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.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
85 PP/01 Solaris build fix for Oracle's LDAP libraries.
86 Bugzilla 1109, patch from Stephen Usher.
88 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
90 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
93 TF/02 Fix a couple more cases where we did not log the error message
94 when unlink() failed. See also change 4.74-TF/03.
96 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
97 lock up or crash if it happened to be inside a call to libc when it
98 got a SIGUSR1 from exiwhat.
100 The SIGUSR1 handler appends the current process status to the process
101 log which is later printed by exiwhat. It used to use the general
102 purpose logging code to do this, but several functions it calls are
103 not safe for signals.
105 The new output code in the SIGUSR1 handler is specific to the process
106 log, and simple enough that it's easy to inspect for signal safety.
107 Removing some special cases also simplifies the general logging code.
108 Removing the spurious timestamps from the process log simplifies
111 TF/04 Improved ratelimit ACL condition.
113 The /noupdate option has been deprecated in favour of /readonly which
114 has clearer semantics. The /leaky, /strict, and /readonly update modes
115 are mutually exclusive. The update mode is no longer included in the
116 database key; it just determines when the database is updated. (This
117 means that when you upgrde Exim will forget old rate measurements.)
119 Exim now checks that the per_* options are used with an update mode that
120 makes sense for the current ACL. For example, when Exim is processing a
121 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
122 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
123 must specify per_mail/readonly. If you omit the update mode it defaults to
124 /leaky where that makes sense (as before) or /readonly where required.
126 The /noupdate option is now undocumented but still supported for
127 backwards compatibility. It is equivalent to /readonly except that in
128 ACLs where /readonly is required you may specify /leaky/noupdate or
129 /strict/noupdate which are treated the same as /readonly.
131 A useful new feature is the /count= option. This is a generalization
132 of the per_byte option, so that you can measure the throughput of other
133 aggregate values. For example, the per_byte option is now equivalent
134 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
136 The per_rcpt option has been generalized using the /count= mechanism
137 (though it's more complicated than the per_byte equivalence). When it is
138 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
139 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
140 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
141 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
142 Note that using per_rcpt with a non-readonly update mode in more than
143 one ACL will cause the recipients to be double-counted. (The per_mail
144 and per_byte options don't have this problem.)
146 The handling of very low rates has changed slightly. If the computed rate
147 is less than the event's count (usually one) then this event is the first
148 after a long gap. In this case the rate is set to the same as this event's
149 count, so that the first message of a spam run is counted properly.
151 The major new feature is a mechanism for counting the rate of unique
152 events. The new per_addr option counts the number of different
153 recipients that someone has sent messages to in the last time period. It
154 behaves like per_rcpt if all the recipient addresses are different, but
155 duplicate recipient addresses do not increase the measured rate. Like
156 the /count= option this is a general mechanism, so the per_addr option
157 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
158 example, measure the rate that a client uses different sender addresses
159 with the options per_mail/unique=$sender_address. There are further
160 details in the main documentation.
162 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
164 TF/06 Removed a few PCRE remnants.
166 TF/07 Automatically extract Exim's version number from tags in the git
167 repository when doing development or release builds.
169 PP/02 Raise smtp_cmd_buffer_size to 16kB.
170 Bugzilla 879. Patch from Paul Fisher.
172 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
173 Heavily based on revision 40f9a89a from Simon Arlott's tree.
176 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
178 PP/05 Variable $av_failed, true if the AV scanner deferred.
179 Bugzilla 1078. Patch from John Horne.
181 PP/06 Stop make process more reliably on build failure.
182 Bugzilla 1087. Patch from Heiko Schlittermann.
184 PP/07 Make maildir_use_size_file an _expandable_ boolean.
185 Bugzilla 1089. Patch from Heiko Schlittermann.
187 PP/08 Handle ${run} returning more data than OS pipe buffer size.
188 Bugzilla 1131. Patch from Holger Weiß.
190 PP/09 Handle IPv6 addresses with SPF.
191 Bugzilla 860. Patch from Wolfgang Breyha.
193 PP/10 GnuTLS: support TLS 1.2 & 1.1.
195 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
198 PP/11 match_* no longer expand right-hand-side by default.
199 New compile-time build option, EXPAND_LISTMATCH_RHS.
200 New expansion conditions, "inlist", "inlisti".
202 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
204 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
206 PP/14 fix log_write() format string regression from TF/03.
207 Bugzilla 1152. Patch from Dmitry Isaikin.
213 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
215 PP/02 Harmonised TLS library version reporting; only show if debugging.
216 Layout now matches that introduced for other libraries in 4.74 PP/03.
218 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
220 PP/04 New "dns_use_edns0" global option.
222 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
225 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
226 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
228 TK/01 Updated PolarSSL code to 0.14.2.
229 Bugzilla 1097. Patch from Andreas Metzler.
231 PP/07 Catch divide-by-zero in ${eval:...}.
234 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
237 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
238 format-string attack -- SECURITY: remote arbitrary code execution.
240 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
241 time unintentionally subject to list matching rules, letting the header
242 cause arbitrary Exim lookups (of items which can occur in lists, *not*
243 arbitrary string expansion). This allowed for information disclosure.
245 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
246 INT_MIN/-1 -- value coerced to INT_MAX.
252 NM/01 Workround for PCRE version dependency in version reporting
255 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
256 This fixes portability to compilers other than gcc, notably
257 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
259 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
260 makefiles for portability to HP-UX and POSIX correctness.
262 PP/01 Permit LOOKUP_foo enabling on the make command-line.
263 Also via indented variable definition in the Makefile.
264 (Debugging by Oliver Heesakkers).
266 PP/02 Restore caching of spamd results with expanded spamd_address.
267 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
269 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
270 Improves build reliability. Fix from: Frank Elsner
272 NM/02 Fix wide character breakage in the rfc2047 coding
273 Fixes bug 1064. Patch from Andrey N. Oktyabrski
275 NM/03 Allow underscore in dnslist lookups
276 Fixes bug 1026. Patch from Graeme Fowler
278 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
279 Code patches from Adam Ciarcinski of NetBSD.
281 NM/04 Fixed exiqgrep to cope with mailq missing size issue
284 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
285 is logged, to avoid truncation. Patch from John Horne.
287 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
288 Patch from Jakob Hirsch.
290 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
291 SQL string expansion failure details.
292 Patch from Andrey Oktyabrski.
294 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
295 Patch from Simon Arlott.
297 PP/09 New lookups functionality failed to compile on old gcc which rejects
298 extern declarations in function scope.
299 Patch from Oliver Fleischmann
301 PP/10 Use sig_atomic_t for flags set from signal handlers.
302 Check getgroups() return and improve debugging.
303 Fixed developed for diagnosis in bug 927 (which turned out to be
306 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
307 Patch from Mark Zealey.
309 PP/12 Bugzilla 1056: Improved spamd server selection.
310 Patch from Mark Zealey.
312 PP/13 Bugzilla 1086: Deal with maildir quota file races.
313 Based on patch from Heiko Schlittermann.
315 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
316 Patch from Uwe Doering, sign-off by Michael Haardt.
318 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
319 variable declaration deep within a block. Bug and patch from
322 PP/15 lookups-Makefile IRIX compatibilty coercion.
324 PP/16 Make DISABLE_DKIM build knob functional.
326 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
327 Patch by Simon Arlott
329 TF/03 Fix valgrind.h portability to C89 compilers that do not support
330 variable argument macros. Our copy now differs from upstream.
336 TF/01 Failure to get a lock on a hints database can have serious
337 consequences so log it to the panic log.
339 TF/02 Log LMTP confirmation messages in the same way as SMTP,
340 controlled using the smtp_confirmation log selector.
342 TF/03 Include the error message when we fail to unlink a spool file.
344 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
345 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
346 for maintaining out-of-tree patches for some time.
348 PP/01 Bugzilla 139: Documentation and portability issues.
349 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
350 Handle per-OS dynamic-module compilation flags.
352 PP/02 Let /dev/null have normal permissions.
353 The 4.73 fixes were a little too stringent and complained about the
354 permissions on /dev/null. Exempt it from some checks.
355 Reported by Andreas M. Kirchwitz.
357 PP/03 Report version information for many libraries, including
358 Exim version information for dynamically loaded libraries. Created
359 version.h, now support a version extension string for distributors
360 who patch heavily. Dynamic module ABI change.
362 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
363 privilege escalation vulnerability whereby the Exim run-time user
364 can cause root to append content of the attacker's choosing to
367 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
370 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
371 If dropping privileges for untrusted macros, we disabled normal logging
372 on the basis that it would fail; for the Exim run-time user, this is not
373 the case, and it resulted in successful deliveries going unlogged.
374 Fixed. Reported by Andreas Metzler.
380 PP/01 Date: & Message-Id: revert to normally being appended to a message,
381 only prepend for the Resent-* case. Fixes regression introduced in
382 Exim 4.70 by NM/22 for Bugzilla 607.
384 PP/02 Include check_rfc2047_length in configure.default because we're seeing
385 increasing numbers of administrators be bitten by this.
387 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
389 PP/03 Bugzilla 994: added openssl_options main configuration option.
391 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
393 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
395 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
397 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
398 without a peer certificate, leading to a segfault because of an
399 assumption that peers always have certificates. Be a little more
400 paranoid. Problem reported by Martin Tscholak.
402 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
403 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
404 NB: ClamAV planning to remove STREAM in "middle of 2010".
405 CL also introduces -bmalware, various -d+acl logging additions and
406 more caution in buffer sizes.
408 PP/09 Implemented reverse_ip expansion operator.
410 PP/10 Bugzilla 937: provide a "debug" ACL control.
412 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
414 PP/12 Bugzilla 973: Implement --version.
416 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
418 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
420 PP/15 Bugzilla 816: support multiple condition rules on Routers.
422 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
423 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
424 ignore trailing whitespace.
426 JJ/02 prevent non-panic DKIM error from being sent to paniclog
428 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
431 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
432 Notification from Dr Andrew Aitchison.
434 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
435 ExtendedDetectionInfo response format.
436 Notification from John Horne.
438 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
441 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
442 XSL and documented dependency on system catalogs, with examples of how
445 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
448 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
449 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
450 configuration file which is writeable by the Exim user or group.
452 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
453 of configuration files to cover files specified with the -C option if
454 they are going to be used with root privileges, not just the default
457 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
458 option (effectively making it always true).
460 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
461 files to be used while preserving root privileges.
463 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
464 that rogue child processes cannot use them.
466 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
467 run-time user, instead of root.
469 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
470 Exim run-time user without dropping privileges.
472 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
473 result string, instead of calling string_vformat() twice with the same
476 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
477 for other users. Others should always drop root privileges if they use
478 -C on the command line, even for a whitelisted configure file.
480 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
482 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
488 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
489 $data_path, and $header_path variables; fixed documentation bugs and
492 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
493 exipick to access non-standard spools, including the "frozen" queue
496 NM/01 Bugzilla 965: Support mysql stored procedures.
497 Patch from Alain Williams
499 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
501 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
502 Patch from Andreas Metzler
504 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
505 Patch from Kirill Miazine
507 NM/05 Bugzilla 671: Added umask to procmail example.
509 JJ/03 installed exipick 20100323.0, fixing doc bug
511 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
512 directory. Notification and patch from Dan Rosenberg.
514 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
516 TK/02 Improve log output when DKIM signing operation fails.
518 MH/01 Treat the transport option dkim_domain as a colon separated
519 list, not as a single string, and sign the message with each element,
520 omitting multiple occurences of the same signer.
522 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
523 Bugzilla 985, 986. Patch by Simon Arlott
525 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
526 Patch by Simon Arlott
528 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
529 MBX locking. Notification from Dan Rosenberg.
535 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
537 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
539 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
541 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
543 NM/04 Bugzilla 915: Flag broken perl installation during build.
549 TK/01 Added patch by Johannes Berg that expands the main option
550 "spamd_address" if it starts with a dollar sign.
552 TK/02 Write list of recipients to X-Envelope-Sender header when building
553 the mbox-format spool file for content scanning (suggested by Jakob
556 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
557 (http://www.dcc-servers.net/) support via dccifd. Activated by
558 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
560 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
561 by Mark Daniel Reidel <mr@df.eu>.
563 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
564 When building exim an external PCRE library is now needed -
565 PCRE is a system library on the majority of modern systems.
566 See entry on PCRE_LIBS in EDITME file.
568 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
569 conversation. Added nologin parameter to request.
570 Patch contributed by Kirill Miazine.
572 TF/01 Do not log submission mode rewrites if they do not change the address.
574 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
576 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
577 log files in place. Contributed by Roberto Lima.
579 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
581 TF/03 Bugzilla 615: When checking the local_parts router precondition
582 after a local_part_suffix or local_part_prefix option, Exim now
583 does not use the address's named list lookup cache, since this
584 contains cached lookups for the whole local part.
586 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
587 Robert Millan. Documentation is in experimental-spec.txt.
589 TF/04 Bugzilla 668: Fix parallel build (make -j).
591 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
593 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
594 Patch provided by Jan Srzednicki.
596 TF/05 Leading white space used to be stripped from $spam_report which
597 wrecked the formatting. Now it is preserved.
599 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
600 that they are available at delivery time.
602 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
604 TF/08 TLS error reporting now respects the incoming_interface and
605 incoming_port log selectors.
607 TF/09 Produce a more useful error message if an SMTP transport's hosts
608 setting expands to an empty string.
610 NM/06 Bugzilla 744: EXPN did not work under TLS.
611 Patch provided by Phil Pennock.
613 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
614 Patch provided by Richard Godbee.
616 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
617 acl_smtp_notquit, added index entry.
619 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
620 Patch provided by Eugene Bujak.
622 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
623 accept(). Patch provided by Maxim Dounin.
625 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
626 Patch provided by Phil Pennock.
628 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
630 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
631 Patch provided by Brad "anomie" Jorsch.
633 NM/14 Bugzilla 622: Added timeout setting to transport filter.
634 Patch provided by Dean Brooks.
636 TK/05 Add native DKIM support (does not depend on external libraries).
638 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
639 Patch provided by Graeme Fowler.
641 NM/16 Bugzilla 851: Documentation example syntax fix.
643 NM/17 Changed NOTICE file to remove references to embedded PCRE.
645 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
648 NM/19 Bugzilla 745: TLS version reporting.
649 Patch provided by Phil Pennock.
651 NM/20 Bugzilla 167: bool: condition support.
652 Patch provided by Phil Pennock.
654 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
655 clients. Patch provided by Phil Pennock.
657 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
658 Patch provided by Brad "anomie" Jorsch.
660 NM/23 Bugzilla 687: Fix misparses in eximstats.
661 Patch provided by Heiko Schlittermann.
663 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
664 Patch provided by Heiko Schlittermann.
666 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
667 plus update to original patch.
669 NM/26 Bugzilla 799: Documentation correction for ratelimit.
671 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
672 Patch provided by David Brownlee.
674 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
676 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
678 NM/30 Bugzilla 888: TLS documentation bugfixes.
680 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
682 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
683 Unlike the original bugzilla I have changed all shell scripts in src tree.
685 NM/33 Bugzilla 898: Transport filter timeout fix.
686 Patch by Todd Rinaldo.
688 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
689 Patch by Serge Demonchaux.
691 NM/35 Bugzilla 39: Base64 decode bug fixes.
692 Patch by Jakob Hirsch.
694 NM/36 Bugzilla 909: Correct connect() call in dcc code.
696 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
698 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
700 NM/39 Bugzilla 911: Fixed MakeLinks build script.
706 TK/01 Add preliminary DKIM support. Currently requires a forked version of
707 ALT-N's libdkim that I have put here:
708 http://duncanthrax.net/exim-experimental/
710 Note to Michael Haardt: I had to rename some vars in sieve.c. They
711 were called 'true' and it seems that C99 defines that as a reserved
712 keyword to be used with 'bool' variable types. That means you could
713 not include C99-style headers which use bools without triggering
714 build errors in sieve.c.
716 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
717 as mailq or other aliases. Changed the --help handling significantly
718 to do whats expected. exim_usage() emits usage/help information.
720 SC/01 Added the -bylocaldomain option to eximstats.
722 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
724 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
726 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
732 PH/01 Another patch from the Sieve maintainer.
734 PH/02 When an IPv6 address is converted to a string for single-key lookup
735 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
736 dots are used instead of colons so that keys in lsearch files need not
737 contain colons. This was done some time before quoting was made available
738 in lsearch files. However, iplsearch files do require colons in IPv6 keys
739 (notated using the quote facility) so as to distinguish them from IPv4
740 keys. This meant that lookups for IP addresses in host lists did not work
741 for iplsearch lookups.
743 This has been fixed by arranging for IPv6 addresses to be expressed with
744 colons if the lookup type is iplsearch. This is not incompatible, because
745 previously such lookups could never work.
747 The situation is now rather anomolous, since one *can* have colons in
748 ordinary lsearch keys. However, making the change in all cases is
749 incompatible and would probably break a number of configurations.
751 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
754 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
755 conversion specification without a maximum field width, thereby enabling
756 a rogue spamd server to cause a buffer overflow. While nobody in their
757 right mind would setup Exim to query an untrusted spamd server, an
758 attacker that gains access to a server running spamd could potentially
759 exploit this vulnerability to run arbitrary code as the Exim user.
761 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
762 $primary_hostname instead of what libspf2 thinks the hosts name is.
764 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
765 a directory entry by the name of the lookup key. Previously, if a
766 symlink pointed to a non-existing file or a file in a directory that
767 Exim lacked permissions to read, a lookup for a key matching that
768 symlink would fail. Now it is enough that a matching directory entry
769 exists, symlink or not. (Bugzilla 503.)
771 PH/03 The body_linecount and body_zerocount variables are now exported in the
774 PH/04 Added the $dnslist_matched variable.
776 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
777 This means they are set thereafter only if the connection becomes
780 PH/06 Added the client_condition to authenticators so that some can be skipped
781 by clients under certain conditions.
783 PH/07 The error message for a badly-placed control=no_multiline_responses left
784 "_responses" off the end of the name.
786 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
788 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
789 (without spaces) instead of just copying the configuration text.
791 PH/10 Added the /noupdate option to the ratelimit ACL condition.
793 PH/11 Added $max_received_linelength.
795 PH/12 Added +ignore_defer and +include_defer to host lists.
797 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
798 way in which PCRE > 7.0 is built.
800 PH/14 Implemented queue_only_load_latch.
802 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
803 MAIL command. The effect was to mangle the value on 64-bit systems.
805 PH/16 Another patch from the Sieve maintainer.
807 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
809 PH/18 If a system quota error occurred while trying to create the file for
810 a maildir delivery, the message "Mailbox is full" was not appended to the
811 bounce if the delivery eventually timed out. Change 4.67/27 below applied
812 only to a quota excession during the actual writing of the file.
814 PH/19 It seems that peer DN values may contain newlines (and other non-printing
815 characters?) which causes problems in log lines. The DN values are now
816 passed through string_printing() before being added to log lines.
818 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
819 and InterBase are left for another time.)
821 PH/21 Added message_body_newlines option.
823 PH/22 Guard against possible overflow in moan_check_errorcopy().
825 PH/23 POSIX allows open() to be a macro; guard against that.
827 PH/24 If the recipient of an error message contained an @ in the local part
828 (suitably quoted, of course), incorrect values were put in $domain and
829 $local_part during the evaluation of errors_copy.
835 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
836 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
839 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
840 issue a MAIL command.
842 PH/02 In an ACL statement such as
844 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
846 if a client was not listed at all, or was listed with a value other than
847 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
848 the condition was not true (as it should be), so access was not denied.
849 The bug was that the ! inversion was incorrectly passed on to the second
850 item. This has been fixed.
852 PH/03 Added additional dnslists conditions == and =& which are different from
853 = and & when the dns lookup returns more than one IP address.
855 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
856 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
858 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
859 FSYNC, which compiles an option called disable_fsync that allows for
860 bypassing fsync(). The documentation is heavily laced with warnings.
862 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
864 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
865 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
866 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
867 including adding "make clean"; (3) Added -fPIC when compiling the test
868 dynamically loaded module, to get rid of a warning.
870 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
871 message fails, move_frozen_messages = true and ignore_bounce_errors_after
872 = 0s. The bug is otherwise harmless.
874 PH/07 There was a bug in the dovecot authenticator such that the value of
875 $auth1 could be overwritten, and so not correctly preserved, after a
876 successful authentication. This usually meant that the value preserved by
877 the server_setid option was incorrect.
879 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
881 PH/09 Installed PCRE release 7.0.
883 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
884 run for batched SMTP input. It is now run at the start of every message
885 in the batch. While fixing this I discovered that the process information
886 (output by running exiwhat) was not always getting set for -bs and -bS
887 input. This is fixed, and it now also says "batched" for BSMTP.
889 PH/11 Added control=no_pipelining.
891 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
892 patch, slightly modified), and move the expansion of helo_data till after
893 the connection is made in the smtp transport (so it can use these
896 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
898 PH/14 Added log_selector = +pid.
900 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
902 PH/16 Add ${if forany and ${if forall.
904 PH/17 Added dsn_from option to vary the From: line in DSNs.
906 PH/18 Flush SMTP output before performing a callout, unless control =
907 no_callout_flush is set.
909 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
910 was true (the default) a successful delivery failed to delete the retry
911 item, thus causing premature timeout of the address. The bug is now
914 PH/20 Added hosts_avoid_pipelining to the smtp transport.
916 PH/21 Long custom messages for fakedefer and fakereject are now split up
917 into multiline reponses in the same way that messages for "deny" and
918 other ACL rejections are.
920 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
921 with slight modification.
923 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
924 draft, changing the syntax and factoring some duplicate code".
926 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
927 for deliveries of the second and subsequent messages over the same SMTP
930 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
931 ${reduce, with only minor "tidies".
933 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
935 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
936 expansion side effects.
938 PH/27 When a message times out after an over-quota error from an Exim-imposed
939 quota, the bounce message says "mailbox is full". This message was not
940 being given when it was a system quota that was exceeded. It now should
943 MH/03 Made $recipients available in local_scan(). local_scan() already has
944 better access to the recipient list through recipients_list[], but
945 $recipients can be useful in postmaster-provided expansion strings.
947 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
948 in the case of a MAIL command with additional options following the
949 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
950 were accidentally chopped off.
952 PH/29 SMTP synchronization checks are implemented when a command is read -
953 there is a check that no more input is waiting when there shouldn't be
954 any. However, for some commands, a delay in an ACL can mean that it is
955 some time before the response is written. In this time, more input might
956 arrive, invalidly. So now there are extra checks after an ACL has run for
957 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
958 pipelining has not been advertised.
960 PH/30 MH's patch to allow iscntrl() characters to be list separators.
962 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
963 returned in the SMTP response when smtp_return_error_details was false.
966 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
967 instead of the C I/O that was originally supplied, because problems were
970 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
971 Exim which did not show up earlier: it was assuming that a call to
972 SSL_CTX_set_info_callback() might give an error value. In fact, there is
973 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
974 was a macro that became an assignment, so it seemed to work. This has
975 changed to a proper function call with a void return, hence the compile
976 error. Exim's code has been fixed.
978 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
981 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
983 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
985 PH/37 If a message is not accepted after it has had an id assigned (e.g.
986 because it turns out to be too big or there is a timeout) there is no
987 "Completed" line in the log. When some messages of this type were
988 selected by exigrep, they were listed as "not completed". Others were
989 picked up by some special patterns. I have improved the selection
990 criteria to be more general.
992 PH/38 The host_find_failed option in the manualroute router can now be set
993 to "ignore", to completely ignore a host whose IP address cannot be
994 found. If all hosts are ignored, the behaviour is controlled by the new
995 host_all_ignored option.
997 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
998 homing or because of multiple MX records with /mx) generated more than
999 one IP address, and the following item turned out to be the local host,
1000 all the secondary addresses of the first item were incorrectly removed
1001 from the list, along with the local host and any following hosts (which
1002 is what is supposed to happen).
1004 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1005 whoever called Exim into the -H file. In the case of the daemon it was
1006 behaving confusingly. When first started, it used values for whoever
1007 started the daemon, but after a SIGHUP it used the Exim user (because it
1008 calls itself on a restart). I have changed the code so that it now always
1011 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1012 message are rejected with the same error (e.g. no authentication or bad
1013 sender address), and a DATA command is nevertheless sent (as can happen
1014 with PIPELINING or a stupid MUA), the error message that was given to the
1015 RCPT commands is included in the rejection of the DATA command. This is
1016 intended to be helpful for MUAs that show only the final error to their
1019 PH/42 Another patch from the Sieve maintainer.
1021 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1022 Eximstats - Fixed some broken HTML links and added missing column headers
1024 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1025 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1027 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1029 PH/43 Yet another patch from the Sieve maintainer.
1031 PH/44 I found a way to check for a TCP/IP connection going away before sending
1032 the response to the final '.' that terminates a message, but only in the
1033 case where the client has not sent further data following the '.'
1034 (unfortunately, this is allowed). However, in many cases there won't be
1035 any further data because there won't be any more messages to send. A call
1036 to select() can be used: if it shows that the input is "ready", there is
1037 either input waiting, or the socket has been closed. An attempt to read
1038 the next input character can distinguish the two cases. Previously, Exim
1039 would have sent an OK response which the client would never have see.
1040 This could lead to message repetition. This fix should cure that, at
1041 least in a lot of common cases.
1043 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1044 advertised in response to EHLO.
1050 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1051 fixed by 4.65/MH/01 (is this a record?) are fixed:
1053 (i) An empty string was always treated as zero by the numeric comparison
1054 operators. This behaviour has been restored.
1056 (ii) It is documented that the numeric comparison operators always treat
1057 their arguments as decimal numbers. This was broken in that numbers
1058 starting with 0 were being interpreted as octal.
1060 While fixing these problems I realized that there was another issue that
1061 hadn't been noticed. Values of message_size_limit (both the global option
1062 and the transport option) were treated as octal if they started with 0.
1063 The documentation was vague. These values are now always treated as
1064 decimal, and I will make that clear in the documentation.
1070 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1071 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1074 MH/01 Don't check that the operands of numeric comparison operators are
1075 integers when their expansion is in "skipping" mode (fixes bug
1076 introduced by 4.64-PH/07).
1078 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1079 child addresses, Exim now panics and dies. Previously, because the count
1080 is held in a short int, deliveries were likely to be lost. As such a
1081 large number of recipients for a single message is ridiculous
1082 (performance will be very, very poor), I have chosen to impose a limit
1083 rather than extend the field.
1089 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1090 leftover -K file (the existence of which was triggered by #402).
1091 While we were at it, introduced process PID as part of the -K
1092 filename. This should rule out race conditions when creating
1095 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1096 processing considerably. Previous code took too long for large mails,
1097 triggering a timeout which in turn triggers #401.
1099 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1100 in the DK code in transports.c. sendfile() is not really portable,
1101 hence the _LINUX specificness.
1103 TF/01 In the add_headers option to the mail command in an Exim filter,
1104 there was a bug that Exim would claim a syntax error in any
1105 header after the first one which had an odd number of characters
1108 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1109 callout verification, Exim cached a "reject" for the entire domain. This
1110 is correct for most verifications, but it is not correct for a recipient
1111 verification with use_sender or use_postmaster set, because in that case
1112 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1113 case of MAIL FROM:<> rejection from other early rejections (e.g.
1114 rejection of HELO). When verifying a recipient using a non-null MAIL
1115 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1116 Whatever the result of the callout, the value of the domain cache is
1117 left unchanged (for any other kind of callout, getting as far as trying
1118 RCPT means that the domain itself is ok).
1120 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1123 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1124 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1125 session, the daemon ignores ECONNECT errors and logs others; it now
1126 ignores EPIPE as well.
1128 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1129 (quoted-printable decoding).
1131 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1132 later the small subsequent patch to fix an introduced bug.
1134 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1136 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1138 PH/08 An error is now given if message_size_limit is specified negative.
1140 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1141 to be given (somewhat) arbitrary names.
1143 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1146 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1147 miscellaneous code fixes
1149 PH/10 Added the log_reject_target ACL modifier to specify where to log
1152 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1153 hostname. This is wrong, because it relates to the incoming message (and
1154 probably the interface on which it is arriving) and not to the outgoing
1155 callout (which could be using a different interface). This has been
1156 changed to use the value of the helo_data option from the smtp transport
1157 instead - this is what is used when a message is actually being sent. If
1158 there is no remote transport (possible with a router that sets up host
1159 addresses), $smtp_active_hostname is used.
1161 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1162 tweaks were necessary in order to get it to work (see also 21 below):
1163 (a) The code assumed that strncpy() returns a negative number on buffer
1164 overflow, which isn't the case. Replaced with Exim's string_format()
1166 (b) There were several signed/unsigned issues. I just did the minimum
1167 hacking in of casts. There is scope for a larger refactoring.
1168 (c) The code used strcasecmp() which is not a standard C function.
1169 Replaced with Exim's strcmpic() function.
1170 (d) The code set only $1; it now sets $auth1 as well.
1171 (e) A simple test gave the error "authentication client didn't specify
1172 service in request". It would seem that Dovecot has changed its
1173 interface. Fortunately there's a specification; I followed it and
1174 changed what the client sends and it appears to be working now.
1176 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1179 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1180 address A is aliased to B and C, where B exists and C does not. Without
1181 -v the output is "A verified" because verification stops after a
1182 successful redirection if more than one address is generated. However,
1183 with -v the child addresses are also verified. Exim was outputting "A
1184 failed to verify" and then showing the successful verification for C,
1185 with its parentage. It now outputs "B failed to verify", showing B's
1186 parentage before showing the successful verification of C.
1188 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1189 look up a TXT record in a specific list after matching in a combined
1192 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1193 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1194 they consult the DNS. I had assumed they would set it the way they
1195 wanted; and indeed my experiments on Linux seem to show that in some
1196 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1197 To be on the safe side, however, I have now made the interface to
1198 host_find_byname() similar to host_find_bydns(), with an argument
1199 containing the DNS resolver options. The host_find_byname() function now
1200 sets these options at its start, just as host_find_bydns() does. The smtp
1201 transport options dns_qualify_single and dns_search_parents are passed to
1202 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1203 of host_find_byname() use the default settings of RES_DEFNAMES
1204 (qualify_single) but not RES_DNSRCH (search_parents).
1206 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1207 spool_read_header() do less string testing, by means of a preliminary
1208 switch on the second character of optional "-foo" lines. (This is
1209 overdue, caused by the large number of possibilities that now exist.
1210 Originally there were few.) While I was there, I also converted the
1211 str(n)cmp tests so they don't re-test the leading "-" and the first
1212 character, in the hope this might squeeze out yet more improvement.
1214 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1215 flag allowing group syntax was set by the header_syntax check but not
1216 turned off, possible causing trouble later; (2) The flag was not being
1217 set at all for the header_verify test, causing "group"-style headers to
1218 be rejected. I have now set it in this case, and also caused header_
1219 verify to ignore an empty address taken from a group. While doing this, I
1220 came across some other cases where the code for allowing group syntax
1221 while scanning a header line wasn't quite right (mostly, not resetting
1222 the flag correctly in the right place). These bugs could have caused
1223 trouble for malformed header lines. I hope it is now all correct.
1225 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1226 with the "reply" argument non-NULL. The code, however (which originally
1227 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1228 but it didn't always do it. This confused somebody who was copying the
1229 code for some other use. I have removed all the tests.
1231 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1232 feature that was used to support insecure browsers during the U.S. crypto
1233 embargo. It requires special client support, and Exim is probably the
1234 only MTA that supported it -- and would never use it because real RSA is
1235 always available. This code has been removed, because it had the bad
1236 effect of slowing Exim down by computing (never used) parameters for the
1237 RSA_EXPORT functionality.
1239 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1240 authenticator to fail if there's a tab character in the incoming data
1241 (there should never be unless someone is messing about, as it's supposed
1242 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1243 if the connection is using TLS or if the remote IP is the same as the
1244 local IP, and the "valid-client-cert option" if a client certificate has
1247 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1248 authenticators. This can be used for authorization after authentication
1249 succeeds. (In the case of plaintext, it servers for both authentication
1252 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1253 if any retry times were supplied.
1255 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1256 connection, where there is no client IP address to check. In this
1257 situation, the verify now always succeeds.
1259 PH/25 Applied John Jetmore's -Mset patch.
1261 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1263 PH/27 In a string expansion for a processed (not raw) header when multiple
1264 headers of the same name were present, leading whitespace was being
1265 removed from all of them, but trailing whitespace was being removed only
1266 from the last one. Now trailing whitespace is removed from each header
1267 before concatenation. Completely empty headers in a concatenation (as
1268 before) are ignored.
1270 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1271 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1273 PH/29 [Removed. This was a change that I later backed out, and forgot to
1274 correct the ChangeLog entry (that I had efficiently created) before
1275 committing the later change.]
1277 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1278 address errors (4xx response to RCPT) over the same connection as other
1279 messages routed to the same hosts. Such deliveries are always "forced",
1280 so retry times are not inspected. This resulted in far too many retries
1281 for the affected addresses. The effect occurred only when there were more
1282 hosts than the hosts_max_try setting in the smtp transport when it had
1283 the 4xx errors. Those hosts that it had tried were not added to the list
1284 of hosts for which the message was waiting, so if all were tried, there
1285 was no problem. Two fixes have been applied:
1287 (i) If there are any address or message errors in an SMTP delivery, none
1288 of the hosts (tried or untried) are now added to the list of hosts
1289 for which the message is waiting, so the message should not be a
1290 candidate for sending over the same connection that was used for a
1291 successful delivery of some other message. This seems entirely
1292 reasonable: after all the message is NOT "waiting for some host".
1293 This is so "obvious" that I'm not sure why it wasn't done
1294 previously. Hope I haven't missed anything, but it can't do any
1295 harm, as the worst effect is to miss an optimization.
1297 (ii) If, despite (i), such a delivery is accidentally attempted, the
1298 routing retry time is respected, so at least it doesn't keep
1299 hammering the server.
1301 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1302 in ${readsocket because some servers need this prod.
1304 PH/32 Added some extra debug output when updating a wait-xxx database.
1306 PH/33 The hint "could be header name not terminated by colon", which has been
1307 given for certain expansion errors for a long time, was not being given
1308 for the ${if def:h_colon_omitted{... case.
1310 PH/34 The spec says: "With one important exception, whenever a domain list is
1311 being scanned, $domain contains the subject domain." There was at least
1312 one case where this was not true.
1314 PH/35 The error "getsockname() failed: connection reset by peer" was being
1315 written to the panic log as well as the main log, but it isn't really
1316 panic-worthy as it just means the connection died rather early on. I have
1317 removed the panic log writing for the ECONNRESET error when getsockname()
1320 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1321 runs only) independently of the message's sender address. This meant
1322 that, if the 4xx error was in fact related to the sender, a different
1323 message to the same recipient with a different sender could confuse
1324 things. In particualar, this can happen when sending to a greylisting
1325 server, but other circumstances could also provoke similar problems.
1326 I have changed the default so that the retry time for these errors is now
1327 based a combination of the sender and recipient addresses. This change
1328 can be overridden by setting address_retry_include_sender=false in the
1331 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1332 remote server are returned as part of bounce messages. This was not
1333 happening for LMTP over a pipe (the lmtp transport), but now it is the
1334 same for both kinds of LMTP.
1336 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1337 in header lines that were in fact CNAMEs. This is no longer the case.
1339 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1340 and queue runs started by the daemon processed all messages. This has
1341 been fixed so that -R and -S can now usefully be given with -q<time>.
1343 PH/40 Import PCRE release 6.7 (fixes some bugs).
1345 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1347 PH/42 Give an error if -q is specified more than once.
1349 PH/43 Renamed the variables $interface_address and $interface_port as
1350 $received_ip_address and $received_port, to make it clear that these
1351 values apply to message reception, and not to the outgoing interface when
1352 a message is delivered. (The old names remain recognized, of course.)
1354 PH/44 There was no timeout on the connect() call when using a Unix domain
1355 socket in the ${readsocket expansion. There now is.
1357 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1358 be meaningful with "accept".
1360 SC/01 Eximstats V1.43
1361 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1363 SC/02 Eximstats V1.44
1364 Use a glob alias rather than an array ref in the generated
1365 parser. This improves both readability and performance.
1367 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1368 Collect SpamAssassin and rejection statistics.
1369 Don't display local sender or destination tables unless
1370 there is data to show.
1371 Added average volumes into the top table text output.
1373 SC/04 Eximstats V1.46
1374 Collect data on the number of addresses (recipients)
1375 as well as the number of messages.
1377 SC/05 Eximstats V1.47
1378 Added 'Message too big' to the list of mail rejection
1379 reasons (thanks to Marco Gaiarin).
1381 SC/06 Eximstats V1.48
1382 Mainlog lines which have GMT offsets and are too short to
1383 have a flag are now skipped.
1385 SC/07 Eximstats V1.49 (Alain Williams)
1386 Added the -emptyok flag.
1388 SC/08 Eximstats V1.50
1389 Fixes for obtaining the IP address from reject messages.
1391 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1392 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1393 whitesspace changes from 4.64-PH/27
1395 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1398 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1399 are found, allow negative numbers in numeric criteria)
1401 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1403 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1404 to match changes made in 4.64-PH/43
1406 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1408 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1409 "domain" is actually an IP address, and force a failure. This locks out
1410 those revolvers/nameservers that support "A-for-A" lookups, in
1411 contravention of the specifications.
1413 PH/48 When a host name was looked up from an IP address, and the subsequent
1414 forward lookup of the name timed out, the host name was left in
1415 $sender_host_name, contrary to the specification.
1417 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1418 restricted to single-key lookups, Exim was not diagnosing an error if
1419 * or *@ was used with a query-style lookup.
1421 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1423 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1424 long ago, but noone interested enough thought of it. Let's just say that
1425 the "1.1" means that there are some new functions that weren't there at
1426 some point in the past.
1428 PH/51 Error processing for expansion failure of helo_data from an smtp
1429 transport during callout processing was broken.
1431 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1432 tested/used via the -bh/-bhc/-bs options.
1434 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1435 bug, fixed in subsequent PCRE releases).
1437 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1438 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1440 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1446 SC/01 Use a glob alias rather than an array ref in eximstats generated
1447 parser. This improves both readability and performance.
1449 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1450 Don't display local sender or destination tables in eximstats unless
1451 there is data to show.
1452 Added average volumes into the eximstats top table text output.
1454 SC/03 Collect data on the number of addresses (recipients) as well
1455 as the number of messages in eximstats.
1457 TF/01 Correct an error in the documentation for the redirect router. Exim
1458 does (usually) call initgroups() when daemonizing.
1460 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1461 with consistent privilege compared to when running as a daemon.
1463 TF/03 Note in the spec that $authenticated_id is not set for local
1464 submissions from trusted users.
1466 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1467 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1469 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1470 by adding some example configuration directives to the default
1471 configuration file. A little bit of work is required to uncomment the
1472 directives and define how usernames and passwords are checked, but
1473 there is now a framework to start from.
1475 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1476 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1477 without this. I don't know how relevant this is to other LDAP libraries.
1479 PH/02 Add the verb name to the "unknown ACL verb" error.
1481 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1483 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1485 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1486 directory not expanded when it should be if an expanded home directory
1487 was set for the address (which is overridden by the transport).
1489 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1492 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1493 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1494 because it is too late at that time, and has no effect.
1496 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1497 security issue with \' (bugzilla #107). I could not use the
1498 PQescapeStringConn() function, because it needs a PGconn value as one of
1501 PH/08 When testing addresses using -bt, indicate those final addresses that
1502 are duplicates that would not cause an additional delivery. At least one
1503 person was confused, thinking that -bt output corresponded to deliveries.
1504 (Suppressing duplicates isn't a good idea as you lose the information
1505 about possibly different redirections that led to the duplicates.)
1507 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1508 systems where poll() doesn't work, in particular OS X.
1510 PH/10 Added more information to debugging output for retry time not reached.
1512 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1513 operations in malware.c.
1515 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1518 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1519 syslog_duplication set false, log lines that would normally be written
1520 both the the main log and to the reject log were not written to syslog at
1523 PH/14 In the default configuration, change the use of "message" in ACL warn
1524 statements to "add_header".
1526 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1527 not followed by a command (e.g. "seen endif").
1529 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1530 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1533 PH/17 Added extra conditions to the default value of delay_warning_condition
1537 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1538 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1539 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1542 The Auto-Submitted: and various List- headers are standardised, whereas I
1543 don't think Precedence: ever was.
1545 PH/18 Refactored debugging code in route_finduser() to show more information,
1546 in particular, the error code if getpwnam() issues one.
1548 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1549 This is apparently needed in addition to the PH/07 change above to avoid
1550 any possible encoding problems.
1552 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1553 but not after initializing Perl.
1555 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1556 output them only if debugging. By default they are written stderr,
1557 apparently, which is not desirable.
1559 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1562 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1565 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1567 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1568 authenticated or an ident call has been made. Suppress the default
1569 values for $authenticated_id and $authenticated_sender (but permit -oMai
1570 and -oMas) when testing with -bh.
1572 PH/24 Re-jigged the order of the tests in the default configuration so that the
1573 tests for valid domains and recipients precede the DNS black list and CSA
1574 tests, on the grounds that those ones are more expensive.
1576 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1577 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1578 This bug exists in every version of Exim that I still have, right back to
1581 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1582 However, an attempt to turn on case-sensitivity in a regex key by
1583 including (?-i) didn't work because the subject string was already
1584 lowercased, and the effects were non-intuitive. It turns out that a
1585 one-line patch can be used to allow (?-i) to work as expected.
1591 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1592 other effects) broke the use of negated acl sub-conditions.
1594 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1597 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1598 "Deny" causes Exim to reject the incoming connection with a 554 error.
1599 Unfortunately, if there is a major crisis, such as a disk failure,
1600 tcp-wrappers gives "deny", whereas what one would like would be some
1601 kind of temporary error. A kludge has been added to help with this.
1602 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1603 554 error is used if errno is still zero or contains ENOENT (which occurs
1604 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1607 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1609 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1610 errors. Otherwise a message that provokes a temporary error (when other
1611 messages do not) can cause a whole host to time out.
1613 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1614 addresses were routed directly to files or pipes from a redirect router.
1615 File deliveries just didn't batch; pipe deliveries might have suffered
1618 PH/06 A failure to get a lock for a hints database would erroneously always say
1619 "Failed to get write lock", even when it was really a read lock.
1621 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1622 of 0600. This has been changed to use the value of the lockfile_mode
1623 option (which defaults to 0600).
1625 PH/08 Applied small patch from the Sieve maintainer.
1627 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1628 folder from quota calculations, a direct delivery into this folder messed
1629 up the contents of the maildirsize file. This was because the regex was
1630 used only to exclude .Trash (or whatever) when the size of the mailbox
1631 was calculated. There was no check that a delivery was happening into an
1632 excluded directory. This bug has been fixed by ignoring all quota
1633 processing for deliveries into excluded directories.
1635 PH/10 Added the maildirfolder_create_regex option to appendfile.
1641 PH/01 The code for finding all the local interface addresses on a FreeBSD
1642 system running IPv6 was broken. This may well have applied to all BSD
1643 systems, as well as to others that have similar system calls. The broken
1644 code found IPv4 interfaces correctly, but gave incorrect values for the
1645 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1646 that it would not match correctly against @[] and not recognize the IPv6
1649 PH/02 The ipliteral router was not recognizing addresses of the form user@
1650 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1652 PH/03 Added disable_ipv6.
1654 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1655 decoded form, because it is most often used to construct To: headers
1656 lines in autoreplies, and the decoded form may well be syntactically
1657 invalid. However, $reply_address has leading white space removed, and all
1658 newlines turned into spaces so that the autoreply transport does not
1661 PH/05 If group was specified without a user on a router, and no group or user
1662 was specified on a transport, the group from the router was ignored.
1664 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1665 for visible compile-time settings that can be used to change these
1666 numbers, for those that want even more. Backwards compatibility with old
1667 spool files has been maintained. However, going back to a previous Exim
1668 release will lost any variables that are in spool files.
1670 PH/07 Two small changes when running in the test harness: increase delay when
1671 passing a TCP/IP connection to a new process, in case the original
1672 process has to generate a bounce, and remove special handling of
1673 127.0.0.2 (sic), which is no longer necessary.
1675 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1676 be the same on different OS.
1678 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1681 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1682 whether --show-vars was specified or not
1684 JJ/02 exipick: Added support for new ACL variable spool format introduced
1687 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1688 syntactically invalid From: or Reply-to: line, and a filter used this to
1689 generate an autoreply, and therefore failed to obtain an address for the
1690 autoreply, Exim could try to deliver to a non-existent relative file
1691 name, causing unrelated and misleading errors. What now happens is that
1692 it logs this as a hard delivery error, but does not attempt to create a
1695 PH/11 The exinext utility has a -C option for testing purposes, but although
1696 the given file was scanned by exinext itself; it wasn't being passed on
1697 when Exim was called.
1699 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1700 an end-of-file indication when reading a command response.
1702 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1703 compiled. In many other places in Exim, IPv6 addresses are always
1704 recognized, so I have changed this. It also means that IPv4 domain
1705 literals of the form [IPV4:n.n.n.n] are now always recognized.
1707 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1708 used if the router is not running as root, for example, when verifying at
1709 ACL time, or when using -bh. The debugging output from this situation was
1710 non-existent - all you got was a failure to exec. I have made two
1713 (a) Failures to set uid/gid, the current directory, or a process leader
1714 in a subprocess such as that created by queryprogram now generate
1715 suitable debugging ouput when -d is set.
1717 (b) The queryprogram router detects when it is not running as root,
1718 outputs suitable debugging information if -d is set, and then runs
1719 the subprocess without attempting to change uid/gid.
1721 PH/15 Minor change to Makefile for building test_host (undocumented testing
1724 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1725 additional section of a DNS packet that returns MX or SRV records.
1726 Instead, it always explicitly searches for A/AAAA records. This avoids
1727 major problems that occur when a DNS server includes only records of one
1728 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1729 fixed another bug: if SRV records were looked up and the corresponding
1730 address records were *not* found in the additional section, the port
1731 values from the SRV records were lost.
1733 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1734 using the correct key (the original address) when searching the retry
1735 rules in order to find which one to use for generating the retry hint.
1737 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1738 adding the default one. Similarly, if it contains a Reply-To: header, the
1739 errors_reply_to option, if set, is not used.
1741 PH/19 When calculating a retry time, Exim used to measure the "time since
1742 failure" by looking at the "first failed" field in the retry record. Now
1743 it does not use this if it is later than than the arrival time of the
1744 message. Instead it uses the arrival time. This makes for better
1745 behaviour in cases where some deliveries succeed, thus re-setting the
1746 "first failed" field. An example is a quota failure for a huge message
1747 when small messages continue to be delivered. Without this change, the
1748 "time since failure" will always be short, possible causing more frequent
1749 delivery attempts for the huge message than are intended.
1750 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1752 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1753 $1, $2, $3) because the numerical variables can be reset during some
1754 expansion items (e.g. "match"), thereby losing the authentication data.
1756 PH/21 Make -bV show the size of off_t variables so that the test suite can
1757 decide whether to run tests for quotas > 2G.
1759 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1760 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1761 filecount value is greater than 2G or if a quota value is greater than 2G
1762 on a system where the size of off_t is not greater than 4, a panic error
1765 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1766 never match. The debug and -bh output now contains an explicit error
1767 message indicating a malformed IPv4 address or mask.
1769 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1770 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1771 PH/23 above applies.
1773 PH/25 Do not write to syslog when running in the test harness. The only
1774 occasion when this arises is a failure to open the main or panic logs
1775 (for which there is an explicit test).
1777 PH/26 Added the /no_tell option to "control=freeze".
1779 PH/27 If a host name lookup failed very early in a connection, for example, if
1780 the IP address matched host_lookup and the reverse lookup yielded a name
1781 that did not have a forward lookup, an error message of the form "no IP
1782 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1783 could be logged. Now it outputs the IP address instead of "NULL".
1785 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1786 allows the sender and the authenticated sender to be set when
1787 submitting a message from within Exim. Since child_open_exim() is
1788 documented for local_scan(), the new function should be too.
1790 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1791 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1792 results in an empty string is now treated as unset.
1794 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1796 PH/31 Added sender_verify_fail logging option.
1798 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1799 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1800 tidied the source and removed it altogether.
1802 PH/33 When a queue run was abandoned because the load average was too high, a
1803 log line was always written; now it is written only if the queue_run log
1804 selector is set. In addition, the log line for abandonment now contains
1805 information about the queue run such as the pid. This is always present
1806 in "start" and "stop" lines but was omitted from the "abandon" line.
1808 PH/34 Omit spaces between a header name and the colon in the error message that
1809 is given when verify = headers_syntax fails (if there are lots of them,
1810 the message gets confusing).
1812 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1813 names, as there are now some PTR records that contain slashes. This check
1814 is only to protect against broken name servers that fall over on strange
1815 characters, so the fact that it applies to all lookups doesn't matter.
1817 PH/36 Now that the new test suite is complete, we can remove some of the
1818 special code in Exim that was needed for the old test suite. For example,
1819 sorting DNS records because real resolvers return them in an arbitrary
1820 order. The new test suite's fake resolver always returns records in the
1823 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1824 bounces) except when queue_only is set, to avoid logging races between
1825 the different processes.
1827 PH/38 Panic-die if .include specifies a non-absolute path.
1829 PH/39 A tweak to the "H" retry rule from its user.
1831 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1832 a label. They prevented compilation on older perls.
1834 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1835 a warning to be raised on newish perls.
1837 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1838 on queue. Changes to match documented behaviour of showing count of
1839 messages matching specified criteria.
1841 PH/40 Changed the default ident timeout from 30s to 5s.
1843 PH/41 Added support for the use of login_cap features, on those BSD systems
1844 that have them, for controlling the resources used by pipe deliveries.
1846 PH/42 The content-scanning code uses fopen() to create files in which to put
1847 message data. Previously it was not paying any attention to the mode of
1848 the files. Exim runs with umask(0) because the rest of the code creates
1849 files with open(), and sets the required mode explicitly. Thus, these
1850 files were ending up world-writeable. This was not a big issue, because,
1851 being within the spool directory, they were not world-accessible. I have
1852 created a function called modefopen, which takes an additional mode
1853 argument. It sets umask(777), creates the file, chmods it to the required
1854 mode, then resets the umask. All the relevant calls to fopen() in the
1855 content scanning code have been changed to use this function.
1857 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1858 to 24 hours. This avoids potential overflow problems when processing G
1859 and H retry rules. I suspect nobody ever tinkers with this value.
1861 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1863 PH/45 When the plaintext authenticator is running as a client, the server's
1864 challenges are checked to ensure they are valid base64 strings. By
1865 default, the authentication attempt is cancelled if an invalid string is
1866 received. Setting client_ignore_invalid_base64 true ignores these errors.
1867 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1868 they are received. Thus, the responses can be made to depend on the
1869 challenges. If an invalid string is ignored, an empty string is placed in
1872 PH/46 Messages that are created by the autoreply transport now contains a
1873 References: header, in accordance with RFCs 2822 and 3834.
1875 PH/47 Added authenticated_sender_force to the smtp transport.
1877 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1879 PH/49 Installed latest patch from the Sieve maintainer.
1881 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1882 was also set, the appendfile transport was unnecessarily scanning a
1883 directory of message files (e.g. for maildir delivery) to find the count
1884 of files (along with the size), even though it did not need this
1885 information. It now does the scan only if it needs to find either the
1886 size of the count of files.
1888 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1890 PH/52 Two bugs concerned with error handling when the smtp transport is
1893 (i) Exim was not creating retry information for temporary errors given
1894 for individual recipients after the DATA command when the smtp transport
1895 was used in LMTP mode. This meant that they could be retried too
1896 frequently, and not timed out correctly.
1898 (ii) Exim was setting the flag that allows error details to be returned
1899 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1900 recipients that were returned after the DATA command.
1902 PH/53 This is related to PH/52, but is more general: for any failing address,
1903 when detailed error information was permitted to be returned to the
1904 sender, but the error was temporary, then after the final timeout, only
1905 "retry timeout exceeded" was returned. Now it returns the full error as
1906 well as "retry timeout exceeded".
1908 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1909 do this, and (what is worse) MTAs that accept it.
1911 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1912 will now be deprecated.
1914 PH/56 New os.c-cygwin from the Cygwin maintainer.
1916 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1917 formats (previously only available in exim formats via -bpr, -bpru,
1918 and -bpra. Now also available in native and exiqgrep formats)
1920 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1921 with very large, slow to parse queues
1923 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1925 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1927 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1928 responses to authentication challenges, though it was showing the
1929 challenges; (ii) I've removed the CR characters from the debug output for
1932 PH/58 Allow for the insertion of a newline as well as a space when a string
1933 is turned into more than one encoded-word during RFC 2047 encoding. The
1934 Sieve code now uses this.
1936 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1937 data_4xx, lost_connection, tls_required.
1939 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1940 message was being sent as an SMTP response.
1942 PH/61 Add -l and -k options to exicyclog.
1944 PH/62 When verifying, if an address was redirected to one new address, so that
1945 verification continued, and the new address failed or deferred after
1946 having set something in $address_data, the value of $address_data was not
1947 passed back to the ACL. This was different to the case when no
1948 redirection occurred. The value is now passed back in both cases.
1950 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1951 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1952 use login_cap.h, so on its own it isn't the distinguishing feature. The
1953 new name refers directly to the setclassresources() function.
1955 PH/65 Added configuration files for NetBSD3.
1957 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1959 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1960 is preferred over IPv4.
1962 PH/68 The bounce_return_message and bounce_return_body options were not being
1963 honoured for bounces generated during the reception of non-SMTP messages.
1964 In particular, this applied to messages rejected by the ACL. This bug has
1965 been fixed. However, if bounce_return_message is true and bounce_return_
1966 body is false, the headers that are returned for a non-SMTP message
1967 include only those that have been read before the error was detected.
1968 (In the case of an ACL rejection, they have all been read.)
1970 PH/69 The HTML version of the specification is now built in a directory called
1971 spec_html instead of spec.html, because the latter looks like a path with
1972 a MIME-type, and this confuses some software.
1974 PH/70 Catch two compiler warnings in sieve.c.
1976 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1977 function verify_get_ident() calls ip_connect() to connect a socket, but
1978 if the "connect()" function timed out, ip_connect() used to close the
1979 socket. However, verify_get_ident() also closes the socket later, and in
1980 between Exim writes to the log, which may get opened at this point. When
1981 the socket was closed in ip_connect(), the log could get the same file
1982 descriptor number as the socket. This naturally causes chaos. The fix is
1983 not to close the socket in ip_connect(); the socket should be closed by
1984 the function that creates it. There was only one place in the code where
1985 this was missing, in the iplookup router, which I don't think anybody now
1986 uses, but I've fixed it anyway.
1988 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1989 well as to direct DNS lookups. Otherwise the handling of names in host
1990 lists is inconsistent and therefore confusing.
1996 PH/01 Two changes to the default runtime configuration:
1998 (1) Move the checks for relay_from_hosts and authenticated clients from
1999 after to before the (commented out) DNS black list checks.
2001 (2) Add control=submission to the relay_from_hosts and authenticated
2002 clients checks, on the grounds that messages accepted by these
2003 statements are most likely to be submissions.
2005 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2007 (1) Generate an error if the third argument for the ${prvs expansion is
2010 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2013 (3) Reset the variables that are obtained from the first argument of
2014 ${prvscheck and used in the second argument before leaving the code,
2015 because their memory is reclaimed, so using them afterwards may do
2018 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2019 one (it's much easier than Tom thought :-).
2021 (5) Because of (4), we can now allow for the use of $prvscheck_result
2022 inside the third argument.
2024 PH/03 For some reason, the default setting of PATH when running a command from
2025 a pipe transport was just "/usr/bin". I have changed it to
2028 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2029 anything to be listed in the output from -bV.
2031 PH/05 When a filter generated an autoreply, the entire To: header line was
2032 quoted in the delivery log line, like this:
2034 => >A.N.Other <ano@some.domain> <original@ddress> ...
2036 This has been changed so that it extracts the operative address. There
2037 may be more than one such address. If so, they are comma-separated, like
2040 => >ano@some.domain,ona@other.domain <original@ddress> ...
2042 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2043 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2044 not being looked up in the rDNS to get a host name, Exim was showing the
2045 IP address twice in Received: lines, even though the IP addresses were
2046 identical. For example:
2048 Received: from [1.2.3.4] (helo=[1.2.3.4])
2050 However, if the real host name was known, it was omitting the HELO data
2051 if it matched the actual IP address. This has been tidied up so that it
2052 doesn't show the same IP address twice.
2054 PH/07 When both +timestamp and +memory debugging was on, the value given by
2055 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2056 called by the debug printing, thereby overwriting the timestamp buffer.
2057 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2059 PH/08 When the original message was included in an autoreply transport, it
2060 always said "this is a copy of the message, including all the headers",
2061 even if body_only or headers_only was set. It now gives an appropriate
2064 PH/09 Applied a patch from the Sieve maintainer which:
2066 o fixes some comments
2067 o adds the (disabled) notify extension core
2068 o adds some debug output for the result of if/elsif tests
2069 o points to the current vacation draft in the documentation
2070 and documents the missing references header update
2074 o fixes a bug in processing the envelope test (when testing
2075 multiple envelope elements, the last element determinted the
2078 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2079 Electronic Mail") by including:
2081 Auto-submitted: auto-generated
2083 in the messages that it generates (bounce messages and others, such as
2084 warnings). In the case of bounce messages for non-SMTP mesages, there was
2085 also a typo: it was using "Auto_submitted" (underscore instead of
2086 hyphen). Since every message generated by Exim is necessarily in response
2087 to another message, thes have all been changed to:
2089 Auto-Submitted: auto-replied
2091 in accordance with these statements in the RFC:
2093 The auto-replied keyword:
2095 - SHOULD be used on messages sent in direct response to another
2096 message by an automatic process,
2098 - MUST NOT be used on manually-generated messages,
2100 - MAY be used on Delivery Status Notifications (DSNs) and Message
2101 Disposition Notifications (MDNs),
2103 - MUST NOT be used on messages generated by automatic or periodic
2104 processes, except for messages which are automatic responses to
2107 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2108 to the default Received: header definition.
2110 PH/12 Added log selector acl_warn_skipped (default on).
2112 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2113 variables because (a) they are in the wrong storage pool and (b) even if
2114 they were copied, it wouldn't work properly because of the caching.
2116 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2117 checking when decoding. Apparently there are clients that generate
2118 overlong encoded strings. Why am I not surprised?
2120 PH/15 If the first argument of "${if match_address" was not empty, but did not
2121 contain an "@" character, Exim crashed. Now it writes a panic log message
2122 and treats the condition as false.
2124 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2126 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2127 extension "envelope-auth". The code is finished and in agreement with
2128 other implementations, but there is no documentation so far and in fact,
2129 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2130 not changing the active code.
2132 Print executed "if" and "elsif" statements when debugging is used. This
2133 helps a great deal to understand what a filter does.
2135 Document more things not specified clearly in RFC3028. I had all this
2136 sorted out, when out of a sudden new issues came to my mind. Oops."
2138 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2141 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2142 canonical form (as documented). However, after a host name lookup from
2143 the IP address, check_host() was doing a simple string comparison with
2144 addresses acquired from the DNS when checking that the found name did
2145 have the original IP as one of its addresses. Since any found IPv6
2146 addresses are likely to be in abbreviated form, the comparison could
2147 fail. Luckily, there already exists a function for doing the comparison
2148 by converting both addresses to binary, so now that is used instead of
2149 the text comparison.
2151 PH/20 There was another similar case to PH/19, when a complete host name was
2152 given in a host list; looking up its IP address could give an abbreviated
2153 form, whereas the current host's name might or might not be abbreviated.
2154 The same fix has been applied.
2160 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2161 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2164 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2165 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2167 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2169 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2170 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2171 Also, for "Auto-Submitted", treat anything other than "no" as
2172 non-personal, in accordance with RFC 3834. (Previously it treated
2173 anything starting "auto-" as non-personal.)
2175 TF/01 The control=submission/name=... option had a problem with syntax
2176 errors if the name included a slash character. The /name= option
2177 now slurps the rest of the string, so it can include any characters
2178 but it must come last in the list of options (after /sender_retain
2181 PH/05 Some modifications to the interface to the fake nameserver for the new
2189 TK/01 Added the "success_on_redirect" address verification option. See
2190 NewStuff for rationale and an example.
2192 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2194 PH/02 Patch to exigrep to allow it to work on syslog lines.
2196 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2197 fread() to skip over the body file's header line, because in Cygwin the
2198 header line is locked and is inaccessible.
2200 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2201 co-exist for some time) to make it clear that it is the Exim ID that is
2202 referenced, not the Message-ID: header line.
2204 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2205 string_format() function, because snprintf() does not exist on all
2208 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2209 ${stat: expansion item.
2211 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2212 protocol synchronization error", to keep the pedants happy.
2214 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2215 well as for IRIX systems, when gcc is being used. See the host.c source
2218 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2220 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2223 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2224 transport and to the smtp transport in LMTP mode.
2226 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2228 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2229 run a filter in a subprocess. This could lead to confusion in subsequent
2230 lookups in the parent process. There should also be a search_tidyup() at
2231 the end of the subprocess.
2233 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2234 only if the host matched helo_try_verify_hosts, which caused the
2235 verification to occur when the EHLO/HELO command was issued. The ACL just
2236 tested the remembered result. Now, if a previous verification attempt has
2237 not happened, "verify = helo" does it there and then.
2239 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2241 TK/03 Fix log output including CR from clamd.
2243 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2244 exist provoked a memory error which could cause a segfault.
2246 PH/15 Installed PCRE 6.2
2248 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2250 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2251 of the problem. Specifically, suggested +O2 rather than +O1 for the
2254 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2256 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2257 record, the port was not correctly logged when the outgoing_port log
2258 selector was set (it logged the transort's default port).
2260 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2261 fallback_hosts, and "hosts" in the smtp transport.
2263 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2264 host errors such as "Connection refused".
2266 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2267 authentication with radiusclient 0.4.9:
2269 - Error returned from rc_read_config was caught wrongly
2270 - Username/password not passed on to radius server due to wrong length.
2272 The presumption is that some radiusclient API changes for 4.51/PH/17
2273 were not taken care of correctly. The code is still untested by me (my
2274 Linux distribution still has 0.3.2 of radiusclient), but it was
2275 contributed by a Radius user.
2277 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2278 expanding the "port" option of the smtp transport.
2280 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2281 while reading a MIME header. Thanks to Tom Hughes for a patch.
2283 PH/24 Include config.h inside local_scan.h so that configuration settings are
2286 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2287 that in an ACL for RCPT (for example), you can examine exactly what was
2290 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2291 commands, but it was not correctly comparing the address with the actual
2292 client host address. Thus, it would show the EHLO address in Received:
2293 header lines when this was not necessary.
2295 PH/27 Added the % operator to ${eval:}.
2297 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2298 it should be ignoring failures (because with -C, for example, it has lost
2299 privilege). It wasn't ignoring creation failures other than "already
2302 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2305 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2306 because an input line was too long, either on its own, or by virtue of
2307 too many continuations, the temporary file was not being removed, and the
2308 return code was incorrect.
2310 PH/31 Missing "BOOL" in function definition in filtertest.c.
2312 PH/32 Applied Sieve patches from the maintainer.
2314 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2316 PH/33 Added "verify = not_blind".
2318 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2319 Local/Makefile (with some defaults set). These are used in built scripts
2320 such as exicyclog, but they have never been used in the exim_install
2321 script (though there are many overriding facilities there). I have
2322 arranged that the exim_install script now takes note of these two
2325 PH/35 Installed configuration files for Dragonfly.
2327 PH/36 When a locally submitted message by a trusted user did not contain a
2328 From: header, and the sender address was obtained from -f or from an SMTP
2329 MAIL command, and the trusted user did not use -F to supply a sender
2330 name, $originator_name was incorrectly used when constructing a From:
2331 header. Furthermore, $originator_name was used for submission mode
2332 messages from external hosts without From: headers in a similar way,
2333 which is clearly wrong.
2335 PH/37 Added control=suppress_local_fixups.
2337 PH/38 When log_selector = +received_sender was set, and the addition of the
2338 sender made the log line's construction buffer exactly full, or one byte
2339 less than full, an overflow happened when the terminating "\n" was
2342 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2343 when the result of a list match is failure because a DNS lookup failed.
2345 PH/40 RM_COMMAND is now used in the building process.
2347 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2348 the "build-* directories that it finds.
2350 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2351 address in a domain literal was a prefix of an interface address.
2353 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2354 when verifying a sender address, unless rewrite_headers is false.
2356 PH/44 Wrote a long comment about why errors_to addresses are verified as
2357 recipients, not senders.
2359 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2360 the ratelimit ACL was added.
2362 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2364 PH/46 Added extra information about PostgreSQL errors to the error string.
2366 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2367 avoiding the need to install special zones in a real server. This is
2368 backwards compatible; if it can't find the fake resolver, it drops back.
2369 Thus, both old and new test suites can be run.
2371 TF/02 Added util/ratelimit.pl
2373 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2374 clock is set back in time.
2376 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2377 Candler <B.Candler@pobox.com>.
2379 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2380 OK for addresses that are the result of redirections.
2382 PH/48 A number of further additions for the benefit of the new test suite,
2383 including a fake gethostbyname() that interfaces to the fake DNS resolver
2386 TF/06 The fix for widen_domains has also been applied to qualify_single and
2387 search_parents which are the other dnslookup options that can cause
2390 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2393 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2395 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2396 service" if it encountered an MX record with an empty target hostname.
2397 The message is now "an MX or SRV record indicated no SMTP service".
2399 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2400 if the DNS of the sending site is misconfigured. This is quite a
2401 common situation. This change restores the behaviour of treating a
2402 helo verification defer as a failure.
2404 PH/51 If self=fail was set on a router, the bounce message did not include the
2405 actual error message.
2411 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2413 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2414 command itself ended in error, the underlying message about the transport
2415 filter timeout was being overwritten with the pipe command error. Now the
2416 underlying error message should be appended to the second error message.
2418 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2420 PH/02 Reduce the amount of output that "make" produces by default. Full output
2421 can still be requested.
2423 PH/03 The warning log line about a condition test deferring for a "warn" verb
2424 was being output only once per connection, rather than after each
2425 occurrence (because it was using the same function as for successful
2426 "warn" verbs). This seems wrong, so I have changed it.
2428 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2429 it should not have, which might have caused a crash in the right
2430 circumstances, but probably never did.
2432 PH/04 Installed a modified version of Tony Finch's patch to make submission
2433 mode fix the return path as well as the Sender: header line, and to
2434 add a /name= option so that you can make the user's friendly name appear
2437 TF/03 Added the control = fakedefer ACL modifier.
2439 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2440 Mark Lowes for thorough testing.
2442 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2444 TK/03 Merged latest SRS patch from Miles Wilton.
2446 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2447 with the definition in sysexits.h (which is #included earlier).
2448 Fortunately, Exim does not actually use EX_OK. The code used to try to
2449 preserve the sysexits.h value, by assumimg that macro definitions were
2450 scanned for macro replacements. I have been disabused of this notion,
2451 so now the code just undefines EX_OK before #including unistd.h.
2453 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2454 in the smtp transport. When a block could not be written in a single
2455 write() function, the timeout was being re-applied to each part-write.
2456 This seems wrong - if the receiver was accepting one byte at a time it
2457 would take for ever. The timeout is now adjusted when this happens. It
2458 doesn't have to be particularly precise.
2460 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2461 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2463 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2464 without a domain if the check to <postmaster@domain> fails.
2466 SC/01 Eximstats: added -xls and the ability to specify output files
2467 (patch written by Frank Heydlauf).
2469 SC/02 Eximstats: use FileHandles for outputing results.
2471 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2473 SC/04 Eximstats: fixed display of large numbers with -nvr option
2475 SC/05 Eximstats: fixed merging of reports with empty tables.
2477 SC/06 Eximstats: added the -include_original_destination flag
2479 SC/07 Eximstats: removed tabs and trailing whitespace.
2481 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2483 TK/06 MBOX spool code: Add real "From " MBOX separator line
2484 so the .eml file is really in mbox format (even though
2485 most programs do not really care). Patch from Alex Miller.
2487 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2488 The latter is generated from $received_to and is only set if the
2489 message has one envelope recipient. SA can use these headers,
2490 obviously out-of-the-box. Patch from Alex Miller.
2492 PH/08 The ${def test on a variable was returning false if the variable's
2493 value was "0", contrary to what the specification has always said!
2494 The result should be true unless the variable is empty.
2496 PH/09 The syntax error of a character other than { following "${if
2497 def:variable_name" (after optional whitespace) was not being diagnosed.
2498 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2499 accidental colon was present, for example, could give incorrect results.
2501 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2502 result is used (not including appendfile, where other changes are about
2505 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2506 This involved changing a lot of size variables from int to off_t. It
2507 should work with maildirs and everything.
2509 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2510 spamd dying while we are connected to it.
2512 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2515 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2516 function for BDB 4.3.
2518 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2520 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2521 the sources for PCRE have been re-arranged and more files are now
2524 PH/15 The code I had for printing potentially long long variables in PH/11
2525 above was not the best (it lost precision). The length of off_t variables
2526 is now inspected at build time, and an appropriate printing format (%ld
2527 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2528 to be "long long int" or "long int". This is needed for the internal
2529 formatting function string_vformat().
2531 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2532 the configuration file to be ":syslog", then the script "guesses" where
2533 the logs files are, rather than using the compiled in default. In our
2534 case the guess is not the same as the compiled default, so the script
2535 suddenly stopped working when I started to use syslog. The patch checks
2536 to see if log_file_path is "". If so, it attempts to read it from exim
2537 with no configuration file to get the compiled in version, before it
2538 falls back to the previous guessing code."
2540 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2541 implementing BATV in an Exim configuration. See NewStuff for the gory
2544 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2545 Makefile that are specific to HP-UX.
2547 PH/18 If the "use_postmaster" option was set for a recipient callout together
2548 with the "random" option, the postmaster address was used as the MAIL
2549 FROM address for the random test, but not for the subsequent recipient
2550 test. It is now used for both.
2552 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2553 patch removes a few documentation additions to RFC 3028, because the
2554 latest draft now contains them. It adds the new en;ascii-case comparator
2555 and a new error check for 8bit text in MIME parts. Comparator and
2556 require names are now matched exactly. I enabled the subaddress
2557 extension, but it is not well tested yet (read: it works for me)."
2559 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2560 rework some of the code of TK/09 above to avoid the hardwired use of
2561 "%lld" and "long long". Replaced the call to snprintf() with a call to
2564 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2565 records point to non-existent hosts", "retry timeout exceeded", and
2566 "retry time not reached for any host after a long failure period".
2568 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2569 experimental DomainKeys support:
2571 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2572 (2) On an error such as an illegally used "control", the wrong name for
2573 the control was given.
2575 These problems did NOT occur unless DomainKeys support was compiled.
2577 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2579 PH/24 Added ${if match_ip condition.
2581 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2582 databases so that it will be absolutely obvious if a crash occurs in the
2583 DB library. This is a regular occurrence (often caused by mis-matched
2586 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2587 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2588 picked up on a user's system that detects such things. There doesn't seem
2589 to be a gcc warning option for this - only an attribute that has to be
2590 put on the function's prototype. It seems that in Fedora Core 4 they have
2591 set this on a number of new functions. No doubt there will be more in due
2594 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2595 specify a transport. However, if an address that was verified by such a
2596 router was the subject of a callout, Exim crashed because it tried to
2597 read the rcpt_include_affixes from the non-existent transport. Now it
2598 just assumes that the setting of that option is false. This bug was
2599 introduced by 4.51/PH/31.
2601 PH/28 Changed -d+all to exclude +memory, because that information is very
2602 rarely of interest, but it makes the output a lot bigger. People tend to
2603 do -d+all out of habit.
2605 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2606 code in os-type was giving problems when libc.so lives in lib64, like on
2609 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2610 aren't the modern standard, and it seems that some systems' include files
2611 don't always have them. Exim was already checking for some of the newer
2612 ones like T_AAAA, and defining it itself. I've added checks for all the
2613 record types that Exim uses.
2615 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2616 not automatically generating a new one, as it is supposed to. This
2617 prevented TLS from working. If the file did exist, but contained invalid
2618 data, a new version was generated, as expected. It was only the case of a
2619 non-existent file that was broken.
2621 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2622 with a change in libdomainkeys > 0.64.
2624 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2625 from DNS. If the selector record carries the flag, it now has
2626 precedence over the domain-wide flag.
2628 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2630 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2631 the use of a port name that isn't defined in /etc/services) occurred, the
2632 message was deferred as in a normal delivery, and thus remained on the
2633 spool, instead of being failed because of the mua_wrapper setting. This
2634 is now fixed, and I tidied up some of the mua_wrapper messages at the
2637 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2638 the messages in a hash of arrays rather than using individual hashes.
2639 This is a bit cleaner and results in dramatic memory savings, albeit
2640 at a slight CPU cost.
2642 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2643 as requested by Marc Sherman.
2645 SC/10 Eximstats: added histograms for user specified patterns as requested
2648 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2650 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2651 fopen() in the content-scanning modules that did not already have it.
2657 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2658 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2660 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2662 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2664 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2665 to test Sieve filters that use "vacation".
2667 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2668 that changes the way the GnuTLS parameters are stored in the cache file.
2669 The new format can be generated externally. For backward compatibility,
2670 if the data in the cache doesn't make sense, Exim assumes it has read an
2671 old-format file, and it generates new data and writes a new file. This
2672 means that you can't go back to an older release without removing the
2675 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2676 work if there are any delivery delays because "one_time" forces the
2677 parent to be marked "delivered", so its unseen clone is never tried
2678 again. For this reason, Exim now forbids the simultaneous setting of
2681 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2682 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2683 entry if you want to know the details. The fix, however, neglected to
2684 consider the case when local delivery batching is involved. The test for
2685 "previously delivered" was not happening when checking to see if an
2686 address could be batched with a previous (undelivered) one; under
2687 certain circumstances this could lead to multiple deliveries to the same
2690 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2691 in its include files, and this causes problems building Exim.
2693 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2694 header_syntax) but Exim was just ignoring anything given after a slash.
2695 In particular, this caused confusion with an attempt to use "verify =
2696 reverse_host_lookup/defer_ok". An error is now given when options are
2697 supplied for verify items that do not have them. (Maybe reverse_host_
2698 lookup should have a defer_ok option, but that's a different point.)
2700 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2701 defined by RFC 821) to 2048, because there were problems with some AUTH
2702 commands, and RFC 1869 says the size should be increased for extended
2703 SMTP commands that take arguments.
2705 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2708 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2709 "unknown" error; now it says that the functionality isn't in the binary.
2711 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2712 an address' error message when a string expansion fails (syntax or
2713 whatever). Otherwise the password may appear in the log. Following change
2714 PH/42 below, there is no longer a chance of it appearing in a bounce
2717 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2719 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2720 of its addresses was ever tried. (Bugzilla bug #2.)
2722 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2723 the result incorrectly in the debug output. (It correctly added a newline
2724 to what was transported.)
2726 TF/01 Added $received_time.
2728 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2729 commented out examples of how to interface to a virus scanner and to
2730 SpamAssassin. Also added commented examples of av_scanner and
2731 spamd_address settings.
2733 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2734 and controls are allowed in which ACLs. There were a couple of minor
2735 errors. Some of the entries in the conditions table (which is a table of
2736 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2737 negation of where the condition IS allowed.
2739 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2741 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2742 header file does not have a version number, so I've had to invent a new
2743 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2744 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2745 radiusclient), but it was contributed by a Radius user.
2747 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2748 files or directories, for OpenSSL.
2750 PH/19 When an Exim process that is running as root has to create an Exim log
2751 file, it does so in a subprocess that runs as exim:exim so as to get the
2752 ownership right at creation (otherwise, other Exim processes might see
2753 the file with the wrong ownership). There was no test for failure of this
2754 fork() call, which would lead to the process getting stuck as it waited
2755 for a non-existent subprocess. Forks do occasionally fail when resources
2756 run out. I reviewed all the other calls to fork(); they all seem to check
2759 PH/20 When checking for unexpected SMTP input at connect time (before writing
2760 the banner), Exim was not dealing correctly with a non-positive return
2761 from the read() function. If the client had disconnected by this time,
2762 the result was a log entry for a synchronization error with an empty
2763 string after "input=" when read() returned zero. If read() returned -1
2764 (an event I could not check), uninitialized data bytes were printed.
2765 There were reports of junk text (parts of files, etc) appearing after
2768 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2770 PH/22 Added support for macro redefinition, and (re)definition in between
2771 driver and ACL definitions.
2773 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2774 forgetting to use the resulting value; it was using the unexpanded value.
2776 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2777 hadn't been configured. The fix is from Juergen Kreileder, who
2778 understands it better than I do:
2780 "Here's what I see happening with three configured cyrus_sasl
2781 authenticators configured (plain, login, cram-md5):
2783 On startup auth_cyrus_sasl_init() gets called for each of these.
2784 This means three calls to sasl_listmech() without a specified mech_list.
2785 => SASL tests which mechs of all available mechs actually work
2786 => three warnings about OTP not working
2787 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2789 With the patch, sasl_listmech() also gets called three times. But now
2790 SASL's mech_list option is set to the server_mech specified in the the
2791 authenticator. Or in other words, the answer from sasl_listmech()
2792 gets limited to just the mech you're testing for (which is different
2794 => the return list contains just 'plain' or 'login', 'cram-md5' or
2795 nothing depending on the value of ob->server_mech.
2797 I've just tested the patch: Authentication still works fine,
2798 unavailable mechs specified in the exim configuration are still
2799 caught, and the auth.log warnings about OTP are gone."
2801 PH/25 When debugging is enabled, the contents of the command line are added
2802 to the debugging output, even when log_selector=+arguments is not
2805 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2806 answer is "GNU", and only if the return is "GNU/something" is the answer
2809 PH/27 $acl_verify_message is now set immediately after the failure of a
2810 verification in an ACL, and so is available in subsequent modifiers. In
2811 particular, the message can be preserved by coding like this:
2813 warn !verify = sender
2814 set acl_m0 = $acl_verify_message
2816 Previously, $acl_verify_message was set only while expanding "message"
2817 and "log_message" when a very denied access.
2819 PH/28 Modified OS/os.c-Linux with
2821 -#ifndef OS_LOAD_AVERAGE
2822 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2824 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2825 nomenclature these days.)
2827 PH/29 Installed patch from the Sieve maintainer that adds the options
2828 sieve_useraddress and sieve_subaddress to the redirect router.
2830 PH/30 In these circumstances:
2831 . Two addresses routed to the same list of hosts;
2832 . First host does not offer TLS;
2833 . First host accepts first address;
2834 . First host gives temporary error to second address;
2835 . Second host offers TLS and a TLS session is established;
2836 . Second host accepts second address.
2837 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2838 and peerdn, if requested) that were in fact used only for the second
2841 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2842 attention to any local part prefix or suffix that was matched by the
2843 router that accepted the address. It now behaves in the same way as it
2844 does for delivery: the affixes are removed from the local part unless
2845 rcpt_include_affixes is set on the transport.
2847 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2848 timeout during the DATA phase of an incoming message.
2850 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2851 applied a patch sanctioned by the Sieve maintainer.
2853 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2854 the uid or gid is negative. A case of a negative gid caused this to be
2855 noticed. The fix allows for either to be negative.
2857 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2858 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2861 PH/36 The change PH/12 above was broken. Fixed it.
2863 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2864 the grounds that routing the same address twice would always produce the
2865 same answer. This might have been true once, but it is certainly no
2866 longer true now. Routing a child address may depend on the previous
2867 routing that produced that child. Some complicated redirection strategies
2868 went wrong when messages had multiple recipients, and made Exim's
2869 behaviour dependent on the order in which the addresses were given.
2871 I have moved the duplicate checking until after the routing is complete.
2872 Exim scans the addresses that are assigned to local and remote
2873 transports, and removes any duplicates. This means that more work will be
2874 done, as duplicates will always all be routed, but duplicates are
2875 presumably rare, so I don't expect this is of any significance.
2877 For deliveries to pipes, files, and autoreplies, the duplicate checking
2878 still happens during the routing process, since they are not going to be
2881 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2882 It corrects a timeout issue with spamd. This is Ian's comment: "The
2883 background is that sometimes spamd either never reads data from a
2884 connection it has accepted, or it never writes response data. The exiscan
2885 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2886 blindly assumes that writes won't block so it may never time out."
2888 PH/39 Allow G after quota size as well as K and M.
2890 PH/40 The value set for $authenticated_id in an authenticator may not contain
2891 binary zeroes or newlines because the value is written to log lines and
2892 to spool files. There was no check on this. Now the value is run through
2893 the string_printing() function so that such characters are converted to
2894 printable escape sequences.
2896 PH/41 $message_linecount is a new variable that contains the total number of
2897 lines in the message. Compare $body_linecount, which is the count for the
2900 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2901 bounce and delay warning messages, except in certain special cases, which
2904 (a) An SMTP error message from a remote host;
2905 (b) A message specified in a :fail: redirection;
2906 (c) A message specified in a "fail" command in a system filter;
2907 (d) A message specified in a FAIL return from the queryprogram router;
2908 (e) A message specified by the cannot_route_message router option.
2910 In these cases only, Exim does include the error details in bounce and
2911 warning messages. There are also a few cases where bland messages such
2912 as "unrouteable address" or "local delivery error" are given.
2914 PH/43 $value is now also set for the "else" part of a ${run expansion.
2916 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2917 being worked on, but at least Exim now implements the latest version to
2920 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2921 process to complete was treated as a delivery failure, a timeout while
2922 writing the message to the pipe was logged, but erroneously treated as a
2923 successful delivery. Such timeouts include transport filter timeouts. For
2924 consistency with the overall process timeout, these timeouts are now
2925 treated as errors, giving rise to delivery failures by default. However,
2926 there is now a new Boolean option for the pipe transport called
2927 timeout_defer, which, if set TRUE, converts the failures into defers for
2928 both kinds of timeout. A transport filter timeout is now identified in
2931 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2932 systems where "make" and "gmake" are different, calling "gmake" at top
2933 level broke things. I've arranged for the value of $(MAKE) to be passed
2934 from the Makefile to this script so that it can call the same version of
2938 A note about Exim versions 4.44 and 4.50
2939 ----------------------------------------
2941 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2942 changes of various kinds. As a consequence, a big documentation update was
2943 needed. This delayed the release for rather longer than seemed good, especially
2944 in the light of a couple of (minor) security issues. Therefore, the changes
2945 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2946 release. So 4.44 and 4.50 are in effect two different branches that both start
2949 I have left the 4.50 change log unchanged; it contains all the changes since
2950 4.43. The change log for 4.44 is below; many of its items are identical to
2951 those for 4.50. This seems to be the most sensible way to preserve the
2952 historical information.
2958 1. Minor wording change to the doc/README.SIEVE file.
2960 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2961 computation of the current number of files was incorrect.
2963 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2964 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2967 4. Give more explanation in the error message when the command for a transport
2968 filter fails to execute.
2970 5. There are several places where Exim runs a non-Exim command in a
2971 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2972 was being done only for the command run by the queryprogram router. It is
2973 now done for all such subprocesses. The other cases are: ${run, transport
2974 filters, and the commands run by the lmtp and pipe transports.
2976 6. Added CONFIGURE_GROUP build-time option.
2978 7. Some older OS have a limit of 256 on the maximum number of file
2979 descriptors. Exim was using setrlimit() to set 1000 as a large value
2980 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2981 systems. I've change it so that if it can't get 1000, it tries for 256.
2983 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2984 was an oversight, and furthermore, ever since the addition of extra
2985 controls (e.g. 4.43/32), the checks on when to allow different forms of
2986 "control" were broken. There should now be diagnostics for all cases when a
2987 control that does not make sense is encountered.
2989 9. Added the /retain_sender option to "control=submission".
2991 10. $recipients is now available in the predata ACL (oversight).
2993 11. Tidy the search cache before the fork to do a delivery from a message
2994 received from the command line. Otherwise the child will trigger a lookup
2995 failure and thereby defer the delivery if it tries to use (for example) a
2996 cached ldap connection that the parent has called unbind on.
2998 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2999 of $address_data from the recipient verification was clobbered by the
3000 sender verification.
3002 13. The value of address_data from a sender verification is now available in
3003 $sender_address_data in subsequent conditions in the ACL statement.
3005 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3007 15. Added a new option "connect=<time>" to callout options, to set a different
3010 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3011 was its contents. (It was OK if the option was not defined at all.)
3013 17. A "Completed" log line is now written for messages that are removed from
3014 the spool by the -Mrm option.
3016 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3017 information about exactly what failed.
3019 19. Added -dd to debug only the daemon process.
3021 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3022 handles timeouts, both on the server side and network timeouts. Renamed the
3023 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3025 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3026 It is now set to "smtps".
3028 22. $host_address is now set to the target address during the checking of
3029 ignore_target_hosts.
3031 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3032 being passed; this would have caused $sender_host_name to have been used if
3033 matching the list had actually called for a host name (not very likely,
3034 since this list is usually IP addresses). A host name is now passed as
3037 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3038 code that specifies a non-restarting handler (typically sigaction() in
3039 modern systems) in an attempt to fix a rare and obscure crash bug.
3041 25. Narrowed the window for a race in the daemon that could cause it to ignore
3042 SIGCHLD signals. This is not a major problem, because they are used only to
3043 wake it up if nothing else does.
3045 26. A malformed maildirsize file could cause Exim to calculate negative values
3046 for the mailbox size or file count. Odd effects could occur as a result.
3047 The maildirsize information is now recalculated if the size or filecount
3050 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3051 support for a long time. Removed HAVE_SYS_VFS_H.
3053 28. Installed the latest version of exipick from John Jetmore.
3055 29. In an address list, if the pattern was not a regular expression, an empty
3056 subject address (from a bounce message) matched only if the pattern was an
3057 empty string. Non-empty patterns were not even tested. This was the wrong
3058 because it is perfectly reasonable to use an empty address as part of a
3059 database query. An empty address is now tested by patterns that are
3060 lookups. However, all the other forms of pattern expect the subject to
3061 contain a local part and a domain, and therefore, for them, an empty
3062 address still always fails if the pattern is not itself empty.
3064 30. Exim went into a mad DNS loop when attempting to do a callout where the
3065 host was specified on an smtp transport, and looking it up yielded more
3066 than one IP address.
3068 31. Re-factored the code for checking spool and log partition space into a
3069 function that finds that data and another that does the check. The former
3070 is then used to implement four new variables: $spool_space, $log_space,
3071 $spool_inodes, and $log_inodes.
3073 32. The RFC2047 encoding function was originally intended for short strings
3074 such as real names; it was not keeping to the 75-character limit for
3075 encoded words that the RFC imposes. It now respects the limit, and
3076 generates multiple encoded words if necessary. To be on the safe side, I
3077 have increased the buffer size for the ${rfc2047: expansion operator from
3080 33. It is now permitted to omit both strings after an "if" condition; if the
3081 condition is true, the result is "true". As before, when the second string
3082 is omitted, a false condition yields an empty string. This makes it less
3083 cumbersome to write custom ACL and router conditions.
3085 34. Failure to deliver a bounce message always caused it to be frozen, even if
3086 there was an errors_to setting on the router. The errors_to setting is now
3089 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3090 canonical form (fully expanded) before being placed in
3091 $sender_host_address.
3093 36. The table in the code that translates DNS record types into text (T_A to
3094 "A" for instance) was missing entries for NS and CNAME. It is just possible
3095 that this could have caused confusion if both these types were looked up
3096 for the same domain, because the text type is used as part of Exim's
3097 per-process caching. But the chance of anyone hitting this buglet seems
3100 37. The dnsdb lookup has been extended in a number of ways.
3102 (1) There is a new type, "zns", which walks up the domain tree until it
3103 finds some nameserver records. It should be used with care.
3105 (2) There is a new type, "mxh", which is like "mx" except that it returns
3106 just the host names, not the priorities.
3108 (3) It is now possible to give a list of domains (or IP addresses) to be
3109 looked up. The behaviour when one of the lookups defers can be
3110 controlled by a keyword.
3112 (4) It is now possible to specify the separator character for use when
3113 multiple records are returned.
3115 38. The dnslists ACL condition has been extended: it is now possible to supply
3116 a list of IP addresses and/or domains to be looked up in a particular DNS
3119 39. Added log_selector=+queue_time_overall.
3121 40. When running the queue in the test harness, wait just a tad after forking a
3122 delivery process, to get repeatability of debugging output.
3124 41. Include certificate and key file names in error message when GnuTLS fails
3125 to set them up, because the GnuTLS error message doesn't include the name
3126 of the failing file when there is a problem reading it.
3128 42. Allow both -bf and -bF in the same test run.
3130 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3132 44. The "Exiscan patch" is now merged into the mainline Exim source.
3134 45. Sometimes the final signoff response after QUIT could fail to get
3135 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3136 < 0 before doing a fflush(). This bug looks as though it goes back to the
3137 introduction of TLS in release 3.20, but "sometimes" must have been rare
3138 because the tests only now provoked it.
3140 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3141 (this can affect the format of dates).
3143 47. exim_tidydb, when checking for the continued existence of a message for
3144 which it has found a message-specific retry record, was not finding
3145 messages that were in split spool directories. Consequently, it was
3146 deleting retry records that should have stayed in existence.
3148 48. Steve fixed some bugs in eximstats.
3150 49. The SPA authentication driver was not abandoning authentication and moving
3151 on to the next authenticator when an expansion was forced to fail,
3152 contradicting the general specification for all authenticators. Instead it
3153 was generating a temporary error. It now behaves as specified.
3155 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3156 (the order specifies the preference for clients). The order is now AES256,
3157 AES128, 3DES, ARCFOUR128.
3159 51. Small patch to Sieve code - explicitly set From: when generating an
3162 52. Exim crashed if a remote delivery caused a very long error message to be
3163 recorded - for instance if somebody sent an entire SpamAssassin report back
3164 as a large number of 550 error lines. This bug was coincidentally fixed by
3165 increasing the size of one of Exim's internal buffers (big_buffer) that
3166 happened as part of the Exiscan merge. However, to be on the safe side, I
3167 have made the code more robust (and fixed the comments that describe what
3170 53. Now that there can be additional text after "Completed" in log lines (if
3171 the queue_time_overall log selector is set), a one-byte patch to exigrep
3172 was needed to allow it to recognize "Completed" as not the last thing in
3175 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3176 patch that reportedly fixes this has been added. I am not expert enough to
3177 create a test for it. This is what the patch creator wrote:
3179 "I found a little strange behaviour of ldap code when working with
3180 Windows 2003 AD Domain, where users was placed in more than one
3181 Organization Units. When I tried to give exim partial DN, the exit code
3182 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3183 But simultaneously result of request was absolutely normal ldap result,
3184 so I produce this patch..."
3186 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3187 REFERENCE, so I have modified the code to exclude the patch when that macro
3190 55. Some experimental protocols are using DNS PTR records for new purposes. The
3191 keys for these records are domain names, not reversed IP addresses. The
3192 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3193 leaves it alone. Component reversal etc. now happens only for IP addresses.
3196 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3198 57. Double the size of the debug message buffer (to 2048) so that more of very
3199 long debug lines gets shown.
3201 58. The exicyclog utility now does better if the number of log files to keep
3202 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3204 59. Two changes related to the smtp_active_hostname option:
3206 (1) $smtp_active_hostname is now available as a variable.
3207 (2) The default for smtp_banner uses $smtp_active_hostname instead
3208 of $primary_hostname.
3210 60. The host_aton() function is supposed to be passed a string that is known
3211 to be a valid IP address. However, in the case of IPv6 addresses, it was
3212 not checking this. This is a hostage to fortune. Exim now panics and dies
3213 if the condition is not met. A case was found where this could be provoked
3214 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3215 components; fortuitously, this particular loophole had already been fixed
3216 by change 4.50/55 above.
3218 If there are any other similar loopholes, the new check in host_aton()
3219 itself should stop them being exploited. The report I received stated that
3220 data on the command line could provoke the exploit when Exim was running as
3221 exim, but did not say which command line option was involved. All I could
3222 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3223 running as the user.
3226 61. There was a buffer overflow vulnerability in the SPA authentication code
3227 (which came originally from the Samba project). I have added a test to the
3228 spa_base64_to_bits() function which I hope fixes it.
3231 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3232 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3234 63. The daemon start-up calls getloadavg() while still root for those OS that
3235 need the first call to be done as root, but it missed one case: when
3236 deliver_queue_load_max is set with deliver_drop_privilege. This is
3237 necessary for the benefit of the queue runner, because there is no re-exec
3238 when deliver_drop_privilege is set.
3240 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3241 This has been fixed.
3243 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3244 was in use, was not putting the data itself into the right store pool;
3245 consequently, it could be overwritten for a subsequent message in the same
3246 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3249 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3251 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3252 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3253 the function were treating the return as a boolean value, which happened to
3254 work because 0=false and not-0=true, but is not correct code.
3256 68. The host_aton() function was not handling scoped IPv6 addresses (those
3257 with, for example, "%eth0" on the end) correctly.
3259 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3260 negated items (that is, ~something) in unsigned ints. Some compilers
3261 apparently mutter when there is no cast.
3263 70. If an address verification called from an ACL failed, and did not produce a
3264 user-specific message (i.e. there was only a "system" message), nothing was
3265 put in $acl_verify_message. In this situation, it now puts the system
3268 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3269 session; change 4.31/43 added the unwanted input to the log line - except
3270 that it did not do this in the start of session case. It now does.
3272 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3273 This is wrong and can cause the other end to generate a synchronization
3274 error if it is another Exim or anything else that does the synchronization
3275 check. A QUIT command is no longer sent after a timeout.
3277 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3278 during host lookups.
3280 74. The defer_ok option of callout verification was not working if it was used
3281 when verifying addresses in header lines, that is, for this case:
3283 verify = header_sender/callout=defer_ok
3285 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3286 those file descriptors could be used for SMTP connections. If anything
3287 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3288 could be sent to the SMTP client, causing chaos. The daemon now opens
3289 stdin, stdout, and stderr to /dev/null when it puts itself into the
3292 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3293 log by default. The user can override this with suitable Perl magic.
3295 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3296 the log message when discard triggers, was not working for the DATA ACL or
3297 for the non-SMTP ACL.
3299 78. Error message wording change in sieve.c.
3301 79. If smtp_accept_max_per_host was set, the number of connections could be
3302 restricted to fewer than expected, because the daemon was trying to set up
3303 a new connection before checking whether the processes handling previous
3304 connections had finished. The check for completed processes is now done
3305 earlier. On busy systems, this bug wouldn't be noticed because something
3306 else would have woken the daemon, and it would have reaped the completed
3309 80. If a message was submitted locally by a user whose login name contained one
3310 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3311 It caused a spool format error. I have fixed the spool reading code. A
3312 related problem was that the "from" clause in the Received: line became
3313 illegal because of the space(s). It is now covered by ${quote_local_part.
3315 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3318 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3320 83. Patch from Sieve maintainer to fix unterminated string problem in
3321 "vacation" handling.
3323 84. Some minor changes to the Linux configuration files to help with other
3324 OS variants using glibc.
3326 85. One more patch for Sieve to update vacation handling to latest spec.
3329 ----------------------------------------------------
3330 See the note above about the 4.44 and 4.50 releases.
3331 ----------------------------------------------------
3337 1. Change 4.43/35 introduced a bug that caused file counts to be
3338 incorrectly computed when quota_filecount was set in an appendfile
3341 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3342 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3345 3. Give more explanation in the error message when the command for a transport
3346 filter fails to execute.
3348 4. There are several places where Exim runs a non-Exim command in a
3349 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3350 was being done only for the command run by the queryprogram router. It is
3351 now done for all such subprocesses. The other cases are: ${run, transport
3352 filters, and the commands run by the lmtp and pipe transports.
3354 5. Some older OS have a limit of 256 on the maximum number of file
3355 descriptors. Exim was using setrlimit() to set 1000 as a large value
3356 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3357 systems. I've change it so that if it can't get 1000, it tries for 256.
3359 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3360 was an oversight, and furthermore, ever since the addition of extra
3361 controls (e.g. 4.43/32), the checks on when to allow different forms of
3362 "control" were broken. There should now be diagnostics for all cases when a
3363 control that does not make sense is encountered.
3365 7. $recipients is now available in the predata ACL (oversight).
3367 8. Tidy the search cache before the fork to do a delivery from a message
3368 received from the command line. Otherwise the child will trigger a lookup
3369 failure and thereby defer the delivery if it tries to use (for example) a
3370 cached ldap connection that the parent has called unbind on.
3372 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3373 of $address_data from the recipient verification was clobbered by the
3374 sender verification.
3376 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3377 was its contents. (It was OK if the option was not defined at all.)
3379 11. A "Completed" log line is now written for messages that are removed from
3380 the spool by the -Mrm option.
3382 12. $host_address is now set to the target address during the checking of
3383 ignore_target_hosts.
3385 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3386 being passed; this would have caused $sender_host_name to have been used if
3387 matching the list had actually called for a host name (not very likely,
3388 since this list is usually IP addresses). A host name is now passed as
3391 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3392 code that specifies a non-restarting handler (typically sigaction() in
3393 modern systems) in an attempt to fix a rare and obscure crash bug.
3395 15. Narrowed the window for a race in the daemon that could cause it to ignore
3396 SIGCHLD signals. This is not a major problem, because they are used only to
3397 wake it up if nothing else does.
3399 16. A malformed maildirsize file could cause Exim to calculate negative values
3400 for the mailbox size or file count. Odd effects could occur as a result.
3401 The maildirsize information is now recalculated if the size or filecount
3404 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3405 support for a long time. Removed HAVE_SYS_VFS_H.
3407 18. Updated exipick to current release from John Jetmore.
3409 19. Allow an empty sender to be matched against a lookup in an address list.
3410 Previously the only cases considered were a regular expression, or an
3413 20. Exim went into a mad DNS lookup loop when doing a callout where the
3414 host was specified on the transport, if the DNS lookup yielded more than
3417 21. The RFC2047 encoding function was originally intended for short strings
3418 such as real names; it was not keeping to the 75-character limit for
3419 encoded words that the RFC imposes. It now respects the limit, and
3420 generates multiple encoded words if necessary. To be on the safe side, I
3421 have increased the buffer size for the ${rfc2047: expansion operator from
3424 22. Failure to deliver a bounce message always caused it to be frozen, even if
3425 there was an errors_to setting on the router. The errors_to setting is now
3428 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3429 canonical form (fully expanded) before being placed in
3430 $sender_host_address.
3432 24. Updated eximstats to version 1.33
3434 25. Include certificate and key file names in error message when GnuTLS fails
3435 to set them up, because the GnuTLS error message doesn't include the name
3436 of the failing file when there is a problem reading it.
3438 26. Expand error message when OpenSSL has problems setting up cert/key files.
3441 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3442 (this can affect the format of dates).
3444 28. exim_tidydb, when checking for the continued existence of a message for
3445 which it has found a message-specific retry record, was not finding
3446 messages that were in split spool directories. Consequently, it was
3447 deleting retry records that should have stayed in existence.
3449 29. eximstats updated to version 1.35
3450 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3451 1.35 - bugfix such that pie charts by volume are generated correctly
3453 30. The SPA authentication driver was not abandoning authentication and moving
3454 on to the next authenticator when an expansion was forced to fail,
3455 contradicting the general specification for all authenticators. Instead it
3456 was generating a temporary error. It now behaves as specified.
3458 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3459 (the order specifies the preference for clients). The order is now AES256,
3460 AES128, 3DES, ARCFOUR128.
3462 31. Small patch to Sieve code - explicitly set From: when generating an
3465 32. Exim crashed if a remote delivery caused a very long error message to be
3466 recorded - for instance if somebody sent an entire SpamAssassin report back
3467 as a large number of 550 error lines. This bug was coincidentally fixed by
3468 increasing the size of one of Exim's internal buffers (big_buffer) that
3469 happened as part of the Exiscan merge. However, to be on the safe side, I
3470 have made the code more robust (and fixed the comments that describe what
3473 33. Some experimental protocols are using DNS PTR records for new purposes. The
3474 keys for these records are domain names, not reversed IP addresses. The
3475 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3476 leaves it alone. Component reversal etc. now happens only for IP addresses.
3479 34. The host_aton() function is supposed to be passed a string that is known
3480 to be a valid IP address. However, in the case of IPv6 addresses, it was
3481 not checking this. This is a hostage to fortune. Exim now panics and dies
3482 if the condition is not met. A case was found where this could be provoked
3483 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3484 components; fortuitously, this particular loophole had already been fixed
3485 by change 4.50/55 or 4.44/33 above.
3487 If there are any other similar loopholes, the new check in host_aton()
3488 itself should stop them being exploited. The report I received stated that
3489 data on the command line could provoke the exploit when Exim was running as
3490 exim, but did not say which command line option was involved. All I could
3491 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3492 running as the user.
3495 35. There was a buffer overflow vulnerability in the SPA authentication code
3496 (which came originally from the Samba project). I have added a test to the
3497 spa_base64_to_bits() function which I hope fixes it.
3500 36. The daemon start-up calls getloadavg() while still root for those OS that
3501 need the first call to be done as root, but it missed one case: when
3502 deliver_queue_load_max is set with deliver_drop_privilege. This is
3503 necessary for the benefit of the queue runner, because there is no re-exec
3504 when deliver_drop_privilege is set.
3506 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3507 was in use, was not putting the data itself into the right store pool;
3508 consequently, it could be overwritten for a subsequent message in the same
3509 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3512 38. Sometimes the final signoff response after QUIT could fail to get
3513 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3514 < 0 before doing a fflush(). This bug looks as though it goes back to the
3515 introduction of TLS in release 3.20, but "sometimes" must have been rare
3516 because the tests only now provoked it.
3522 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3523 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3524 Now it returns an integer. A number of calls were still expecting a T/F
3525 return. Fortuitously, in all cases, the tests worked in OK situations,
3526 which is the norm. However, things would have gone wrong on any write
3527 failures on the smtp file descriptor. This function is used when sending
3528 messages over SMTP and also when doing verify callouts.
3530 2. When Exim is called to do synchronous delivery of a locally submitted
3531 message (the -odf or -odi options), it no longer closes stderr before doing
3534 3. Implemented the mua_wrapper option.
3536 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3538 5. Implemented the functions header_remove(), header_testname(),
3539 header_add_at_position(), and receive_remove_recipient(), and exported them
3542 6. If an ACL "warn" statement specified the addition of headers, Exim already
3543 inserted X-ACL-Warn: at the start if there was no header name. However, it
3544 was not making this test for the second and subsequent header lines if
3545 there were newlines in the string. This meant that an invalid header could
3546 be inserted if Exim was badly configured.
3548 7. Allow an ACL "warn" statement to add header lines at the start or after all
3549 the Received: headers, as well as at the end.
3551 8. Added the rcpt_4xx retry error code.
3553 9. Added postmaster_mailfrom=xxx to callout verification option.
3555 10. Added mailfrom=xxxx to the callout verification option, for verify=
3558 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3559 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3561 12. Inserted some casts to stop certain compilers warning when using pointer
3562 differences as field lengths or precisions in printf-type calls (mostly
3563 affecting debugging statements).
3565 13. Added optional readline() support for -be (dynamically loaded).
3567 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3568 same clock tick as a message's arrival, so that its received time was the
3569 same as the "first fail" time on the retry record, and that message
3570 remained on the queue past the ultimate address timeout, every queue runner
3571 would try a delivery (because it was past the ultimate address timeout) but
3572 after another failure, the ultimate address timeout, which should have then
3573 bounced the address, did not kick in. This was a "< instead of <=" error;
3574 in most cases the first failure would have been in the next clock tick
3575 after the received time, and all would be well.
3577 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3578 being recognized when the domain list was tested by the match_domain
3579 condition in an expansion string.
3581 16. Added the ${str2b64: operator.
3583 17. Exim was always calling setrlimit() to set a large limit for the number of
3584 processes, without checking whether the existing limit was already
3585 adequate. (It did check for the limit on file descriptors.) Furthermore,
3586 errors from getrlimit() and setrlimit() were being ignored. Now they are
3587 logged to the main and panic logs, but Exim does carry on, to try to do its
3588 job under whatever limits there are.
3590 18. Imported PCRE 5.0.
3592 19. Trivial typo in log message " temporarily refused connection" (the leading
3595 20. If the log selector return_path_on_delivery was set and an address was
3596 redirected to /dev/null, the delivery process crashed because it assumed
3597 that a return path would always be set for a "successful" delivery. In this
3598 case, the whole delivery is bypassed as an optimization, and therefore no
3601 21. Internal re-arrangement: the function for sending a challenge and reading
3602 a response while authentication was assuming a zero-terminated challenge
3603 string. It's now changed to take a pointer and a length, to allow for
3604 binary data in such strings.
3606 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3608 23. Exim was not respecting finduser_retries when seeking the login of the
3609 uid under which it was called; it was always trying 10 times. (The default
3610 setting of finduser_retries is zero.) Also, it was sleeping after the final
3611 failure, which is pointless.
3613 24. Implemented tls_on_connect_ports.
3615 25. Implemented acl_smtp_predata.
3617 26. If the domain in control=submission is set empty, Exim assumes that the
3618 authenticated id is a complete email address when it generates From: or
3619 Sender: header lines.
3621 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3622 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3623 chown and chgrp in /bin and hostname in /usr/bin.
3625 28. Exim was keeping the "process log" file open after each use, just as it
3626 does for the main log. This opens the possibility of it remaining open for
3627 long periods when the USR1 signal hits a daemon. Occasional processlog
3628 errors were reported, that could have been caused by this. Anyway, it seems
3629 much more sensible not to leave this file open at all, so that is what now
3632 29. The long-running daemon process does not normally write to the log once it
3633 has entered its main loop, and it closes the log before doing so. This is
3634 so that log files can straightforwardly be renamed and moved. However,
3635 there are a couple of unusual error situations where the daemon does write
3636 log entries, and I had neglected to close the log afterwards.
3638 30. The text of an SMTP error response that was received during a remote
3639 delivery was being truncated at 512 bytes. This is too short for some of
3640 the long messages that one sometimes sees. I've increased the limit to
3643 31. It is now possible to make retry rules that apply only when a message has a
3644 specific sender, in particular, an empty sender.
3646 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3647 it possible to be selective about when SMTP synchronization is enforced.
3649 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3651 32. Implemented hosts_connection_nolog.
3653 33. Added an ACL for QUIT.
3655 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3658 35. Added mailbox_size and mailbox_filecount to appendfile.
3660 36. Added control = no_multiline_responses to ACLs.
3662 37. There was a bug in the logic of the code that waits for the clock to tick
3663 in the case where the clock went backwards by a substantial amount such
3664 that the microsecond fraction of "now" was more than the microsecond
3665 fraction of "then" (but the whole seconds number was less).
3667 38. Added support for the libradius Radius client library this is found on
3668 FreeBSD (previously only the radiusclient library was supported).
3674 1. When certain lookups returned multiple values in the form name=value, the
3675 quoting of the values was not always being done properly. Specifically:
3676 (a) If the value started with a double quote, but contained no whitespace,
3678 (b) If the value contained whitespace other than a space character (i.e.
3679 tabs or newlines or carriage returns) it was not quoted.
3680 This fix has been applied to the mysql and pgsql lookups by writing a
3681 separate quoting function and calling it from the lookup code. The fix
3682 should probably also be applied to nisplus, ibase and oracle lookups, but
3683 since I cannot test any of those, I have not disturbed their existing code.
3685 2. A hit in the callout cache for a specific address caused a log line with no
3686 reason for rejecting RCPT. Now it says "Previous (cached) callout
3687 verification failure".
3689 3. There was an off-by-one bug in the queryprogram router. An over-long
3690 return line was truncated at 256 instead of 255 characters, thereby
3691 overflowing its buffer with the terminating zero. As well as fixing this, I
3692 have increased the buffer size to 1024 (and made a note to document this).
3694 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3695 when Exim is waiting for an SMTP response from a remote server, Exim
3696 restarts its select() call on the socket, thereby resetting its timeout.
3697 This is not a problem when such interrupts are rare. Somebody set up a cron
3698 job to run exiwhat every 2 minutes, which is less than the normal select()
3699 timeout (5 or 10 minutes). This meant that the select() timeout never
3700 kicked in because it was always reset. I have fixed this by comparing the
3701 time when an interrupt arrives with the time at the start of the first call
3702 to select(). If more time than the timeout has elapsed, the interrupt is
3703 treated as a timeout.
3705 5. Some internal re-factoring in preparation for the addition of Sieve
3706 extensions (by MH). In particular, the "personal" test is moved to a
3707 separate function, and given an option for scanning Cc: and Bcc: (which is
3708 not set for Exim filters).
3710 6. When Exim created an email address using the login of the caller as the
3711 local part (e.g. when creating a From: or Sender: header line), it was not
3712 quoting the local part when it contained special characters such as @.
3714 7. Installed new OpenBSD configuration files.
3716 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3717 try to make them clearer.
3719 9. Callout options, other than the timeout value, were being ignored when
3720 verifying sender addresses in header lines. For example, when using
3722 verify = header_sender/callout=no_cache
3724 the cache was (incorrectly) being used.
3726 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3727 only the Cygwin environment.
3729 11. When return_path_on_delivery was set as a log selector, if different remote
3730 addresses in the same message used different return paths and parallel
3731 remote delivery occurred, the wrong values would sometimes be logged.
3732 (Whenever a remote delivery process finished, the return path value from
3733 the most recently started remote delivery process was logged.)
3735 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3736 lines when AUTH and/or TLS are in use. This is the "received protocol"
3737 field. Exim used to use "asmtp" for authenticated SMTP, without any
3738 indication (in the protocol name) for TLS use. Now it follows the RFC and
3739 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3740 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3741 names appear in log lines as well as in Received: header lines.
3743 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3744 extensions, and comparison tests, and to fix some bugs.
3746 14. Changes to the "personal" filter test:
3748 (1) The test was buggy in that it was just doing the equivalent of
3749 "contains" tests on header lines. For example, if a user's address was
3750 anne@some.where, the "personal" test would incorrectly be true for
3752 To: susanne@some.where
3754 This test is now done by extracting each address from the header in turn,
3755 and checking the entire address. Other tests that are part of "personal"
3756 are now done using regular expressions (for example, to check local parts
3757 of addresses in From: header lines).
3759 (2) The list of non-personal local parts in From: addresses has been
3760 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3761 taken from the Sieve specification recommendations.
3763 (3) If the message contains any header line starting with "List-" it is
3764 treated as non-personal.
3766 (4) The test for "circular" in the Subject: header line has been removed
3767 because it now seems ill-conceived.
3769 15. Minor typos in src/EDITME comments corrected.
3771 16. Installed latest exipick from John Jetmore.
3773 17. If headers_add on a router specified a text string that was too long for
3774 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3775 of string_sprintf() is now avoided.
3777 18. $message_body_size was not set (it was always zero) when running the DATA
3778 ACL and the local_scan() function.
3780 19. For the "mail" command in an Exim filter, no default was being set for
3781 the once_repeat time, causing a random time value to be used if "once" was
3782 specified. (If the value happened to be <= 0, no repeat happened.) The
3783 default is now 0s, meaning "never repeat". The "vacation" command was OK
3784 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3785 (I found it when inspecting the code).
3787 20. There is now an overall timeout for performing a callout verification. It
3788 defaults to 4 times the callout timeout, which applies to individual SMTP
3789 commands during the callout. The overall timeout applies when there is more
3790 than one host that can be tried. The timeout is checked before trying the
3791 next host. This prevents very long delays if there are a large number of
3792 hosts and all are timing out (e.g. when the network connections are timing
3793 out). The value of the overall timeout can be changed by specifying an
3794 additional sub-option for "callout", called "maxwait". For example:
3796 verify = sender/callout=5s,maxwait=20s
3798 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3799 to the end before writing, but this should make it even safer).
3801 22. Exim was forgetting that it had advertised PIPELINING for the second and
3802 subsequent messages on an SMTP connection. It was also not resetting its
3803 memory on STARTTLS and an internal HELO.
3805 23. When Exim logs an SMTP synchronization error within a session, it now
3806 records whether PIPELINING has been advertised or not.
3808 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3809 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3810 rather than long int.
3812 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3814 26. Added the never_mail option to autoreply.
3820 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3821 crash if the getsockname() call failed; this can happen if a connection is
3822 closed very soon after it is established. The problem was simply in the
3823 order in which certain operations were done, causing Exim to try to write
3824 to the SMTP stream before it had set up the file descriptor. The bug has
3825 been fixed by making things happen in the correct order.
3831 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3832 before the connection was closed, thus losing the rejection response.
3834 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3835 some early Solaris releases, but causes trouble in current releases where
3836 socklen_t is defined.
3838 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3841 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3844 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3845 that releasing the top of it at the end releases what was used for sub-
3846 expansions (unless the block got too big). However, discard this block if
3847 the first thing is a variable or header, so that we can use its block when
3848 it is dynamic (useful for very large $message_headers, for example).
3850 6. Lookups now cache *every* query, not just the most recent. A new, separate
3851 store pool is used for this. It can be recovered when all lookup caches are
3852 flushed. Lookups now release memory at the end of their result strings.
3853 This has involved some general refactoring of the lookup sources.
3855 7. Some code has been added to the store_xxx() functions to reduce the amount
3856 of flapping under certain conditions.
3858 8. log_incoming_interface used to affect only the <= reception log lines. Now
3859 it causes the local interface and port to be added to several more SMTP log
3860 lines, for example "SMTP connection from", and rejection lines.
3862 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3864 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3866 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3867 could be overwritten at the end of the current message (or the start of a
3868 new message if it was set in a HELO ACL). The value is now preserved for
3869 the duration of the SMTP connection.
3871 12. If a transport had a headers_rewrite setting, and a matching header line
3872 contained an unqualified address, that address was qualified, even if it
3873 did not match any rewriting rules. The underlying bug was that the values
3874 of the flags that permit the existence of unqualified sender and recipient
3875 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3876 non-local messages, and by -bnq for local messages) were not being
3877 preserved with the message after it was received.
3879 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3880 "next input=" as part of the text comprising the host identity instead of
3881 the correct text. The code was using the same buffer for two different
3882 strings. However, depending on which order the printing function evaluated
3883 its arguments, the bug did not always show up. Under Linux, for example, my
3884 test suite worked just fine.
3886 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3887 doesn't work with some older versions of Perl. It has been changed to "my",
3888 which in any case is probably the better facility to use.
3890 15. A really picky compiler found some instances of statements for creating
3891 error messages that either had too many or two few arguments for the format
3894 16. The size of the buffer for calls to the DNS resolver has been increased
3895 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3896 for addresses that have a lot of PTR records. This alleviates a problem; it
3897 does not fully solve it.
3899 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3900 buffer now truncates the list and logs the incident, which is the same
3901 action as happens when Exim is looking up a host name and its aliases.
3902 Previously in this situation something unpredictable would happen;
3903 sometimes it was "internal error: store_reset failed".
3905 18. If a server dropped the connection unexpectedly when an Exim client was
3906 using GnuTLS and trying to read a response, the client delivery process
3907 crashed while trying to generate an error log message.
3909 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3910 string, for example:
3912 warn message = H1: something\nH2: something
3914 the text was added as a single header line from Exim's point of view
3915 though it ended up OK in the delivered message. However, searching for the
3916 second and subsequent header lines using $h_h2: did not work. This has been
3917 fixed. Similarly, if a system filter added multiple headers in this way,
3918 the routers could not see them.
3920 20. Expanded the error message when iplsearch is called with an invalid key to
3921 suggest using net-iplsearch in a host list.
3923 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3924 ACLs are no longer actually imposed (and a message to that effect is
3927 22. If a "gecos" field in a passwd entry contained escaped characters, in
3928 particular, if it contained a \" sequence, Exim got it wrong when building
3929 a From: or a Sender: header from that name. A second bug also caused
3930 incorrect handling when an unquoted " was present following a character
3931 that needed quoting.
3933 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3934 was not being matched caselessly.
3936 24. Arranged for all hyphens in the exim.8 source to be escaped with
3939 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3940 properly. Recipient callout cache records were still being keyed to include
3941 the sender, even when use_sender was set false. This led to far more
3942 callouts that were necessary. The sender is no longer included in the key
3943 when use_sender is false.
3945 26. Added "control = submission" modifier to ACLs.
3947 27. Added the ${base62d: operator to decode base 62 numbers.
3949 28. dnsdb lookups can now access SRV records.
3951 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3952 the configuration file.
3954 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3955 (-v) mode. This makes the output for a verbose queue run more intelligible.
3957 31. Added a use_postmaster feature to recipient callouts.
3959 32. Added the $body_zerocount variable, containing the number of binary zero
3960 bytes in the message body.
3962 33. The time of last modification of the "new" subdirectory is now used as the
3963 "mailbox time last read" when there is a quota error for a maildir
3966 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3968 35. Added +ignore_unknown as a special item in host lists.
3970 36. Code for decoding IPv6 addresses in host lists is now included, even if
3971 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3972 address was recognized as an IP address, but was then not correctly decoded
3973 into binary, causing unexpected and incorrect effects when compared with
3980 1. Very minor rewording of debugging text in manualroute to say "list of
3981 hosts" instead of "hostlist".
3983 2. If verify=header_syntax was set, and a header line with an unqualified
3984 address (no domain) and a large number of spaces between the end of the
3985 name and the colon was received, the reception process suffered a buffer
3986 overflow, and (when I tested it) crashed. This was caused by some obsolete
3987 code that should have been removed. The fix is to remove it!
3989 3. When running in the test harness, delay a bit after writing a bounce
3990 message to get a bit more predictability in the log output.
3992 4. Added a call to search_tidyup() just before forking a reception process. In
3993 theory, someone could use a lookup in the expansion of smtp_accept_max_
3994 per_host which, without the tidyup, could leave open a database connection.
3996 5. Added the variables $recipient_data and $sender_data which get set from a
3997 lookup success in an ACL "recipients" or "senders" condition, or a router
3998 "senders" option, similar to $domain_data and $local_part_data.
4000 6. Moved the writing of debug_print from before to after the "senders" test
4003 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4004 problems for message scanning, either using a data ACL, or using
4005 local_scan() because the Received: header was not generated till after they
4006 were called (in order to set the time as the time of reception completion).
4007 I have revised the way this works. The header is now generated after the
4008 body is received, but before the ACL or local_scan() are called. After they
4009 are run, the timestamp in the header is updated.
4015 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4016 before starting a queue runner without re-exec. This happened only when
4017 deliver_drop_privilege was set or when the Exim user was set to root. The
4018 effect of the bug was that timeouts during subsequent deliveries caused
4019 crashes instead of being properly handled. The handler is now left at its
4020 default (and expected) setting.
4022 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4023 message, again when deliver_drop_privilege is set or Exim is run as root.
4024 The bug described in (1) was not present in this case, but the tidying up
4025 of the other signals was missing. I have made the two cases consistent.
4027 3. The ignore_target_hosts setting on a manualroute router was being ignored
4028 for hosts that were looked up using the /MX notation.
4030 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4033 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4034 operated on the sender address. After changing the $sender_address to <>
4035 for the sender address verify, Exim was re-instated it as the original
4036 (before rewriting) address, but remembering that it had rewritten it, so it
4037 wasn't rewriting it again. This bug also had the effect of breaking the
4038 sender address verification caching when the sender address was rewritten.
4040 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4041 This has been changed so that if the ip literal address matches
4042 ignore_target_hosts, the router declines.
4044 7. Added expansion conditions match_domain, match_address, and match_local_
4045 part (NOT match_host).
4047 8. The placeholder for the Received: header didn't have a length field set.
4049 9. Added code to Exim itself and to exim_lock to test for a specific race
4050 condition that could lead to file corruption when using MBX delivery. The
4051 issue is with the lockfile that is created in /tmp. If this file is removed
4052 after a process has opened it but before that process has acquired a lock,
4053 there is the potential for a second process to recreate the file and also
4054 acquire a lock. This could lead to two Exim processes writing to the file
4055 at the same time. The added code performs the same test as UW imapd; it
4056 checks after acquiring the lock that its file descriptor still refers to
4057 the same named file.
4059 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4060 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4063 11. Added the smtp_active_hostname option. If used, this will typically be made
4064 to depend on the incoming interface address. Because $interface_address is
4065 not set up until the daemon has forked a reception process, error responses
4066 that can happen earlier (such as "too many connections") no longer contain
4069 12. If an expansion in a condition on a "warn" statement fails because a lookup
4070 defers, the "warn" statement is abandoned, and the next ACL statement is
4071 processed. Previously this caused the whole ACL to be aborted.
4073 13. Added the iplsearch lookup type.
4075 14. Added ident_timeout as a log selector.
4077 15. Added tls_certificate_verified as a log selector.
4079 16. Added a global option tls_require_ciphers (compare the smtp transport
4080 option of the same name). This controls incoming TLS connections.
4082 17. I finally figured out how to make tls_require_ciphers do a similar thing
4083 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4084 before starting the TLS session.
4086 18. Tabs are now shown as \t in -bP output.
4088 19. If the log selector return_path_on_delivery was set, Exim crashed when
4089 bouncing a message because it had too many Received: header lines.
4091 20. If two routers both had headers_remove settings, and the first one included
4092 a superfluous trailing colon, the final name in the first list and the
4093 first name in the second list were incorrectly joined into one item (with a
4094 colon in the middle).
4100 1. Added -C and -D options to the exinext utility, mainly to make it easier
4101 to include in the automated testing, but these could be helpful when
4102 multiple configurations are in use.
4104 2. The exinext utility was not formatting the output nicely when there was
4105 an alternate port involved in the retry record key, nor when there was a
4106 message id as well (for retries that were specific to a specific message
4107 and a specific host). It was also confused by IPv6 addresses, because of
4108 the additional colons they contain. I have fixed the IPv4 problem, and
4109 patched it up to do a reasonable job for IPv6.
4111 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4112 delivery, the log line now contains "pipelined" if PIPELINING was used.
4114 4. An SMTP transport process used to panic and die if the bind() call to set
4115 an explicit outgoing interface failed. This has been changed; it is now
4116 treated in the same way as a connect() failure.
4118 5. A reference to $sender_host_name in the part of a conditional expansion
4119 that was being skipped was still causing a DNS lookup. This no longer
4122 6. The def: expansion condition was not recognizing references to header lines
4123 that used bh_ and bheader_.
4125 7. Added the _cache feature to named lists.
4127 8. The code for checking quota_filecount in the appendfile transport was
4128 allowing one more file than it should have been.
4130 9. For compatibility with Sendmail, the command line option
4138 and sets the incoming protocol and host name (for trusted callers). The
4139 host name and its colon can be omitted when only the protocol is to be set.
4140 Note the Exim already has two private options, -pd and -ps, that refer to
4141 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4142 "s", but I don't think that's a major issue.
4144 10. A number of refactoring changes to the code, none of which should affect
4147 (a) The number of logging options was getting close to filling up the
4148 32-bit word that was used as a bit map. I have split them into two classes:
4149 those that are passed in the argument to log_write(), and those that are
4150 only ever tested independently outside of that function. These are now in
4151 separate 32-bit words, so there is plenty of room for expansion again.
4152 There is no change in the user interface or the logging behaviour.
4154 (b) When building, for example, log lines, the code previously used a
4155 macro that called string_cat() twice, in order to add two strings. This is
4156 not really sufficiently general. Furthermore, there was one instance where
4157 it was actually wrong because one of the argument was used twice, and in
4158 one call a function was used. (As it happened, calling the function twice
4159 did not affect the overall behaviour.) The macro has been replaced by a
4160 function that can join an arbitrary number of extra strings onto a growing
4163 (c) The code for expansion conditions now uses a table and a binary chop
4164 instead of a serial search (which was left over from when there were very
4165 few conditions). Also, it now recognizes conditions like "pam" even when
4166 the relevant support is not compiled in: a suitably worded error message is
4167 given if an attempt is made to use such a condition.
4169 11. Added ${time_interval:xxxxx}.
4171 12. A bug was causing one of the ddress fields not to be passed back correctly
4172 from remote delivery subprocesses. The field in question was not being
4173 subsequently used, so this caused to problems in practice.
4175 13. Added new log selectors queue_time and deliver_time.
4177 14. Might have fixed a bug in maildirsizefile handling that threw up
4178 "unexpected character" debug warnings, and recalculated the data
4179 unnecessarily. In any case, I expanded the warning message to give more
4182 15. Added the message "Restricted characters in address" to the statements in
4183 the default ACL that block characters like @ and % in local parts.
4185 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4186 Three changes have been made:
4188 (a) There was a serious bug; a negative response to MAIL caused the whole
4189 recipient domain to be cached as invalid, thereby blocking all messages
4190 to all local parts at the same domain, from all senders. This bug has
4191 been fixed. The domain is no longer cached after a negative response to
4192 MAIL if the sender used is not empty.
4194 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4197 (c) A new callout option, "use_sender" has been added for people who want
4198 the modified behaviour.
4204 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4207 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4208 indeed breaks things for older releases.
4210 3. Added additional logging to the case where there is a problem reading data
4211 from a filter that is running in a subprocess using a pipe, in order to
4212 try to track down a specific problem.
4214 4. Testing facility fudge: when running in the test harness and attempting
4215 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4216 getting "No route to host". Convert this to a timeout.
4218 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4221 6. Some OS don't have socklen_t but use size_t instead. This affects the
4222 fifth argument of getsockopt() amongst other things. This is now
4223 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4224 can be set for individual OS. I have set it for SunOS5, OSF1, and
4225 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4226 some earlier ones do not.
4228 7. Change 4.30/15 was not doing the test caselessly.
4230 8. The standard form for an IPv6 address literal was being rejected by address
4231 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4232 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4233 this, as well as the form without the "IPv6" on the front (but only when
4234 address literals are enabled, of course).
4236 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4238 10. Exim crashed if a message with an empty sender address specified by -f
4239 encountered a router with an errors_to setting. This could be provoked only
4240 by a command such as
4244 where an empty string was supplied; "<>" did not hit this bug.
4246 11. Installed PCRE release 4.5.
4248 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4249 remained set. It is now erased.
4251 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4252 times from message ids (which are base 36 rather than the normal 62).
4254 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4255 were being counted as actual protocol errors, and logged if the log
4256 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4257 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4258 and DATA following a set of rejected RCPTs do not count as protocol errors.
4259 In other words, Exim assumes they were pipelined, though this may not
4260 actually be the case. Of course, in all cases the client gets an
4261 appropriate error code.
4263 15. If a lookup fails in an ACL condition, a message about the failure may
4264 be available; it is used if testing the ACL cannot continue, because most
4265 such messages specify what the cause of the deferral is. However, some
4266 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4267 that caused an old message to be retained and used if a later statement
4268 caused a defer, replacing the real cause of the deferral.
4270 16. If an IP address had so many PTR records that the DNS lookup buffer
4271 was not large enough to hold them, Exim could crash while trying to process
4272 the truncated data. It now detects and logs this case.
4274 17. Further to 4.21/58, another change has been made: if (and only if) the
4275 first line of a message (the first header line) ends with CRLF, a bare LF
4276 in a subsequent header line has a space inserted after it, so as not to
4277 terminate the header.
4279 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4280 unnecessarily, used atoi() instead of strtol(), and didn't check the
4281 termination when getting file sizes from file names by regex.
4283 19. Completely re-implemented the support for maildirsize files, in the light
4284 of a number of problems with the previous contributed implementation
4285 (4.30/29). In particular:
4287 . If the quota is zero, the maildirsize file is maintained, but no quota is
4290 . If the maildir directory does not exist, it is created before any attempt
4291 to write a maildirsize file.
4293 . The quota value in the file is just a cache; if the quota is changed in
4294 the transport, the new value overrides.
4296 . A regular expression is available for excluding directories from the
4299 20. The autoreply transport checks the characters in options that define the
4300 message's headers; it allows continued headers, but it was checking with
4301 isspace() after an embedded newline instead of explicitly looking for a
4304 21. If all the "regular" hosts to which an address was routed had passed their
4305 expiry times, and had not reached their retry times, the address was
4306 bounced, even if fallback hosts were defined. Now Exim should go on to try
4309 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4310 equivalent code in the SMTP transport. Some hosts send humungous responses
4311 to HELO/EHLO, more than 1024 it seems.
4313 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4314 gives compiler warnings in some environments. I've now done it "properly",
4317 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4318 (because of problems with the built-in one) was declared to return uschar *
4319 instead of char *, causing compiler failure.
4321 25. Fixed a file descriptor leak when processing alias/forward files.
4323 26. Fixed a minor format string issue in dbfn.c.
4325 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4327 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4328 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4331 29. When a sender address is verified, it is cached, to save repeating the test
4332 when there is more than one recipient in a message. However, when the
4333 verification involves a callout, it is possible for different callout
4334 options to be set for different recipients. It is too complicated to keep
4335 track of this in the cache, so now Exim always runs a verification when a
4336 callout is required, relying on the callout cache for the optimization.
4337 The overhead is duplication of the address routing, but this should not be
4340 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4341 to be verified with callout=postmaster, and the main callout worked but the
4342 postmaster check failed, the verification correctly failed. However, if a
4343 subsequent RCPT command asked for sender verification *without* the
4344 postmaster check, incorrect caching caused this verification also to fail,
4347 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4348 it was not caching the DNS options (qualify_single, search_parents) that
4349 were used when the lookup failed. A subsequent lookup with different
4350 options therefore always gave the same answer, though there were cases
4351 where it should not have. (Example: a "domains = !$mx_any" option on a
4352 dnslookup router: the "domains" option is always processed without any
4353 widening, but the router might have qualify_single set.) Now Exim uses the
4354 cached value only when the same options are set.
4356 32. Added John Jetmore's "exipick" utility to the distribution.
4358 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4359 than a timeout (e.g. a certificate is required, and is not provided), an
4360 Exim server now closes the connection immediately. Previously it waited for
4361 the client to close - but if the client is SSL, it seems that they each
4362 wait for each other, leading to a delay before one of them times out.
4364 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4365 maintained 0.8.x compatibility because I don't think many are using it, and
4366 it is clearly obsolete.
4368 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4371 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4372 client certificate was expired. A simple patch fixes this, though I don't
4373 understand the full logic of why the verify callback is called multiple
4376 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4377 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4378 which causes problems with some clients (such as the Certicom SSL Plus
4379 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4380 disables the coutermeasure allowing Eudora to connect."
4382 38. Exim was not checking that a write() to a log file succeeded. This could
4383 lead to Bad Things if a log got too big, in particular if it hit a file
4384 size limit. Exim now panics and dies if it cannot write to a log file, just
4385 as it does if it cannot open a log file.
4387 39. Modified OS/Makefile-Linux so that it now contains
4389 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4391 The two -D definitions ensure that Exim is compiled with large file
4392 support, which makes it possible to handle log files that are bigger than
4395 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4396 instance) a domain was checked against a named list that involved a lookup,
4397 causing $domain_data to be set, then another domain was checked against the
4398 same list, then the first domain was re-checked, the value of $domain_data
4399 after the final check could be wrong. In particular, if the second check
4400 failed, it could be set empty. This bug probably also applied to
4403 41. The strip_trailing_dot option was not being applied to the address given
4404 with the -f command-line option.
4406 42. The code for reading a message's header from the spool was incrementing
4407 $received_count, but never initializing it. This meant that the value was
4408 incorrect (doubled) while delivering a message in the same process in which
4409 it was received. In the most common configuration of Exim, this never
4410 happens - a fresh exec is done - but it can happen when
4411 deliver_drop_privilege is set.
4413 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4414 it now includes up to 150 characters of the unexpected data in the log
4417 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4418 and building data strings. The size of both of these buffers was 10 000
4419 bytes - far larger than anybody would *ever* want, thought I. Needless to
4420 say, somebody hit the limit. I have increased the maximum line length to
4421 20 000 and the maximum data length of concatenated lines to 100 000. I have
4422 also fixed two bugs, because there was no checking on these buffers. Tsk,
4423 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4424 buffer is too small.
4426 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4427 lsearch lookups. Now it does.
4429 46. When parsing a route_list item in a manualroute router, a fixed-length
4430 buffer was used for the list of hosts. I made this 1024 bytes long,
4431 thinking that nobody would ever have a list of hosts that long. Wrong.
4432 Somebody had a whole pile of complicated expansion conditions, and the
4433 string was silently truncated, leading to an expansion error. It turns out
4434 that it is easier to change to an unlimited length (owing to other changes
4435 that have happened since this code was originally written) than to build
4436 structure for giving a limitation error. The length of the item that
4437 expands into the list of hosts is now unlimited.
4439 47. The lsearch lookup could not handle data where the length of text line was
4440 more than 4095 characters. Such lines were truncated, leading to shortened
4441 data being returned. It should now handle lines of any length.
4443 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4444 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4447 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4448 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4449 no longer happens in comment lines. A list of replacements is now placed
4450 at the head of all of the source files, except those whose only change is
4451 to replace PERL_COMMAND in the very first #! line.
4453 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4454 messages on the queue, with a bottom-up merge sort, using code contributed
4455 by Michael Haardt. This should make operations like -bp somewhat faster on
4456 large queues. It won't affect queue runners, except when queue_run_in_order
4459 51. Installed eximstats 1.31 in the distribution.
4461 52. Added support for SRV lookups to the dnslookup router.
4463 53. If an ACL referred to $message_body or $message_body_end, the value was not
4464 reset for any messages that followed in the same SMTP session.
4466 54. The store-handling optimization for building very long strings was not
4467 differentiating between the different store pools. I don't think this
4468 actually made any difference in practice, but I've tidied it.
4470 55. While running the routers to verify a sender address, $sender_address
4471 was still set to the sender address. This is wrong, because when routing to
4472 send a bounce to the sender, it would be empty. Therefore, I have changed
4473 it so that, while verifying a sender address, $sender_address is set to <>.
4474 (There is no change to what happens when verifying a recipient address.)
4476 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4477 target A or AAAA records (if not already returned) without resetting the
4478 qualify_single or search_parents options of the DNS resolver. These are
4479 inappropriate in this case because the targets of MX and SRV records must
4480 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4481 target that, when qualified, matched something in the local domain. These
4482 two options are now turned off when doing these lookups.
4484 57. It seems that at least some releases of Reiserfs (which does not have the
4485 concept of a fixed number of inodes) returns zero and not -1 for the
4486 number of available inodes. This interacted badly with check_spool_inodes,
4487 which assumed that -1 was the "no such thing" setting. What I have done is
4488 to check that the total number of inodes is greater than zero before doing
4489 the test of how many are available.
4491 58. When a "warn" ACL statement has a log_message modifier, the message is
4492 remembered, and not repeated. This is to avoid a lot of repetition when a
4493 message has many recipients that cause the same warning to be written.
4494 Howewer, Exim was preserving the list of already written lines for an
4495 entire SMTP session, which doesn't seem right. The memory is now reset if a
4496 new message is started.
4498 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4499 debugging output unless log_rewrite was also set.
4501 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4502 of (int)(handle) when we know that handle contains (void *)(-1).
4504 61. The Exim daemon panic-logs an error return when it closes the incoming
4505 connection. However "connection reset by peer" seems to be common, and
4506 isn't really an error worthy of noting specially, so that particular error
4509 62. When Exim is trying to find all the local interfaces, it used to panic and
4510 die if the ioctl to get the interface flags failed. However, it seems that
4511 on at least one OS (Solaris 9) it is possible to have an interface that is
4512 included in the list of interfaces, but for which you get a failure error
4513 for this call. This happens when the interface is not "plumbed" into a
4514 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4515 failure of the "get flags" call assumes that the interface is down.
4517 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4518 makes life easier for people who are doing arithmetic on fields extracted
4519 from dates, where you often get leading zeros that should not be
4520 interpreted as octal.
4522 64. Added qualify_domain to the redirect router, to override the global
4525 65. If a pathologically long header line contained very many addresses (the
4526 report of this problem mentioned 10 000) and each of them was rewritten,
4527 Exim could use up a very large amount of memory. (It kept on making new
4528 copies of the header line as it rewrote, and never released the old ones.)
4529 At the expense of a bit more processing, the header rewriting function has
4530 been changed so that it no longer eats memory in this way.
4532 66. The generation of the Received: header has been moved from the time that a
4533 message starts to be received, to the time that it finishes. The timestamp
4534 in the Received: header should now be very close to that of the <= log
4535 line. There are two side-effects of this change:
4537 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4538 logged header lines no longer include the local Received: line, because
4539 it has not yet been created. The same applies to a copy of the message
4540 that is returned to a non-SMTP sender when a message is rejected.
4542 (b) When a filter file is tested using -bf, no additional Received: header
4543 is added to the test message. After some thought, I decided that this
4546 This change does not affect the value of $received_for. It is still set
4547 after address rewriting, but before local_scan() is called.
4549 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4551 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4552 gave an unhelpful panic error message, and a defer error. I have managed to
4553 change this behaviour so that it now rejects any supplied certificate,
4554 which seems right, as the list of acceptable certificates is empty.
4556 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4557 gave an unhelpful defer error. I have not managed to make this reject any
4558 supplied certificates, but the error message it gives is "no certificate
4559 supplied", which is not helpful.
4561 70. exigrep's output now also includes lines that are not associated with any
4562 message, but which match the given pattern. Implemented by a patch from
4563 Martin Sluka, which also tidied up the Perl a bit.
4565 71. Recipient callout verification, like sender verification, was using <> in
4566 the MAIL FROM command. This isn't really the right thing, since the actual
4567 sender may affect whether the remote host accepts the recipient or not. I
4568 have changed it to use the actual sender in the callout; this means that
4569 the cache record is now keyed on a recipient/sender pair, not just the
4570 recipient address. There doesn't seem to be a real danger of callout loops,
4571 since a callout by the remote host to check the sender would use <>.
4572 [SEE ABOVE: changed after hitting problems.]
4574 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4575 temporary errors. However, in the case of such a code being given after
4576 the end of a data transmission (i.e. after ".") Exim was failing to write
4577 a retry record for the message. (Yes, there was some broken host that was
4578 actually sending 8xx at this point.)
4580 73. An unknown lookup type in a host list could cause Exim to panic-die when
4581 the list was checked. (An example that provoked this was putting <; in the
4582 middle of a list instead of at the start.) If this happened during a DATA
4583 ACL check, a -D file could be left lying around. This kind of configuration
4584 error no longer causes Exim to die; instead it causes a defer errror. The
4585 incident is still logged to the main and panic logs.
4587 74. Buglet left over from Exim 3 conversion. The message "too many messages
4588 in one connection" was written to the rejectlog but not the mainlog, except
4589 when address rewriting (yes!) was being logged.
4591 75. Added write_rejectlog option.
4593 76. When a system filter was run not as root (that is, when system_filter_user
4594 was set), the values of the $n variables were not being returned to the
4595 main process; thus, they were not subsequently available in the $sn
4598 77. Added +return_path_on_delivery log selector.
4600 78. A connection timeout was being treated differently from recipients deferred
4601 when testing hosts_max_try with a message that was older than the host's
4602 retry timeout. (The host should not be counted, thus allowing all hosts to
4603 be tried at least once before bouncing.) This may have been the cause of an
4604 occasionally reported bug whereby a message would remain on the queue
4605 longer than the retry timeout, but would be bounced if a delivery was
4606 forced. I say "may" because I never totally pinned down the problem;
4607 setting up timeout/retry tests is difficult. See also the next item.
4609 79. The ultimate address timeout was not being applied to errors that involved
4610 a combination of host plus message (for example, a timeout on a MAIL
4611 command). When an address resolved to a number of possible hosts, and they
4612 were not all tried for each delivery (e.g. because of hosts_max_try), a
4613 message could remain on the queue longer than the retry timeout.
4615 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4618 81. Fixed an obscure SMTP outgoing bug which required at least the following
4619 conditions: (a) there was another message waiting for the same server;
4620 (b) the server returned 5xx to all RCPT commands in the first message so
4621 that the message was not completed; (c) the server dropped the connection
4622 or gave a negative response to the RSET that Exim sends to abort the
4623 transaction. The observed case was a dropped connection after DATA that had
4624 been sent in pipelining mode. That is, the server had advertised PIPELINING
4625 but was not implementing it correctly. The effect of the bug was incorrect
4626 behaviour, such as trying another host, and this could lead to a crash.
4632 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4633 and daemon.c were passed as pointers to ints; they should have been
4634 pointers to socklen_t variables (which are typically unsigned ints).
4636 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4639 3. Fixed a really odd bug that affected only the testing scheme; patching a
4640 certain fixed string in the binary changed the value of another string that
4641 happened to be identical to the end of the original first string.
4643 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4644 name", it returns that address as the IP address. On some operating
4645 systems (e.g. Solaris), it also passes back the IP address string as the
4646 "host name". However, on others (e.g. Linux), it passes back an empty
4647 string. Exim wasn't checking for this, and was changing the host name to an
4648 empty string, assuming it had been canonicized.
4650 5. Although rare, it is permitted to have more than one PTR record for a given
4651 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4652 all the names associated with an address, because they do in Solaris.
4653 However, it seems that they do not in Linux for data that comes from the
4654 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4655 I found this out when I moved to a new Linux workstation and tried to run
4656 the Exim test suite.
4658 To get round this problem I have changed the code so that it now does its
4659 own call to the DNS to look up PTR records when searching for a host name.
4660 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4661 addresses that are only in /etc/hosts are still found.
4663 This behaviour is, however, controlled by an option called host_lookup_
4664 order, which defaults to "bydns:byaddr". If people want to use the other
4665 order, or indeed, just use one or the other means of lookup, they can
4666 specify it in this variable.
4668 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4669 some operating systems, this comes back from gethostbyaddr() as an empty
4670 string, and this is what Exim used to test for. However, it seems that in
4671 other systems, "." is yielded. Exim now tests for this case too.
4673 7. The values of check_spool_space and check_log_space are now held internally
4674 as a number of kilobytes instead of an absolute number of bytes. If a
4675 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4676 kilobyte. This means that much larger values can be stored.
4678 8. Exim monitor: an attempt to get the action menu when not actually pointing
4679 at a message produces an empty menu entitled "No message selected". This
4680 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4681 no entries in it ("Shell widget menu has zero width and/or height"). So I
4682 have added a single, blank menu entry in this case.
4684 9. Added ${quote_local_part.
4686 10. MIME decoding is now applied to the contents of Subject: header lines when
4689 11. Now that a reference to $sender_host_address automatically causes a reverse
4690 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4691 host lookup before query-style lookups in lists that might use this
4692 variable. This has therefore been abolished, and the "net-" prefix is no
4693 longer necessary for query-style lookups.
4695 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4696 have been a typo for LFLAGS, so it has been changed.
4698 13. The install script calls Exim with "-C /dev/null" in order to find the
4699 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4700 to be output. Howeve, since Exim outputs its version number before the
4701 error, it didn't break the script. It just looked ugly. I fixed this by
4702 always allowing "-C /dev/null" if the caller is root.
4704 14. Ignore overlarge ACL variable number when reading spool file - insurance
4705 against a later release with more variables having written the file.
4707 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4708 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4709 this, as well as the form without the "IPv6" on the front.
4711 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4712 OS/Makefile-Darwin file.
4714 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4715 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4717 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4718 variable. However, some OS use ints for inodes. I've added cast to long int
4719 to get rid of the compiler warning.
4721 19. I had forgotten to lock out "/../" in configuration file names when
4722 ALT_CONFIG_PREFIX was set.
4724 20. Routers used for verification do not need to specify transports. However,
4725 if such a router generated a host list, and callout was configured, Exim
4726 crashed, because it could not find a port number from the (non-existent)
4727 transport. It now assumes port 25 in this circumstance.
4729 21. Added the -t option to exigrep.
4731 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4732 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4733 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4734 them is set, LOOKUP_LSEARCH is forced.
4736 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4738 24. Added sender and host information to the "rejected by local_scan()" log
4739 line; previously there was no indication of these.
4741 25. Added .include_if_exists.
4743 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4744 Linux. It turns out that not all file systems support this. Apparently some
4745 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4746 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4747 not supported on the file descriptor, is now ignored when Exim is trying to
4748 sync a directory. This applies only to Linux.
4750 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4752 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4753 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4756 29. Added support for maildirsize files from supplied patch (modified a bit).
4758 30. The use of :fail: followed by an empty string could lead Exim to respond to
4759 sender verification failures with (e.g.):
4761 550 Verification failed for <xxx>
4762 550 Sender verify failed
4764 where the first response line was missing the '-' that indicates it is not
4765 the final line of the response.
4767 31. The loop for finding the name of the user that called Exim had a hardwired
4768 limit of 10; it now uses the value of finduser_retries, which is used for
4769 all other user lookups.
4771 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4774 33. Exim was neglecting to zero errno before one call of strtol() when
4775 expanding a string and expecting an integer value. On some systems this
4776 resulted in spurious "integer overflow" errors. Also, it was casting the
4777 result into an int without checking.
4779 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4780 did not work. The code looks as if it has *never* worked, though it appears
4781 to have been documented since at least releast 1.62. I have made it work.
4783 35. The "timeout_DNS" error in retry rules, also documented since at least
4784 1.62, also never worked. As it isn't clear exactly what this means, and
4785 clearly it isn't a major issue, I have abolished the feature by treating it
4786 as "timeout", and writing a warning to the main and panic logs.
4788 36. The display of retry rules for -brt wasn't always showing the error code
4791 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4792 timeout_connect_A, timeout_connect_MX.
4794 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4795 to the empty sender.
4797 39. The daemon was not analysing the content of -oX till after it had closed
4798 stderr and disconnected from the controlling terminal. This meant that any
4799 syntax errors were only noted on the panic log, and the return code from
4800 the command was 0. By re-arranging the code a little, I've made the
4801 decoding happen first, so such errors now appear on stderr, and the return
4802 code is 1. However, the actual setting up of the sockets still happens in
4803 the disconnected process, so errors there are still only recorded on the
4806 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4807 connections (as happens on some IP stacks) was logged at start up time as
4808 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4809 it from "IPv6 and IPv4", which means that two separate sockets are being
4812 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4813 says whether AF_INET or AF_INET6 was passed as an argument.
4815 42. Exiwhat output was messed up when time zones were included in log
4818 43. Exiwhat now gives more information about the daemon's listening ports,
4819 and whether -tls-on-connect was used.
4821 44. The "port" option of the smtp transport is now expanded.
4823 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4824 silently ignored. Now an error message is written to the main and panic
4827 46. There's a new ACL modifier called "logwrite" which writes to a log file
4828 as soon as it is encountered.
4830 47. Added $local_user_uid and $local_user_gid at routing time.
4832 48. Exim crashed when trying to verify a sender address that was being
4835 49. Exim was recognizing only a space character after ".include". It now also
4836 recognizes a tab character.
4838 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4839 extracting the relevant information from the specification. The man page no
4840 longer contains scrambled data for the -d option, and I've added a section
4841 at the front about calling Exim under different names.
4843 51. Added "extra_headers" argument to the "mail" command in filter files.
4845 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4848 53. Installed eximstats 1.29.
4850 54. Added transport_filter_timeout as a generic transport option.
4852 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4853 Cc: header lines. This was required by RFC 822, but it not required by RFC
4856 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4857 incoming messages that did not have them. Now it does so only if the
4858 message originates locally, that is, if there is no associated remote host
4859 address. When Resent- header lines are present, this applies to the Resent-
4860 lines rather than the non-Resent- lines.
4862 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4863 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4865 58. Messages for configuration errors now include the name of the main
4866 configuration file - useful now that there may be more than one file in a
4867 list (.included file names were always shown).
4869 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4870 for those rare installations that do not start the daemon as root or run it
4871 setuid root. I've cut out the call to initgroups() if the daemon is not
4874 60. The Exim user and group can now be bound into the binary as text strings
4875 that are looked up at the start of Exim's processing.
4877 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4879 62. Added $mailstore_basename variable.
4881 63. Installed patch to sieve.c from Michael Haardt.
4883 64. When Exim failed to open the panic log after failing to open the main log,
4884 the original message it was trying to log was written to stderr and debug
4885 output, but if they were not available (the usual case in production), it
4886 was lost. Now it is written to syslog before the two lines that record the
4887 failures to open the logs.
4889 65. Users' Exim filters run in subprocesses under the user's uid. It is
4890 possible for a "deliver" command or an alias in a "personal" command to
4891 provoke an address rewrite. If logging of address rewriting is configured,
4892 this fails because the process is not running as root or exim. There may be
4893 a better way of dealing with this, but for the moment (because 4.30 needs
4894 to be released), I have disabled address rewrite logging when running a
4895 filter in a non-root, non-exim process.
4901 1. The buildconfig auxiliary program wasn't quoting the value set for
4902 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4903 not defined. This bug was masked in 4.22 by the effect that was fixed in
4906 2. Some messages that were rejected after a message id was allocated were
4907 shown as "incomplete" by exigrep. It no longer does this for messages that
4908 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4910 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4911 have allow_domain_literals set, the ID did not get logged in the <= line.
4912 Domain literals are now always recognized in Message-ID: header lines.
4914 4. The first argument for a ${extract expansion item is the key name or field
4915 number. Leading and trailing spaces in this item were not being ignored,
4916 causing some misleading effects.
4918 5. When deliver_drop_privilege was set, single queue runner processes started
4919 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4920 same command in the process it spins off) were not dropping privilege.
4922 6. When the daemon running as "exim" started a queue runner, it always
4923 re-executed Exim in the spun-off process. This is a waste of effort when
4924 deliver_drop_privilege is set. The new process now just calls the
4925 queue-runner function directly.
4931 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4934 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4935 ignored. Though the use of -oP was forcing the writing of a pid file, it
4936 was always written to the default place.
4938 3. If the message "no IP address found for host xxxx" is generated during
4939 incoming verification, it is now followed by identification of the incoming
4940 connection (so you can more easily find what provoked it).
4942 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4944 5. Added some features to "harden" Exim a bit more against certain attacks:
4946 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4947 be put in Local/Makefile. This is like the never_users runtime option,
4948 but it cannot be overridden. The default setting is "root".
4950 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4951 prefix string with which any file named in a -C command line option
4954 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4955 is retained for -C and -D only if the caller of Exim is root. Without
4956 it, the exim user may also use -C and -D and retain privilege.
4958 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4959 command line option is disabled.
4961 6. Macro names set by the -D option must start with an upper case letter, just
4962 like macro names defined in the configuration file.
4964 7. Added "dereference=" facility to LDAP.
4966 8. Two instances of the typo "uknown" in the source files are fixed.
4968 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4969 the Configure-Makefile script screwed up while processing it.
4971 10. Incorporated PCRE 4.4.
4973 11. The SMTP synchronization check was not operating right at the start of an
4974 SMTP session. For example, it could not catch a HELO sent before the client
4975 waited for the greeting. There is now a check for outstanding input at the
4976 point when the greeting is written. Because of the duplex, asynchronous
4977 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4978 way, but not yet received, when the check is performed.
4980 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4981 on TCP/IP sockets, because this apparently causes some broken clients to
4984 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4985 unchanged) from the Cygwin maintainer.
4987 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4988 was supported instead of testing for mbx. Effectively a typo.
4990 15. The spa authenticator server code was not checking that the input it
4991 received was valid base64.
4993 16. The debug output line for the "set" modifier in ACLs was not showing the
4994 name of the variable that was being set.
4996 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4998 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4999 lookup on its own. Something else was needed to trigger the lookup. For
5000 example, a match in host_lookup or the need for a host name in a host list.
5001 Now, if $sender_host_name is referenced and the host name has not yet been
5002 looked up, a lookup is performed. If the lookup fails, the variable remains
5003 empty, and $host_lookup_failed is set to "1".
5005 19. Added "eqi" as a case-independent comparison operator.
5007 20. The saslauthd authentication condition could segfault if neither service
5008 nor realm was specified.
5010 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5011 error message that was logged was misleading, and incoming SMTP
5012 connections were dropped. The message is now more accurate, and temporary
5013 errors are given to SMTP connections.
5015 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5016 (except RSET) with a 421 error, until QUIT is received. However, it was
5017 failing to send a response to QUIT.
5019 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5020 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5021 ACL is now run after the helo_try_verify_hosts code.
5023 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5024 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5025 case-independent, so other case variants are also recognized). Apparently
5026 some systems use these upper case variants.
5028 25. If more than two messages were waiting for the same host, and a transport
5029 filter was specified for the transport, Exim sent two messages over the
5030 same TCP/IP connection, and then failed with "socket operation on non-
5031 socket" when it tried to send the third.
5033 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5035 27. The extern definition of crypt16() in expand.c was not being excluded when
5036 the OS had its own crypt16() function.
5038 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5039 as a preferred synonym for return_size_limit, both as an option and as an
5042 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5044 30. Changed the default configuration ACL to relax the local part checking rule
5045 for addresses that are not in any local domains. For these addresses,
5046 slashes and pipe symbols are allowed within local parts, but the sequence
5047 /../ is explicitly forbidden.
5049 31. SPA server authentication was not clearing the challenge buffer before
5052 32. log_message in a "warn" ACL statement was writing to the reject log as
5053 well as to the main log, which contradicts the documentation and doesn't
5054 seem right (because no rejection is happening). So I have stopped it.
5056 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5057 However, I am unable to do any testing of this.
5059 34. Fixed an infelicity in the appendfile transport. When checking directories
5060 for a mailbox, to see if any needed to be created, it was accidentally
5061 using path names with one or more superfluous leading slashes; tracing
5062 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5064 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5065 contents are added to the log line that is written for every discarded
5066 recipient. (Previously a log_message setting was ignored.)
5068 36. The ${quote: operator now quotes the string if it is empty.
5070 37. The install script runs exim in order to find its version number. If for
5071 some reason other than non-existence or emptiness, which it checks, it
5072 could not run './exim', it was installing it with an empty version number,
5073 i.e. as "exim-". This error state is now caught, and the installation is
5076 38. An argument was missing from the function that creates an error message
5077 when Exim fails to connect to the socket for saslauthd authentication.
5078 This could cause Exim to crash, or give a corrupted message.
5080 39. Added isip, isip4, and isip6 to ${if conditions.
5082 40. The ACL variables $acl_xx are now saved with the message, and can be
5083 accessed later in routers, transports, and filters.
5085 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5086 strings in the file are not string-expanded.
5088 42. If a MAIL command specified a SIZE value that was too large to fit into an
5089 int variable, the check against message_size_limit failed. Such values are
5090 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5091 day this will have to be increased, but I don't think I want to be around
5092 when emails are that large.
5099 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5100 iconv() is not standard in FreeBSD.
5102 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5103 IPv6 enabled. The observed symptom was a segmentation fault on return from
5104 the function os_common_find_running_interfaces() in src/os.c.
5106 3. In the check_special_case() function in daemon.c I had used "errno" as an
5107 argument name, which causes warnings on some systems. This was basically a
5108 typo, since it was named "eno" in the comments!
5110 4. The code that waits for the clock to tick (at a resolution of some fraction
5111 of a second) so as to ensure message-id uniqueness was always waiting for
5112 at least one whole tick, when it could have waited for less. [This is
5113 almost certainly not relevant at current processor speeds, where it is
5114 unlikely to ever wait at all. But we try to future-proof.]
5116 5. The function that sleeps for a time interval that includes fractions of a
5117 second contained a race. It did not block SIGALRM between setting the
5118 timer, and suspending (a couple of lines later). If the interval was short
5119 and the sigsuspend() was delayed until after it had expired, the suspension
5120 never ended. On busy systems this could lead to processes getting stuck for
5123 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5124 process, before it forks any delivery processes. The open lookup caching
5125 mechanism meant that the open file or database connection was passed into
5126 the delivery process. The problem was that delivery processes always tidy
5127 up cached lookup data. This could cause a problem for the next delivery
5128 process started by the queue runner, because the external queue runner
5129 process does not know about the closure. So the next delivery process
5130 still has data in the lookup cache. In the case of a file lookup, there was
5131 no problem because closing a file descriptor in a subprocess doesn't affect
5132 the parent. However, if the lookup was caching a connection to a database,
5133 the connection was closed, and the second delivery process was likely to
5134 see errors such as "PGSQL: query failed: server closed the connection
5135 unexpectedly". The problem has been fixed by closing all cached lookups
5136 in a queue runner before running a delivery process.
5138 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5139 seem to have the "const" qualifier which it has on other OS. I've
5142 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5143 *of the same type* (client or server) with the same public name that an
5144 error should be diagnosed.
5146 9. When Exim looked up a host name for an IP address, but failed to find the
5147 original IP address when looking up the host name (a safety check), it
5148 output the message "<ip address> does not match any IP for NULL", which was
5149 confusing, to say the least. The bug was that the host name should have
5150 appeared instead of "NULL".
5152 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5153 user that is built into the binary, and the -C or -D options is used, root
5154 privilege is dropped before the configuration file is read. In addition,
5155 logging is switched to stderr instead of the normal log files. If the
5156 configuration then re-defines the Exim user, the unprivileged environment
5157 is probably not what is expected, so Exim logs a panic warning message (but
5160 However, if deliver_drop_privilege is set, the unprivileged state may well
5161 be exactly what is intended, so the warning has been cut out in that case,
5162 and Exim is allowed to try to write to its normal log files.
5168 1. smtp_return_error_details was not giving details for temporary sender
5169 or receiver verification errors.
5171 2. Diagnose a configuration error if two authenticators have the same public
5174 3. Exim used not to create the message log file for a message until the first
5175 delivery attempt. This could be confusing when incoming messages were held
5176 for policy or load reasons. The message log file is now created at the time
5177 the message is received, and an initial "Received" line is written to it.
5179 4. The automatically generated man page for command line options had a minor
5180 bug that caused no ill effects; however, a more serious problem was that
5181 the procedure for building the man page automatically didn't always
5182 operate. Consequently, release 4.20 contains an out-of-date version. This
5183 shouldn't happen again.
5185 5. When building Exim with embedded Perl support, the script that builds the
5186 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5187 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5189 6. The freeze_tell option was not being used for messages that were frozen on
5190 arrival, either by an ACL or by local_scan().
5192 7. Added the smtp_incomplete_transaction log selector.
5194 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5195 was accepting AUTH without a new EHLO.
5197 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5198 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5201 10. Logging of TCP/IP connections (when configured) now happens in the main
5202 daemon process instead of the child process, so that the TCP/IP connection
5203 count is more accurate (but it can never be perfect).
5205 11. The use of "drop" in a nested ACL was not being handled correctly in the
5206 outer ACL. Now, if condition failure induced by the nested "drop" causes
5207 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5208 or "require"), the connection is dropped.
5210 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5211 that yield "discard" can now be used with an "accept" or a "discard" verb,
5212 but an error is generated for any others (because I can't see a useful way
5213 to define what should happen).
5215 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5216 are now processed in the same way as lines in the Exim configuration file.
5217 In particular, continuation lines are supported.
5219 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5221 15. Added -ti meaning -t -i.
5223 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5224 domains, and warn by logging if others are found.
5226 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5227 returned from GIFCONF when Exim is trying to find the list of interfaces on
5228 a host. The code in os.c has been modified to copy each ifreq to an aligned
5229 structure in all cases.
5231 Also, in some cases, the returned ifreq's were being copied to a 'struct
5232 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5233 means the last couple of bytes of an IPv6 address could be chopped if the
5234 ifreq contained only a normal sockaddr (14 bytes storage).
5236 18. Named domain lists were not supported in the hosts_treat_as_local option.
5237 An entry such as +xxxx was not recognized, and was treated as a literal
5240 19. Ensure that header lines added by a DATA ACL are included in the reject log
5241 if the ACL subsequently rejects the message.
5243 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5244 MD5 (which is deprecated).
5246 21. When testing a filter file using -bf, Exim was writing a message when it
5247 took the sender from a "From " line in the message, but it was not doing so
5248 when it took $return_path from a Return-Path: header line. It now does.
5250 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5251 with a valid header line field name (a series of printing characters
5252 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5254 23. Changed "disc" in the source to "disk" to conform to the documentation and
5255 the book and for uniformity.
5257 24. Ignore Sendmail's -Ooption=value command line item.
5259 25. When execve() failed while trying to run a command in a pipe transport,
5260 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5261 could be confused with a return value of 69 from the command itself. This
5262 has been changed to 127, the value the shell returns if it is asked to run
5263 a non-existent command. The wording for the related log line suggests a
5264 non-existent command as the problem.
5266 26. If received_header_text expands to an empty string, do not add a Received:
5267 header line to the message. (Well, it adds a token one on the spool, but
5268 marks it "old" so that it doesn't get used or transmitted.)
5270 27. Installed eximstats 1.28 (addition of -nt option).
5272 28. There was no check for failure on the call to getsockname() in the daemon
5273 code. This can fail if there is a shortage of resources on the system, with
5274 ENOMEM, for example. A temporary error is now given on failure.
5276 29. Contrary to the C standard, it seems that in some environments, the
5277 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5278 program. Exim now does this explicitly; it affects the formatting of
5279 timestamps using strftime().
5281 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5282 complaints. I've now initialized all the variables, to avoid this.
5284 32. Header lines added by a system filter were not being "seen" during
5285 transport-time rewrites.
5287 33. The info_callback() function passed to OpenSSL is set up with type void
5288 (*)(SSL *, int, int), as described somewhere. However, when calling the
5289 function (actually a macro) that sets it up, the type void(*)() is
5290 expected. I've put in a cast to prevent warnings from picky compilers.
5292 34. If a DNS black list lookup found a CNAME record, but there were no A
5293 records associated with the domain it pointed at, Exim crashed.
5295 35. If a DNS black list lookup returned more than one A record, Exim ignored
5296 all but the first. It now scans all returned addresses if a particular IP
5297 value is being sought. In this situation, the contents of the
5298 $dnslist_value variable are a list of all the addresses, separated by a
5301 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5302 to accept a host name and all its aliases if the forward lookup for any of
5303 them yielded the IP address of the incoming connection. Now it accepts only
5304 those names whose forward lookup yields the correct IP address. Any other
5305 names are discarded. This closes a loophole whereby a rogue DNS
5306 administrator could create reverse DNS records to break through a
5307 wildcarded host restriction in an ACL.
5309 37. If a user filter or a system filter that ran in a subprocess used any of
5310 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5311 the wrong values were passed to the pipe command ($thisaddress had the
5312 value of $0, $0 had the value of $1, etc). This bug was introduced by
5313 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5315 38. Improved the line breaking for long SMTP error messages from ACLs.
5316 Previously, if there was no break point between 40 and 75 characters, Exim
5317 left the rest of the message alone. Two changes have been made: (a) I've
5318 reduced the minimum length to 35 characters; (b) if it can't find a break
5319 point between 35 and 75 characters, it looks ahead and uses the first one
5320 that it finds. This may give the occasional overlong line, but at least the
5321 remaining text gets split now.
5323 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5324 file descriptors might be low, and that setting 1000 would always raise it.
5325 It turns out that in some environments, the limit is already over 1000 and
5326 that lowering it causes trouble. So now Exim takes care not to decrease it.
5328 40. When delivering a message, the value of $return_path is set to $sender_
5329 address at the start of routing (routers may change the value). By an
5330 oversight, this default was not being set up when an address was tested by
5331 -bt or -bv, which affected the outcome if any router or filter referred to
5334 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5335 log only when "message" or "log_message" are set. However, if one of the
5336 conditions was an address verification, or a call to a nested ACL, the
5337 messages generated by the underlying test were being passed through. This
5338 no longer happens. The underlying message is available in $acl_verify_
5339 message for both "message" and "log_message" expansions, so it can be
5340 passed through if needed.
5342 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5343 new expansion $bh_ to give the encoded byte string without charset
5344 translation. Translation happens only if iconv() is available; HAVE_ICONV
5345 indicates this at build time. HEADERS_CHARSET gives the charset to
5346 translate to; headers_charset can change it in the configuration, and
5347 "headers charset" can change it in an individual filter file.
5349 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5350 that creates RFC 2047 encoded "words" labels them as that charset instead
5351 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5352 expansion operator; (ii) when Exim creates a From: line for a local
5353 message; (iii) when a header line is rewritten to include a "phrase" part.
5355 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5356 buggy, causing it to skip the first lines of messages whose message ID
5357 ended in 'D'. This would not have bitten before Exim release 4.14, because
5358 message IDs were unlikely to end in 'D' before then. The effect was to have
5359 incorrect size information for certain domains.
5361 45. #include "config.h" was missing at the start of the crypt16.c module. This
5362 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5365 46. If there was a timeout during a "random" callout check, Exim treated it as
5366 a failure of the random address, and carried on sending RSET and the real
5367 address. If the delay was just some slowness somewhere, the response to the
5368 original RCPT would be taken as a response to RSET and so on, causing
5369 mayhem of various kinds.
5371 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5372 when I implemented it. It didn't allow for the fact that some option values
5373 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5374 the right test for positive values.
5376 48. Domain names in DNS records are case-independent. Exim always looks them up
5377 in lower case. Some resolvers return domain names in exactly the case they
5378 appear in the zone file, that is, they may contain uppercase letters. Not
5379 all resolvers do this - some return always lower case. Exim was treating a
5380 change of case by a resolver as a change of domain, similar to a widening
5381 of a domain abbreviation. This triggered its re-routing code and so it was
5382 trying to route what was effectively the same domain again. This normally
5383 caused routing to fail (because the router wouldn't handle the domain
5384 twice). Now Exim checks for this case specially, and just changes the
5385 casing of the domain that it ultimately uses when it transmits the message
5388 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5391 50. If a filter generated a file delivery with a non-absolute name (possible if
5392 no home directory exists for the router), the forbid_file option was not
5395 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5396 the existing equality matching.
5398 52. Exim was using ints instead of ino_t variables in some places where it was
5399 dealing with inode numbers.
5401 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5402 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5403 called TMPDIR, and if it finds it is different, it changes its value.
5405 54. The smtp_printf() function is now made available to local_scan() so
5406 additional output lines can be written before returning. There is also an
5407 smtp_fflush() function to enable the detection of a dropped connection.
5408 The variables smtp_input and smtp_batched_input are exported to
5411 55. Changed the default runtime configuration: the message "Unknown user"
5412 has been removed from the ACL, and instead placed on the localuser router,
5413 using the cannot_route_message feature. This means that any verification
5414 failures that generate their own messages won't get overridden. Similarly,
5415 the "Unrouteable address" message that was in the ACL for unverifiable
5416 relay addresses has also been removed.
5418 56. Added hosts_avoid_esmtp to the smtp transport.
5420 57. The exicyclog script was not checking for the esoteric option
5421 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5422 will work only if exicyclog is run under the appropriate euid.
5424 58. Following a discussion on the list, the rules by which Exim recognises line
5425 endings on incoming messages have been changed. The -dropcr and drop_cr
5426 options are now no-ops, retained only for backwards compatibility. The
5427 following line terminators are recognized: LF CRLF CR. However, special
5428 processing applies to CR:
5430 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5431 nor a local message in the state where . is a terminator.
5433 (ii) If a bare CR is encountered in a header line, an extra space is added
5434 after the line terminator so as not to end the header. The reasoning
5435 behind this is that bare CRs in header lines are most likely either
5436 to be mistakes, or people trying to play silly games.
5438 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5439 was being incorrectly given as 18 bytes larger than it should have been.
5440 This is a VOB (very old bug).
5442 60. This may never have affected anything current, but just in case it has:
5443 When the local host is found other than at the start of a list of hosts,
5444 the local host, those with the same MX, and any that follow, are discarded.
5445 When the list in question was part of a longer list of hosts, the following
5446 hosts (not currently being processed) were also being discarded. This no
5447 longer happens. I'm not sure if this situation could ever has previously
5450 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5453 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5454 header. This is something that is recommended in a new Internet Draft, and
5455 is something that is documented as being done by Sendmail. There are two
5456 possible values. For messages generated by the autoreply transport, Exim
5459 Auto-Submitted: auto-replied
5461 whereas for all other generated messages (e.g. bounces) it adds
5463 Auto-Submitted: auto-generated
5465 63. The "personal" condition in filters now includes a test for the
5466 Auto-Submitted: header. If it contains the string "auto-" the message it
5467 not considered personal.
5469 64. Added rcpt_include_affixes as a generic transport option.
5471 65. Added queue_only_override (default true).
5473 66. Added the syslog_duplication option.
5475 67. If what should have been the first header line of a message consisted of
5476 a space followed by a colon, Exim was mis-interpreting it as a header line.
5477 It isn't of course - it is syntactically invalid and should therefore be
5478 treated as the start of the message body. The misbehaviour could have
5479 caused a number of strange effects, including loss of data in subsequent
5480 header lines, and spool format errors.
5482 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5483 client host had authenticated. This control can now be exercised by an ACL
5484 for more flexibility.
5486 69. By default, callouts do not happen when testing with -bh. There is now a
5487 variant, -bhc, which does actually run the callout code, including
5488 consulting and updating the callout cache.
5490 70. Added support for saslauthd authentication, courtesy of Alexander
5493 71. If statvfs() failed on the spool or log directories while checking their
5494 size for availability, Exim confusingly gave the error "space shortage".
5495 Furthermore, in debugging mode it crashed with a floating point exception.
5496 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5497 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5498 serious problem, Exim now writes to the main and panic logs when this
5499 happens, with details of the failure. It then refuses to accept the
5500 incoming message, giving the message "spool directory problem" or "log
5501 directory problem" with a 421 code for SMTP messages.
5503 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5504 0, 1, and 2 exist, because some OS complain for execs without them (see
5505 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5506 descriptors. However, the code omitted to check that the open succeeded,
5507 causing mysterious errors if for some reason the permissions on /dev/null
5508 got screwed. Now Exim writes a message to the main and panic logs, and
5509 bombs out if it can't open /dev/null.
5511 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5512 interact so that it is all more flexible. It is supposed to remain
5513 backwards compatible. Also added extra_local_interfaces.
5515 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5516 to bomb out with an assertion failure - to the client this appears as a
5517 connection drop. This problem occurs in the part of the code that was taken
5518 from the Samba project. Fortunately, the assertion is in a very simple
5519 function, so I have fixed this by reproducing the function inline in the
5520 one place where it is called, and arranging for authentication to fail
5521 instead of killing the process with assert().
5523 75. The SPA client code was not working when the server requested OEM rather
5524 than Unicode encoding.
5526 76. Added code to make require_files with a specific uid setting more usable in
5527 the case where statting the file as root fails - usually a non-root-mounted
5528 NFS file system. When this happens and the failure is EACCES, Exim now
5529 forks a subprocess and does the per-uid checking as the relevant uid.
5531 77. Added process_log_path.
5533 78. If log_file_path was not explicitly set, a setting of check_log_space or
5534 check_log_inodes was ignored.
5536 79. If a space check for the spool or log partitions fails, the incident is now
5537 logged. Of course, in the latter case the data may get lost...
5539 80. Added the %p formatting code to string_format() so that it can be used to
5540 print addresses in debug_print(). Adjusted all the address printing in the
5541 debugging in store.c to use %p rather than %d.
5543 81. There was a concern that a line of code in smtp_in.c could overflow a
5544 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5545 initially expressed, the concern was not well-founded, because trailing
5546 spaces are removed early. However, if the trailing spaces were followed by
5547 a NULL, they did not get removed, so the overflow was possible. Two fixes
5550 (a) I re-wrote the offending code in a cleaner fashion.
5551 (b) If an incoming SMTP command contains a NULL character, it is rejected
5554 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5555 runs initgroups(), so that if the Exim user is in any additional groups,
5556 they will be used during message reception.
5562 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob