.
1991 PH/03 For some reason, the default setting of PATH when running a command from
1992 a pipe transport was just "/usr/bin". I have changed it to
1995 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1996 anything to be listed in the output from -bV.
1998 PH/05 When a filter generated an autoreply, the entire To: header line was
1999 quoted in the delivery log line, like this:
2001 => >A.N.Other <ano@some.domain> <original@ddress> ...
2003 This has been changed so that it extracts the operative address. There
2004 may be more than one such address. If so, they are comma-separated, like
2007 => >ano@some.domain,ona@other.domain <original@ddress> ...
2009 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2010 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2011 not being looked up in the rDNS to get a host name, Exim was showing the
2012 IP address twice in Received: lines, even though the IP addresses were
2013 identical. For example:
2015 Received: from [1.2.3.4] (helo=[1.2.3.4])
2017 However, if the real host name was known, it was omitting the HELO data
2018 if it matched the actual IP address. This has been tidied up so that it
2019 doesn't show the same IP address twice.
2021 PH/07 When both +timestamp and +memory debugging was on, the value given by
2022 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2023 called by the debug printing, thereby overwriting the timestamp buffer.
2024 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2026 PH/08 When the original message was included in an autoreply transport, it
2027 always said "this is a copy of the message, including all the headers",
2028 even if body_only or headers_only was set. It now gives an appropriate
2031 PH/09 Applied a patch from the Sieve maintainer which:
2033 o fixes some comments
2034 o adds the (disabled) notify extension core
2035 o adds some debug output for the result of if/elsif tests
2036 o points to the current vacation draft in the documentation
2037 and documents the missing references header update
2041 o fixes a bug in processing the envelope test (when testing
2042 multiple envelope elements, the last element determinted the
2045 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2046 Electronic Mail") by including:
2048 Auto-submitted: auto-generated
2050 in the messages that it generates (bounce messages and others, such as
2051 warnings). In the case of bounce messages for non-SMTP mesages, there was
2052 also a typo: it was using "Auto_submitted" (underscore instead of
2053 hyphen). Since every message generated by Exim is necessarily in response
2054 to another message, thes have all been changed to:
2056 Auto-Submitted: auto-replied
2058 in accordance with these statements in the RFC:
2060 The auto-replied keyword:
2062 - SHOULD be used on messages sent in direct response to another
2063 message by an automatic process,
2065 - MUST NOT be used on manually-generated messages,
2067 - MAY be used on Delivery Status Notifications (DSNs) and Message
2068 Disposition Notifications (MDNs),
2070 - MUST NOT be used on messages generated by automatic or periodic
2071 processes, except for messages which are automatic responses to
2074 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2075 to the default Received: header definition.
2077 PH/12 Added log selector acl_warn_skipped (default on).
2079 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2080 variables because (a) they are in the wrong storage pool and (b) even if
2081 they were copied, it wouldn't work properly because of the caching.
2083 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2084 checking when decoding. Apparently there are clients that generate
2085 overlong encoded strings. Why am I not surprised?
2087 PH/15 If the first argument of "${if match_address" was not empty, but did not
2088 contain an "@" character, Exim crashed. Now it writes a panic log message
2089 and treats the condition as false.
2091 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2093 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2094 extension "envelope-auth". The code is finished and in agreement with
2095 other implementations, but there is no documentation so far and in fact,
2096 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2097 not changing the active code.
2099 Print executed "if" and "elsif" statements when debugging is used. This
2100 helps a great deal to understand what a filter does.
2102 Document more things not specified clearly in RFC3028. I had all this
2103 sorted out, when out of a sudden new issues came to my mind. Oops."
2105 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2108 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2109 canonical form (as documented). However, after a host name lookup from
2110 the IP address, check_host() was doing a simple string comparison with
2111 addresses acquired from the DNS when checking that the found name did
2112 have the original IP as one of its addresses. Since any found IPv6
2113 addresses are likely to be in abbreviated form, the comparison could
2114 fail. Luckily, there already exists a function for doing the comparison
2115 by converting both addresses to binary, so now that is used instead of
2116 the text comparison.
2118 PH/20 There was another similar case to PH/19, when a complete host name was
2119 given in a host list; looking up its IP address could give an abbreviated
2120 form, whereas the current host's name might or might not be abbreviated.
2121 The same fix has been applied.
2127 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2128 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2131 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2132 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2134 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2136 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2137 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2138 Also, for "Auto-Submitted", treat anything other than "no" as
2139 non-personal, in accordance with RFC 3834. (Previously it treated
2140 anything starting "auto-" as non-personal.)
2142 TF/01 The control=submission/name=... option had a problem with syntax
2143 errors if the name included a slash character. The /name= option
2144 now slurps the rest of the string, so it can include any characters
2145 but it must come last in the list of options (after /sender_retain
2148 PH/05 Some modifications to the interface to the fake nameserver for the new
2156 TK/01 Added the "success_on_redirect" address verification option. See
2157 NewStuff for rationale and an example.
2159 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2161 PH/02 Patch to exigrep to allow it to work on syslog lines.
2163 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2164 fread() to skip over the body file's header line, because in Cygwin the
2165 header line is locked and is inaccessible.
2167 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2168 co-exist for some time) to make it clear that it is the Exim ID that is
2169 referenced, not the Message-ID: header line.
2171 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2172 string_format() function, because snprintf() does not exist on all
2175 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2176 ${stat: expansion item.
2178 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2179 protocol synchronization error", to keep the pedants happy.
2181 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2182 well as for IRIX systems, when gcc is being used. See the host.c source
2185 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2187 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2190 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2191 transport and to the smtp transport in LMTP mode.
2193 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2195 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2196 run a filter in a subprocess. This could lead to confusion in subsequent
2197 lookups in the parent process. There should also be a search_tidyup() at
2198 the end of the subprocess.
2200 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2201 only if the host matched helo_try_verify_hosts, which caused the
2202 verification to occur when the EHLO/HELO command was issued. The ACL just
2203 tested the remembered result. Now, if a previous verification attempt has
2204 not happened, "verify = helo" does it there and then.
2206 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2208 TK/03 Fix log output including CR from clamd.
2210 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2211 exist provoked a memory error which could cause a segfault.
2213 PH/15 Installed PCRE 6.2
2215 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2217 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2218 of the problem. Specifically, suggested +O2 rather than +O1 for the
2221 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2223 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2224 record, the port was not correctly logged when the outgoing_port log
2225 selector was set (it logged the transort's default port).
2227 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2228 fallback_hosts, and "hosts" in the smtp transport.
2230 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2231 host errors such as "Connection refused".
2233 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2234 authentication with radiusclient 0.4.9:
2236 - Error returned from rc_read_config was caught wrongly
2237 - Username/password not passed on to radius server due to wrong length.
2239 The presumption is that some radiusclient API changes for 4.51/PH/17
2240 were not taken care of correctly. The code is still untested by me (my
2241 Linux distribution still has 0.3.2 of radiusclient), but it was
2242 contributed by a Radius user.
2244 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2245 expanding the "port" option of the smtp transport.
2247 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2248 while reading a MIME header. Thanks to Tom Hughes for a patch.
2250 PH/24 Include config.h inside local_scan.h so that configuration settings are
2253 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2254 that in an ACL for RCPT (for example), you can examine exactly what was
2257 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2258 commands, but it was not correctly comparing the address with the actual
2259 client host address. Thus, it would show the EHLO address in Received:
2260 header lines when this was not necessary.
2262 PH/27 Added the % operator to ${eval:}.
2264 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2265 it should be ignoring failures (because with -C, for example, it has lost
2266 privilege). It wasn't ignoring creation failures other than "already
2269 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2272 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2273 because an input line was too long, either on its own, or by virtue of
2274 too many continuations, the temporary file was not being removed, and the
2275 return code was incorrect.
2277 PH/31 Missing "BOOL" in function definition in filtertest.c.
2279 PH/32 Applied Sieve patches from the maintainer.
2281 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2283 PH/33 Added "verify = not_blind".
2285 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2286 Local/Makefile (with some defaults set). These are used in built scripts
2287 such as exicyclog, but they have never been used in the exim_install
2288 script (though there are many overriding facilities there). I have
2289 arranged that the exim_install script now takes note of these two
2292 PH/35 Installed configuration files for Dragonfly.
2294 PH/36 When a locally submitted message by a trusted user did not contain a
2295 From: header, and the sender address was obtained from -f or from an SMTP
2296 MAIL command, and the trusted user did not use -F to supply a sender
2297 name, $originator_name was incorrectly used when constructing a From:
2298 header. Furthermore, $originator_name was used for submission mode
2299 messages from external hosts without From: headers in a similar way,
2300 which is clearly wrong.
2302 PH/37 Added control=suppress_local_fixups.
2304 PH/38 When log_selector = +received_sender was set, and the addition of the
2305 sender made the log line's construction buffer exactly full, or one byte
2306 less than full, an overflow happened when the terminating "\n" was
2309 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2310 when the result of a list match is failure because a DNS lookup failed.
2312 PH/40 RM_COMMAND is now used in the building process.
2314 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2315 the "build-* directories that it finds.
2317 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2318 address in a domain literal was a prefix of an interface address.
2320 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2321 when verifying a sender address, unless rewrite_headers is false.
2323 PH/44 Wrote a long comment about why errors_to addresses are verified as
2324 recipients, not senders.
2326 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2327 the ratelimit ACL was added.
2329 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2331 PH/46 Added extra information about PostgreSQL errors to the error string.
2333 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2334 avoiding the need to install special zones in a real server. This is
2335 backwards compatible; if it can't find the fake resolver, it drops back.
2336 Thus, both old and new test suites can be run.
2338 TF/02 Added util/ratelimit.pl
2340 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2341 clock is set back in time.
2343 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2344 Candler <B.Candler@pobox.com>.
2346 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2347 OK for addresses that are the result of redirections.
2349 PH/48 A number of further additions for the benefit of the new test suite,
2350 including a fake gethostbyname() that interfaces to the fake DNS resolver
2353 TF/06 The fix for widen_domains has also been applied to qualify_single and
2354 search_parents which are the other dnslookup options that can cause
2357 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2360 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2362 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2363 service" if it encountered an MX record with an empty target hostname.
2364 The message is now "an MX or SRV record indicated no SMTP service".
2366 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2367 if the DNS of the sending site is misconfigured. This is quite a
2368 common situation. This change restores the behaviour of treating a
2369 helo verification defer as a failure.
2371 PH/51 If self=fail was set on a router, the bounce message did not include the
2372 actual error message.
2378 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2380 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2381 command itself ended in error, the underlying message about the transport
2382 filter timeout was being overwritten with the pipe command error. Now the
2383 underlying error message should be appended to the second error message.
2385 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2387 PH/02 Reduce the amount of output that "make" produces by default. Full output
2388 can still be requested.
2390 PH/03 The warning log line about a condition test deferring for a "warn" verb
2391 was being output only once per connection, rather than after each
2392 occurrence (because it was using the same function as for successful
2393 "warn" verbs). This seems wrong, so I have changed it.
2395 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2396 it should not have, which might have caused a crash in the right
2397 circumstances, but probably never did.
2399 PH/04 Installed a modified version of Tony Finch's patch to make submission
2400 mode fix the return path as well as the Sender: header line, and to
2401 add a /name= option so that you can make the user's friendly name appear
2404 TF/03 Added the control = fakedefer ACL modifier.
2406 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2407 Mark Lowes for thorough testing.
2409 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2411 TK/03 Merged latest SRS patch from Miles Wilton.
2413 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2414 with the definition in sysexits.h (which is #included earlier).
2415 Fortunately, Exim does not actually use EX_OK. The code used to try to
2416 preserve the sysexits.h value, by assumimg that macro definitions were
2417 scanned for macro replacements. I have been disabused of this notion,
2418 so now the code just undefines EX_OK before #including unistd.h.
2420 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2421 in the smtp transport. When a block could not be written in a single
2422 write() function, the timeout was being re-applied to each part-write.
2423 This seems wrong - if the receiver was accepting one byte at a time it
2424 would take for ever. The timeout is now adjusted when this happens. It
2425 doesn't have to be particularly precise.
2427 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2428 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2430 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2431 without a domain if the check to <postmaster@domain> fails.
2433 SC/01 Eximstats: added -xls and the ability to specify output files
2434 (patch written by Frank Heydlauf).
2436 SC/02 Eximstats: use FileHandles for outputing results.
2438 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2440 SC/04 Eximstats: fixed display of large numbers with -nvr option
2442 SC/05 Eximstats: fixed merging of reports with empty tables.
2444 SC/06 Eximstats: added the -include_original_destination flag
2446 SC/07 Eximstats: removed tabs and trailing whitespace.
2448 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2450 TK/06 MBOX spool code: Add real "From " MBOX separator line
2451 so the .eml file is really in mbox format (even though
2452 most programs do not really care). Patch from Alex Miller.
2454 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2455 The latter is generated from $received_to and is only set if the
2456 message has one envelope recipient. SA can use these headers,
2457 obviously out-of-the-box. Patch from Alex Miller.
2459 PH/08 The ${def test on a variable was returning false if the variable's
2460 value was "0", contrary to what the specification has always said!
2461 The result should be true unless the variable is empty.
2463 PH/09 The syntax error of a character other than { following "${if
2464 def:variable_name" (after optional whitespace) was not being diagnosed.
2465 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2466 accidental colon was present, for example, could give incorrect results.
2468 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2469 result is used (not including appendfile, where other changes are about
2472 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2473 This involved changing a lot of size variables from int to off_t. It
2474 should work with maildirs and everything.
2476 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2477 spamd dying while we are connected to it.
2479 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2482 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2483 function for BDB 4.3.
2485 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2487 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2488 the sources for PCRE have been re-arranged and more files are now
2491 PH/15 The code I had for printing potentially long long variables in PH/11
2492 above was not the best (it lost precision). The length of off_t variables
2493 is now inspected at build time, and an appropriate printing format (%ld
2494 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2495 to be "long long int" or "long int". This is needed for the internal
2496 formatting function string_vformat().
2498 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2499 the configuration file to be ":syslog", then the script "guesses" where
2500 the logs files are, rather than using the compiled in default. In our
2501 case the guess is not the same as the compiled default, so the script
2502 suddenly stopped working when I started to use syslog. The patch checks
2503 to see if log_file_path is "". If so, it attempts to read it from exim
2504 with no configuration file to get the compiled in version, before it
2505 falls back to the previous guessing code."
2507 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2508 implementing BATV in an Exim configuration. See NewStuff for the gory
2511 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2512 Makefile that are specific to HP-UX.
2514 PH/18 If the "use_postmaster" option was set for a recipient callout together
2515 with the "random" option, the postmaster address was used as the MAIL
2516 FROM address for the random test, but not for the subsequent recipient
2517 test. It is now used for both.
2519 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2520 patch removes a few documentation additions to RFC 3028, because the
2521 latest draft now contains them. It adds the new en;ascii-case comparator
2522 and a new error check for 8bit text in MIME parts. Comparator and
2523 require names are now matched exactly. I enabled the subaddress
2524 extension, but it is not well tested yet (read: it works for me)."
2526 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2527 rework some of the code of TK/09 above to avoid the hardwired use of
2528 "%lld" and "long long". Replaced the call to snprintf() with a call to
2531 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2532 records point to non-existent hosts", "retry timeout exceeded", and
2533 "retry time not reached for any host after a long failure period".
2535 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2536 experimental DomainKeys support:
2538 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2539 (2) On an error such as an illegally used "control", the wrong name for
2540 the control was given.
2542 These problems did NOT occur unless DomainKeys support was compiled.
2544 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2546 PH/24 Added ${if match_ip condition.
2548 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2549 databases so that it will be absolutely obvious if a crash occurs in the
2550 DB library. This is a regular occurrence (often caused by mis-matched
2553 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2554 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2555 picked up on a user's system that detects such things. There doesn't seem
2556 to be a gcc warning option for this - only an attribute that has to be
2557 put on the function's prototype. It seems that in Fedora Core 4 they have
2558 set this on a number of new functions. No doubt there will be more in due
2561 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2562 specify a transport. However, if an address that was verified by such a
2563 router was the subject of a callout, Exim crashed because it tried to
2564 read the rcpt_include_affixes from the non-existent transport. Now it
2565 just assumes that the setting of that option is false. This bug was
2566 introduced by 4.51/PH/31.
2568 PH/28 Changed -d+all to exclude +memory, because that information is very
2569 rarely of interest, but it makes the output a lot bigger. People tend to
2570 do -d+all out of habit.
2572 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2573 code in os-type was giving problems when libc.so lives in lib64, like on
2576 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2577 aren't the modern standard, and it seems that some systems' include files
2578 don't always have them. Exim was already checking for some of the newer
2579 ones like T_AAAA, and defining it itself. I've added checks for all the
2580 record types that Exim uses.
2582 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2583 not automatically generating a new one, as it is supposed to. This
2584 prevented TLS from working. If the file did exist, but contained invalid
2585 data, a new version was generated, as expected. It was only the case of a
2586 non-existent file that was broken.
2588 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2589 with a change in libdomainkeys > 0.64.
2591 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2592 from DNS. If the selector record carries the flag, it now has
2593 precedence over the domain-wide flag.
2595 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2597 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2598 the use of a port name that isn't defined in /etc/services) occurred, the
2599 message was deferred as in a normal delivery, and thus remained on the
2600 spool, instead of being failed because of the mua_wrapper setting. This
2601 is now fixed, and I tidied up some of the mua_wrapper messages at the
2604 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2605 the messages in a hash of arrays rather than using individual hashes.
2606 This is a bit cleaner and results in dramatic memory savings, albeit
2607 at a slight CPU cost.
2609 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2610 as requested by Marc Sherman.
2612 SC/10 Eximstats: added histograms for user specified patterns as requested
2615 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2617 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2618 fopen() in the content-scanning modules that did not already have it.
2624 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2625 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2627 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2629 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2631 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2632 to test Sieve filters that use "vacation".
2634 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2635 that changes the way the GnuTLS parameters are stored in the cache file.
2636 The new format can be generated externally. For backward compatibility,
2637 if the data in the cache doesn't make sense, Exim assumes it has read an
2638 old-format file, and it generates new data and writes a new file. This
2639 means that you can't go back to an older release without removing the
2642 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2643 work if there are any delivery delays because "one_time" forces the
2644 parent to be marked "delivered", so its unseen clone is never tried
2645 again. For this reason, Exim now forbids the simultaneous setting of
2648 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2649 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2650 entry if you want to know the details. The fix, however, neglected to
2651 consider the case when local delivery batching is involved. The test for
2652 "previously delivered" was not happening when checking to see if an
2653 address could be batched with a previous (undelivered) one; under
2654 certain circumstances this could lead to multiple deliveries to the same
2657 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2658 in its include files, and this causes problems building Exim.
2660 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2661 header_syntax) but Exim was just ignoring anything given after a slash.
2662 In particular, this caused confusion with an attempt to use "verify =
2663 reverse_host_lookup/defer_ok". An error is now given when options are
2664 supplied for verify items that do not have them. (Maybe reverse_host_
2665 lookup should have a defer_ok option, but that's a different point.)
2667 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2668 defined by RFC 821) to 2048, because there were problems with some AUTH
2669 commands, and RFC 1869 says the size should be increased for extended
2670 SMTP commands that take arguments.
2672 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2675 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2676 "unknown" error; now it says that the functionality isn't in the binary.
2678 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2679 an address' error message when a string expansion fails (syntax or
2680 whatever). Otherwise the password may appear in the log. Following change
2681 PH/42 below, there is no longer a chance of it appearing in a bounce
2684 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2686 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2687 of its addresses was ever tried. (Bugzilla bug #2.)
2689 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2690 the result incorrectly in the debug output. (It correctly added a newline
2691 to what was transported.)
2693 TF/01 Added $received_time.
2695 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2696 commented out examples of how to interface to a virus scanner and to
2697 SpamAssassin. Also added commented examples of av_scanner and
2698 spamd_address settings.
2700 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2701 and controls are allowed in which ACLs. There were a couple of minor
2702 errors. Some of the entries in the conditions table (which is a table of
2703 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2704 negation of where the condition IS allowed.
2706 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2708 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2709 header file does not have a version number, so I've had to invent a new
2710 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2711 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2712 radiusclient), but it was contributed by a Radius user.
2714 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2715 files or directories, for OpenSSL.
2717 PH/19 When an Exim process that is running as root has to create an Exim log
2718 file, it does so in a subprocess that runs as exim:exim so as to get the
2719 ownership right at creation (otherwise, other Exim processes might see
2720 the file with the wrong ownership). There was no test for failure of this
2721 fork() call, which would lead to the process getting stuck as it waited
2722 for a non-existent subprocess. Forks do occasionally fail when resources
2723 run out. I reviewed all the other calls to fork(); they all seem to check
2726 PH/20 When checking for unexpected SMTP input at connect time (before writing
2727 the banner), Exim was not dealing correctly with a non-positive return
2728 from the read() function. If the client had disconnected by this time,
2729 the result was a log entry for a synchronization error with an empty
2730 string after "input=" when read() returned zero. If read() returned -1
2731 (an event I could not check), uninitialized data bytes were printed.
2732 There were reports of junk text (parts of files, etc) appearing after
2735 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2737 PH/22 Added support for macro redefinition, and (re)definition in between
2738 driver and ACL definitions.
2740 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2741 forgetting to use the resulting value; it was using the unexpanded value.
2743 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2744 hadn't been configured. The fix is from Juergen Kreileder, who
2745 understands it better than I do:
2747 "Here's what I see happening with three configured cyrus_sasl
2748 authenticators configured (plain, login, cram-md5):
2750 On startup auth_cyrus_sasl_init() gets called for each of these.
2751 This means three calls to sasl_listmech() without a specified mech_list.
2752 => SASL tests which mechs of all available mechs actually work
2753 => three warnings about OTP not working
2754 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2756 With the patch, sasl_listmech() also gets called three times. But now
2757 SASL's mech_list option is set to the server_mech specified in the the
2758 authenticator. Or in other words, the answer from sasl_listmech()
2759 gets limited to just the mech you're testing for (which is different
2761 => the return list contains just 'plain' or 'login', 'cram-md5' or
2762 nothing depending on the value of ob->server_mech.
2764 I've just tested the patch: Authentication still works fine,
2765 unavailable mechs specified in the exim configuration are still
2766 caught, and the auth.log warnings about OTP are gone."
2768 PH/25 When debugging is enabled, the contents of the command line are added
2769 to the debugging output, even when log_selector=+arguments is not
2772 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2773 answer is "GNU", and only if the return is "GNU/something" is the answer
2776 PH/27 $acl_verify_message is now set immediately after the failure of a
2777 verification in an ACL, and so is available in subsequent modifiers. In
2778 particular, the message can be preserved by coding like this:
2780 warn !verify = sender
2781 set acl_m0 = $acl_verify_message
2783 Previously, $acl_verify_message was set only while expanding "message"
2784 and "log_message" when a very denied access.
2786 PH/28 Modified OS/os.c-Linux with
2788 -#ifndef OS_LOAD_AVERAGE
2789 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2791 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2792 nomenclature these days.)
2794 PH/29 Installed patch from the Sieve maintainer that adds the options
2795 sieve_useraddress and sieve_subaddress to the redirect router.
2797 PH/30 In these circumstances:
2798 . Two addresses routed to the same list of hosts;
2799 . First host does not offer TLS;
2800 . First host accepts first address;
2801 . First host gives temporary error to second address;
2802 . Second host offers TLS and a TLS session is established;
2803 . Second host accepts second address.
2804 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2805 and peerdn, if requested) that were in fact used only for the second
2808 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2809 attention to any local part prefix or suffix that was matched by the
2810 router that accepted the address. It now behaves in the same way as it
2811 does for delivery: the affixes are removed from the local part unless
2812 rcpt_include_affixes is set on the transport.
2814 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2815 timeout during the DATA phase of an incoming message.
2817 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2818 applied a patch sanctioned by the Sieve maintainer.
2820 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2821 the uid or gid is negative. A case of a negative gid caused this to be
2822 noticed. The fix allows for either to be negative.
2824 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2825 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2828 PH/36 The change PH/12 above was broken. Fixed it.
2830 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2831 the grounds that routing the same address twice would always produce the
2832 same answer. This might have been true once, but it is certainly no
2833 longer true now. Routing a child address may depend on the previous
2834 routing that produced that child. Some complicated redirection strategies
2835 went wrong when messages had multiple recipients, and made Exim's
2836 behaviour dependent on the order in which the addresses were given.
2838 I have moved the duplicate checking until after the routing is complete.
2839 Exim scans the addresses that are assigned to local and remote
2840 transports, and removes any duplicates. This means that more work will be
2841 done, as duplicates will always all be routed, but duplicates are
2842 presumably rare, so I don't expect this is of any significance.
2844 For deliveries to pipes, files, and autoreplies, the duplicate checking
2845 still happens during the routing process, since they are not going to be
2848 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2849 It corrects a timeout issue with spamd. This is Ian's comment: "The
2850 background is that sometimes spamd either never reads data from a
2851 connection it has accepted, or it never writes response data. The exiscan
2852 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2853 blindly assumes that writes won't block so it may never time out."
2855 PH/39 Allow G after quota size as well as K and M.
2857 PH/40 The value set for $authenticated_id in an authenticator may not contain
2858 binary zeroes or newlines because the value is written to log lines and
2859 to spool files. There was no check on this. Now the value is run through
2860 the string_printing() function so that such characters are converted to
2861 printable escape sequences.
2863 PH/41 $message_linecount is a new variable that contains the total number of
2864 lines in the message. Compare $body_linecount, which is the count for the
2867 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2868 bounce and delay warning messages, except in certain special cases, which
2871 (a) An SMTP error message from a remote host;
2872 (b) A message specified in a :fail: redirection;
2873 (c) A message specified in a "fail" command in a system filter;
2874 (d) A message specified in a FAIL return from the queryprogram router;
2875 (e) A message specified by the cannot_route_message router option.
2877 In these cases only, Exim does include the error details in bounce and
2878 warning messages. There are also a few cases where bland messages such
2879 as "unrouteable address" or "local delivery error" are given.
2881 PH/43 $value is now also set for the "else" part of a ${run expansion.
2883 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2884 being worked on, but at least Exim now implements the latest version to
2887 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2888 process to complete was treated as a delivery failure, a timeout while
2889 writing the message to the pipe was logged, but erroneously treated as a
2890 successful delivery. Such timeouts include transport filter timeouts. For
2891 consistency with the overall process timeout, these timeouts are now
2892 treated as errors, giving rise to delivery failures by default. However,
2893 there is now a new Boolean option for the pipe transport called
2894 timeout_defer, which, if set TRUE, converts the failures into defers for
2895 both kinds of timeout. A transport filter timeout is now identified in
2898 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2899 systems where "make" and "gmake" are different, calling "gmake" at top
2900 level broke things. I've arranged for the value of $(MAKE) to be passed
2901 from the Makefile to this script so that it can call the same version of
2905 A note about Exim versions 4.44 and 4.50
2906 ----------------------------------------
2908 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2909 changes of various kinds. As a consequence, a big documentation update was
2910 needed. This delayed the release for rather longer than seemed good, especially
2911 in the light of a couple of (minor) security issues. Therefore, the changes
2912 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2913 release. So 4.44 and 4.50 are in effect two different branches that both start
2916 I have left the 4.50 change log unchanged; it contains all the changes since
2917 4.43. The change log for 4.44 is below; many of its items are identical to
2918 those for 4.50. This seems to be the most sensible way to preserve the
2919 historical information.
2925 1. Minor wording change to the doc/README.SIEVE file.
2927 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2928 computation of the current number of files was incorrect.
2930 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2931 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2934 4. Give more explanation in the error message when the command for a transport
2935 filter fails to execute.
2937 5. There are several places where Exim runs a non-Exim command in a
2938 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2939 was being done only for the command run by the queryprogram router. It is
2940 now done for all such subprocesses. The other cases are: ${run, transport
2941 filters, and the commands run by the lmtp and pipe transports.
2943 6. Added CONFIGURE_GROUP build-time option.
2945 7. Some older OS have a limit of 256 on the maximum number of file
2946 descriptors. Exim was using setrlimit() to set 1000 as a large value
2947 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2948 systems. I've change it so that if it can't get 1000, it tries for 256.
2950 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2951 was an oversight, and furthermore, ever since the addition of extra
2952 controls (e.g. 4.43/32), the checks on when to allow different forms of
2953 "control" were broken. There should now be diagnostics for all cases when a
2954 control that does not make sense is encountered.
2956 9. Added the /retain_sender option to "control=submission".
2958 10. $recipients is now available in the predata ACL (oversight).
2960 11. Tidy the search cache before the fork to do a delivery from a message
2961 received from the command line. Otherwise the child will trigger a lookup
2962 failure and thereby defer the delivery if it tries to use (for example) a
2963 cached ldap connection that the parent has called unbind on.
2965 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2966 of $address_data from the recipient verification was clobbered by the
2967 sender verification.
2969 13. The value of address_data from a sender verification is now available in
2970 $sender_address_data in subsequent conditions in the ACL statement.
2972 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2974 15. Added a new option "connect=<time>" to callout options, to set a different
2977 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2978 was its contents. (It was OK if the option was not defined at all.)
2980 17. A "Completed" log line is now written for messages that are removed from
2981 the spool by the -Mrm option.
2983 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2984 information about exactly what failed.
2986 19. Added -dd to debug only the daemon process.
2988 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2989 handles timeouts, both on the server side and network timeouts. Renamed the
2990 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2992 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2993 It is now set to "smtps".
2995 22. $host_address is now set to the target address during the checking of
2996 ignore_target_hosts.
2998 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2999 being passed; this would have caused $sender_host_name to have been used if
3000 matching the list had actually called for a host name (not very likely,
3001 since this list is usually IP addresses). A host name is now passed as
3004 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3005 code that specifies a non-restarting handler (typically sigaction() in
3006 modern systems) in an attempt to fix a rare and obscure crash bug.
3008 25. Narrowed the window for a race in the daemon that could cause it to ignore
3009 SIGCHLD signals. This is not a major problem, because they are used only to
3010 wake it up if nothing else does.
3012 26. A malformed maildirsize file could cause Exim to calculate negative values
3013 for the mailbox size or file count. Odd effects could occur as a result.
3014 The maildirsize information is now recalculated if the size or filecount
3017 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3018 support for a long time. Removed HAVE_SYS_VFS_H.
3020 28. Installed the latest version of exipick from John Jetmore.
3022 29. In an address list, if the pattern was not a regular expression, an empty
3023 subject address (from a bounce message) matched only if the pattern was an
3024 empty string. Non-empty patterns were not even tested. This was the wrong
3025 because it is perfectly reasonable to use an empty address as part of a
3026 database query. An empty address is now tested by patterns that are
3027 lookups. However, all the other forms of pattern expect the subject to
3028 contain a local part and a domain, and therefore, for them, an empty
3029 address still always fails if the pattern is not itself empty.
3031 30. Exim went into a mad DNS loop when attempting to do a callout where the
3032 host was specified on an smtp transport, and looking it up yielded more
3033 than one IP address.
3035 31. Re-factored the code for checking spool and log partition space into a
3036 function that finds that data and another that does the check. The former
3037 is then used to implement four new variables: $spool_space, $log_space,
3038 $spool_inodes, and $log_inodes.
3040 32. The RFC2047 encoding function was originally intended for short strings
3041 such as real names; it was not keeping to the 75-character limit for
3042 encoded words that the RFC imposes. It now respects the limit, and
3043 generates multiple encoded words if necessary. To be on the safe side, I
3044 have increased the buffer size for the ${rfc2047: expansion operator from
3047 33. It is now permitted to omit both strings after an "if" condition; if the
3048 condition is true, the result is "true". As before, when the second string
3049 is omitted, a false condition yields an empty string. This makes it less
3050 cumbersome to write custom ACL and router conditions.
3052 34. Failure to deliver a bounce message always caused it to be frozen, even if
3053 there was an errors_to setting on the router. The errors_to setting is now
3056 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3057 canonical form (fully expanded) before being placed in
3058 $sender_host_address.
3060 36. The table in the code that translates DNS record types into text (T_A to
3061 "A" for instance) was missing entries for NS and CNAME. It is just possible
3062 that this could have caused confusion if both these types were looked up
3063 for the same domain, because the text type is used as part of Exim's
3064 per-process caching. But the chance of anyone hitting this buglet seems
3067 37. The dnsdb lookup has been extended in a number of ways.
3069 (1) There is a new type, "zns", which walks up the domain tree until it
3070 finds some nameserver records. It should be used with care.
3072 (2) There is a new type, "mxh", which is like "mx" except that it returns
3073 just the host names, not the priorities.
3075 (3) It is now possible to give a list of domains (or IP addresses) to be
3076 looked up. The behaviour when one of the lookups defers can be
3077 controlled by a keyword.
3079 (4) It is now possible to specify the separator character for use when
3080 multiple records are returned.
3082 38. The dnslists ACL condition has been extended: it is now possible to supply
3083 a list of IP addresses and/or domains to be looked up in a particular DNS
3086 39. Added log_selector=+queue_time_overall.
3088 40. When running the queue in the test harness, wait just a tad after forking a
3089 delivery process, to get repeatability of debugging output.
3091 41. Include certificate and key file names in error message when GnuTLS fails
3092 to set them up, because the GnuTLS error message doesn't include the name
3093 of the failing file when there is a problem reading it.
3095 42. Allow both -bf and -bF in the same test run.
3097 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3099 44. The "Exiscan patch" is now merged into the mainline Exim source.
3101 45. Sometimes the final signoff response after QUIT could fail to get
3102 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3103 < 0 before doing a fflush(). This bug looks as though it goes back to the
3104 introduction of TLS in release 3.20, but "sometimes" must have been rare
3105 because the tests only now provoked it.
3107 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3108 (this can affect the format of dates).
3110 47. exim_tidydb, when checking for the continued existence of a message for
3111 which it has found a message-specific retry record, was not finding
3112 messages that were in split spool directories. Consequently, it was
3113 deleting retry records that should have stayed in existence.
3115 48. Steve fixed some bugs in eximstats.
3117 49. The SPA authentication driver was not abandoning authentication and moving
3118 on to the next authenticator when an expansion was forced to fail,
3119 contradicting the general specification for all authenticators. Instead it
3120 was generating a temporary error. It now behaves as specified.
3122 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3123 (the order specifies the preference for clients). The order is now AES256,
3124 AES128, 3DES, ARCFOUR128.
3126 51. Small patch to Sieve code - explicitly set From: when generating an
3129 52. Exim crashed if a remote delivery caused a very long error message to be
3130 recorded - for instance if somebody sent an entire SpamAssassin report back
3131 as a large number of 550 error lines. This bug was coincidentally fixed by
3132 increasing the size of one of Exim's internal buffers (big_buffer) that
3133 happened as part of the Exiscan merge. However, to be on the safe side, I
3134 have made the code more robust (and fixed the comments that describe what
3137 53. Now that there can be additional text after "Completed" in log lines (if
3138 the queue_time_overall log selector is set), a one-byte patch to exigrep
3139 was needed to allow it to recognize "Completed" as not the last thing in
3142 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3143 patch that reportedly fixes this has been added. I am not expert enough to
3144 create a test for it. This is what the patch creator wrote:
3146 "I found a little strange behaviour of ldap code when working with
3147 Windows 2003 AD Domain, where users was placed in more than one
3148 Organization Units. When I tried to give exim partial DN, the exit code
3149 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3150 But simultaneously result of request was absolutely normal ldap result,
3151 so I produce this patch..."
3153 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3154 REFERENCE, so I have modified the code to exclude the patch when that macro
3157 55. Some experimental protocols are using DNS PTR records for new purposes. The
3158 keys for these records are domain names, not reversed IP addresses. The
3159 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3160 leaves it alone. Component reversal etc. now happens only for IP addresses.
3163 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3165 57. Double the size of the debug message buffer (to 2048) so that more of very
3166 long debug lines gets shown.
3168 58. The exicyclog utility now does better if the number of log files to keep
3169 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3171 59. Two changes related to the smtp_active_hostname option:
3173 (1) $smtp_active_hostname is now available as a variable.
3174 (2) The default for smtp_banner uses $smtp_active_hostname instead
3175 of $primary_hostname.
3177 60. The host_aton() function is supposed to be passed a string that is known
3178 to be a valid IP address. However, in the case of IPv6 addresses, it was
3179 not checking this. This is a hostage to fortune. Exim now panics and dies
3180 if the condition is not met. A case was found where this could be provoked
3181 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3182 components; fortuitously, this particular loophole had already been fixed
3183 by change 4.50/55 above.
3185 If there are any other similar loopholes, the new check in host_aton()
3186 itself should stop them being exploited. The report I received stated that
3187 data on the command line could provoke the exploit when Exim was running as
3188 exim, but did not say which command line option was involved. All I could
3189 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3190 running as the user.
3193 61. There was a buffer overflow vulnerability in the SPA authentication code
3194 (which came originally from the Samba project). I have added a test to the
3195 spa_base64_to_bits() function which I hope fixes it.
3198 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3199 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3201 63. The daemon start-up calls getloadavg() while still root for those OS that
3202 need the first call to be done as root, but it missed one case: when
3203 deliver_queue_load_max is set with deliver_drop_privilege. This is
3204 necessary for the benefit of the queue runner, because there is no re-exec
3205 when deliver_drop_privilege is set.
3207 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3208 This has been fixed.
3210 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3211 was in use, was not putting the data itself into the right store pool;
3212 consequently, it could be overwritten for a subsequent message in the same
3213 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3216 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3218 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3219 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3220 the function were treating the return as a boolean value, which happened to
3221 work because 0=false and not-0=true, but is not correct code.
3223 68. The host_aton() function was not handling scoped IPv6 addresses (those
3224 with, for example, "%eth0" on the end) correctly.
3226 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3227 negated items (that is, ~something) in unsigned ints. Some compilers
3228 apparently mutter when there is no cast.
3230 70. If an address verification called from an ACL failed, and did not produce a
3231 user-specific message (i.e. there was only a "system" message), nothing was
3232 put in $acl_verify_message. In this situation, it now puts the system
3235 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3236 session; change 4.31/43 added the unwanted input to the log line - except
3237 that it did not do this in the start of session case. It now does.
3239 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3240 This is wrong and can cause the other end to generate a synchronization
3241 error if it is another Exim or anything else that does the synchronization
3242 check. A QUIT command is no longer sent after a timeout.
3244 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3245 during host lookups.
3247 74. The defer_ok option of callout verification was not working if it was used
3248 when verifying addresses in header lines, that is, for this case:
3250 verify = header_sender/callout=defer_ok
3252 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3253 those file descriptors could be used for SMTP connections. If anything
3254 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3255 could be sent to the SMTP client, causing chaos. The daemon now opens
3256 stdin, stdout, and stderr to /dev/null when it puts itself into the
3259 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3260 log by default. The user can override this with suitable Perl magic.
3262 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3263 the log message when discard triggers, was not working for the DATA ACL or
3264 for the non-SMTP ACL.
3266 78. Error message wording change in sieve.c.
3268 79. If smtp_accept_max_per_host was set, the number of connections could be
3269 restricted to fewer than expected, because the daemon was trying to set up
3270 a new connection before checking whether the processes handling previous
3271 connections had finished. The check for completed processes is now done
3272 earlier. On busy systems, this bug wouldn't be noticed because something
3273 else would have woken the daemon, and it would have reaped the completed
3276 80. If a message was submitted locally by a user whose login name contained one
3277 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3278 It caused a spool format error. I have fixed the spool reading code. A
3279 related problem was that the "from" clause in the Received: line became
3280 illegal because of the space(s). It is now covered by ${quote_local_part.
3282 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3285 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3287 83. Patch from Sieve maintainer to fix unterminated string problem in
3288 "vacation" handling.
3290 84. Some minor changes to the Linux configuration files to help with other
3291 OS variants using glibc.
3293 85. One more patch for Sieve to update vacation handling to latest spec.
3296 ----------------------------------------------------
3297 See the note above about the 4.44 and 4.50 releases.
3298 ----------------------------------------------------
3304 1. Change 4.43/35 introduced a bug that caused file counts to be
3305 incorrectly computed when quota_filecount was set in an appendfile
3308 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3309 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3312 3. Give more explanation in the error message when the command for a transport
3313 filter fails to execute.
3315 4. There are several places where Exim runs a non-Exim command in a
3316 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3317 was being done only for the command run by the queryprogram router. It is
3318 now done for all such subprocesses. The other cases are: ${run, transport
3319 filters, and the commands run by the lmtp and pipe transports.
3321 5. Some older OS have a limit of 256 on the maximum number of file
3322 descriptors. Exim was using setrlimit() to set 1000 as a large value
3323 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3324 systems. I've change it so that if it can't get 1000, it tries for 256.
3326 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3327 was an oversight, and furthermore, ever since the addition of extra
3328 controls (e.g. 4.43/32), the checks on when to allow different forms of
3329 "control" were broken. There should now be diagnostics for all cases when a
3330 control that does not make sense is encountered.
3332 7. $recipients is now available in the predata ACL (oversight).
3334 8. Tidy the search cache before the fork to do a delivery from a message
3335 received from the command line. Otherwise the child will trigger a lookup
3336 failure and thereby defer the delivery if it tries to use (for example) a
3337 cached ldap connection that the parent has called unbind on.
3339 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3340 of $address_data from the recipient verification was clobbered by the
3341 sender verification.
3343 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3344 was its contents. (It was OK if the option was not defined at all.)
3346 11. A "Completed" log line is now written for messages that are removed from
3347 the spool by the -Mrm option.
3349 12. $host_address is now set to the target address during the checking of
3350 ignore_target_hosts.
3352 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3353 being passed; this would have caused $sender_host_name to have been used if
3354 matching the list had actually called for a host name (not very likely,
3355 since this list is usually IP addresses). A host name is now passed as
3358 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3359 code that specifies a non-restarting handler (typically sigaction() in
3360 modern systems) in an attempt to fix a rare and obscure crash bug.
3362 15. Narrowed the window for a race in the daemon that could cause it to ignore
3363 SIGCHLD signals. This is not a major problem, because they are used only to
3364 wake it up if nothing else does.
3366 16. A malformed maildirsize file could cause Exim to calculate negative values
3367 for the mailbox size or file count. Odd effects could occur as a result.
3368 The maildirsize information is now recalculated if the size or filecount
3371 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3372 support for a long time. Removed HAVE_SYS_VFS_H.
3374 18. Updated exipick to current release from John Jetmore.
3376 19. Allow an empty sender to be matched against a lookup in an address list.
3377 Previously the only cases considered were a regular expression, or an
3380 20. Exim went into a mad DNS lookup loop when doing a callout where the
3381 host was specified on the transport, if the DNS lookup yielded more than
3384 21. The RFC2047 encoding function was originally intended for short strings
3385 such as real names; it was not keeping to the 75-character limit for
3386 encoded words that the RFC imposes. It now respects the limit, and
3387 generates multiple encoded words if necessary. To be on the safe side, I
3388 have increased the buffer size for the ${rfc2047: expansion operator from
3391 22. Failure to deliver a bounce message always caused it to be frozen, even if
3392 there was an errors_to setting on the router. The errors_to setting is now
3395 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3396 canonical form (fully expanded) before being placed in
3397 $sender_host_address.
3399 24. Updated eximstats to version 1.33
3401 25. Include certificate and key file names in error message when GnuTLS fails
3402 to set them up, because the GnuTLS error message doesn't include the name
3403 of the failing file when there is a problem reading it.
3405 26. Expand error message when OpenSSL has problems setting up cert/key files.
3408 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3409 (this can affect the format of dates).
3411 28. exim_tidydb, when checking for the continued existence of a message for
3412 which it has found a message-specific retry record, was not finding
3413 messages that were in split spool directories. Consequently, it was
3414 deleting retry records that should have stayed in existence.
3416 29. eximstats updated to version 1.35
3417 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3418 1.35 - bugfix such that pie charts by volume are generated correctly
3420 30. The SPA authentication driver was not abandoning authentication and moving
3421 on to the next authenticator when an expansion was forced to fail,
3422 contradicting the general specification for all authenticators. Instead it
3423 was generating a temporary error. It now behaves as specified.
3425 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3426 (the order specifies the preference for clients). The order is now AES256,
3427 AES128, 3DES, ARCFOUR128.
3429 31. Small patch to Sieve code - explicitly set From: when generating an
3432 32. Exim crashed if a remote delivery caused a very long error message to be
3433 recorded - for instance if somebody sent an entire SpamAssassin report back
3434 as a large number of 550 error lines. This bug was coincidentally fixed by
3435 increasing the size of one of Exim's internal buffers (big_buffer) that
3436 happened as part of the Exiscan merge. However, to be on the safe side, I
3437 have made the code more robust (and fixed the comments that describe what
3440 33. Some experimental protocols are using DNS PTR records for new purposes. The
3441 keys for these records are domain names, not reversed IP addresses. The
3442 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3443 leaves it alone. Component reversal etc. now happens only for IP addresses.
3446 34. The host_aton() function is supposed to be passed a string that is known
3447 to be a valid IP address. However, in the case of IPv6 addresses, it was
3448 not checking this. This is a hostage to fortune. Exim now panics and dies
3449 if the condition is not met. A case was found where this could be provoked
3450 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3451 components; fortuitously, this particular loophole had already been fixed
3452 by change 4.50/55 or 4.44/33 above.
3454 If there are any other similar loopholes, the new check in host_aton()
3455 itself should stop them being exploited. The report I received stated that
3456 data on the command line could provoke the exploit when Exim was running as
3457 exim, but did not say which command line option was involved. All I could
3458 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3459 running as the user.
3462 35. There was a buffer overflow vulnerability in the SPA authentication code
3463 (which came originally from the Samba project). I have added a test to the
3464 spa_base64_to_bits() function which I hope fixes it.
3467 36. The daemon start-up calls getloadavg() while still root for those OS that
3468 need the first call to be done as root, but it missed one case: when
3469 deliver_queue_load_max is set with deliver_drop_privilege. This is
3470 necessary for the benefit of the queue runner, because there is no re-exec
3471 when deliver_drop_privilege is set.
3473 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3474 was in use, was not putting the data itself into the right store pool;
3475 consequently, it could be overwritten for a subsequent message in the same
3476 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3479 38. Sometimes the final signoff response after QUIT could fail to get
3480 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3481 < 0 before doing a fflush(). This bug looks as though it goes back to the
3482 introduction of TLS in release 3.20, but "sometimes" must have been rare
3483 because the tests only now provoked it.
3489 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3490 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3491 Now it returns an integer. A number of calls were still expecting a T/F
3492 return. Fortuitously, in all cases, the tests worked in OK situations,
3493 which is the norm. However, things would have gone wrong on any write
3494 failures on the smtp file descriptor. This function is used when sending
3495 messages over SMTP and also when doing verify callouts.
3497 2. When Exim is called to do synchronous delivery of a locally submitted
3498 message (the -odf or -odi options), it no longer closes stderr before doing
3501 3. Implemented the mua_wrapper option.
3503 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3505 5. Implemented the functions header_remove(), header_testname(),
3506 header_add_at_position(), and receive_remove_recipient(), and exported them
3509 6. If an ACL "warn" statement specified the addition of headers, Exim already
3510 inserted X-ACL-Warn: at the start if there was no header name. However, it
3511 was not making this test for the second and subsequent header lines if
3512 there were newlines in the string. This meant that an invalid header could
3513 be inserted if Exim was badly configured.
3515 7. Allow an ACL "warn" statement to add header lines at the start or after all
3516 the Received: headers, as well as at the end.
3518 8. Added the rcpt_4xx retry error code.
3520 9. Added postmaster_mailfrom=xxx to callout verification option.
3522 10. Added mailfrom=xxxx to the callout verification option, for verify=
3525 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3526 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3528 12. Inserted some casts to stop certain compilers warning when using pointer
3529 differences as field lengths or precisions in printf-type calls (mostly
3530 affecting debugging statements).
3532 13. Added optional readline() support for -be (dynamically loaded).
3534 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3535 same clock tick as a message's arrival, so that its received time was the
3536 same as the "first fail" time on the retry record, and that message
3537 remained on the queue past the ultimate address timeout, every queue runner
3538 would try a delivery (because it was past the ultimate address timeout) but
3539 after another failure, the ultimate address timeout, which should have then
3540 bounced the address, did not kick in. This was a "< instead of <=" error;
3541 in most cases the first failure would have been in the next clock tick
3542 after the received time, and all would be well.
3544 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3545 being recognized when the domain list was tested by the match_domain
3546 condition in an expansion string.
3548 16. Added the ${str2b64: operator.
3550 17. Exim was always calling setrlimit() to set a large limit for the number of
3551 processes, without checking whether the existing limit was already
3552 adequate. (It did check for the limit on file descriptors.) Furthermore,
3553 errors from getrlimit() and setrlimit() were being ignored. Now they are
3554 logged to the main and panic logs, but Exim does carry on, to try to do its
3555 job under whatever limits there are.
3557 18. Imported PCRE 5.0.
3559 19. Trivial typo in log message " temporarily refused connection" (the leading
3562 20. If the log selector return_path_on_delivery was set and an address was
3563 redirected to /dev/null, the delivery process crashed because it assumed
3564 that a return path would always be set for a "successful" delivery. In this
3565 case, the whole delivery is bypassed as an optimization, and therefore no
3568 21. Internal re-arrangement: the function for sending a challenge and reading
3569 a response while authentication was assuming a zero-terminated challenge
3570 string. It's now changed to take a pointer and a length, to allow for
3571 binary data in such strings.
3573 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3575 23. Exim was not respecting finduser_retries when seeking the login of the
3576 uid under which it was called; it was always trying 10 times. (The default
3577 setting of finduser_retries is zero.) Also, it was sleeping after the final
3578 failure, which is pointless.
3580 24. Implemented tls_on_connect_ports.
3582 25. Implemented acl_smtp_predata.
3584 26. If the domain in control=submission is set empty, Exim assumes that the
3585 authenticated id is a complete email address when it generates From: or
3586 Sender: header lines.
3588 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3589 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3590 chown and chgrp in /bin and hostname in /usr/bin.
3592 28. Exim was keeping the "process log" file open after each use, just as it
3593 does for the main log. This opens the possibility of it remaining open for
3594 long periods when the USR1 signal hits a daemon. Occasional processlog
3595 errors were reported, that could have been caused by this. Anyway, it seems
3596 much more sensible not to leave this file open at all, so that is what now
3599 29. The long-running daemon process does not normally write to the log once it
3600 has entered its main loop, and it closes the log before doing so. This is
3601 so that log files can straightforwardly be renamed and moved. However,
3602 there are a couple of unusual error situations where the daemon does write
3603 log entries, and I had neglected to close the log afterwards.
3605 30. The text of an SMTP error response that was received during a remote
3606 delivery was being truncated at 512 bytes. This is too short for some of
3607 the long messages that one sometimes sees. I've increased the limit to
3610 31. It is now possible to make retry rules that apply only when a message has a
3611 specific sender, in particular, an empty sender.
3613 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3614 it possible to be selective about when SMTP synchronization is enforced.
3616 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3618 32. Implemented hosts_connection_nolog.
3620 33. Added an ACL for QUIT.
3622 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3625 35. Added mailbox_size and mailbox_filecount to appendfile.
3627 36. Added control = no_multiline_responses to ACLs.
3629 37. There was a bug in the logic of the code that waits for the clock to tick
3630 in the case where the clock went backwards by a substantial amount such
3631 that the microsecond fraction of "now" was more than the microsecond
3632 fraction of "then" (but the whole seconds number was less).
3634 38. Added support for the libradius Radius client library this is found on
3635 FreeBSD (previously only the radiusclient library was supported).
3641 1. When certain lookups returned multiple values in the form name=value, the
3642 quoting of the values was not always being done properly. Specifically:
3643 (a) If the value started with a double quote, but contained no whitespace,
3645 (b) If the value contained whitespace other than a space character (i.e.
3646 tabs or newlines or carriage returns) it was not quoted.
3647 This fix has been applied to the mysql and pgsql lookups by writing a
3648 separate quoting function and calling it from the lookup code. The fix
3649 should probably also be applied to nisplus, ibase and oracle lookups, but
3650 since I cannot test any of those, I have not disturbed their existing code.
3652 2. A hit in the callout cache for a specific address caused a log line with no
3653 reason for rejecting RCPT. Now it says "Previous (cached) callout
3654 verification failure".
3656 3. There was an off-by-one bug in the queryprogram router. An over-long
3657 return line was truncated at 256 instead of 255 characters, thereby
3658 overflowing its buffer with the terminating zero. As well as fixing this, I
3659 have increased the buffer size to 1024 (and made a note to document this).
3661 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3662 when Exim is waiting for an SMTP response from a remote server, Exim
3663 restarts its select() call on the socket, thereby resetting its timeout.
3664 This is not a problem when such interrupts are rare. Somebody set up a cron
3665 job to run exiwhat every 2 minutes, which is less than the normal select()
3666 timeout (5 or 10 minutes). This meant that the select() timeout never
3667 kicked in because it was always reset. I have fixed this by comparing the
3668 time when an interrupt arrives with the time at the start of the first call
3669 to select(). If more time than the timeout has elapsed, the interrupt is
3670 treated as a timeout.
3672 5. Some internal re-factoring in preparation for the addition of Sieve
3673 extensions (by MH). In particular, the "personal" test is moved to a
3674 separate function, and given an option for scanning Cc: and Bcc: (which is
3675 not set for Exim filters).
3677 6. When Exim created an email address using the login of the caller as the
3678 local part (e.g. when creating a From: or Sender: header line), it was not
3679 quoting the local part when it contained special characters such as @.
3681 7. Installed new OpenBSD configuration files.
3683 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3684 try to make them clearer.
3686 9. Callout options, other than the timeout value, were being ignored when
3687 verifying sender addresses in header lines. For example, when using
3689 verify = header_sender/callout=no_cache
3691 the cache was (incorrectly) being used.
3693 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3694 only the Cygwin environment.
3696 11. When return_path_on_delivery was set as a log selector, if different remote
3697 addresses in the same message used different return paths and parallel
3698 remote delivery occurred, the wrong values would sometimes be logged.
3699 (Whenever a remote delivery process finished, the return path value from
3700 the most recently started remote delivery process was logged.)
3702 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3703 lines when AUTH and/or TLS are in use. This is the "received protocol"
3704 field. Exim used to use "asmtp" for authenticated SMTP, without any
3705 indication (in the protocol name) for TLS use. Now it follows the RFC and
3706 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3707 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3708 names appear in log lines as well as in Received: header lines.
3710 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3711 extensions, and comparison tests, and to fix some bugs.
3713 14. Changes to the "personal" filter test:
3715 (1) The test was buggy in that it was just doing the equivalent of
3716 "contains" tests on header lines. For example, if a user's address was
3717 anne@some.where, the "personal" test would incorrectly be true for
3719 To: susanne@some.where
3721 This test is now done by extracting each address from the header in turn,
3722 and checking the entire address. Other tests that are part of "personal"
3723 are now done using regular expressions (for example, to check local parts
3724 of addresses in From: header lines).
3726 (2) The list of non-personal local parts in From: addresses has been
3727 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3728 taken from the Sieve specification recommendations.
3730 (3) If the message contains any header line starting with "List-" it is
3731 treated as non-personal.
3733 (4) The test for "circular" in the Subject: header line has been removed
3734 because it now seems ill-conceived.
3736 15. Minor typos in src/EDITME comments corrected.
3738 16. Installed latest exipick from John Jetmore.
3740 17. If headers_add on a router specified a text string that was too long for
3741 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3742 of string_sprintf() is now avoided.
3744 18. $message_body_size was not set (it was always zero) when running the DATA
3745 ACL and the local_scan() function.
3747 19. For the "mail" command in an Exim filter, no default was being set for
3748 the once_repeat time, causing a random time value to be used if "once" was
3749 specified. (If the value happened to be <= 0, no repeat happened.) The
3750 default is now 0s, meaning "never repeat". The "vacation" command was OK
3751 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3752 (I found it when inspecting the code).
3754 20. There is now an overall timeout for performing a callout verification. It
3755 defaults to 4 times the callout timeout, which applies to individual SMTP
3756 commands during the callout. The overall timeout applies when there is more
3757 than one host that can be tried. The timeout is checked before trying the
3758 next host. This prevents very long delays if there are a large number of
3759 hosts and all are timing out (e.g. when the network connections are timing
3760 out). The value of the overall timeout can be changed by specifying an
3761 additional sub-option for "callout", called "maxwait". For example:
3763 verify = sender/callout=5s,maxwait=20s
3765 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3766 to the end before writing, but this should make it even safer).
3768 22. Exim was forgetting that it had advertised PIPELINING for the second and
3769 subsequent messages on an SMTP connection. It was also not resetting its
3770 memory on STARTTLS and an internal HELO.
3772 23. When Exim logs an SMTP synchronization error within a session, it now
3773 records whether PIPELINING has been advertised or not.
3775 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3776 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3777 rather than long int.
3779 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3781 26. Added the never_mail option to autoreply.
3787 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3788 crash if the getsockname() call failed; this can happen if a connection is
3789 closed very soon after it is established. The problem was simply in the
3790 order in which certain operations were done, causing Exim to try to write
3791 to the SMTP stream before it had set up the file descriptor. The bug has
3792 been fixed by making things happen in the correct order.
3798 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3799 before the connection was closed, thus losing the rejection response.
3801 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3802 some early Solaris releases, but causes trouble in current releases where
3803 socklen_t is defined.
3805 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3808 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3811 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3812 that releasing the top of it at the end releases what was used for sub-
3813 expansions (unless the block got too big). However, discard this block if
3814 the first thing is a variable or header, so that we can use its block when
3815 it is dynamic (useful for very large $message_headers, for example).
3817 6. Lookups now cache *every* query, not just the most recent. A new, separate
3818 store pool is used for this. It can be recovered when all lookup caches are
3819 flushed. Lookups now release memory at the end of their result strings.
3820 This has involved some general refactoring of the lookup sources.
3822 7. Some code has been added to the store_xxx() functions to reduce the amount
3823 of flapping under certain conditions.
3825 8. log_incoming_interface used to affect only the <= reception log lines. Now
3826 it causes the local interface and port to be added to several more SMTP log
3827 lines, for example "SMTP connection from", and rejection lines.
3829 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3831 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3833 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3834 could be overwritten at the end of the current message (or the start of a
3835 new message if it was set in a HELO ACL). The value is now preserved for
3836 the duration of the SMTP connection.
3838 12. If a transport had a headers_rewrite setting, and a matching header line
3839 contained an unqualified address, that address was qualified, even if it
3840 did not match any rewriting rules. The underlying bug was that the values
3841 of the flags that permit the existence of unqualified sender and recipient
3842 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3843 non-local messages, and by -bnq for local messages) were not being
3844 preserved with the message after it was received.
3846 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3847 "next input=" as part of the text comprising the host identity instead of
3848 the correct text. The code was using the same buffer for two different
3849 strings. However, depending on which order the printing function evaluated
3850 its arguments, the bug did not always show up. Under Linux, for example, my
3851 test suite worked just fine.
3853 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3854 doesn't work with some older versions of Perl. It has been changed to "my",
3855 which in any case is probably the better facility to use.
3857 15. A really picky compiler found some instances of statements for creating
3858 error messages that either had too many or two few arguments for the format
3861 16. The size of the buffer for calls to the DNS resolver has been increased
3862 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3863 for addresses that have a lot of PTR records. This alleviates a problem; it
3864 does not fully solve it.
3866 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3867 buffer now truncates the list and logs the incident, which is the same
3868 action as happens when Exim is looking up a host name and its aliases.
3869 Previously in this situation something unpredictable would happen;
3870 sometimes it was "internal error: store_reset failed".
3872 18. If a server dropped the connection unexpectedly when an Exim client was
3873 using GnuTLS and trying to read a response, the client delivery process
3874 crashed while trying to generate an error log message.
3876 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3877 string, for example:
3879 warn message = H1: something\nH2: something
3881 the text was added as a single header line from Exim's point of view
3882 though it ended up OK in the delivered message. However, searching for the
3883 second and subsequent header lines using $h_h2: did not work. This has been
3884 fixed. Similarly, if a system filter added multiple headers in this way,
3885 the routers could not see them.
3887 20. Expanded the error message when iplsearch is called with an invalid key to
3888 suggest using net-iplsearch in a host list.
3890 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3891 ACLs are no longer actually imposed (and a message to that effect is
3894 22. If a "gecos" field in a passwd entry contained escaped characters, in
3895 particular, if it contained a \" sequence, Exim got it wrong when building
3896 a From: or a Sender: header from that name. A second bug also caused
3897 incorrect handling when an unquoted " was present following a character
3898 that needed quoting.
3900 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3901 was not being matched caselessly.
3903 24. Arranged for all hyphens in the exim.8 source to be escaped with
3906 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3907 properly. Recipient callout cache records were still being keyed to include
3908 the sender, even when use_sender was set false. This led to far more
3909 callouts that were necessary. The sender is no longer included in the key
3910 when use_sender is false.
3912 26. Added "control = submission" modifier to ACLs.
3914 27. Added the ${base62d: operator to decode base 62 numbers.
3916 28. dnsdb lookups can now access SRV records.
3918 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3919 the configuration file.
3921 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3922 (-v) mode. This makes the output for a verbose queue run more intelligible.
3924 31. Added a use_postmaster feature to recipient callouts.
3926 32. Added the $body_zerocount variable, containing the number of binary zero
3927 bytes in the message body.
3929 33. The time of last modification of the "new" subdirectory is now used as the
3930 "mailbox time last read" when there is a quota error for a maildir
3933 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3935 35. Added +ignore_unknown as a special item in host lists.
3937 36. Code for decoding IPv6 addresses in host lists is now included, even if
3938 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3939 address was recognized as an IP address, but was then not correctly decoded
3940 into binary, causing unexpected and incorrect effects when compared with
3947 1. Very minor rewording of debugging text in manualroute to say "list of
3948 hosts" instead of "hostlist".
3950 2. If verify=header_syntax was set, and a header line with an unqualified
3951 address (no domain) and a large number of spaces between the end of the
3952 name and the colon was received, the reception process suffered a buffer
3953 overflow, and (when I tested it) crashed. This was caused by some obsolete
3954 code that should have been removed. The fix is to remove it!
3956 3. When running in the test harness, delay a bit after writing a bounce
3957 message to get a bit more predictability in the log output.
3959 4. Added a call to search_tidyup() just before forking a reception process. In
3960 theory, someone could use a lookup in the expansion of smtp_accept_max_
3961 per_host which, without the tidyup, could leave open a database connection.
3963 5. Added the variables $recipient_data and $sender_data which get set from a
3964 lookup success in an ACL "recipients" or "senders" condition, or a router
3965 "senders" option, similar to $domain_data and $local_part_data.
3967 6. Moved the writing of debug_print from before to after the "senders" test
3970 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3971 problems for message scanning, either using a data ACL, or using
3972 local_scan() because the Received: header was not generated till after they
3973 were called (in order to set the time as the time of reception completion).
3974 I have revised the way this works. The header is now generated after the
3975 body is received, but before the ACL or local_scan() are called. After they
3976 are run, the timestamp in the header is updated.
3982 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3983 before starting a queue runner without re-exec. This happened only when
3984 deliver_drop_privilege was set or when the Exim user was set to root. The
3985 effect of the bug was that timeouts during subsequent deliveries caused
3986 crashes instead of being properly handled. The handler is now left at its
3987 default (and expected) setting.
3989 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3990 message, again when deliver_drop_privilege is set or Exim is run as root.
3991 The bug described in (1) was not present in this case, but the tidying up
3992 of the other signals was missing. I have made the two cases consistent.
3994 3. The ignore_target_hosts setting on a manualroute router was being ignored
3995 for hosts that were looked up using the /MX notation.
3997 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4000 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4001 operated on the sender address. After changing the $sender_address to <>
4002 for the sender address verify, Exim was re-instated it as the original
4003 (before rewriting) address, but remembering that it had rewritten it, so it
4004 wasn't rewriting it again. This bug also had the effect of breaking the
4005 sender address verification caching when the sender address was rewritten.
4007 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4008 This has been changed so that if the ip literal address matches
4009 ignore_target_hosts, the router declines.
4011 7. Added expansion conditions match_domain, match_address, and match_local_
4012 part (NOT match_host).
4014 8. The placeholder for the Received: header didn't have a length field set.
4016 9. Added code to Exim itself and to exim_lock to test for a specific race
4017 condition that could lead to file corruption when using MBX delivery. The
4018 issue is with the lockfile that is created in /tmp. If this file is removed
4019 after a process has opened it but before that process has acquired a lock,
4020 there is the potential for a second process to recreate the file and also
4021 acquire a lock. This could lead to two Exim processes writing to the file
4022 at the same time. The added code performs the same test as UW imapd; it
4023 checks after acquiring the lock that its file descriptor still refers to
4024 the same named file.
4026 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4027 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4030 11. Added the smtp_active_hostname option. If used, this will typically be made
4031 to depend on the incoming interface address. Because $interface_address is
4032 not set up until the daemon has forked a reception process, error responses
4033 that can happen earlier (such as "too many connections") no longer contain
4036 12. If an expansion in a condition on a "warn" statement fails because a lookup
4037 defers, the "warn" statement is abandoned, and the next ACL statement is
4038 processed. Previously this caused the whole ACL to be aborted.
4040 13. Added the iplsearch lookup type.
4042 14. Added ident_timeout as a log selector.
4044 15. Added tls_certificate_verified as a log selector.
4046 16. Added a global option tls_require_ciphers (compare the smtp transport
4047 option of the same name). This controls incoming TLS connections.
4049 17. I finally figured out how to make tls_require_ciphers do a similar thing
4050 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4051 before starting the TLS session.
4053 18. Tabs are now shown as \t in -bP output.
4055 19. If the log selector return_path_on_delivery was set, Exim crashed when
4056 bouncing a message because it had too many Received: header lines.
4058 20. If two routers both had headers_remove settings, and the first one included
4059 a superfluous trailing colon, the final name in the first list and the
4060 first name in the second list were incorrectly joined into one item (with a
4061 colon in the middle).
4067 1. Added -C and -D options to the exinext utility, mainly to make it easier
4068 to include in the automated testing, but these could be helpful when
4069 multiple configurations are in use.
4071 2. The exinext utility was not formatting the output nicely when there was
4072 an alternate port involved in the retry record key, nor when there was a
4073 message id as well (for retries that were specific to a specific message
4074 and a specific host). It was also confused by IPv6 addresses, because of
4075 the additional colons they contain. I have fixed the IPv4 problem, and
4076 patched it up to do a reasonable job for IPv6.
4078 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4079 delivery, the log line now contains "pipelined" if PIPELINING was used.
4081 4. An SMTP transport process used to panic and die if the bind() call to set
4082 an explicit outgoing interface failed. This has been changed; it is now
4083 treated in the same way as a connect() failure.
4085 5. A reference to $sender_host_name in the part of a conditional expansion
4086 that was being skipped was still causing a DNS lookup. This no longer
4089 6. The def: expansion condition was not recognizing references to header lines
4090 that used bh_ and bheader_.
4092 7. Added the _cache feature to named lists.
4094 8. The code for checking quota_filecount in the appendfile transport was
4095 allowing one more file than it should have been.
4097 9. For compatibility with Sendmail, the command line option
4105 and sets the incoming protocol and host name (for trusted callers). The
4106 host name and its colon can be omitted when only the protocol is to be set.
4107 Note the Exim already has two private options, -pd and -ps, that refer to
4108 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4109 "s", but I don't think that's a major issue.
4111 10. A number of refactoring changes to the code, none of which should affect
4114 (a) The number of logging options was getting close to filling up the
4115 32-bit word that was used as a bit map. I have split them into two classes:
4116 those that are passed in the argument to log_write(), and those that are
4117 only ever tested independently outside of that function. These are now in
4118 separate 32-bit words, so there is plenty of room for expansion again.
4119 There is no change in the user interface or the logging behaviour.
4121 (b) When building, for example, log lines, the code previously used a
4122 macro that called string_cat() twice, in order to add two strings. This is
4123 not really sufficiently general. Furthermore, there was one instance where
4124 it was actually wrong because one of the argument was used twice, and in
4125 one call a function was used. (As it happened, calling the function twice
4126 did not affect the overall behaviour.) The macro has been replaced by a
4127 function that can join an arbitrary number of extra strings onto a growing
4130 (c) The code for expansion conditions now uses a table and a binary chop
4131 instead of a serial search (which was left over from when there were very
4132 few conditions). Also, it now recognizes conditions like "pam" even when
4133 the relevant support is not compiled in: a suitably worded error message is
4134 given if an attempt is made to use such a condition.
4136 11. Added ${time_interval:xxxxx}.
4138 12. A bug was causing one of the ddress fields not to be passed back correctly
4139 from remote delivery subprocesses. The field in question was not being
4140 subsequently used, so this caused to problems in practice.
4142 13. Added new log selectors queue_time and deliver_time.
4144 14. Might have fixed a bug in maildirsizefile handling that threw up
4145 "unexpected character" debug warnings, and recalculated the data
4146 unnecessarily. In any case, I expanded the warning message to give more
4149 15. Added the message "Restricted characters in address" to the statements in
4150 the default ACL that block characters like @ and % in local parts.
4152 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4153 Three changes have been made:
4155 (a) There was a serious bug; a negative response to MAIL caused the whole
4156 recipient domain to be cached as invalid, thereby blocking all messages
4157 to all local parts at the same domain, from all senders. This bug has
4158 been fixed. The domain is no longer cached after a negative response to
4159 MAIL if the sender used is not empty.
4161 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4164 (c) A new callout option, "use_sender" has been added for people who want
4165 the modified behaviour.
4171 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4174 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4175 indeed breaks things for older releases.
4177 3. Added additional logging to the case where there is a problem reading data
4178 from a filter that is running in a subprocess using a pipe, in order to
4179 try to track down a specific problem.
4181 4. Testing facility fudge: when running in the test harness and attempting
4182 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4183 getting "No route to host". Convert this to a timeout.
4185 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4188 6. Some OS don't have socklen_t but use size_t instead. This affects the
4189 fifth argument of getsockopt() amongst other things. This is now
4190 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4191 can be set for individual OS. I have set it for SunOS5, OSF1, and
4192 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4193 some earlier ones do not.
4195 7. Change 4.30/15 was not doing the test caselessly.
4197 8. The standard form for an IPv6 address literal was being rejected by address
4198 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4199 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4200 this, as well as the form without the "IPv6" on the front (but only when
4201 address literals are enabled, of course).
4203 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4205 10. Exim crashed if a message with an empty sender address specified by -f
4206 encountered a router with an errors_to setting. This could be provoked only
4207 by a command such as
4211 where an empty string was supplied; "<>" did not hit this bug.
4213 11. Installed PCRE release 4.5.
4215 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4216 remained set. It is now erased.
4218 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4219 times from message ids (which are base 36 rather than the normal 62).
4221 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4222 were being counted as actual protocol errors, and logged if the log
4223 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4224 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4225 and DATA following a set of rejected RCPTs do not count as protocol errors.
4226 In other words, Exim assumes they were pipelined, though this may not
4227 actually be the case. Of course, in all cases the client gets an
4228 appropriate error code.
4230 15. If a lookup fails in an ACL condition, a message about the failure may
4231 be available; it is used if testing the ACL cannot continue, because most
4232 such messages specify what the cause of the deferral is. However, some
4233 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4234 that caused an old message to be retained and used if a later statement
4235 caused a defer, replacing the real cause of the deferral.
4237 16. If an IP address had so many PTR records that the DNS lookup buffer
4238 was not large enough to hold them, Exim could crash while trying to process
4239 the truncated data. It now detects and logs this case.
4241 17. Further to 4.21/58, another change has been made: if (and only if) the
4242 first line of a message (the first header line) ends with CRLF, a bare LF
4243 in a subsequent header line has a space inserted after it, so as not to
4244 terminate the header.
4246 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4247 unnecessarily, used atoi() instead of strtol(), and didn't check the
4248 termination when getting file sizes from file names by regex.
4250 19. Completely re-implemented the support for maildirsize files, in the light
4251 of a number of problems with the previous contributed implementation
4252 (4.30/29). In particular:
4254 . If the quota is zero, the maildirsize file is maintained, but no quota is
4257 . If the maildir directory does not exist, it is created before any attempt
4258 to write a maildirsize file.
4260 . The quota value in the file is just a cache; if the quota is changed in
4261 the transport, the new value overrides.
4263 . A regular expression is available for excluding directories from the
4266 20. The autoreply transport checks the characters in options that define the
4267 message's headers; it allows continued headers, but it was checking with
4268 isspace() after an embedded newline instead of explicitly looking for a
4271 21. If all the "regular" hosts to which an address was routed had passed their
4272 expiry times, and had not reached their retry times, the address was
4273 bounced, even if fallback hosts were defined. Now Exim should go on to try
4276 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4277 equivalent code in the SMTP transport. Some hosts send humungous responses
4278 to HELO/EHLO, more than 1024 it seems.
4280 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4281 gives compiler warnings in some environments. I've now done it "properly",
4284 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4285 (because of problems with the built-in one) was declared to return uschar *
4286 instead of char *, causing compiler failure.
4288 25. Fixed a file descriptor leak when processing alias/forward files.
4290 26. Fixed a minor format string issue in dbfn.c.
4292 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4294 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4295 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4298 29. When a sender address is verified, it is cached, to save repeating the test
4299 when there is more than one recipient in a message. However, when the
4300 verification involves a callout, it is possible for different callout
4301 options to be set for different recipients. It is too complicated to keep
4302 track of this in the cache, so now Exim always runs a verification when a
4303 callout is required, relying on the callout cache for the optimization.
4304 The overhead is duplication of the address routing, but this should not be
4307 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4308 to be verified with callout=postmaster, and the main callout worked but the
4309 postmaster check failed, the verification correctly failed. However, if a
4310 subsequent RCPT command asked for sender verification *without* the
4311 postmaster check, incorrect caching caused this verification also to fail,
4314 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4315 it was not caching the DNS options (qualify_single, search_parents) that
4316 were used when the lookup failed. A subsequent lookup with different
4317 options therefore always gave the same answer, though there were cases
4318 where it should not have. (Example: a "domains = !$mx_any" option on a
4319 dnslookup router: the "domains" option is always processed without any
4320 widening, but the router might have qualify_single set.) Now Exim uses the
4321 cached value only when the same options are set.
4323 32. Added John Jetmore's "exipick" utility to the distribution.
4325 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4326 than a timeout (e.g. a certificate is required, and is not provided), an
4327 Exim server now closes the connection immediately. Previously it waited for
4328 the client to close - but if the client is SSL, it seems that they each
4329 wait for each other, leading to a delay before one of them times out.
4331 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4332 maintained 0.8.x compatibility because I don't think many are using it, and
4333 it is clearly obsolete.
4335 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4338 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4339 client certificate was expired. A simple patch fixes this, though I don't
4340 understand the full logic of why the verify callback is called multiple
4343 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4344 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4345 which causes problems with some clients (such as the Certicom SSL Plus
4346 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4347 disables the coutermeasure allowing Eudora to connect."
4349 38. Exim was not checking that a write() to a log file succeeded. This could
4350 lead to Bad Things if a log got too big, in particular if it hit a file
4351 size limit. Exim now panics and dies if it cannot write to a log file, just
4352 as it does if it cannot open a log file.
4354 39. Modified OS/Makefile-Linux so that it now contains
4356 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4358 The two -D definitions ensure that Exim is compiled with large file
4359 support, which makes it possible to handle log files that are bigger than
4362 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4363 instance) a domain was checked against a named list that involved a lookup,
4364 causing $domain_data to be set, then another domain was checked against the
4365 same list, then the first domain was re-checked, the value of $domain_data
4366 after the final check could be wrong. In particular, if the second check
4367 failed, it could be set empty. This bug probably also applied to
4370 41. The strip_trailing_dot option was not being applied to the address given
4371 with the -f command-line option.
4373 42. The code for reading a message's header from the spool was incrementing
4374 $received_count, but never initializing it. This meant that the value was
4375 incorrect (doubled) while delivering a message in the same process in which
4376 it was received. In the most common configuration of Exim, this never
4377 happens - a fresh exec is done - but it can happen when
4378 deliver_drop_privilege is set.
4380 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4381 it now includes up to 150 characters of the unexpected data in the log
4384 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4385 and building data strings. The size of both of these buffers was 10 000
4386 bytes - far larger than anybody would *ever* want, thought I. Needless to
4387 say, somebody hit the limit. I have increased the maximum line length to
4388 20 000 and the maximum data length of concatenated lines to 100 000. I have
4389 also fixed two bugs, because there was no checking on these buffers. Tsk,
4390 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4391 buffer is too small.
4393 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4394 lsearch lookups. Now it does.
4396 46. When parsing a route_list item in a manualroute router, a fixed-length
4397 buffer was used for the list of hosts. I made this 1024 bytes long,
4398 thinking that nobody would ever have a list of hosts that long. Wrong.
4399 Somebody had a whole pile of complicated expansion conditions, and the
4400 string was silently truncated, leading to an expansion error. It turns out
4401 that it is easier to change to an unlimited length (owing to other changes
4402 that have happened since this code was originally written) than to build
4403 structure for giving a limitation error. The length of the item that
4404 expands into the list of hosts is now unlimited.
4406 47. The lsearch lookup could not handle data where the length of text line was
4407 more than 4095 characters. Such lines were truncated, leading to shortened
4408 data being returned. It should now handle lines of any length.
4410 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4411 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4414 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4415 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4416 no longer happens in comment lines. A list of replacements is now placed
4417 at the head of all of the source files, except those whose only change is
4418 to replace PERL_COMMAND in the very first #! line.
4420 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4421 messages on the queue, with a bottom-up merge sort, using code contributed
4422 by Michael Haardt. This should make operations like -bp somewhat faster on
4423 large queues. It won't affect queue runners, except when queue_run_in_order
4426 51. Installed eximstats 1.31 in the distribution.
4428 52. Added support for SRV lookups to the dnslookup router.
4430 53. If an ACL referred to $message_body or $message_body_end, the value was not
4431 reset for any messages that followed in the same SMTP session.
4433 54. The store-handling optimization for building very long strings was not
4434 differentiating between the different store pools. I don't think this
4435 actually made any difference in practice, but I've tidied it.
4437 55. While running the routers to verify a sender address, $sender_address
4438 was still set to the sender address. This is wrong, because when routing to
4439 send a bounce to the sender, it would be empty. Therefore, I have changed
4440 it so that, while verifying a sender address, $sender_address is set to <>.
4441 (There is no change to what happens when verifying a recipient address.)
4443 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4444 target A or AAAA records (if not already returned) without resetting the
4445 qualify_single or search_parents options of the DNS resolver. These are
4446 inappropriate in this case because the targets of MX and SRV records must
4447 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4448 target that, when qualified, matched something in the local domain. These
4449 two options are now turned off when doing these lookups.
4451 57. It seems that at least some releases of Reiserfs (which does not have the
4452 concept of a fixed number of inodes) returns zero and not -1 for the
4453 number of available inodes. This interacted badly with check_spool_inodes,
4454 which assumed that -1 was the "no such thing" setting. What I have done is
4455 to check that the total number of inodes is greater than zero before doing
4456 the test of how many are available.
4458 58. When a "warn" ACL statement has a log_message modifier, the message is
4459 remembered, and not repeated. This is to avoid a lot of repetition when a
4460 message has many recipients that cause the same warning to be written.
4461 Howewer, Exim was preserving the list of already written lines for an
4462 entire SMTP session, which doesn't seem right. The memory is now reset if a
4463 new message is started.
4465 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4466 debugging output unless log_rewrite was also set.
4468 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4469 of (int)(handle) when we know that handle contains (void *)(-1).
4471 61. The Exim daemon panic-logs an error return when it closes the incoming
4472 connection. However "connection reset by peer" seems to be common, and
4473 isn't really an error worthy of noting specially, so that particular error
4476 62. When Exim is trying to find all the local interfaces, it used to panic and
4477 die if the ioctl to get the interface flags failed. However, it seems that
4478 on at least one OS (Solaris 9) it is possible to have an interface that is
4479 included in the list of interfaces, but for which you get a failure error
4480 for this call. This happens when the interface is not "plumbed" into a
4481 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4482 failure of the "get flags" call assumes that the interface is down.
4484 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4485 makes life easier for people who are doing arithmetic on fields extracted
4486 from dates, where you often get leading zeros that should not be
4487 interpreted as octal.
4489 64. Added qualify_domain to the redirect router, to override the global
4492 65. If a pathologically long header line contained very many addresses (the
4493 report of this problem mentioned 10 000) and each of them was rewritten,
4494 Exim could use up a very large amount of memory. (It kept on making new
4495 copies of the header line as it rewrote, and never released the old ones.)
4496 At the expense of a bit more processing, the header rewriting function has
4497 been changed so that it no longer eats memory in this way.
4499 66. The generation of the Received: header has been moved from the time that a
4500 message starts to be received, to the time that it finishes. The timestamp
4501 in the Received: header should now be very close to that of the <= log
4502 line. There are two side-effects of this change:
4504 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4505 logged header lines no longer include the local Received: line, because
4506 it has not yet been created. The same applies to a copy of the message
4507 that is returned to a non-SMTP sender when a message is rejected.
4509 (b) When a filter file is tested using -bf, no additional Received: header
4510 is added to the test message. After some thought, I decided that this
4513 This change does not affect the value of $received_for. It is still set
4514 after address rewriting, but before local_scan() is called.
4516 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4518 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4519 gave an unhelpful panic error message, and a defer error. I have managed to
4520 change this behaviour so that it now rejects any supplied certificate,
4521 which seems right, as the list of acceptable certificates is empty.
4523 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4524 gave an unhelpful defer error. I have not managed to make this reject any
4525 supplied certificates, but the error message it gives is "no certificate
4526 supplied", which is not helpful.
4528 70. exigrep's output now also includes lines that are not associated with any
4529 message, but which match the given pattern. Implemented by a patch from
4530 Martin Sluka, which also tidied up the Perl a bit.
4532 71. Recipient callout verification, like sender verification, was using <> in
4533 the MAIL FROM command. This isn't really the right thing, since the actual
4534 sender may affect whether the remote host accepts the recipient or not. I
4535 have changed it to use the actual sender in the callout; this means that
4536 the cache record is now keyed on a recipient/sender pair, not just the
4537 recipient address. There doesn't seem to be a real danger of callout loops,
4538 since a callout by the remote host to check the sender would use <>.
4539 [SEE ABOVE: changed after hitting problems.]
4541 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4542 temporary errors. However, in the case of such a code being given after
4543 the end of a data transmission (i.e. after ".") Exim was failing to write
4544 a retry record for the message. (Yes, there was some broken host that was
4545 actually sending 8xx at this point.)
4547 73. An unknown lookup type in a host list could cause Exim to panic-die when
4548 the list was checked. (An example that provoked this was putting <; in the
4549 middle of a list instead of at the start.) If this happened during a DATA
4550 ACL check, a -D file could be left lying around. This kind of configuration
4551 error no longer causes Exim to die; instead it causes a defer errror. The
4552 incident is still logged to the main and panic logs.
4554 74. Buglet left over from Exim 3 conversion. The message "too many messages
4555 in one connection" was written to the rejectlog but not the mainlog, except
4556 when address rewriting (yes!) was being logged.
4558 75. Added write_rejectlog option.
4560 76. When a system filter was run not as root (that is, when system_filter_user
4561 was set), the values of the $n variables were not being returned to the
4562 main process; thus, they were not subsequently available in the $sn
4565 77. Added +return_path_on_delivery log selector.
4567 78. A connection timeout was being treated differently from recipients deferred
4568 when testing hosts_max_try with a message that was older than the host's
4569 retry timeout. (The host should not be counted, thus allowing all hosts to
4570 be tried at least once before bouncing.) This may have been the cause of an
4571 occasionally reported bug whereby a message would remain on the queue
4572 longer than the retry timeout, but would be bounced if a delivery was
4573 forced. I say "may" because I never totally pinned down the problem;
4574 setting up timeout/retry tests is difficult. See also the next item.
4576 79. The ultimate address timeout was not being applied to errors that involved
4577 a combination of host plus message (for example, a timeout on a MAIL
4578 command). When an address resolved to a number of possible hosts, and they
4579 were not all tried for each delivery (e.g. because of hosts_max_try), a
4580 message could remain on the queue longer than the retry timeout.
4582 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4585 81. Fixed an obscure SMTP outgoing bug which required at least the following
4586 conditions: (a) there was another message waiting for the same server;
4587 (b) the server returned 5xx to all RCPT commands in the first message so
4588 that the message was not completed; (c) the server dropped the connection
4589 or gave a negative response to the RSET that Exim sends to abort the
4590 transaction. The observed case was a dropped connection after DATA that had
4591 been sent in pipelining mode. That is, the server had advertised PIPELINING
4592 but was not implementing it correctly. The effect of the bug was incorrect
4593 behaviour, such as trying another host, and this could lead to a crash.
4599 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4600 and daemon.c were passed as pointers to ints; they should have been
4601 pointers to socklen_t variables (which are typically unsigned ints).
4603 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4606 3. Fixed a really odd bug that affected only the testing scheme; patching a
4607 certain fixed string in the binary changed the value of another string that
4608 happened to be identical to the end of the original first string.
4610 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4611 name", it returns that address as the IP address. On some operating
4612 systems (e.g. Solaris), it also passes back the IP address string as the
4613 "host name". However, on others (e.g. Linux), it passes back an empty
4614 string. Exim wasn't checking for this, and was changing the host name to an
4615 empty string, assuming it had been canonicized.
4617 5. Although rare, it is permitted to have more than one PTR record for a given
4618 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4619 all the names associated with an address, because they do in Solaris.
4620 However, it seems that they do not in Linux for data that comes from the
4621 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4622 I found this out when I moved to a new Linux workstation and tried to run
4623 the Exim test suite.
4625 To get round this problem I have changed the code so that it now does its
4626 own call to the DNS to look up PTR records when searching for a host name.
4627 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4628 addresses that are only in /etc/hosts are still found.
4630 This behaviour is, however, controlled by an option called host_lookup_
4631 order, which defaults to "bydns:byaddr". If people want to use the other
4632 order, or indeed, just use one or the other means of lookup, they can
4633 specify it in this variable.
4635 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4636 some operating systems, this comes back from gethostbyaddr() as an empty
4637 string, and this is what Exim used to test for. However, it seems that in
4638 other systems, "." is yielded. Exim now tests for this case too.
4640 7. The values of check_spool_space and check_log_space are now held internally
4641 as a number of kilobytes instead of an absolute number of bytes. If a
4642 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4643 kilobyte. This means that much larger values can be stored.
4645 8. Exim monitor: an attempt to get the action menu when not actually pointing
4646 at a message produces an empty menu entitled "No message selected". This
4647 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4648 no entries in it ("Shell widget menu has zero width and/or height"). So I
4649 have added a single, blank menu entry in this case.
4651 9. Added ${quote_local_part.
4653 10. MIME decoding is now applied to the contents of Subject: header lines when
4656 11. Now that a reference to $sender_host_address automatically causes a reverse
4657 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4658 host lookup before query-style lookups in lists that might use this
4659 variable. This has therefore been abolished, and the "net-" prefix is no
4660 longer necessary for query-style lookups.
4662 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4663 have been a typo for LFLAGS, so it has been changed.
4665 13. The install script calls Exim with "-C /dev/null" in order to find the
4666 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4667 to be output. Howeve, since Exim outputs its version number before the
4668 error, it didn't break the script. It just looked ugly. I fixed this by
4669 always allowing "-C /dev/null" if the caller is root.
4671 14. Ignore overlarge ACL variable number when reading spool file - insurance
4672 against a later release with more variables having written the file.
4674 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4675 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4676 this, as well as the form without the "IPv6" on the front.
4678 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4679 OS/Makefile-Darwin file.
4681 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4682 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4684 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4685 variable. However, some OS use ints for inodes. I've added cast to long int
4686 to get rid of the compiler warning.
4688 19. I had forgotten to lock out "/../" in configuration file names when
4689 ALT_CONFIG_PREFIX was set.
4691 20. Routers used for verification do not need to specify transports. However,
4692 if such a router generated a host list, and callout was configured, Exim
4693 crashed, because it could not find a port number from the (non-existent)
4694 transport. It now assumes port 25 in this circumstance.
4696 21. Added the -t option to exigrep.
4698 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4699 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4700 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4701 them is set, LOOKUP_LSEARCH is forced.
4703 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4705 24. Added sender and host information to the "rejected by local_scan()" log
4706 line; previously there was no indication of these.
4708 25. Added .include_if_exists.
4710 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4711 Linux. It turns out that not all file systems support this. Apparently some
4712 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4713 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4714 not supported on the file descriptor, is now ignored when Exim is trying to
4715 sync a directory. This applies only to Linux.
4717 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4719 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4720 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4723 29. Added support for maildirsize files from supplied patch (modified a bit).
4725 30. The use of :fail: followed by an empty string could lead Exim to respond to
4726 sender verification failures with (e.g.):
4728 550 Verification failed for <xxx>
4729 550 Sender verify failed
4731 where the first response line was missing the '-' that indicates it is not
4732 the final line of the response.
4734 31. The loop for finding the name of the user that called Exim had a hardwired
4735 limit of 10; it now uses the value of finduser_retries, which is used for
4736 all other user lookups.
4738 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4741 33. Exim was neglecting to zero errno before one call of strtol() when
4742 expanding a string and expecting an integer value. On some systems this
4743 resulted in spurious "integer overflow" errors. Also, it was casting the
4744 result into an int without checking.
4746 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4747 did not work. The code looks as if it has *never* worked, though it appears
4748 to have been documented since at least releast 1.62. I have made it work.
4750 35. The "timeout_DNS" error in retry rules, also documented since at least
4751 1.62, also never worked. As it isn't clear exactly what this means, and
4752 clearly it isn't a major issue, I have abolished the feature by treating it
4753 as "timeout", and writing a warning to the main and panic logs.
4755 36. The display of retry rules for -brt wasn't always showing the error code
4758 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4759 timeout_connect_A, timeout_connect_MX.
4761 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4762 to the empty sender.
4764 39. The daemon was not analysing the content of -oX till after it had closed
4765 stderr and disconnected from the controlling terminal. This meant that any
4766 syntax errors were only noted on the panic log, and the return code from
4767 the command was 0. By re-arranging the code a little, I've made the
4768 decoding happen first, so such errors now appear on stderr, and the return
4769 code is 1. However, the actual setting up of the sockets still happens in
4770 the disconnected process, so errors there are still only recorded on the
4773 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4774 connections (as happens on some IP stacks) was logged at start up time as
4775 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4776 it from "IPv6 and IPv4", which means that two separate sockets are being
4779 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4780 says whether AF_INET or AF_INET6 was passed as an argument.
4782 42. Exiwhat output was messed up when time zones were included in log
4785 43. Exiwhat now gives more information about the daemon's listening ports,
4786 and whether -tls-on-connect was used.
4788 44. The "port" option of the smtp transport is now expanded.
4790 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4791 silently ignored. Now an error message is written to the main and panic
4794 46. There's a new ACL modifier called "logwrite" which writes to a log file
4795 as soon as it is encountered.
4797 47. Added $local_user_uid and $local_user_gid at routing time.
4799 48. Exim crashed when trying to verify a sender address that was being
4802 49. Exim was recognizing only a space character after ".include". It now also
4803 recognizes a tab character.
4805 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4806 extracting the relevant information from the specification. The man page no
4807 longer contains scrambled data for the -d option, and I've added a section
4808 at the front about calling Exim under different names.
4810 51. Added "extra_headers" argument to the "mail" command in filter files.
4812 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4815 53. Installed eximstats 1.29.
4817 54. Added transport_filter_timeout as a generic transport option.
4819 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4820 Cc: header lines. This was required by RFC 822, but it not required by RFC
4823 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4824 incoming messages that did not have them. Now it does so only if the
4825 message originates locally, that is, if there is no associated remote host
4826 address. When Resent- header lines are present, this applies to the Resent-
4827 lines rather than the non-Resent- lines.
4829 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4830 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4832 58. Messages for configuration errors now include the name of the main
4833 configuration file - useful now that there may be more than one file in a
4834 list (.included file names were always shown).
4836 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4837 for those rare installations that do not start the daemon as root or run it
4838 setuid root. I've cut out the call to initgroups() if the daemon is not
4841 60. The Exim user and group can now be bound into the binary as text strings
4842 that are looked up at the start of Exim's processing.
4844 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4846 62. Added $mailstore_basename variable.
4848 63. Installed patch to sieve.c from Michael Haardt.
4850 64. When Exim failed to open the panic log after failing to open the main log,
4851 the original message it was trying to log was written to stderr and debug
4852 output, but if they were not available (the usual case in production), it
4853 was lost. Now it is written to syslog before the two lines that record the
4854 failures to open the logs.
4856 65. Users' Exim filters run in subprocesses under the user's uid. It is
4857 possible for a "deliver" command or an alias in a "personal" command to
4858 provoke an address rewrite. If logging of address rewriting is configured,
4859 this fails because the process is not running as root or exim. There may be
4860 a better way of dealing with this, but for the moment (because 4.30 needs
4861 to be released), I have disabled address rewrite logging when running a
4862 filter in a non-root, non-exim process.
4868 1. The buildconfig auxiliary program wasn't quoting the value set for
4869 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4870 not defined. This bug was masked in 4.22 by the effect that was fixed in
4873 2. Some messages that were rejected after a message id was allocated were
4874 shown as "incomplete" by exigrep. It no longer does this for messages that
4875 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4877 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4878 have allow_domain_literals set, the ID did not get logged in the <= line.
4879 Domain literals are now always recognized in Message-ID: header lines.
4881 4. The first argument for a ${extract expansion item is the key name or field
4882 number. Leading and trailing spaces in this item were not being ignored,
4883 causing some misleading effects.
4885 5. When deliver_drop_privilege was set, single queue runner processes started
4886 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4887 same command in the process it spins off) were not dropping privilege.
4889 6. When the daemon running as "exim" started a queue runner, it always
4890 re-executed Exim in the spun-off process. This is a waste of effort when
4891 deliver_drop_privilege is set. The new process now just calls the
4892 queue-runner function directly.
4898 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4901 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4902 ignored. Though the use of -oP was forcing the writing of a pid file, it
4903 was always written to the default place.
4905 3. If the message "no IP address found for host xxxx" is generated during
4906 incoming verification, it is now followed by identification of the incoming
4907 connection (so you can more easily find what provoked it).
4909 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4911 5. Added some features to "harden" Exim a bit more against certain attacks:
4913 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4914 be put in Local/Makefile. This is like the never_users runtime option,
4915 but it cannot be overridden. The default setting is "root".
4917 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4918 prefix string with which any file named in a -C command line option
4921 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4922 is retained for -C and -D only if the caller of Exim is root. Without
4923 it, the exim user may also use -C and -D and retain privilege.
4925 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4926 command line option is disabled.
4928 6. Macro names set by the -D option must start with an upper case letter, just
4929 like macro names defined in the configuration file.
4931 7. Added "dereference=" facility to LDAP.
4933 8. Two instances of the typo "uknown" in the source files are fixed.
4935 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4936 the Configure-Makefile script screwed up while processing it.
4938 10. Incorporated PCRE 4.4.
4940 11. The SMTP synchronization check was not operating right at the start of an
4941 SMTP session. For example, it could not catch a HELO sent before the client
4942 waited for the greeting. There is now a check for outstanding input at the
4943 point when the greeting is written. Because of the duplex, asynchronous
4944 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4945 way, but not yet received, when the check is performed.
4947 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4948 on TCP/IP sockets, because this apparently causes some broken clients to
4951 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4952 unchanged) from the Cygwin maintainer.
4954 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4955 was supported instead of testing for mbx. Effectively a typo.
4957 15. The spa authenticator server code was not checking that the input it
4958 received was valid base64.
4960 16. The debug output line for the "set" modifier in ACLs was not showing the
4961 name of the variable that was being set.
4963 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4965 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4966 lookup on its own. Something else was needed to trigger the lookup. For
4967 example, a match in host_lookup or the need for a host name in a host list.
4968 Now, if $sender_host_name is referenced and the host name has not yet been
4969 looked up, a lookup is performed. If the lookup fails, the variable remains
4970 empty, and $host_lookup_failed is set to "1".
4972 19. Added "eqi" as a case-independent comparison operator.
4974 20. The saslauthd authentication condition could segfault if neither service
4975 nor realm was specified.
4977 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4978 error message that was logged was misleading, and incoming SMTP
4979 connections were dropped. The message is now more accurate, and temporary
4980 errors are given to SMTP connections.
4982 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4983 (except RSET) with a 421 error, until QUIT is received. However, it was
4984 failing to send a response to QUIT.
4986 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4987 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4988 ACL is now run after the helo_try_verify_hosts code.
4990 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4991 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4992 case-independent, so other case variants are also recognized). Apparently
4993 some systems use these upper case variants.
4995 25. If more than two messages were waiting for the same host, and a transport
4996 filter was specified for the transport, Exim sent two messages over the
4997 same TCP/IP connection, and then failed with "socket operation on non-
4998 socket" when it tried to send the third.
5000 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5002 27. The extern definition of crypt16() in expand.c was not being excluded when
5003 the OS had its own crypt16() function.
5005 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5006 as a preferred synonym for return_size_limit, both as an option and as an
5009 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5011 30. Changed the default configuration ACL to relax the local part checking rule
5012 for addresses that are not in any local domains. For these addresses,
5013 slashes and pipe symbols are allowed within local parts, but the sequence
5014 /../ is explicitly forbidden.
5016 31. SPA server authentication was not clearing the challenge buffer before
5019 32. log_message in a "warn" ACL statement was writing to the reject log as
5020 well as to the main log, which contradicts the documentation and doesn't
5021 seem right (because no rejection is happening). So I have stopped it.
5023 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5024 However, I am unable to do any testing of this.
5026 34. Fixed an infelicity in the appendfile transport. When checking directories
5027 for a mailbox, to see if any needed to be created, it was accidentally
5028 using path names with one or more superfluous leading slashes; tracing
5029 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5031 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5032 contents are added to the log line that is written for every discarded
5033 recipient. (Previously a log_message setting was ignored.)
5035 36. The ${quote: operator now quotes the string if it is empty.
5037 37. The install script runs exim in order to find its version number. If for
5038 some reason other than non-existence or emptiness, which it checks, it
5039 could not run './exim', it was installing it with an empty version number,
5040 i.e. as "exim-". This error state is now caught, and the installation is
5043 38. An argument was missing from the function that creates an error message
5044 when Exim fails to connect to the socket for saslauthd authentication.
5045 This could cause Exim to crash, or give a corrupted message.
5047 39. Added isip, isip4, and isip6 to ${if conditions.
5049 40. The ACL variables $acl_xx are now saved with the message, and can be
5050 accessed later in routers, transports, and filters.
5052 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5053 strings in the file are not string-expanded.
5055 42. If a MAIL command specified a SIZE value that was too large to fit into an
5056 int variable, the check against message_size_limit failed. Such values are
5057 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5058 day this will have to be increased, but I don't think I want to be around
5059 when emails are that large.
5066 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5067 iconv() is not standard in FreeBSD.
5069 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5070 IPv6 enabled. The observed symptom was a segmentation fault on return from
5071 the function os_common_find_running_interfaces() in src/os.c.
5073 3. In the check_special_case() function in daemon.c I had used "errno" as an
5074 argument name, which causes warnings on some systems. This was basically a
5075 typo, since it was named "eno" in the comments!
5077 4. The code that waits for the clock to tick (at a resolution of some fraction
5078 of a second) so as to ensure message-id uniqueness was always waiting for
5079 at least one whole tick, when it could have waited for less. [This is
5080 almost certainly not relevant at current processor speeds, where it is
5081 unlikely to ever wait at all. But we try to future-proof.]
5083 5. The function that sleeps for a time interval that includes fractions of a
5084 second contained a race. It did not block SIGALRM between setting the
5085 timer, and suspending (a couple of lines later). If the interval was short
5086 and the sigsuspend() was delayed until after it had expired, the suspension
5087 never ended. On busy systems this could lead to processes getting stuck for
5090 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5091 process, before it forks any delivery processes. The open lookup caching
5092 mechanism meant that the open file or database connection was passed into
5093 the delivery process. The problem was that delivery processes always tidy
5094 up cached lookup data. This could cause a problem for the next delivery
5095 process started by the queue runner, because the external queue runner
5096 process does not know about the closure. So the next delivery process
5097 still has data in the lookup cache. In the case of a file lookup, there was
5098 no problem because closing a file descriptor in a subprocess doesn't affect
5099 the parent. However, if the lookup was caching a connection to a database,
5100 the connection was closed, and the second delivery process was likely to
5101 see errors such as "PGSQL: query failed: server closed the connection
5102 unexpectedly". The problem has been fixed by closing all cached lookups
5103 in a queue runner before running a delivery process.
5105 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5106 seem to have the "const" qualifier which it has on other OS. I've
5109 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5110 *of the same type* (client or server) with the same public name that an
5111 error should be diagnosed.
5113 9. When Exim looked up a host name for an IP address, but failed to find the
5114 original IP address when looking up the host name (a safety check), it
5115 output the message "<ip address> does not match any IP for NULL", which was
5116 confusing, to say the least. The bug was that the host name should have
5117 appeared instead of "NULL".
5119 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5120 user that is built into the binary, and the -C or -D options is used, root
5121 privilege is dropped before the configuration file is read. In addition,
5122 logging is switched to stderr instead of the normal log files. If the
5123 configuration then re-defines the Exim user, the unprivileged environment
5124 is probably not what is expected, so Exim logs a panic warning message (but
5127 However, if deliver_drop_privilege is set, the unprivileged state may well
5128 be exactly what is intended, so the warning has been cut out in that case,
5129 and Exim is allowed to try to write to its normal log files.
5135 1. smtp_return_error_details was not giving details for temporary sender
5136 or receiver verification errors.
5138 2. Diagnose a configuration error if two authenticators have the same public
5141 3. Exim used not to create the message log file for a message until the first
5142 delivery attempt. This could be confusing when incoming messages were held
5143 for policy or load reasons. The message log file is now created at the time
5144 the message is received, and an initial "Received" line is written to it.
5146 4. The automatically generated man page for command line options had a minor
5147 bug that caused no ill effects; however, a more serious problem was that
5148 the procedure for building the man page automatically didn't always
5149 operate. Consequently, release 4.20 contains an out-of-date version. This
5150 shouldn't happen again.
5152 5. When building Exim with embedded Perl support, the script that builds the
5153 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5154 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5156 6. The freeze_tell option was not being used for messages that were frozen on
5157 arrival, either by an ACL or by local_scan().
5159 7. Added the smtp_incomplete_transaction log selector.
5161 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5162 was accepting AUTH without a new EHLO.
5164 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5165 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5168 10. Logging of TCP/IP connections (when configured) now happens in the main
5169 daemon process instead of the child process, so that the TCP/IP connection
5170 count is more accurate (but it can never be perfect).
5172 11. The use of "drop" in a nested ACL was not being handled correctly in the
5173 outer ACL. Now, if condition failure induced by the nested "drop" causes
5174 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5175 or "require"), the connection is dropped.
5177 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5178 that yield "discard" can now be used with an "accept" or a "discard" verb,
5179 but an error is generated for any others (because I can't see a useful way
5180 to define what should happen).
5182 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5183 are now processed in the same way as lines in the Exim configuration file.
5184 In particular, continuation lines are supported.
5186 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5188 15. Added -ti meaning -t -i.
5190 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5191 domains, and warn by logging if others are found.
5193 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5194 returned from GIFCONF when Exim is trying to find the list of interfaces on
5195 a host. The code in os.c has been modified to copy each ifreq to an aligned
5196 structure in all cases.
5198 Also, in some cases, the returned ifreq's were being copied to a 'struct
5199 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5200 means the last couple of bytes of an IPv6 address could be chopped if the
5201 ifreq contained only a normal sockaddr (14 bytes storage).
5203 18. Named domain lists were not supported in the hosts_treat_as_local option.
5204 An entry such as +xxxx was not recognized, and was treated as a literal
5207 19. Ensure that header lines added by a DATA ACL are included in the reject log
5208 if the ACL subsequently rejects the message.
5210 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5211 MD5 (which is deprecated).
5213 21. When testing a filter file using -bf, Exim was writing a message when it
5214 took the sender from a "From " line in the message, but it was not doing so
5215 when it took $return_path from a Return-Path: header line. It now does.
5217 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5218 with a valid header line field name (a series of printing characters
5219 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5221 23. Changed "disc" in the source to "disk" to conform to the documentation and
5222 the book and for uniformity.
5224 24. Ignore Sendmail's -Ooption=value command line item.
5226 25. When execve() failed while trying to run a command in a pipe transport,
5227 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5228 could be confused with a return value of 69 from the command itself. This
5229 has been changed to 127, the value the shell returns if it is asked to run
5230 a non-existent command. The wording for the related log line suggests a
5231 non-existent command as the problem.
5233 26. If received_header_text expands to an empty string, do not add a Received:
5234 header line to the message. (Well, it adds a token one on the spool, but
5235 marks it "old" so that it doesn't get used or transmitted.)
5237 27. Installed eximstats 1.28 (addition of -nt option).
5239 28. There was no check for failure on the call to getsockname() in the daemon
5240 code. This can fail if there is a shortage of resources on the system, with
5241 ENOMEM, for example. A temporary error is now given on failure.
5243 29. Contrary to the C standard, it seems that in some environments, the
5244 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5245 program. Exim now does this explicitly; it affects the formatting of
5246 timestamps using strftime().
5248 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5249 complaints. I've now initialized all the variables, to avoid this.
5251 32. Header lines added by a system filter were not being "seen" during
5252 transport-time rewrites.
5254 33. The info_callback() function passed to OpenSSL is set up with type void
5255 (*)(SSL *, int, int), as described somewhere. However, when calling the
5256 function (actually a macro) that sets it up, the type void(*)() is
5257 expected. I've put in a cast to prevent warnings from picky compilers.
5259 34. If a DNS black list lookup found a CNAME record, but there were no A
5260 records associated with the domain it pointed at, Exim crashed.
5262 35. If a DNS black list lookup returned more than one A record, Exim ignored
5263 all but the first. It now scans all returned addresses if a particular IP
5264 value is being sought. In this situation, the contents of the
5265 $dnslist_value variable are a list of all the addresses, separated by a
5268 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5269 to accept a host name and all its aliases if the forward lookup for any of
5270 them yielded the IP address of the incoming connection. Now it accepts only
5271 those names whose forward lookup yields the correct IP address. Any other
5272 names are discarded. This closes a loophole whereby a rogue DNS
5273 administrator could create reverse DNS records to break through a
5274 wildcarded host restriction in an ACL.
5276 37. If a user filter or a system filter that ran in a subprocess used any of
5277 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5278 the wrong values were passed to the pipe command ($thisaddress had the
5279 value of $0, $0 had the value of $1, etc). This bug was introduced by
5280 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5282 38. Improved the line breaking for long SMTP error messages from ACLs.
5283 Previously, if there was no break point between 40 and 75 characters, Exim
5284 left the rest of the message alone. Two changes have been made: (a) I've
5285 reduced the minimum length to 35 characters; (b) if it can't find a break
5286 point between 35 and 75 characters, it looks ahead and uses the first one
5287 that it finds. This may give the occasional overlong line, but at least the
5288 remaining text gets split now.
5290 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5291 file descriptors might be low, and that setting 1000 would always raise it.
5292 It turns out that in some environments, the limit is already over 1000 and
5293 that lowering it causes trouble. So now Exim takes care not to decrease it.
5295 40. When delivering a message, the value of $return_path is set to $sender_
5296 address at the start of routing (routers may change the value). By an
5297 oversight, this default was not being set up when an address was tested by
5298 -bt or -bv, which affected the outcome if any router or filter referred to
5301 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5302 log only when "message" or "log_message" are set. However, if one of the
5303 conditions was an address verification, or a call to a nested ACL, the
5304 messages generated by the underlying test were being passed through. This
5305 no longer happens. The underlying message is available in $acl_verify_
5306 message for both "message" and "log_message" expansions, so it can be
5307 passed through if needed.
5309 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5310 new expansion $bh_ to give the encoded byte string without charset
5311 translation. Translation happens only if iconv() is available; HAVE_ICONV
5312 indicates this at build time. HEADERS_CHARSET gives the charset to
5313 translate to; headers_charset can change it in the configuration, and
5314 "headers charset" can change it in an individual filter file.
5316 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5317 that creates RFC 2047 encoded "words" labels them as that charset instead
5318 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5319 expansion operator; (ii) when Exim creates a From: line for a local
5320 message; (iii) when a header line is rewritten to include a "phrase" part.
5322 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5323 buggy, causing it to skip the first lines of messages whose message ID
5324 ended in 'D'. This would not have bitten before Exim release 4.14, because
5325 message IDs were unlikely to end in 'D' before then. The effect was to have
5326 incorrect size information for certain domains.
5328 45. #include "config.h" was missing at the start of the crypt16.c module. This
5329 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5332 46. If there was a timeout during a "random" callout check, Exim treated it as
5333 a failure of the random address, and carried on sending RSET and the real
5334 address. If the delay was just some slowness somewhere, the response to the
5335 original RCPT would be taken as a response to RSET and so on, causing
5336 mayhem of various kinds.
5338 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5339 when I implemented it. It didn't allow for the fact that some option values
5340 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5341 the right test for positive values.
5343 48. Domain names in DNS records are case-independent. Exim always looks them up
5344 in lower case. Some resolvers return domain names in exactly the case they
5345 appear in the zone file, that is, they may contain uppercase letters. Not
5346 all resolvers do this - some return always lower case. Exim was treating a
5347 change of case by a resolver as a change of domain, similar to a widening
5348 of a domain abbreviation. This triggered its re-routing code and so it was
5349 trying to route what was effectively the same domain again. This normally
5350 caused routing to fail (because the router wouldn't handle the domain
5351 twice). Now Exim checks for this case specially, and just changes the
5352 casing of the domain that it ultimately uses when it transmits the message
5355 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5358 50. If a filter generated a file delivery with a non-absolute name (possible if
5359 no home directory exists for the router), the forbid_file option was not
5362 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5363 the existing equality matching.
5365 52. Exim was using ints instead of ino_t variables in some places where it was
5366 dealing with inode numbers.
5368 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5369 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5370 called TMPDIR, and if it finds it is different, it changes its value.
5372 54. The smtp_printf() function is now made available to local_scan() so
5373 additional output lines can be written before returning. There is also an
5374 smtp_fflush() function to enable the detection of a dropped connection.
5375 The variables smtp_input and smtp_batched_input are exported to
5378 55. Changed the default runtime configuration: the message "Unknown user"
5379 has been removed from the ACL, and instead placed on the localuser router,
5380 using the cannot_route_message feature. This means that any verification
5381 failures that generate their own messages won't get overridden. Similarly,
5382 the "Unrouteable address" message that was in the ACL for unverifiable
5383 relay addresses has also been removed.
5385 56. Added hosts_avoid_esmtp to the smtp transport.
5387 57. The exicyclog script was not checking for the esoteric option
5388 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5389 will work only if exicyclog is run under the appropriate euid.
5391 58. Following a discussion on the list, the rules by which Exim recognises line
5392 endings on incoming messages have been changed. The -dropcr and drop_cr
5393 options are now no-ops, retained only for backwards compatibility. The
5394 following line terminators are recognized: LF CRLF CR. However, special
5395 processing applies to CR:
5397 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5398 nor a local message in the state where . is a terminator.
5400 (ii) If a bare CR is encountered in a header line, an extra space is added
5401 after the line terminator so as not to end the header. The reasoning
5402 behind this is that bare CRs in header lines are most likely either
5403 to be mistakes, or people trying to play silly games.
5405 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5406 was being incorrectly given as 18 bytes larger than it should have been.
5407 This is a VOB (very old bug).
5409 60. This may never have affected anything current, but just in case it has:
5410 When the local host is found other than at the start of a list of hosts,
5411 the local host, those with the same MX, and any that follow, are discarded.
5412 When the list in question was part of a longer list of hosts, the following
5413 hosts (not currently being processed) were also being discarded. This no
5414 longer happens. I'm not sure if this situation could ever has previously
5417 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5420 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5421 header. This is something that is recommended in a new Internet Draft, and
5422 is something that is documented as being done by Sendmail. There are two
5423 possible values. For messages generated by the autoreply transport, Exim
5426 Auto-Submitted: auto-replied
5428 whereas for all other generated messages (e.g. bounces) it adds
5430 Auto-Submitted: auto-generated
5432 63. The "personal" condition in filters now includes a test for the
5433 Auto-Submitted: header. If it contains the string "auto-" the message it
5434 not considered personal.
5436 64. Added rcpt_include_affixes as a generic transport option.
5438 65. Added queue_only_override (default true).
5440 66. Added the syslog_duplication option.
5442 67. If what should have been the first header line of a message consisted of
5443 a space followed by a colon, Exim was mis-interpreting it as a header line.
5444 It isn't of course - it is syntactically invalid and should therefore be
5445 treated as the start of the message body. The misbehaviour could have
5446 caused a number of strange effects, including loss of data in subsequent
5447 header lines, and spool format errors.
5449 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5450 client host had authenticated. This control can now be exercised by an ACL
5451 for more flexibility.
5453 69. By default, callouts do not happen when testing with -bh. There is now a
5454 variant, -bhc, which does actually run the callout code, including
5455 consulting and updating the callout cache.
5457 70. Added support for saslauthd authentication, courtesy of Alexander
5460 71. If statvfs() failed on the spool or log directories while checking their
5461 size for availability, Exim confusingly gave the error "space shortage".
5462 Furthermore, in debugging mode it crashed with a floating point exception.
5463 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5464 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5465 serious problem, Exim now writes to the main and panic logs when this
5466 happens, with details of the failure. It then refuses to accept the
5467 incoming message, giving the message "spool directory problem" or "log
5468 directory problem" with a 421 code for SMTP messages.
5470 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5471 0, 1, and 2 exist, because some OS complain for execs without them (see
5472 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5473 descriptors. However, the code omitted to check that the open succeeded,
5474 causing mysterious errors if for some reason the permissions on /dev/null
5475 got screwed. Now Exim writes a message to the main and panic logs, and
5476 bombs out if it can't open /dev/null.
5478 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5479 interact so that it is all more flexible. It is supposed to remain
5480 backwards compatible. Also added extra_local_interfaces.
5482 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5483 to bomb out with an assertion failure - to the client this appears as a
5484 connection drop. This problem occurs in the part of the code that was taken
5485 from the Samba project. Fortunately, the assertion is in a very simple
5486 function, so I have fixed this by reproducing the function inline in the
5487 one place where it is called, and arranging for authentication to fail
5488 instead of killing the process with assert().
5490 75. The SPA client code was not working when the server requested OEM rather
5491 than Unicode encoding.
5493 76. Added code to make require_files with a specific uid setting more usable in
5494 the case where statting the file as root fails - usually a non-root-mounted
5495 NFS file system. When this happens and the failure is EACCES, Exim now
5496 forks a subprocess and does the per-uid checking as the relevant uid.
5498 77. Added process_log_path.
5500 78. If log_file_path was not explicitly set, a setting of check_log_space or
5501 check_log_inodes was ignored.
5503 79. If a space check for the spool or log partitions fails, the incident is now
5504 logged. Of course, in the latter case the data may get lost...
5506 80. Added the %p formatting code to string_format() so that it can be used to
5507 print addresses in debug_print(). Adjusted all the address printing in the
5508 debugging in store.c to use %p rather than %d.
5510 81. There was a concern that a line of code in smtp_in.c could overflow a
5511 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5512 initially expressed, the concern was not well-founded, because trailing
5513 spaces are removed early. However, if the trailing spaces were followed by
5514 a NULL, they did not get removed, so the overflow was possible. Two fixes
5517 (a) I re-wrote the offending code in a cleaner fashion.
5518 (b) If an incoming SMTP command contains a NULL character, it is rejected
5521 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5522 runs initgroups(), so that if the Exim user is in any additional groups,
5523 they will be used during message reception.
5529 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob