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.
80 PP/01 Solaris build fix for Oracle's LDAP libraries.
81 Bugzilla 1109, patch from Stephen Usher.
83 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
85 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
88 TF/02 Fix a couple more cases where we did not log the error message
89 when unlink() failed. See also change 4.74-TF/03.
91 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
92 lock up or crash if it happened to be inside a call to libc when it
93 got a SIGUSR1 from exiwhat.
95 The SIGUSR1 handler appends the current process status to the process
96 log which is later printed by exiwhat. It used to use the general
97 purpose logging code to do this, but several functions it calls are
100 The new output code in the SIGUSR1 handler is specific to the process
101 log, and simple enough that it's easy to inspect for signal safety.
102 Removing some special cases also simplifies the general logging code.
103 Removing the spurious timestamps from the process log simplifies
106 TF/04 Improved ratelimit ACL condition.
108 The /noupdate option has been deprecated in favour of /readonly which
109 has clearer semantics. The /leaky, /strict, and /readonly update modes
110 are mutually exclusive. The update mode is no longer included in the
111 database key; it just determines when the database is updated. (This
112 means that when you upgrde Exim will forget old rate measurements.)
114 Exim now checks that the per_* options are used with an update mode that
115 makes sense for the current ACL. For example, when Exim is processing a
116 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
117 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
118 must specify per_mail/readonly. If you omit the update mode it defaults to
119 /leaky where that makes sense (as before) or /readonly where required.
121 The /noupdate option is now undocumented but still supported for
122 backwards compatibility. It is equivalent to /readonly except that in
123 ACLs where /readonly is required you may specify /leaky/noupdate or
124 /strict/noupdate which are treated the same as /readonly.
126 A useful new feature is the /count= option. This is a generalization
127 of the per_byte option, so that you can measure the throughput of other
128 aggregate values. For example, the per_byte option is now equivalent
129 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
131 The per_rcpt option has been generalized using the /count= mechanism
132 (though it's more complicated than the per_byte equivalence). When it is
133 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
134 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
135 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
136 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
137 Note that using per_rcpt with a non-readonly update mode in more than
138 one ACL will cause the recipients to be double-counted. (The per_mail
139 and per_byte options don't have this problem.)
141 The handling of very low rates has changed slightly. If the computed rate
142 is less than the event's count (usually one) then this event is the first
143 after a long gap. In this case the rate is set to the same as this event's
144 count, so that the first message of a spam run is counted properly.
146 The major new feature is a mechanism for counting the rate of unique
147 events. The new per_addr option counts the number of different
148 recipients that someone has sent messages to in the last time period. It
149 behaves like per_rcpt if all the recipient addresses are different, but
150 duplicate recipient addresses do not increase the measured rate. Like
151 the /count= option this is a general mechanism, so the per_addr option
152 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
153 example, measure the rate that a client uses different sender addresses
154 with the options per_mail/unique=$sender_address. There are further
155 details in the main documentation.
157 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
159 TF/06 Removed a few PCRE remnants.
161 TF/07 Automatically extract Exim's version number from tags in the git
162 repository when doing development or release builds.
164 PP/02 Raise smtp_cmd_buffer_size to 16kB.
165 Bugzilla 879. Patch from Paul Fisher.
167 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
168 Heavily based on revision 40f9a89a from Simon Arlott's tree.
171 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
173 PP/05 Variable $av_failed, true if the AV scanner deferred.
174 Bugzilla 1078. Patch from John Horne.
176 PP/06 Stop make process more reliably on build failure.
177 Bugzilla 1087. Patch from Heiko Schlittermann.
179 PP/07 Make maildir_use_size_file an _expandable_ boolean.
180 Bugzilla 1089. Patch from Heiko Schlittermann.
182 PP/08 Handle ${run} returning more data than OS pipe buffer size.
183 Bugzilla 1131. Patch from Holger Weiß.
185 PP/09 Handle IPv6 addresses with SPF.
186 Bugzilla 860. Patch from Wolfgang Breyha.
188 PP/10 GnuTLS: support TLS 1.2 & 1.1.
190 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
193 PP/11 match_* no longer expand right-hand-side by default.
194 New compile-time build option, EXPAND_LISTMATCH_RHS.
195 New expansion conditions, "inlist", "inlisti".
197 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
199 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
201 PP/14 fix log_write() format string regression from TF/03.
202 Bugzilla 1152. Patch from Dmitry Isaikin.
208 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
210 PP/02 Harmonised TLS library version reporting; only show if debugging.
211 Layout now matches that introduced for other libraries in 4.74 PP/03.
213 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
215 PP/04 New "dns_use_edns0" global option.
217 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
220 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
221 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
223 TK/01 Updated PolarSSL code to 0.14.2.
224 Bugzilla 1097. Patch from Andreas Metzler.
226 PP/07 Catch divide-by-zero in ${eval:...}.
229 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
232 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
233 format-string attack -- SECURITY: remote arbitrary code execution.
235 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
236 time unintentionally subject to list matching rules, letting the header
237 cause arbitrary Exim lookups (of items which can occur in lists, *not*
238 arbitrary string expansion). This allowed for information disclosure.
240 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
241 INT_MIN/-1 -- value coerced to INT_MAX.
247 NM/01 Workround for PCRE version dependency in version reporting
250 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
251 This fixes portability to compilers other than gcc, notably
252 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
254 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
255 makefiles for portability to HP-UX and POSIX correctness.
257 PP/01 Permit LOOKUP_foo enabling on the make command-line.
258 Also via indented variable definition in the Makefile.
259 (Debugging by Oliver Heesakkers).
261 PP/02 Restore caching of spamd results with expanded spamd_address.
262 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
264 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
265 Improves build reliability. Fix from: Frank Elsner
267 NM/02 Fix wide character breakage in the rfc2047 coding
268 Fixes bug 1064. Patch from Andrey N. Oktyabrski
270 NM/03 Allow underscore in dnslist lookups
271 Fixes bug 1026. Patch from Graeme Fowler
273 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
274 Code patches from Adam Ciarcinski of NetBSD.
276 NM/04 Fixed exiqgrep to cope with mailq missing size issue
279 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
280 is logged, to avoid truncation. Patch from John Horne.
282 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
283 Patch from Jakob Hirsch.
285 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
286 SQL string expansion failure details.
287 Patch from Andrey Oktyabrski.
289 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
290 Patch from Simon Arlott.
292 PP/09 New lookups functionality failed to compile on old gcc which rejects
293 extern declarations in function scope.
294 Patch from Oliver Fleischmann
296 PP/10 Use sig_atomic_t for flags set from signal handlers.
297 Check getgroups() return and improve debugging.
298 Fixed developed for diagnosis in bug 927 (which turned out to be
301 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
302 Patch from Mark Zealey.
304 PP/12 Bugzilla 1056: Improved spamd server selection.
305 Patch from Mark Zealey.
307 PP/13 Bugzilla 1086: Deal with maildir quota file races.
308 Based on patch from Heiko Schlittermann.
310 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
311 Patch from Uwe Doering, sign-off by Michael Haardt.
313 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
314 variable declaration deep within a block. Bug and patch from
317 PP/15 lookups-Makefile IRIX compatibilty coercion.
319 PP/16 Make DISABLE_DKIM build knob functional.
321 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
322 Patch by Simon Arlott
324 TF/03 Fix valgrind.h portability to C89 compilers that do not support
325 variable argument macros. Our copy now differs from upstream.
331 TF/01 Failure to get a lock on a hints database can have serious
332 consequences so log it to the panic log.
334 TF/02 Log LMTP confirmation messages in the same way as SMTP,
335 controlled using the smtp_confirmation log selector.
337 TF/03 Include the error message when we fail to unlink a spool file.
339 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
340 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
341 for maintaining out-of-tree patches for some time.
343 PP/01 Bugzilla 139: Documentation and portability issues.
344 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
345 Handle per-OS dynamic-module compilation flags.
347 PP/02 Let /dev/null have normal permissions.
348 The 4.73 fixes were a little too stringent and complained about the
349 permissions on /dev/null. Exempt it from some checks.
350 Reported by Andreas M. Kirchwitz.
352 PP/03 Report version information for many libraries, including
353 Exim version information for dynamically loaded libraries. Created
354 version.h, now support a version extension string for distributors
355 who patch heavily. Dynamic module ABI change.
357 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
358 privilege escalation vulnerability whereby the Exim run-time user
359 can cause root to append content of the attacker's choosing to
362 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
365 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
366 If dropping privileges for untrusted macros, we disabled normal logging
367 on the basis that it would fail; for the Exim run-time user, this is not
368 the case, and it resulted in successful deliveries going unlogged.
369 Fixed. Reported by Andreas Metzler.
375 PP/01 Date: & Message-Id: revert to normally being appended to a message,
376 only prepend for the Resent-* case. Fixes regression introduced in
377 Exim 4.70 by NM/22 for Bugzilla 607.
379 PP/02 Include check_rfc2047_length in configure.default because we're seeing
380 increasing numbers of administrators be bitten by this.
382 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
384 PP/03 Bugzilla 994: added openssl_options main configuration option.
386 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
388 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
390 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
392 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
393 without a peer certificate, leading to a segfault because of an
394 assumption that peers always have certificates. Be a little more
395 paranoid. Problem reported by Martin Tscholak.
397 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
398 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
399 NB: ClamAV planning to remove STREAM in "middle of 2010".
400 CL also introduces -bmalware, various -d+acl logging additions and
401 more caution in buffer sizes.
403 PP/09 Implemented reverse_ip expansion operator.
405 PP/10 Bugzilla 937: provide a "debug" ACL control.
407 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
409 PP/12 Bugzilla 973: Implement --version.
411 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
413 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
415 PP/15 Bugzilla 816: support multiple condition rules on Routers.
417 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
418 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
419 ignore trailing whitespace.
421 JJ/02 prevent non-panic DKIM error from being sent to paniclog
423 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
426 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
427 Notification from Dr Andrew Aitchison.
429 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
430 ExtendedDetectionInfo response format.
431 Notification from John Horne.
433 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
436 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
437 XSL and documented dependency on system catalogs, with examples of how
440 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
443 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
444 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
445 configuration file which is writeable by the Exim user or group.
447 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
448 of configuration files to cover files specified with the -C option if
449 they are going to be used with root privileges, not just the default
452 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
453 option (effectively making it always true).
455 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
456 files to be used while preserving root privileges.
458 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
459 that rogue child processes cannot use them.
461 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
462 run-time user, instead of root.
464 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
465 Exim run-time user without dropping privileges.
467 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
468 result string, instead of calling string_vformat() twice with the same
471 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
472 for other users. Others should always drop root privileges if they use
473 -C on the command line, even for a whitelisted configure file.
475 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
477 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
483 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
484 $data_path, and $header_path variables; fixed documentation bugs and
487 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
488 exipick to access non-standard spools, including the "frozen" queue
491 NM/01 Bugzilla 965: Support mysql stored procedures.
492 Patch from Alain Williams
494 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
496 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
497 Patch from Andreas Metzler
499 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
500 Patch from Kirill Miazine
502 NM/05 Bugzilla 671: Added umask to procmail example.
504 JJ/03 installed exipick 20100323.0, fixing doc bug
506 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
507 directory. Notification and patch from Dan Rosenberg.
509 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
511 TK/02 Improve log output when DKIM signing operation fails.
513 MH/01 Treat the transport option dkim_domain as a colon separated
514 list, not as a single string, and sign the message with each element,
515 omitting multiple occurences of the same signer.
517 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
518 Bugzilla 985, 986. Patch by Simon Arlott
520 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
521 Patch by Simon Arlott
523 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
524 MBX locking. Notification from Dan Rosenberg.
530 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
532 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
534 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
536 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
538 NM/04 Bugzilla 915: Flag broken perl installation during build.
544 TK/01 Added patch by Johannes Berg that expands the main option
545 "spamd_address" if it starts with a dollar sign.
547 TK/02 Write list of recipients to X-Envelope-Sender header when building
548 the mbox-format spool file for content scanning (suggested by Jakob
551 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
552 (http://www.dcc-servers.net/) support via dccifd. Activated by
553 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
555 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
556 by Mark Daniel Reidel <mr@df.eu>.
558 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
559 When building exim an external PCRE library is now needed -
560 PCRE is a system library on the majority of modern systems.
561 See entry on PCRE_LIBS in EDITME file.
563 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
564 conversation. Added nologin parameter to request.
565 Patch contributed by Kirill Miazine.
567 TF/01 Do not log submission mode rewrites if they do not change the address.
569 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
571 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
572 log files in place. Contributed by Roberto Lima.
574 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
576 TF/03 Bugzilla 615: When checking the local_parts router precondition
577 after a local_part_suffix or local_part_prefix option, Exim now
578 does not use the address's named list lookup cache, since this
579 contains cached lookups for the whole local part.
581 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
582 Robert Millan. Documentation is in experimental-spec.txt.
584 TF/04 Bugzilla 668: Fix parallel build (make -j).
586 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
588 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
589 Patch provided by Jan Srzednicki.
591 TF/05 Leading white space used to be stripped from $spam_report which
592 wrecked the formatting. Now it is preserved.
594 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
595 that they are available at delivery time.
597 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
599 TF/08 TLS error reporting now respects the incoming_interface and
600 incoming_port log selectors.
602 TF/09 Produce a more useful error message if an SMTP transport's hosts
603 setting expands to an empty string.
605 NM/06 Bugzilla 744: EXPN did not work under TLS.
606 Patch provided by Phil Pennock.
608 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
609 Patch provided by Richard Godbee.
611 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
612 acl_smtp_notquit, added index entry.
614 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
615 Patch provided by Eugene Bujak.
617 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
618 accept(). Patch provided by Maxim Dounin.
620 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
621 Patch provided by Phil Pennock.
623 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
625 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
626 Patch provided by Brad "anomie" Jorsch.
628 NM/14 Bugzilla 622: Added timeout setting to transport filter.
629 Patch provided by Dean Brooks.
631 TK/05 Add native DKIM support (does not depend on external libraries).
633 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
634 Patch provided by Graeme Fowler.
636 NM/16 Bugzilla 851: Documentation example syntax fix.
638 NM/17 Changed NOTICE file to remove references to embedded PCRE.
640 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
643 NM/19 Bugzilla 745: TLS version reporting.
644 Patch provided by Phil Pennock.
646 NM/20 Bugzilla 167: bool: condition support.
647 Patch provided by Phil Pennock.
649 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
650 clients. Patch provided by Phil Pennock.
652 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
653 Patch provided by Brad "anomie" Jorsch.
655 NM/23 Bugzilla 687: Fix misparses in eximstats.
656 Patch provided by Heiko Schlittermann.
658 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
659 Patch provided by Heiko Schlittermann.
661 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
662 plus update to original patch.
664 NM/26 Bugzilla 799: Documentation correction for ratelimit.
666 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
667 Patch provided by David Brownlee.
669 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
671 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
673 NM/30 Bugzilla 888: TLS documentation bugfixes.
675 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
677 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
678 Unlike the original bugzilla I have changed all shell scripts in src tree.
680 NM/33 Bugzilla 898: Transport filter timeout fix.
681 Patch by Todd Rinaldo.
683 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
684 Patch by Serge Demonchaux.
686 NM/35 Bugzilla 39: Base64 decode bug fixes.
687 Patch by Jakob Hirsch.
689 NM/36 Bugzilla 909: Correct connect() call in dcc code.
691 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
693 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
695 NM/39 Bugzilla 911: Fixed MakeLinks build script.
701 TK/01 Add preliminary DKIM support. Currently requires a forked version of
702 ALT-N's libdkim that I have put here:
703 http://duncanthrax.net/exim-experimental/
705 Note to Michael Haardt: I had to rename some vars in sieve.c. They
706 were called 'true' and it seems that C99 defines that as a reserved
707 keyword to be used with 'bool' variable types. That means you could
708 not include C99-style headers which use bools without triggering
709 build errors in sieve.c.
711 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
712 as mailq or other aliases. Changed the --help handling significantly
713 to do whats expected. exim_usage() emits usage/help information.
715 SC/01 Added the -bylocaldomain option to eximstats.
717 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
719 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
721 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
727 PH/01 Another patch from the Sieve maintainer.
729 PH/02 When an IPv6 address is converted to a string for single-key lookup
730 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
731 dots are used instead of colons so that keys in lsearch files need not
732 contain colons. This was done some time before quoting was made available
733 in lsearch files. However, iplsearch files do require colons in IPv6 keys
734 (notated using the quote facility) so as to distinguish them from IPv4
735 keys. This meant that lookups for IP addresses in host lists did not work
736 for iplsearch lookups.
738 This has been fixed by arranging for IPv6 addresses to be expressed with
739 colons if the lookup type is iplsearch. This is not incompatible, because
740 previously such lookups could never work.
742 The situation is now rather anomolous, since one *can* have colons in
743 ordinary lsearch keys. However, making the change in all cases is
744 incompatible and would probably break a number of configurations.
746 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
749 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
750 conversion specification without a maximum field width, thereby enabling
751 a rogue spamd server to cause a buffer overflow. While nobody in their
752 right mind would setup Exim to query an untrusted spamd server, an
753 attacker that gains access to a server running spamd could potentially
754 exploit this vulnerability to run arbitrary code as the Exim user.
756 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
757 $primary_hostname instead of what libspf2 thinks the hosts name is.
759 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
760 a directory entry by the name of the lookup key. Previously, if a
761 symlink pointed to a non-existing file or a file in a directory that
762 Exim lacked permissions to read, a lookup for a key matching that
763 symlink would fail. Now it is enough that a matching directory entry
764 exists, symlink or not. (Bugzilla 503.)
766 PH/03 The body_linecount and body_zerocount variables are now exported in the
769 PH/04 Added the $dnslist_matched variable.
771 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
772 This means they are set thereafter only if the connection becomes
775 PH/06 Added the client_condition to authenticators so that some can be skipped
776 by clients under certain conditions.
778 PH/07 The error message for a badly-placed control=no_multiline_responses left
779 "_responses" off the end of the name.
781 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
783 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
784 (without spaces) instead of just copying the configuration text.
786 PH/10 Added the /noupdate option to the ratelimit ACL condition.
788 PH/11 Added $max_received_linelength.
790 PH/12 Added +ignore_defer and +include_defer to host lists.
792 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
793 way in which PCRE > 7.0 is built.
795 PH/14 Implemented queue_only_load_latch.
797 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
798 MAIL command. The effect was to mangle the value on 64-bit systems.
800 PH/16 Another patch from the Sieve maintainer.
802 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
804 PH/18 If a system quota error occurred while trying to create the file for
805 a maildir delivery, the message "Mailbox is full" was not appended to the
806 bounce if the delivery eventually timed out. Change 4.67/27 below applied
807 only to a quota excession during the actual writing of the file.
809 PH/19 It seems that peer DN values may contain newlines (and other non-printing
810 characters?) which causes problems in log lines. The DN values are now
811 passed through string_printing() before being added to log lines.
813 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
814 and InterBase are left for another time.)
816 PH/21 Added message_body_newlines option.
818 PH/22 Guard against possible overflow in moan_check_errorcopy().
820 PH/23 POSIX allows open() to be a macro; guard against that.
822 PH/24 If the recipient of an error message contained an @ in the local part
823 (suitably quoted, of course), incorrect values were put in $domain and
824 $local_part during the evaluation of errors_copy.
830 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
831 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
834 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
835 issue a MAIL command.
837 PH/02 In an ACL statement such as
839 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
841 if a client was not listed at all, or was listed with a value other than
842 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
843 the condition was not true (as it should be), so access was not denied.
844 The bug was that the ! inversion was incorrectly passed on to the second
845 item. This has been fixed.
847 PH/03 Added additional dnslists conditions == and =& which are different from
848 = and & when the dns lookup returns more than one IP address.
850 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
851 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
853 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
854 FSYNC, which compiles an option called disable_fsync that allows for
855 bypassing fsync(). The documentation is heavily laced with warnings.
857 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
859 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
860 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
861 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
862 including adding "make clean"; (3) Added -fPIC when compiling the test
863 dynamically loaded module, to get rid of a warning.
865 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
866 message fails, move_frozen_messages = true and ignore_bounce_errors_after
867 = 0s. The bug is otherwise harmless.
869 PH/07 There was a bug in the dovecot authenticator such that the value of
870 $auth1 could be overwritten, and so not correctly preserved, after a
871 successful authentication. This usually meant that the value preserved by
872 the server_setid option was incorrect.
874 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
876 PH/09 Installed PCRE release 7.0.
878 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
879 run for batched SMTP input. It is now run at the start of every message
880 in the batch. While fixing this I discovered that the process information
881 (output by running exiwhat) was not always getting set for -bs and -bS
882 input. This is fixed, and it now also says "batched" for BSMTP.
884 PH/11 Added control=no_pipelining.
886 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
887 patch, slightly modified), and move the expansion of helo_data till after
888 the connection is made in the smtp transport (so it can use these
891 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
893 PH/14 Added log_selector = +pid.
895 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
897 PH/16 Add ${if forany and ${if forall.
899 PH/17 Added dsn_from option to vary the From: line in DSNs.
901 PH/18 Flush SMTP output before performing a callout, unless control =
902 no_callout_flush is set.
904 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
905 was true (the default) a successful delivery failed to delete the retry
906 item, thus causing premature timeout of the address. The bug is now
909 PH/20 Added hosts_avoid_pipelining to the smtp transport.
911 PH/21 Long custom messages for fakedefer and fakereject are now split up
912 into multiline reponses in the same way that messages for "deny" and
913 other ACL rejections are.
915 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
916 with slight modification.
918 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
919 draft, changing the syntax and factoring some duplicate code".
921 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
922 for deliveries of the second and subsequent messages over the same SMTP
925 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
926 ${reduce, with only minor "tidies".
928 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
930 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
931 expansion side effects.
933 PH/27 When a message times out after an over-quota error from an Exim-imposed
934 quota, the bounce message says "mailbox is full". This message was not
935 being given when it was a system quota that was exceeded. It now should
938 MH/03 Made $recipients available in local_scan(). local_scan() already has
939 better access to the recipient list through recipients_list[], but
940 $recipients can be useful in postmaster-provided expansion strings.
942 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
943 in the case of a MAIL command with additional options following the
944 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
945 were accidentally chopped off.
947 PH/29 SMTP synchronization checks are implemented when a command is read -
948 there is a check that no more input is waiting when there shouldn't be
949 any. However, for some commands, a delay in an ACL can mean that it is
950 some time before the response is written. In this time, more input might
951 arrive, invalidly. So now there are extra checks after an ACL has run for
952 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
953 pipelining has not been advertised.
955 PH/30 MH's patch to allow iscntrl() characters to be list separators.
957 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
958 returned in the SMTP response when smtp_return_error_details was false.
961 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
962 instead of the C I/O that was originally supplied, because problems were
965 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
966 Exim which did not show up earlier: it was assuming that a call to
967 SSL_CTX_set_info_callback() might give an error value. In fact, there is
968 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
969 was a macro that became an assignment, so it seemed to work. This has
970 changed to a proper function call with a void return, hence the compile
971 error. Exim's code has been fixed.
973 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
976 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
978 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
980 PH/37 If a message is not accepted after it has had an id assigned (e.g.
981 because it turns out to be too big or there is a timeout) there is no
982 "Completed" line in the log. When some messages of this type were
983 selected by exigrep, they were listed as "not completed". Others were
984 picked up by some special patterns. I have improved the selection
985 criteria to be more general.
987 PH/38 The host_find_failed option in the manualroute router can now be set
988 to "ignore", to completely ignore a host whose IP address cannot be
989 found. If all hosts are ignored, the behaviour is controlled by the new
990 host_all_ignored option.
992 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
993 homing or because of multiple MX records with /mx) generated more than
994 one IP address, and the following item turned out to be the local host,
995 all the secondary addresses of the first item were incorrectly removed
996 from the list, along with the local host and any following hosts (which
997 is what is supposed to happen).
999 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1000 whoever called Exim into the -H file. In the case of the daemon it was
1001 behaving confusingly. When first started, it used values for whoever
1002 started the daemon, but after a SIGHUP it used the Exim user (because it
1003 calls itself on a restart). I have changed the code so that it now always
1006 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1007 message are rejected with the same error (e.g. no authentication or bad
1008 sender address), and a DATA command is nevertheless sent (as can happen
1009 with PIPELINING or a stupid MUA), the error message that was given to the
1010 RCPT commands is included in the rejection of the DATA command. This is
1011 intended to be helpful for MUAs that show only the final error to their
1014 PH/42 Another patch from the Sieve maintainer.
1016 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1017 Eximstats - Fixed some broken HTML links and added missing column headers
1019 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1020 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1022 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1024 PH/43 Yet another patch from the Sieve maintainer.
1026 PH/44 I found a way to check for a TCP/IP connection going away before sending
1027 the response to the final '.' that terminates a message, but only in the
1028 case where the client has not sent further data following the '.'
1029 (unfortunately, this is allowed). However, in many cases there won't be
1030 any further data because there won't be any more messages to send. A call
1031 to select() can be used: if it shows that the input is "ready", there is
1032 either input waiting, or the socket has been closed. An attempt to read
1033 the next input character can distinguish the two cases. Previously, Exim
1034 would have sent an OK response which the client would never have see.
1035 This could lead to message repetition. This fix should cure that, at
1036 least in a lot of common cases.
1038 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1039 advertised in response to EHLO.
1045 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1046 fixed by 4.65/MH/01 (is this a record?) are fixed:
1048 (i) An empty string was always treated as zero by the numeric comparison
1049 operators. This behaviour has been restored.
1051 (ii) It is documented that the numeric comparison operators always treat
1052 their arguments as decimal numbers. This was broken in that numbers
1053 starting with 0 were being interpreted as octal.
1055 While fixing these problems I realized that there was another issue that
1056 hadn't been noticed. Values of message_size_limit (both the global option
1057 and the transport option) were treated as octal if they started with 0.
1058 The documentation was vague. These values are now always treated as
1059 decimal, and I will make that clear in the documentation.
1065 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1066 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1069 MH/01 Don't check that the operands of numeric comparison operators are
1070 integers when their expansion is in "skipping" mode (fixes bug
1071 introduced by 4.64-PH/07).
1073 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1074 child addresses, Exim now panics and dies. Previously, because the count
1075 is held in a short int, deliveries were likely to be lost. As such a
1076 large number of recipients for a single message is ridiculous
1077 (performance will be very, very poor), I have chosen to impose a limit
1078 rather than extend the field.
1084 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1085 leftover -K file (the existence of which was triggered by #402).
1086 While we were at it, introduced process PID as part of the -K
1087 filename. This should rule out race conditions when creating
1090 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1091 processing considerably. Previous code took too long for large mails,
1092 triggering a timeout which in turn triggers #401.
1094 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1095 in the DK code in transports.c. sendfile() is not really portable,
1096 hence the _LINUX specificness.
1098 TF/01 In the add_headers option to the mail command in an Exim filter,
1099 there was a bug that Exim would claim a syntax error in any
1100 header after the first one which had an odd number of characters
1103 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1104 callout verification, Exim cached a "reject" for the entire domain. This
1105 is correct for most verifications, but it is not correct for a recipient
1106 verification with use_sender or use_postmaster set, because in that case
1107 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1108 case of MAIL FROM:<> rejection from other early rejections (e.g.
1109 rejection of HELO). When verifying a recipient using a non-null MAIL
1110 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1111 Whatever the result of the callout, the value of the domain cache is
1112 left unchanged (for any other kind of callout, getting as far as trying
1113 RCPT means that the domain itself is ok).
1115 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1118 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1119 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1120 session, the daemon ignores ECONNECT errors and logs others; it now
1121 ignores EPIPE as well.
1123 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1124 (quoted-printable decoding).
1126 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1127 later the small subsequent patch to fix an introduced bug.
1129 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1131 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1133 PH/08 An error is now given if message_size_limit is specified negative.
1135 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1136 to be given (somewhat) arbitrary names.
1138 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1141 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1142 miscellaneous code fixes
1144 PH/10 Added the log_reject_target ACL modifier to specify where to log
1147 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1148 hostname. This is wrong, because it relates to the incoming message (and
1149 probably the interface on which it is arriving) and not to the outgoing
1150 callout (which could be using a different interface). This has been
1151 changed to use the value of the helo_data option from the smtp transport
1152 instead - this is what is used when a message is actually being sent. If
1153 there is no remote transport (possible with a router that sets up host
1154 addresses), $smtp_active_hostname is used.
1156 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1157 tweaks were necessary in order to get it to work (see also 21 below):
1158 (a) The code assumed that strncpy() returns a negative number on buffer
1159 overflow, which isn't the case. Replaced with Exim's string_format()
1161 (b) There were several signed/unsigned issues. I just did the minimum
1162 hacking in of casts. There is scope for a larger refactoring.
1163 (c) The code used strcasecmp() which is not a standard C function.
1164 Replaced with Exim's strcmpic() function.
1165 (d) The code set only $1; it now sets $auth1 as well.
1166 (e) A simple test gave the error "authentication client didn't specify
1167 service in request". It would seem that Dovecot has changed its
1168 interface. Fortunately there's a specification; I followed it and
1169 changed what the client sends and it appears to be working now.
1171 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1174 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1175 address A is aliased to B and C, where B exists and C does not. Without
1176 -v the output is "A verified" because verification stops after a
1177 successful redirection if more than one address is generated. However,
1178 with -v the child addresses are also verified. Exim was outputting "A
1179 failed to verify" and then showing the successful verification for C,
1180 with its parentage. It now outputs "B failed to verify", showing B's
1181 parentage before showing the successful verification of C.
1183 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1184 look up a TXT record in a specific list after matching in a combined
1187 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1188 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1189 they consult the DNS. I had assumed they would set it the way they
1190 wanted; and indeed my experiments on Linux seem to show that in some
1191 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1192 To be on the safe side, however, I have now made the interface to
1193 host_find_byname() similar to host_find_bydns(), with an argument
1194 containing the DNS resolver options. The host_find_byname() function now
1195 sets these options at its start, just as host_find_bydns() does. The smtp
1196 transport options dns_qualify_single and dns_search_parents are passed to
1197 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1198 of host_find_byname() use the default settings of RES_DEFNAMES
1199 (qualify_single) but not RES_DNSRCH (search_parents).
1201 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1202 spool_read_header() do less string testing, by means of a preliminary
1203 switch on the second character of optional "-foo" lines. (This is
1204 overdue, caused by the large number of possibilities that now exist.
1205 Originally there were few.) While I was there, I also converted the
1206 str(n)cmp tests so they don't re-test the leading "-" and the first
1207 character, in the hope this might squeeze out yet more improvement.
1209 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1210 flag allowing group syntax was set by the header_syntax check but not
1211 turned off, possible causing trouble later; (2) The flag was not being
1212 set at all for the header_verify test, causing "group"-style headers to
1213 be rejected. I have now set it in this case, and also caused header_
1214 verify to ignore an empty address taken from a group. While doing this, I
1215 came across some other cases where the code for allowing group syntax
1216 while scanning a header line wasn't quite right (mostly, not resetting
1217 the flag correctly in the right place). These bugs could have caused
1218 trouble for malformed header lines. I hope it is now all correct.
1220 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1221 with the "reply" argument non-NULL. The code, however (which originally
1222 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1223 but it didn't always do it. This confused somebody who was copying the
1224 code for some other use. I have removed all the tests.
1226 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1227 feature that was used to support insecure browsers during the U.S. crypto
1228 embargo. It requires special client support, and Exim is probably the
1229 only MTA that supported it -- and would never use it because real RSA is
1230 always available. This code has been removed, because it had the bad
1231 effect of slowing Exim down by computing (never used) parameters for the
1232 RSA_EXPORT functionality.
1234 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1235 authenticator to fail if there's a tab character in the incoming data
1236 (there should never be unless someone is messing about, as it's supposed
1237 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1238 if the connection is using TLS or if the remote IP is the same as the
1239 local IP, and the "valid-client-cert option" if a client certificate has
1242 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1243 authenticators. This can be used for authorization after authentication
1244 succeeds. (In the case of plaintext, it servers for both authentication
1247 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1248 if any retry times were supplied.
1250 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1251 connection, where there is no client IP address to check. In this
1252 situation, the verify now always succeeds.
1254 PH/25 Applied John Jetmore's -Mset patch.
1256 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1258 PH/27 In a string expansion for a processed (not raw) header when multiple
1259 headers of the same name were present, leading whitespace was being
1260 removed from all of them, but trailing whitespace was being removed only
1261 from the last one. Now trailing whitespace is removed from each header
1262 before concatenation. Completely empty headers in a concatenation (as
1263 before) are ignored.
1265 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1266 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1268 PH/29 [Removed. This was a change that I later backed out, and forgot to
1269 correct the ChangeLog entry (that I had efficiently created) before
1270 committing the later change.]
1272 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1273 address errors (4xx response to RCPT) over the same connection as other
1274 messages routed to the same hosts. Such deliveries are always "forced",
1275 so retry times are not inspected. This resulted in far too many retries
1276 for the affected addresses. The effect occurred only when there were more
1277 hosts than the hosts_max_try setting in the smtp transport when it had
1278 the 4xx errors. Those hosts that it had tried were not added to the list
1279 of hosts for which the message was waiting, so if all were tried, there
1280 was no problem. Two fixes have been applied:
1282 (i) If there are any address or message errors in an SMTP delivery, none
1283 of the hosts (tried or untried) are now added to the list of hosts
1284 for which the message is waiting, so the message should not be a
1285 candidate for sending over the same connection that was used for a
1286 successful delivery of some other message. This seems entirely
1287 reasonable: after all the message is NOT "waiting for some host".
1288 This is so "obvious" that I'm not sure why it wasn't done
1289 previously. Hope I haven't missed anything, but it can't do any
1290 harm, as the worst effect is to miss an optimization.
1292 (ii) If, despite (i), such a delivery is accidentally attempted, the
1293 routing retry time is respected, so at least it doesn't keep
1294 hammering the server.
1296 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1297 in ${readsocket because some servers need this prod.
1299 PH/32 Added some extra debug output when updating a wait-xxx database.
1301 PH/33 The hint "could be header name not terminated by colon", which has been
1302 given for certain expansion errors for a long time, was not being given
1303 for the ${if def:h_colon_omitted{... case.
1305 PH/34 The spec says: "With one important exception, whenever a domain list is
1306 being scanned, $domain contains the subject domain." There was at least
1307 one case where this was not true.
1309 PH/35 The error "getsockname() failed: connection reset by peer" was being
1310 written to the panic log as well as the main log, but it isn't really
1311 panic-worthy as it just means the connection died rather early on. I have
1312 removed the panic log writing for the ECONNRESET error when getsockname()
1315 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1316 runs only) independently of the message's sender address. This meant
1317 that, if the 4xx error was in fact related to the sender, a different
1318 message to the same recipient with a different sender could confuse
1319 things. In particualar, this can happen when sending to a greylisting
1320 server, but other circumstances could also provoke similar problems.
1321 I have changed the default so that the retry time for these errors is now
1322 based a combination of the sender and recipient addresses. This change
1323 can be overridden by setting address_retry_include_sender=false in the
1326 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1327 remote server are returned as part of bounce messages. This was not
1328 happening for LMTP over a pipe (the lmtp transport), but now it is the
1329 same for both kinds of LMTP.
1331 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1332 in header lines that were in fact CNAMEs. This is no longer the case.
1334 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1335 and queue runs started by the daemon processed all messages. This has
1336 been fixed so that -R and -S can now usefully be given with -q<time>.
1338 PH/40 Import PCRE release 6.7 (fixes some bugs).
1340 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1342 PH/42 Give an error if -q is specified more than once.
1344 PH/43 Renamed the variables $interface_address and $interface_port as
1345 $received_ip_address and $received_port, to make it clear that these
1346 values apply to message reception, and not to the outgoing interface when
1347 a message is delivered. (The old names remain recognized, of course.)
1349 PH/44 There was no timeout on the connect() call when using a Unix domain
1350 socket in the ${readsocket expansion. There now is.
1352 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1353 be meaningful with "accept".
1355 SC/01 Eximstats V1.43
1356 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1358 SC/02 Eximstats V1.44
1359 Use a glob alias rather than an array ref in the generated
1360 parser. This improves both readability and performance.
1362 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1363 Collect SpamAssassin and rejection statistics.
1364 Don't display local sender or destination tables unless
1365 there is data to show.
1366 Added average volumes into the top table text output.
1368 SC/04 Eximstats V1.46
1369 Collect data on the number of addresses (recipients)
1370 as well as the number of messages.
1372 SC/05 Eximstats V1.47
1373 Added 'Message too big' to the list of mail rejection
1374 reasons (thanks to Marco Gaiarin).
1376 SC/06 Eximstats V1.48
1377 Mainlog lines which have GMT offsets and are too short to
1378 have a flag are now skipped.
1380 SC/07 Eximstats V1.49 (Alain Williams)
1381 Added the -emptyok flag.
1383 SC/08 Eximstats V1.50
1384 Fixes for obtaining the IP address from reject messages.
1386 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1387 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1388 whitesspace changes from 4.64-PH/27
1390 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1393 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1394 are found, allow negative numbers in numeric criteria)
1396 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1398 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1399 to match changes made in 4.64-PH/43
1401 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1403 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1404 "domain" is actually an IP address, and force a failure. This locks out
1405 those revolvers/nameservers that support "A-for-A" lookups, in
1406 contravention of the specifications.
1408 PH/48 When a host name was looked up from an IP address, and the subsequent
1409 forward lookup of the name timed out, the host name was left in
1410 $sender_host_name, contrary to the specification.
1412 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1413 restricted to single-key lookups, Exim was not diagnosing an error if
1414 * or *@ was used with a query-style lookup.
1416 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1418 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1419 long ago, but noone interested enough thought of it. Let's just say that
1420 the "1.1" means that there are some new functions that weren't there at
1421 some point in the past.
1423 PH/51 Error processing for expansion failure of helo_data from an smtp
1424 transport during callout processing was broken.
1426 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1427 tested/used via the -bh/-bhc/-bs options.
1429 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1430 bug, fixed in subsequent PCRE releases).
1432 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1433 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1435 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1441 SC/01 Use a glob alias rather than an array ref in eximstats generated
1442 parser. This improves both readability and performance.
1444 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1445 Don't display local sender or destination tables in eximstats unless
1446 there is data to show.
1447 Added average volumes into the eximstats top table text output.
1449 SC/03 Collect data on the number of addresses (recipients) as well
1450 as the number of messages in eximstats.
1452 TF/01 Correct an error in the documentation for the redirect router. Exim
1453 does (usually) call initgroups() when daemonizing.
1455 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1456 with consistent privilege compared to when running as a daemon.
1458 TF/03 Note in the spec that $authenticated_id is not set for local
1459 submissions from trusted users.
1461 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1462 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1464 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1465 by adding some example configuration directives to the default
1466 configuration file. A little bit of work is required to uncomment the
1467 directives and define how usernames and passwords are checked, but
1468 there is now a framework to start from.
1470 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1471 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1472 without this. I don't know how relevant this is to other LDAP libraries.
1474 PH/02 Add the verb name to the "unknown ACL verb" error.
1476 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1478 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1480 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1481 directory not expanded when it should be if an expanded home directory
1482 was set for the address (which is overridden by the transport).
1484 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1487 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1488 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1489 because it is too late at that time, and has no effect.
1491 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1492 security issue with \' (bugzilla #107). I could not use the
1493 PQescapeStringConn() function, because it needs a PGconn value as one of
1496 PH/08 When testing addresses using -bt, indicate those final addresses that
1497 are duplicates that would not cause an additional delivery. At least one
1498 person was confused, thinking that -bt output corresponded to deliveries.
1499 (Suppressing duplicates isn't a good idea as you lose the information
1500 about possibly different redirections that led to the duplicates.)
1502 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1503 systems where poll() doesn't work, in particular OS X.
1505 PH/10 Added more information to debugging output for retry time not reached.
1507 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1508 operations in malware.c.
1510 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1513 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1514 syslog_duplication set false, log lines that would normally be written
1515 both the the main log and to the reject log were not written to syslog at
1518 PH/14 In the default configuration, change the use of "message" in ACL warn
1519 statements to "add_header".
1521 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1522 not followed by a command (e.g. "seen endif").
1524 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1525 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1528 PH/17 Added extra conditions to the default value of delay_warning_condition
1532 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1533 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1534 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1537 The Auto-Submitted: and various List- headers are standardised, whereas I
1538 don't think Precedence: ever was.
1540 PH/18 Refactored debugging code in route_finduser() to show more information,
1541 in particular, the error code if getpwnam() issues one.
1543 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1544 This is apparently needed in addition to the PH/07 change above to avoid
1545 any possible encoding problems.
1547 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1548 but not after initializing Perl.
1550 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1551 output them only if debugging. By default they are written stderr,
1552 apparently, which is not desirable.
1554 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1557 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1560 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1562 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1563 authenticated or an ident call has been made. Suppress the default
1564 values for $authenticated_id and $authenticated_sender (but permit -oMai
1565 and -oMas) when testing with -bh.
1567 PH/24 Re-jigged the order of the tests in the default configuration so that the
1568 tests for valid domains and recipients precede the DNS black list and CSA
1569 tests, on the grounds that those ones are more expensive.
1571 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1572 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1573 This bug exists in every version of Exim that I still have, right back to
1576 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1577 However, an attempt to turn on case-sensitivity in a regex key by
1578 including (?-i) didn't work because the subject string was already
1579 lowercased, and the effects were non-intuitive. It turns out that a
1580 one-line patch can be used to allow (?-i) to work as expected.
1586 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1587 other effects) broke the use of negated acl sub-conditions.
1589 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1592 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1593 "Deny" causes Exim to reject the incoming connection with a 554 error.
1594 Unfortunately, if there is a major crisis, such as a disk failure,
1595 tcp-wrappers gives "deny", whereas what one would like would be some
1596 kind of temporary error. A kludge has been added to help with this.
1597 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1598 554 error is used if errno is still zero or contains ENOENT (which occurs
1599 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1602 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1604 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1605 errors. Otherwise a message that provokes a temporary error (when other
1606 messages do not) can cause a whole host to time out.
1608 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1609 addresses were routed directly to files or pipes from a redirect router.
1610 File deliveries just didn't batch; pipe deliveries might have suffered
1613 PH/06 A failure to get a lock for a hints database would erroneously always say
1614 "Failed to get write lock", even when it was really a read lock.
1616 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1617 of 0600. This has been changed to use the value of the lockfile_mode
1618 option (which defaults to 0600).
1620 PH/08 Applied small patch from the Sieve maintainer.
1622 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1623 folder from quota calculations, a direct delivery into this folder messed
1624 up the contents of the maildirsize file. This was because the regex was
1625 used only to exclude .Trash (or whatever) when the size of the mailbox
1626 was calculated. There was no check that a delivery was happening into an
1627 excluded directory. This bug has been fixed by ignoring all quota
1628 processing for deliveries into excluded directories.
1630 PH/10 Added the maildirfolder_create_regex option to appendfile.
1636 PH/01 The code for finding all the local interface addresses on a FreeBSD
1637 system running IPv6 was broken. This may well have applied to all BSD
1638 systems, as well as to others that have similar system calls. The broken
1639 code found IPv4 interfaces correctly, but gave incorrect values for the
1640 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1641 that it would not match correctly against @[] and not recognize the IPv6
1644 PH/02 The ipliteral router was not recognizing addresses of the form user@
1645 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1647 PH/03 Added disable_ipv6.
1649 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1650 decoded form, because it is most often used to construct To: headers
1651 lines in autoreplies, and the decoded form may well be syntactically
1652 invalid. However, $reply_address has leading white space removed, and all
1653 newlines turned into spaces so that the autoreply transport does not
1656 PH/05 If group was specified without a user on a router, and no group or user
1657 was specified on a transport, the group from the router was ignored.
1659 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1660 for visible compile-time settings that can be used to change these
1661 numbers, for those that want even more. Backwards compatibility with old
1662 spool files has been maintained. However, going back to a previous Exim
1663 release will lost any variables that are in spool files.
1665 PH/07 Two small changes when running in the test harness: increase delay when
1666 passing a TCP/IP connection to a new process, in case the original
1667 process has to generate a bounce, and remove special handling of
1668 127.0.0.2 (sic), which is no longer necessary.
1670 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1671 be the same on different OS.
1673 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1676 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1677 whether --show-vars was specified or not
1679 JJ/02 exipick: Added support for new ACL variable spool format introduced
1682 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1683 syntactically invalid From: or Reply-to: line, and a filter used this to
1684 generate an autoreply, and therefore failed to obtain an address for the
1685 autoreply, Exim could try to deliver to a non-existent relative file
1686 name, causing unrelated and misleading errors. What now happens is that
1687 it logs this as a hard delivery error, but does not attempt to create a
1690 PH/11 The exinext utility has a -C option for testing purposes, but although
1691 the given file was scanned by exinext itself; it wasn't being passed on
1692 when Exim was called.
1694 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1695 an end-of-file indication when reading a command response.
1697 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1698 compiled. In many other places in Exim, IPv6 addresses are always
1699 recognized, so I have changed this. It also means that IPv4 domain
1700 literals of the form [IPV4:n.n.n.n] are now always recognized.
1702 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1703 used if the router is not running as root, for example, when verifying at
1704 ACL time, or when using -bh. The debugging output from this situation was
1705 non-existent - all you got was a failure to exec. I have made two
1708 (a) Failures to set uid/gid, the current directory, or a process leader
1709 in a subprocess such as that created by queryprogram now generate
1710 suitable debugging ouput when -d is set.
1712 (b) The queryprogram router detects when it is not running as root,
1713 outputs suitable debugging information if -d is set, and then runs
1714 the subprocess without attempting to change uid/gid.
1716 PH/15 Minor change to Makefile for building test_host (undocumented testing
1719 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1720 additional section of a DNS packet that returns MX or SRV records.
1721 Instead, it always explicitly searches for A/AAAA records. This avoids
1722 major problems that occur when a DNS server includes only records of one
1723 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1724 fixed another bug: if SRV records were looked up and the corresponding
1725 address records were *not* found in the additional section, the port
1726 values from the SRV records were lost.
1728 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1729 using the correct key (the original address) when searching the retry
1730 rules in order to find which one to use for generating the retry hint.
1732 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1733 adding the default one. Similarly, if it contains a Reply-To: header, the
1734 errors_reply_to option, if set, is not used.
1736 PH/19 When calculating a retry time, Exim used to measure the "time since
1737 failure" by looking at the "first failed" field in the retry record. Now
1738 it does not use this if it is later than than the arrival time of the
1739 message. Instead it uses the arrival time. This makes for better
1740 behaviour in cases where some deliveries succeed, thus re-setting the
1741 "first failed" field. An example is a quota failure for a huge message
1742 when small messages continue to be delivered. Without this change, the
1743 "time since failure" will always be short, possible causing more frequent
1744 delivery attempts for the huge message than are intended.
1745 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1747 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1748 $1, $2, $3) because the numerical variables can be reset during some
1749 expansion items (e.g. "match"), thereby losing the authentication data.
1751 PH/21 Make -bV show the size of off_t variables so that the test suite can
1752 decide whether to run tests for quotas > 2G.
1754 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1755 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1756 filecount value is greater than 2G or if a quota value is greater than 2G
1757 on a system where the size of off_t is not greater than 4, a panic error
1760 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1761 never match. The debug and -bh output now contains an explicit error
1762 message indicating a malformed IPv4 address or mask.
1764 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1765 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1766 PH/23 above applies.
1768 PH/25 Do not write to syslog when running in the test harness. The only
1769 occasion when this arises is a failure to open the main or panic logs
1770 (for which there is an explicit test).
1772 PH/26 Added the /no_tell option to "control=freeze".
1774 PH/27 If a host name lookup failed very early in a connection, for example, if
1775 the IP address matched host_lookup and the reverse lookup yielded a name
1776 that did not have a forward lookup, an error message of the form "no IP
1777 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1778 could be logged. Now it outputs the IP address instead of "NULL".
1780 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1781 allows the sender and the authenticated sender to be set when
1782 submitting a message from within Exim. Since child_open_exim() is
1783 documented for local_scan(), the new function should be too.
1785 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1786 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1787 results in an empty string is now treated as unset.
1789 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1791 PH/31 Added sender_verify_fail logging option.
1793 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1794 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1795 tidied the source and removed it altogether.
1797 PH/33 When a queue run was abandoned because the load average was too high, a
1798 log line was always written; now it is written only if the queue_run log
1799 selector is set. In addition, the log line for abandonment now contains
1800 information about the queue run such as the pid. This is always present
1801 in "start" and "stop" lines but was omitted from the "abandon" line.
1803 PH/34 Omit spaces between a header name and the colon in the error message that
1804 is given when verify = headers_syntax fails (if there are lots of them,
1805 the message gets confusing).
1807 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1808 names, as there are now some PTR records that contain slashes. This check
1809 is only to protect against broken name servers that fall over on strange
1810 characters, so the fact that it applies to all lookups doesn't matter.
1812 PH/36 Now that the new test suite is complete, we can remove some of the
1813 special code in Exim that was needed for the old test suite. For example,
1814 sorting DNS records because real resolvers return them in an arbitrary
1815 order. The new test suite's fake resolver always returns records in the
1818 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1819 bounces) except when queue_only is set, to avoid logging races between
1820 the different processes.
1822 PH/38 Panic-die if .include specifies a non-absolute path.
1824 PH/39 A tweak to the "H" retry rule from its user.
1826 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1827 a label. They prevented compilation on older perls.
1829 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1830 a warning to be raised on newish perls.
1832 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1833 on queue. Changes to match documented behaviour of showing count of
1834 messages matching specified criteria.
1836 PH/40 Changed the default ident timeout from 30s to 5s.
1838 PH/41 Added support for the use of login_cap features, on those BSD systems
1839 that have them, for controlling the resources used by pipe deliveries.
1841 PH/42 The content-scanning code uses fopen() to create files in which to put
1842 message data. Previously it was not paying any attention to the mode of
1843 the files. Exim runs with umask(0) because the rest of the code creates
1844 files with open(), and sets the required mode explicitly. Thus, these
1845 files were ending up world-writeable. This was not a big issue, because,
1846 being within the spool directory, they were not world-accessible. I have
1847 created a function called modefopen, which takes an additional mode
1848 argument. It sets umask(777), creates the file, chmods it to the required
1849 mode, then resets the umask. All the relevant calls to fopen() in the
1850 content scanning code have been changed to use this function.
1852 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1853 to 24 hours. This avoids potential overflow problems when processing G
1854 and H retry rules. I suspect nobody ever tinkers with this value.
1856 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1858 PH/45 When the plaintext authenticator is running as a client, the server's
1859 challenges are checked to ensure they are valid base64 strings. By
1860 default, the authentication attempt is cancelled if an invalid string is
1861 received. Setting client_ignore_invalid_base64 true ignores these errors.
1862 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1863 they are received. Thus, the responses can be made to depend on the
1864 challenges. If an invalid string is ignored, an empty string is placed in
1867 PH/46 Messages that are created by the autoreply transport now contains a
1868 References: header, in accordance with RFCs 2822 and 3834.
1870 PH/47 Added authenticated_sender_force to the smtp transport.
1872 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1874 PH/49 Installed latest patch from the Sieve maintainer.
1876 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1877 was also set, the appendfile transport was unnecessarily scanning a
1878 directory of message files (e.g. for maildir delivery) to find the count
1879 of files (along with the size), even though it did not need this
1880 information. It now does the scan only if it needs to find either the
1881 size of the count of files.
1883 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1885 PH/52 Two bugs concerned with error handling when the smtp transport is
1888 (i) Exim was not creating retry information for temporary errors given
1889 for individual recipients after the DATA command when the smtp transport
1890 was used in LMTP mode. This meant that they could be retried too
1891 frequently, and not timed out correctly.
1893 (ii) Exim was setting the flag that allows error details to be returned
1894 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1895 recipients that were returned after the DATA command.
1897 PH/53 This is related to PH/52, but is more general: for any failing address,
1898 when detailed error information was permitted to be returned to the
1899 sender, but the error was temporary, then after the final timeout, only
1900 "retry timeout exceeded" was returned. Now it returns the full error as
1901 well as "retry timeout exceeded".
1903 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1904 do this, and (what is worse) MTAs that accept it.
1906 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1907 will now be deprecated.
1909 PH/56 New os.c-cygwin from the Cygwin maintainer.
1911 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1912 formats (previously only available in exim formats via -bpr, -bpru,
1913 and -bpra. Now also available in native and exiqgrep formats)
1915 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1916 with very large, slow to parse queues
1918 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1920 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1922 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1923 responses to authentication challenges, though it was showing the
1924 challenges; (ii) I've removed the CR characters from the debug output for
1927 PH/58 Allow for the insertion of a newline as well as a space when a string
1928 is turned into more than one encoded-word during RFC 2047 encoding. The
1929 Sieve code now uses this.
1931 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1932 data_4xx, lost_connection, tls_required.
1934 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1935 message was being sent as an SMTP response.
1937 PH/61 Add -l and -k options to exicyclog.
1939 PH/62 When verifying, if an address was redirected to one new address, so that
1940 verification continued, and the new address failed or deferred after
1941 having set something in $address_data, the value of $address_data was not
1942 passed back to the ACL. This was different to the case when no
1943 redirection occurred. The value is now passed back in both cases.
1945 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1946 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1947 use login_cap.h, so on its own it isn't the distinguishing feature. The
1948 new name refers directly to the setclassresources() function.
1950 PH/65 Added configuration files for NetBSD3.
1952 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1954 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1955 is preferred over IPv4.
1957 PH/68 The bounce_return_message and bounce_return_body options were not being
1958 honoured for bounces generated during the reception of non-SMTP messages.
1959 In particular, this applied to messages rejected by the ACL. This bug has
1960 been fixed. However, if bounce_return_message is true and bounce_return_
1961 body is false, the headers that are returned for a non-SMTP message
1962 include only those that have been read before the error was detected.
1963 (In the case of an ACL rejection, they have all been read.)
1965 PH/69 The HTML version of the specification is now built in a directory called
1966 spec_html instead of spec.html, because the latter looks like a path with
1967 a MIME-type, and this confuses some software.
1969 PH/70 Catch two compiler warnings in sieve.c.
1971 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1972 function verify_get_ident() calls ip_connect() to connect a socket, but
1973 if the "connect()" function timed out, ip_connect() used to close the
1974 socket. However, verify_get_ident() also closes the socket later, and in
1975 between Exim writes to the log, which may get opened at this point. When
1976 the socket was closed in ip_connect(), the log could get the same file
1977 descriptor number as the socket. This naturally causes chaos. The fix is
1978 not to close the socket in ip_connect(); the socket should be closed by
1979 the function that creates it. There was only one place in the code where
1980 this was missing, in the iplookup router, which I don't think anybody now
1981 uses, but I've fixed it anyway.
1983 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1984 well as to direct DNS lookups. Otherwise the handling of names in host
1985 lists is inconsistent and therefore confusing.
1991 PH/01 Two changes to the default runtime configuration:
1993 (1) Move the checks for relay_from_hosts and authenticated clients from
1994 after to before the (commented out) DNS black list checks.
1996 (2) Add control=submission to the relay_from_hosts and authenticated
1997 clients checks, on the grounds that messages accepted by these
1998 statements are most likely to be submissions.
2000 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2002 (1) Generate an error if the third argument for the ${prvs expansion is
2005 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2008 (3) Reset the variables that are obtained from the first argument of
2009 ${prvscheck and used in the second argument before leaving the code,
2010 because their memory is reclaimed, so using them afterwards may do
2013 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2014 one (it's much easier than Tom thought :-).
2016 (5) Because of (4), we can now allow for the use of $prvscheck_result
2017 inside the third argument.
2019 PH/03 For some reason, the default setting of PATH when running a command from
2020 a pipe transport was just "/usr/bin". I have changed it to
2023 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2024 anything to be listed in the output from -bV.
2026 PH/05 When a filter generated an autoreply, the entire To: header line was
2027 quoted in the delivery log line, like this:
2029 => >A.N.Other <ano@some.domain> <original@ddress> ...
2031 This has been changed so that it extracts the operative address. There
2032 may be more than one such address. If so, they are comma-separated, like
2035 => >ano@some.domain,ona@other.domain <original@ddress> ...
2037 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2038 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2039 not being looked up in the rDNS to get a host name, Exim was showing the
2040 IP address twice in Received: lines, even though the IP addresses were
2041 identical. For example:
2043 Received: from [1.2.3.4] (helo=[1.2.3.4])
2045 However, if the real host name was known, it was omitting the HELO data
2046 if it matched the actual IP address. This has been tidied up so that it
2047 doesn't show the same IP address twice.
2049 PH/07 When both +timestamp and +memory debugging was on, the value given by
2050 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2051 called by the debug printing, thereby overwriting the timestamp buffer.
2052 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2054 PH/08 When the original message was included in an autoreply transport, it
2055 always said "this is a copy of the message, including all the headers",
2056 even if body_only or headers_only was set. It now gives an appropriate
2059 PH/09 Applied a patch from the Sieve maintainer which:
2061 o fixes some comments
2062 o adds the (disabled) notify extension core
2063 o adds some debug output for the result of if/elsif tests
2064 o points to the current vacation draft in the documentation
2065 and documents the missing references header update
2069 o fixes a bug in processing the envelope test (when testing
2070 multiple envelope elements, the last element determinted the
2073 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2074 Electronic Mail") by including:
2076 Auto-submitted: auto-generated
2078 in the messages that it generates (bounce messages and others, such as
2079 warnings). In the case of bounce messages for non-SMTP mesages, there was
2080 also a typo: it was using "Auto_submitted" (underscore instead of
2081 hyphen). Since every message generated by Exim is necessarily in response
2082 to another message, thes have all been changed to:
2084 Auto-Submitted: auto-replied
2086 in accordance with these statements in the RFC:
2088 The auto-replied keyword:
2090 - SHOULD be used on messages sent in direct response to another
2091 message by an automatic process,
2093 - MUST NOT be used on manually-generated messages,
2095 - MAY be used on Delivery Status Notifications (DSNs) and Message
2096 Disposition Notifications (MDNs),
2098 - MUST NOT be used on messages generated by automatic or periodic
2099 processes, except for messages which are automatic responses to
2102 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2103 to the default Received: header definition.
2105 PH/12 Added log selector acl_warn_skipped (default on).
2107 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2108 variables because (a) they are in the wrong storage pool and (b) even if
2109 they were copied, it wouldn't work properly because of the caching.
2111 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2112 checking when decoding. Apparently there are clients that generate
2113 overlong encoded strings. Why am I not surprised?
2115 PH/15 If the first argument of "${if match_address" was not empty, but did not
2116 contain an "@" character, Exim crashed. Now it writes a panic log message
2117 and treats the condition as false.
2119 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2121 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2122 extension "envelope-auth". The code is finished and in agreement with
2123 other implementations, but there is no documentation so far and in fact,
2124 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2125 not changing the active code.
2127 Print executed "if" and "elsif" statements when debugging is used. This
2128 helps a great deal to understand what a filter does.
2130 Document more things not specified clearly in RFC3028. I had all this
2131 sorted out, when out of a sudden new issues came to my mind. Oops."
2133 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2136 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2137 canonical form (as documented). However, after a host name lookup from
2138 the IP address, check_host() was doing a simple string comparison with
2139 addresses acquired from the DNS when checking that the found name did
2140 have the original IP as one of its addresses. Since any found IPv6
2141 addresses are likely to be in abbreviated form, the comparison could
2142 fail. Luckily, there already exists a function for doing the comparison
2143 by converting both addresses to binary, so now that is used instead of
2144 the text comparison.
2146 PH/20 There was another similar case to PH/19, when a complete host name was
2147 given in a host list; looking up its IP address could give an abbreviated
2148 form, whereas the current host's name might or might not be abbreviated.
2149 The same fix has been applied.
2155 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2156 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2159 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2160 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2162 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2164 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2165 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2166 Also, for "Auto-Submitted", treat anything other than "no" as
2167 non-personal, in accordance with RFC 3834. (Previously it treated
2168 anything starting "auto-" as non-personal.)
2170 TF/01 The control=submission/name=... option had a problem with syntax
2171 errors if the name included a slash character. The /name= option
2172 now slurps the rest of the string, so it can include any characters
2173 but it must come last in the list of options (after /sender_retain
2176 PH/05 Some modifications to the interface to the fake nameserver for the new
2184 TK/01 Added the "success_on_redirect" address verification option. See
2185 NewStuff for rationale and an example.
2187 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2189 PH/02 Patch to exigrep to allow it to work on syslog lines.
2191 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2192 fread() to skip over the body file's header line, because in Cygwin the
2193 header line is locked and is inaccessible.
2195 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2196 co-exist for some time) to make it clear that it is the Exim ID that is
2197 referenced, not the Message-ID: header line.
2199 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2200 string_format() function, because snprintf() does not exist on all
2203 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2204 ${stat: expansion item.
2206 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2207 protocol synchronization error", to keep the pedants happy.
2209 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2210 well as for IRIX systems, when gcc is being used. See the host.c source
2213 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2215 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2218 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2219 transport and to the smtp transport in LMTP mode.
2221 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2223 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2224 run a filter in a subprocess. This could lead to confusion in subsequent
2225 lookups in the parent process. There should also be a search_tidyup() at
2226 the end of the subprocess.
2228 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2229 only if the host matched helo_try_verify_hosts, which caused the
2230 verification to occur when the EHLO/HELO command was issued. The ACL just
2231 tested the remembered result. Now, if a previous verification attempt has
2232 not happened, "verify = helo" does it there and then.
2234 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2236 TK/03 Fix log output including CR from clamd.
2238 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2239 exist provoked a memory error which could cause a segfault.
2241 PH/15 Installed PCRE 6.2
2243 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2245 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2246 of the problem. Specifically, suggested +O2 rather than +O1 for the
2249 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2251 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2252 record, the port was not correctly logged when the outgoing_port log
2253 selector was set (it logged the transort's default port).
2255 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2256 fallback_hosts, and "hosts" in the smtp transport.
2258 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2259 host errors such as "Connection refused".
2261 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2262 authentication with radiusclient 0.4.9:
2264 - Error returned from rc_read_config was caught wrongly
2265 - Username/password not passed on to radius server due to wrong length.
2267 The presumption is that some radiusclient API changes for 4.51/PH/17
2268 were not taken care of correctly. The code is still untested by me (my
2269 Linux distribution still has 0.3.2 of radiusclient), but it was
2270 contributed by a Radius user.
2272 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2273 expanding the "port" option of the smtp transport.
2275 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2276 while reading a MIME header. Thanks to Tom Hughes for a patch.
2278 PH/24 Include config.h inside local_scan.h so that configuration settings are
2281 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2282 that in an ACL for RCPT (for example), you can examine exactly what was
2285 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2286 commands, but it was not correctly comparing the address with the actual
2287 client host address. Thus, it would show the EHLO address in Received:
2288 header lines when this was not necessary.
2290 PH/27 Added the % operator to ${eval:}.
2292 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2293 it should be ignoring failures (because with -C, for example, it has lost
2294 privilege). It wasn't ignoring creation failures other than "already
2297 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2300 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2301 because an input line was too long, either on its own, or by virtue of
2302 too many continuations, the temporary file was not being removed, and the
2303 return code was incorrect.
2305 PH/31 Missing "BOOL" in function definition in filtertest.c.
2307 PH/32 Applied Sieve patches from the maintainer.
2309 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2311 PH/33 Added "verify = not_blind".
2313 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2314 Local/Makefile (with some defaults set). These are used in built scripts
2315 such as exicyclog, but they have never been used in the exim_install
2316 script (though there are many overriding facilities there). I have
2317 arranged that the exim_install script now takes note of these two
2320 PH/35 Installed configuration files for Dragonfly.
2322 PH/36 When a locally submitted message by a trusted user did not contain a
2323 From: header, and the sender address was obtained from -f or from an SMTP
2324 MAIL command, and the trusted user did not use -F to supply a sender
2325 name, $originator_name was incorrectly used when constructing a From:
2326 header. Furthermore, $originator_name was used for submission mode
2327 messages from external hosts without From: headers in a similar way,
2328 which is clearly wrong.
2330 PH/37 Added control=suppress_local_fixups.
2332 PH/38 When log_selector = +received_sender was set, and the addition of the
2333 sender made the log line's construction buffer exactly full, or one byte
2334 less than full, an overflow happened when the terminating "\n" was
2337 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2338 when the result of a list match is failure because a DNS lookup failed.
2340 PH/40 RM_COMMAND is now used in the building process.
2342 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2343 the "build-* directories that it finds.
2345 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2346 address in a domain literal was a prefix of an interface address.
2348 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2349 when verifying a sender address, unless rewrite_headers is false.
2351 PH/44 Wrote a long comment about why errors_to addresses are verified as
2352 recipients, not senders.
2354 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2355 the ratelimit ACL was added.
2357 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2359 PH/46 Added extra information about PostgreSQL errors to the error string.
2361 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2362 avoiding the need to install special zones in a real server. This is
2363 backwards compatible; if it can't find the fake resolver, it drops back.
2364 Thus, both old and new test suites can be run.
2366 TF/02 Added util/ratelimit.pl
2368 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2369 clock is set back in time.
2371 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2372 Candler <B.Candler@pobox.com>.
2374 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2375 OK for addresses that are the result of redirections.
2377 PH/48 A number of further additions for the benefit of the new test suite,
2378 including a fake gethostbyname() that interfaces to the fake DNS resolver
2381 TF/06 The fix for widen_domains has also been applied to qualify_single and
2382 search_parents which are the other dnslookup options that can cause
2385 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2388 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2390 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2391 service" if it encountered an MX record with an empty target hostname.
2392 The message is now "an MX or SRV record indicated no SMTP service".
2394 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2395 if the DNS of the sending site is misconfigured. This is quite a
2396 common situation. This change restores the behaviour of treating a
2397 helo verification defer as a failure.
2399 PH/51 If self=fail was set on a router, the bounce message did not include the
2400 actual error message.
2406 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2408 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2409 command itself ended in error, the underlying message about the transport
2410 filter timeout was being overwritten with the pipe command error. Now the
2411 underlying error message should be appended to the second error message.
2413 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2415 PH/02 Reduce the amount of output that "make" produces by default. Full output
2416 can still be requested.
2418 PH/03 The warning log line about a condition test deferring for a "warn" verb
2419 was being output only once per connection, rather than after each
2420 occurrence (because it was using the same function as for successful
2421 "warn" verbs). This seems wrong, so I have changed it.
2423 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2424 it should not have, which might have caused a crash in the right
2425 circumstances, but probably never did.
2427 PH/04 Installed a modified version of Tony Finch's patch to make submission
2428 mode fix the return path as well as the Sender: header line, and to
2429 add a /name= option so that you can make the user's friendly name appear
2432 TF/03 Added the control = fakedefer ACL modifier.
2434 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2435 Mark Lowes for thorough testing.
2437 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2439 TK/03 Merged latest SRS patch from Miles Wilton.
2441 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2442 with the definition in sysexits.h (which is #included earlier).
2443 Fortunately, Exim does not actually use EX_OK. The code used to try to
2444 preserve the sysexits.h value, by assumimg that macro definitions were
2445 scanned for macro replacements. I have been disabused of this notion,
2446 so now the code just undefines EX_OK before #including unistd.h.
2448 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2449 in the smtp transport. When a block could not be written in a single
2450 write() function, the timeout was being re-applied to each part-write.
2451 This seems wrong - if the receiver was accepting one byte at a time it
2452 would take for ever. The timeout is now adjusted when this happens. It
2453 doesn't have to be particularly precise.
2455 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2456 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2458 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2459 without a domain if the check to <postmaster@domain> fails.
2461 SC/01 Eximstats: added -xls and the ability to specify output files
2462 (patch written by Frank Heydlauf).
2464 SC/02 Eximstats: use FileHandles for outputing results.
2466 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2468 SC/04 Eximstats: fixed display of large numbers with -nvr option
2470 SC/05 Eximstats: fixed merging of reports with empty tables.
2472 SC/06 Eximstats: added the -include_original_destination flag
2474 SC/07 Eximstats: removed tabs and trailing whitespace.
2476 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2478 TK/06 MBOX spool code: Add real "From " MBOX separator line
2479 so the .eml file is really in mbox format (even though
2480 most programs do not really care). Patch from Alex Miller.
2482 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2483 The latter is generated from $received_to and is only set if the
2484 message has one envelope recipient. SA can use these headers,
2485 obviously out-of-the-box. Patch from Alex Miller.
2487 PH/08 The ${def test on a variable was returning false if the variable's
2488 value was "0", contrary to what the specification has always said!
2489 The result should be true unless the variable is empty.
2491 PH/09 The syntax error of a character other than { following "${if
2492 def:variable_name" (after optional whitespace) was not being diagnosed.
2493 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2494 accidental colon was present, for example, could give incorrect results.
2496 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2497 result is used (not including appendfile, where other changes are about
2500 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2501 This involved changing a lot of size variables from int to off_t. It
2502 should work with maildirs and everything.
2504 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2505 spamd dying while we are connected to it.
2507 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2510 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2511 function for BDB 4.3.
2513 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2515 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2516 the sources for PCRE have been re-arranged and more files are now
2519 PH/15 The code I had for printing potentially long long variables in PH/11
2520 above was not the best (it lost precision). The length of off_t variables
2521 is now inspected at build time, and an appropriate printing format (%ld
2522 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2523 to be "long long int" or "long int". This is needed for the internal
2524 formatting function string_vformat().
2526 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2527 the configuration file to be ":syslog", then the script "guesses" where
2528 the logs files are, rather than using the compiled in default. In our
2529 case the guess is not the same as the compiled default, so the script
2530 suddenly stopped working when I started to use syslog. The patch checks
2531 to see if log_file_path is "". If so, it attempts to read it from exim
2532 with no configuration file to get the compiled in version, before it
2533 falls back to the previous guessing code."
2535 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2536 implementing BATV in an Exim configuration. See NewStuff for the gory
2539 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2540 Makefile that are specific to HP-UX.
2542 PH/18 If the "use_postmaster" option was set for a recipient callout together
2543 with the "random" option, the postmaster address was used as the MAIL
2544 FROM address for the random test, but not for the subsequent recipient
2545 test. It is now used for both.
2547 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2548 patch removes a few documentation additions to RFC 3028, because the
2549 latest draft now contains them. It adds the new en;ascii-case comparator
2550 and a new error check for 8bit text in MIME parts. Comparator and
2551 require names are now matched exactly. I enabled the subaddress
2552 extension, but it is not well tested yet (read: it works for me)."
2554 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2555 rework some of the code of TK/09 above to avoid the hardwired use of
2556 "%lld" and "long long". Replaced the call to snprintf() with a call to
2559 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2560 records point to non-existent hosts", "retry timeout exceeded", and
2561 "retry time not reached for any host after a long failure period".
2563 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2564 experimental DomainKeys support:
2566 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2567 (2) On an error such as an illegally used "control", the wrong name for
2568 the control was given.
2570 These problems did NOT occur unless DomainKeys support was compiled.
2572 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2574 PH/24 Added ${if match_ip condition.
2576 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2577 databases so that it will be absolutely obvious if a crash occurs in the
2578 DB library. This is a regular occurrence (often caused by mis-matched
2581 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2582 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2583 picked up on a user's system that detects such things. There doesn't seem
2584 to be a gcc warning option for this - only an attribute that has to be
2585 put on the function's prototype. It seems that in Fedora Core 4 they have
2586 set this on a number of new functions. No doubt there will be more in due
2589 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2590 specify a transport. However, if an address that was verified by such a
2591 router was the subject of a callout, Exim crashed because it tried to
2592 read the rcpt_include_affixes from the non-existent transport. Now it
2593 just assumes that the setting of that option is false. This bug was
2594 introduced by 4.51/PH/31.
2596 PH/28 Changed -d+all to exclude +memory, because that information is very
2597 rarely of interest, but it makes the output a lot bigger. People tend to
2598 do -d+all out of habit.
2600 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2601 code in os-type was giving problems when libc.so lives in lib64, like on
2604 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2605 aren't the modern standard, and it seems that some systems' include files
2606 don't always have them. Exim was already checking for some of the newer
2607 ones like T_AAAA, and defining it itself. I've added checks for all the
2608 record types that Exim uses.
2610 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2611 not automatically generating a new one, as it is supposed to. This
2612 prevented TLS from working. If the file did exist, but contained invalid
2613 data, a new version was generated, as expected. It was only the case of a
2614 non-existent file that was broken.
2616 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2617 with a change in libdomainkeys > 0.64.
2619 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2620 from DNS. If the selector record carries the flag, it now has
2621 precedence over the domain-wide flag.
2623 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2625 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2626 the use of a port name that isn't defined in /etc/services) occurred, the
2627 message was deferred as in a normal delivery, and thus remained on the
2628 spool, instead of being failed because of the mua_wrapper setting. This
2629 is now fixed, and I tidied up some of the mua_wrapper messages at the
2632 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2633 the messages in a hash of arrays rather than using individual hashes.
2634 This is a bit cleaner and results in dramatic memory savings, albeit
2635 at a slight CPU cost.
2637 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2638 as requested by Marc Sherman.
2640 SC/10 Eximstats: added histograms for user specified patterns as requested
2643 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2645 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2646 fopen() in the content-scanning modules that did not already have it.
2652 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2653 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2655 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2657 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2659 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2660 to test Sieve filters that use "vacation".
2662 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2663 that changes the way the GnuTLS parameters are stored in the cache file.
2664 The new format can be generated externally. For backward compatibility,
2665 if the data in the cache doesn't make sense, Exim assumes it has read an
2666 old-format file, and it generates new data and writes a new file. This
2667 means that you can't go back to an older release without removing the
2670 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2671 work if there are any delivery delays because "one_time" forces the
2672 parent to be marked "delivered", so its unseen clone is never tried
2673 again. For this reason, Exim now forbids the simultaneous setting of
2676 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2677 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2678 entry if you want to know the details. The fix, however, neglected to
2679 consider the case when local delivery batching is involved. The test for
2680 "previously delivered" was not happening when checking to see if an
2681 address could be batched with a previous (undelivered) one; under
2682 certain circumstances this could lead to multiple deliveries to the same
2685 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2686 in its include files, and this causes problems building Exim.
2688 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2689 header_syntax) but Exim was just ignoring anything given after a slash.
2690 In particular, this caused confusion with an attempt to use "verify =
2691 reverse_host_lookup/defer_ok". An error is now given when options are
2692 supplied for verify items that do not have them. (Maybe reverse_host_
2693 lookup should have a defer_ok option, but that's a different point.)
2695 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2696 defined by RFC 821) to 2048, because there were problems with some AUTH
2697 commands, and RFC 1869 says the size should be increased for extended
2698 SMTP commands that take arguments.
2700 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2703 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2704 "unknown" error; now it says that the functionality isn't in the binary.
2706 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2707 an address' error message when a string expansion fails (syntax or
2708 whatever). Otherwise the password may appear in the log. Following change
2709 PH/42 below, there is no longer a chance of it appearing in a bounce
2712 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2714 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2715 of its addresses was ever tried. (Bugzilla bug #2.)
2717 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2718 the result incorrectly in the debug output. (It correctly added a newline
2719 to what was transported.)
2721 TF/01 Added $received_time.
2723 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2724 commented out examples of how to interface to a virus scanner and to
2725 SpamAssassin. Also added commented examples of av_scanner and
2726 spamd_address settings.
2728 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2729 and controls are allowed in which ACLs. There were a couple of minor
2730 errors. Some of the entries in the conditions table (which is a table of
2731 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2732 negation of where the condition IS allowed.
2734 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2736 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2737 header file does not have a version number, so I've had to invent a new
2738 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2739 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2740 radiusclient), but it was contributed by a Radius user.
2742 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2743 files or directories, for OpenSSL.
2745 PH/19 When an Exim process that is running as root has to create an Exim log
2746 file, it does so in a subprocess that runs as exim:exim so as to get the
2747 ownership right at creation (otherwise, other Exim processes might see
2748 the file with the wrong ownership). There was no test for failure of this
2749 fork() call, which would lead to the process getting stuck as it waited
2750 for a non-existent subprocess. Forks do occasionally fail when resources
2751 run out. I reviewed all the other calls to fork(); they all seem to check
2754 PH/20 When checking for unexpected SMTP input at connect time (before writing
2755 the banner), Exim was not dealing correctly with a non-positive return
2756 from the read() function. If the client had disconnected by this time,
2757 the result was a log entry for a synchronization error with an empty
2758 string after "input=" when read() returned zero. If read() returned -1
2759 (an event I could not check), uninitialized data bytes were printed.
2760 There were reports of junk text (parts of files, etc) appearing after
2763 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2765 PH/22 Added support for macro redefinition, and (re)definition in between
2766 driver and ACL definitions.
2768 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2769 forgetting to use the resulting value; it was using the unexpanded value.
2771 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2772 hadn't been configured. The fix is from Juergen Kreileder, who
2773 understands it better than I do:
2775 "Here's what I see happening with three configured cyrus_sasl
2776 authenticators configured (plain, login, cram-md5):
2778 On startup auth_cyrus_sasl_init() gets called for each of these.
2779 This means three calls to sasl_listmech() without a specified mech_list.
2780 => SASL tests which mechs of all available mechs actually work
2781 => three warnings about OTP not working
2782 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2784 With the patch, sasl_listmech() also gets called three times. But now
2785 SASL's mech_list option is set to the server_mech specified in the the
2786 authenticator. Or in other words, the answer from sasl_listmech()
2787 gets limited to just the mech you're testing for (which is different
2789 => the return list contains just 'plain' or 'login', 'cram-md5' or
2790 nothing depending on the value of ob->server_mech.
2792 I've just tested the patch: Authentication still works fine,
2793 unavailable mechs specified in the exim configuration are still
2794 caught, and the auth.log warnings about OTP are gone."
2796 PH/25 When debugging is enabled, the contents of the command line are added
2797 to the debugging output, even when log_selector=+arguments is not
2800 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2801 answer is "GNU", and only if the return is "GNU/something" is the answer
2804 PH/27 $acl_verify_message is now set immediately after the failure of a
2805 verification in an ACL, and so is available in subsequent modifiers. In
2806 particular, the message can be preserved by coding like this:
2808 warn !verify = sender
2809 set acl_m0 = $acl_verify_message
2811 Previously, $acl_verify_message was set only while expanding "message"
2812 and "log_message" when a very denied access.
2814 PH/28 Modified OS/os.c-Linux with
2816 -#ifndef OS_LOAD_AVERAGE
2817 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2819 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2820 nomenclature these days.)
2822 PH/29 Installed patch from the Sieve maintainer that adds the options
2823 sieve_useraddress and sieve_subaddress to the redirect router.
2825 PH/30 In these circumstances:
2826 . Two addresses routed to the same list of hosts;
2827 . First host does not offer TLS;
2828 . First host accepts first address;
2829 . First host gives temporary error to second address;
2830 . Second host offers TLS and a TLS session is established;
2831 . Second host accepts second address.
2832 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2833 and peerdn, if requested) that were in fact used only for the second
2836 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2837 attention to any local part prefix or suffix that was matched by the
2838 router that accepted the address. It now behaves in the same way as it
2839 does for delivery: the affixes are removed from the local part unless
2840 rcpt_include_affixes is set on the transport.
2842 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2843 timeout during the DATA phase of an incoming message.
2845 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2846 applied a patch sanctioned by the Sieve maintainer.
2848 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2849 the uid or gid is negative. A case of a negative gid caused this to be
2850 noticed. The fix allows for either to be negative.
2852 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2853 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2856 PH/36 The change PH/12 above was broken. Fixed it.
2858 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2859 the grounds that routing the same address twice would always produce the
2860 same answer. This might have been true once, but it is certainly no
2861 longer true now. Routing a child address may depend on the previous
2862 routing that produced that child. Some complicated redirection strategies
2863 went wrong when messages had multiple recipients, and made Exim's
2864 behaviour dependent on the order in which the addresses were given.
2866 I have moved the duplicate checking until after the routing is complete.
2867 Exim scans the addresses that are assigned to local and remote
2868 transports, and removes any duplicates. This means that more work will be
2869 done, as duplicates will always all be routed, but duplicates are
2870 presumably rare, so I don't expect this is of any significance.
2872 For deliveries to pipes, files, and autoreplies, the duplicate checking
2873 still happens during the routing process, since they are not going to be
2876 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2877 It corrects a timeout issue with spamd. This is Ian's comment: "The
2878 background is that sometimes spamd either never reads data from a
2879 connection it has accepted, or it never writes response data. The exiscan
2880 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2881 blindly assumes that writes won't block so it may never time out."
2883 PH/39 Allow G after quota size as well as K and M.
2885 PH/40 The value set for $authenticated_id in an authenticator may not contain
2886 binary zeroes or newlines because the value is written to log lines and
2887 to spool files. There was no check on this. Now the value is run through
2888 the string_printing() function so that such characters are converted to
2889 printable escape sequences.
2891 PH/41 $message_linecount is a new variable that contains the total number of
2892 lines in the message. Compare $body_linecount, which is the count for the
2895 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2896 bounce and delay warning messages, except in certain special cases, which
2899 (a) An SMTP error message from a remote host;
2900 (b) A message specified in a :fail: redirection;
2901 (c) A message specified in a "fail" command in a system filter;
2902 (d) A message specified in a FAIL return from the queryprogram router;
2903 (e) A message specified by the cannot_route_message router option.
2905 In these cases only, Exim does include the error details in bounce and
2906 warning messages. There are also a few cases where bland messages such
2907 as "unrouteable address" or "local delivery error" are given.
2909 PH/43 $value is now also set for the "else" part of a ${run expansion.
2911 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2912 being worked on, but at least Exim now implements the latest version to
2915 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2916 process to complete was treated as a delivery failure, a timeout while
2917 writing the message to the pipe was logged, but erroneously treated as a
2918 successful delivery. Such timeouts include transport filter timeouts. For
2919 consistency with the overall process timeout, these timeouts are now
2920 treated as errors, giving rise to delivery failures by default. However,
2921 there is now a new Boolean option for the pipe transport called
2922 timeout_defer, which, if set TRUE, converts the failures into defers for
2923 both kinds of timeout. A transport filter timeout is now identified in
2926 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2927 systems where "make" and "gmake" are different, calling "gmake" at top
2928 level broke things. I've arranged for the value of $(MAKE) to be passed
2929 from the Makefile to this script so that it can call the same version of
2933 A note about Exim versions 4.44 and 4.50
2934 ----------------------------------------
2936 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2937 changes of various kinds. As a consequence, a big documentation update was
2938 needed. This delayed the release for rather longer than seemed good, especially
2939 in the light of a couple of (minor) security issues. Therefore, the changes
2940 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2941 release. So 4.44 and 4.50 are in effect two different branches that both start
2944 I have left the 4.50 change log unchanged; it contains all the changes since
2945 4.43. The change log for 4.44 is below; many of its items are identical to
2946 those for 4.50. This seems to be the most sensible way to preserve the
2947 historical information.
2953 1. Minor wording change to the doc/README.SIEVE file.
2955 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2956 computation of the current number of files was incorrect.
2958 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2959 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2962 4. Give more explanation in the error message when the command for a transport
2963 filter fails to execute.
2965 5. There are several places where Exim runs a non-Exim command in a
2966 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2967 was being done only for the command run by the queryprogram router. It is
2968 now done for all such subprocesses. The other cases are: ${run, transport
2969 filters, and the commands run by the lmtp and pipe transports.
2971 6. Added CONFIGURE_GROUP build-time option.
2973 7. Some older OS have a limit of 256 on the maximum number of file
2974 descriptors. Exim was using setrlimit() to set 1000 as a large value
2975 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2976 systems. I've change it so that if it can't get 1000, it tries for 256.
2978 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2979 was an oversight, and furthermore, ever since the addition of extra
2980 controls (e.g. 4.43/32), the checks on when to allow different forms of
2981 "control" were broken. There should now be diagnostics for all cases when a
2982 control that does not make sense is encountered.
2984 9. Added the /retain_sender option to "control=submission".
2986 10. $recipients is now available in the predata ACL (oversight).
2988 11. Tidy the search cache before the fork to do a delivery from a message
2989 received from the command line. Otherwise the child will trigger a lookup
2990 failure and thereby defer the delivery if it tries to use (for example) a
2991 cached ldap connection that the parent has called unbind on.
2993 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2994 of $address_data from the recipient verification was clobbered by the
2995 sender verification.
2997 13. The value of address_data from a sender verification is now available in
2998 $sender_address_data in subsequent conditions in the ACL statement.
3000 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3002 15. Added a new option "connect=<time>" to callout options, to set a different
3005 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3006 was its contents. (It was OK if the option was not defined at all.)
3008 17. A "Completed" log line is now written for messages that are removed from
3009 the spool by the -Mrm option.
3011 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3012 information about exactly what failed.
3014 19. Added -dd to debug only the daemon process.
3016 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3017 handles timeouts, both on the server side and network timeouts. Renamed the
3018 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3020 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3021 It is now set to "smtps".
3023 22. $host_address is now set to the target address during the checking of
3024 ignore_target_hosts.
3026 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3027 being passed; this would have caused $sender_host_name to have been used if
3028 matching the list had actually called for a host name (not very likely,
3029 since this list is usually IP addresses). A host name is now passed as
3032 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3033 code that specifies a non-restarting handler (typically sigaction() in
3034 modern systems) in an attempt to fix a rare and obscure crash bug.
3036 25. Narrowed the window for a race in the daemon that could cause it to ignore
3037 SIGCHLD signals. This is not a major problem, because they are used only to
3038 wake it up if nothing else does.
3040 26. A malformed maildirsize file could cause Exim to calculate negative values
3041 for the mailbox size or file count. Odd effects could occur as a result.
3042 The maildirsize information is now recalculated if the size or filecount
3045 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3046 support for a long time. Removed HAVE_SYS_VFS_H.
3048 28. Installed the latest version of exipick from John Jetmore.
3050 29. In an address list, if the pattern was not a regular expression, an empty
3051 subject address (from a bounce message) matched only if the pattern was an
3052 empty string. Non-empty patterns were not even tested. This was the wrong
3053 because it is perfectly reasonable to use an empty address as part of a
3054 database query. An empty address is now tested by patterns that are
3055 lookups. However, all the other forms of pattern expect the subject to
3056 contain a local part and a domain, and therefore, for them, an empty
3057 address still always fails if the pattern is not itself empty.
3059 30. Exim went into a mad DNS loop when attempting to do a callout where the
3060 host was specified on an smtp transport, and looking it up yielded more
3061 than one IP address.
3063 31. Re-factored the code for checking spool and log partition space into a
3064 function that finds that data and another that does the check. The former
3065 is then used to implement four new variables: $spool_space, $log_space,
3066 $spool_inodes, and $log_inodes.
3068 32. The RFC2047 encoding function was originally intended for short strings
3069 such as real names; it was not keeping to the 75-character limit for
3070 encoded words that the RFC imposes. It now respects the limit, and
3071 generates multiple encoded words if necessary. To be on the safe side, I
3072 have increased the buffer size for the ${rfc2047: expansion operator from
3075 33. It is now permitted to omit both strings after an "if" condition; if the
3076 condition is true, the result is "true". As before, when the second string
3077 is omitted, a false condition yields an empty string. This makes it less
3078 cumbersome to write custom ACL and router conditions.
3080 34. Failure to deliver a bounce message always caused it to be frozen, even if
3081 there was an errors_to setting on the router. The errors_to setting is now
3084 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3085 canonical form (fully expanded) before being placed in
3086 $sender_host_address.
3088 36. The table in the code that translates DNS record types into text (T_A to
3089 "A" for instance) was missing entries for NS and CNAME. It is just possible
3090 that this could have caused confusion if both these types were looked up
3091 for the same domain, because the text type is used as part of Exim's
3092 per-process caching. But the chance of anyone hitting this buglet seems
3095 37. The dnsdb lookup has been extended in a number of ways.
3097 (1) There is a new type, "zns", which walks up the domain tree until it
3098 finds some nameserver records. It should be used with care.
3100 (2) There is a new type, "mxh", which is like "mx" except that it returns
3101 just the host names, not the priorities.
3103 (3) It is now possible to give a list of domains (or IP addresses) to be
3104 looked up. The behaviour when one of the lookups defers can be
3105 controlled by a keyword.
3107 (4) It is now possible to specify the separator character for use when
3108 multiple records are returned.
3110 38. The dnslists ACL condition has been extended: it is now possible to supply
3111 a list of IP addresses and/or domains to be looked up in a particular DNS
3114 39. Added log_selector=+queue_time_overall.
3116 40. When running the queue in the test harness, wait just a tad after forking a
3117 delivery process, to get repeatability of debugging output.
3119 41. Include certificate and key file names in error message when GnuTLS fails
3120 to set them up, because the GnuTLS error message doesn't include the name
3121 of the failing file when there is a problem reading it.
3123 42. Allow both -bf and -bF in the same test run.
3125 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3127 44. The "Exiscan patch" is now merged into the mainline Exim source.
3129 45. Sometimes the final signoff response after QUIT could fail to get
3130 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3131 < 0 before doing a fflush(). This bug looks as though it goes back to the
3132 introduction of TLS in release 3.20, but "sometimes" must have been rare
3133 because the tests only now provoked it.
3135 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3136 (this can affect the format of dates).
3138 47. exim_tidydb, when checking for the continued existence of a message for
3139 which it has found a message-specific retry record, was not finding
3140 messages that were in split spool directories. Consequently, it was
3141 deleting retry records that should have stayed in existence.
3143 48. Steve fixed some bugs in eximstats.
3145 49. The SPA authentication driver was not abandoning authentication and moving
3146 on to the next authenticator when an expansion was forced to fail,
3147 contradicting the general specification for all authenticators. Instead it
3148 was generating a temporary error. It now behaves as specified.
3150 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3151 (the order specifies the preference for clients). The order is now AES256,
3152 AES128, 3DES, ARCFOUR128.
3154 51. Small patch to Sieve code - explicitly set From: when generating an
3157 52. Exim crashed if a remote delivery caused a very long error message to be
3158 recorded - for instance if somebody sent an entire SpamAssassin report back
3159 as a large number of 550 error lines. This bug was coincidentally fixed by
3160 increasing the size of one of Exim's internal buffers (big_buffer) that
3161 happened as part of the Exiscan merge. However, to be on the safe side, I
3162 have made the code more robust (and fixed the comments that describe what
3165 53. Now that there can be additional text after "Completed" in log lines (if
3166 the queue_time_overall log selector is set), a one-byte patch to exigrep
3167 was needed to allow it to recognize "Completed" as not the last thing in
3170 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3171 patch that reportedly fixes this has been added. I am not expert enough to
3172 create a test for it. This is what the patch creator wrote:
3174 "I found a little strange behaviour of ldap code when working with
3175 Windows 2003 AD Domain, where users was placed in more than one
3176 Organization Units. When I tried to give exim partial DN, the exit code
3177 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3178 But simultaneously result of request was absolutely normal ldap result,
3179 so I produce this patch..."
3181 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3182 REFERENCE, so I have modified the code to exclude the patch when that macro
3185 55. Some experimental protocols are using DNS PTR records for new purposes. The
3186 keys for these records are domain names, not reversed IP addresses. The
3187 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3188 leaves it alone. Component reversal etc. now happens only for IP addresses.
3191 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3193 57. Double the size of the debug message buffer (to 2048) so that more of very
3194 long debug lines gets shown.
3196 58. The exicyclog utility now does better if the number of log files to keep
3197 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3199 59. Two changes related to the smtp_active_hostname option:
3201 (1) $smtp_active_hostname is now available as a variable.
3202 (2) The default for smtp_banner uses $smtp_active_hostname instead
3203 of $primary_hostname.
3205 60. The host_aton() function is supposed to be passed a string that is known
3206 to be a valid IP address. However, in the case of IPv6 addresses, it was
3207 not checking this. This is a hostage to fortune. Exim now panics and dies
3208 if the condition is not met. A case was found where this could be provoked
3209 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3210 components; fortuitously, this particular loophole had already been fixed
3211 by change 4.50/55 above.
3213 If there are any other similar loopholes, the new check in host_aton()
3214 itself should stop them being exploited. The report I received stated that
3215 data on the command line could provoke the exploit when Exim was running as
3216 exim, but did not say which command line option was involved. All I could
3217 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3218 running as the user.
3221 61. There was a buffer overflow vulnerability in the SPA authentication code
3222 (which came originally from the Samba project). I have added a test to the
3223 spa_base64_to_bits() function which I hope fixes it.
3226 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3227 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3229 63. The daemon start-up calls getloadavg() while still root for those OS that
3230 need the first call to be done as root, but it missed one case: when
3231 deliver_queue_load_max is set with deliver_drop_privilege. This is
3232 necessary for the benefit of the queue runner, because there is no re-exec
3233 when deliver_drop_privilege is set.
3235 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3236 This has been fixed.
3238 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3239 was in use, was not putting the data itself into the right store pool;
3240 consequently, it could be overwritten for a subsequent message in the same
3241 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3244 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3246 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3247 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3248 the function were treating the return as a boolean value, which happened to
3249 work because 0=false and not-0=true, but is not correct code.
3251 68. The host_aton() function was not handling scoped IPv6 addresses (those
3252 with, for example, "%eth0" on the end) correctly.
3254 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3255 negated items (that is, ~something) in unsigned ints. Some compilers
3256 apparently mutter when there is no cast.
3258 70. If an address verification called from an ACL failed, and did not produce a
3259 user-specific message (i.e. there was only a "system" message), nothing was
3260 put in $acl_verify_message. In this situation, it now puts the system
3263 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3264 session; change 4.31/43 added the unwanted input to the log line - except
3265 that it did not do this in the start of session case. It now does.
3267 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3268 This is wrong and can cause the other end to generate a synchronization
3269 error if it is another Exim or anything else that does the synchronization
3270 check. A QUIT command is no longer sent after a timeout.
3272 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3273 during host lookups.
3275 74. The defer_ok option of callout verification was not working if it was used
3276 when verifying addresses in header lines, that is, for this case:
3278 verify = header_sender/callout=defer_ok
3280 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3281 those file descriptors could be used for SMTP connections. If anything
3282 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3283 could be sent to the SMTP client, causing chaos. The daemon now opens
3284 stdin, stdout, and stderr to /dev/null when it puts itself into the
3287 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3288 log by default. The user can override this with suitable Perl magic.
3290 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3291 the log message when discard triggers, was not working for the DATA ACL or
3292 for the non-SMTP ACL.
3294 78. Error message wording change in sieve.c.
3296 79. If smtp_accept_max_per_host was set, the number of connections could be
3297 restricted to fewer than expected, because the daemon was trying to set up
3298 a new connection before checking whether the processes handling previous
3299 connections had finished. The check for completed processes is now done
3300 earlier. On busy systems, this bug wouldn't be noticed because something
3301 else would have woken the daemon, and it would have reaped the completed
3304 80. If a message was submitted locally by a user whose login name contained one
3305 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3306 It caused a spool format error. I have fixed the spool reading code. A
3307 related problem was that the "from" clause in the Received: line became
3308 illegal because of the space(s). It is now covered by ${quote_local_part.
3310 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3313 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3315 83. Patch from Sieve maintainer to fix unterminated string problem in
3316 "vacation" handling.
3318 84. Some minor changes to the Linux configuration files to help with other
3319 OS variants using glibc.
3321 85. One more patch for Sieve to update vacation handling to latest spec.
3324 ----------------------------------------------------
3325 See the note above about the 4.44 and 4.50 releases.
3326 ----------------------------------------------------
3332 1. Change 4.43/35 introduced a bug that caused file counts to be
3333 incorrectly computed when quota_filecount was set in an appendfile
3336 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3337 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3340 3. Give more explanation in the error message when the command for a transport
3341 filter fails to execute.
3343 4. There are several places where Exim runs a non-Exim command in a
3344 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3345 was being done only for the command run by the queryprogram router. It is
3346 now done for all such subprocesses. The other cases are: ${run, transport
3347 filters, and the commands run by the lmtp and pipe transports.
3349 5. Some older OS have a limit of 256 on the maximum number of file
3350 descriptors. Exim was using setrlimit() to set 1000 as a large value
3351 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3352 systems. I've change it so that if it can't get 1000, it tries for 256.
3354 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3355 was an oversight, and furthermore, ever since the addition of extra
3356 controls (e.g. 4.43/32), the checks on when to allow different forms of
3357 "control" were broken. There should now be diagnostics for all cases when a
3358 control that does not make sense is encountered.
3360 7. $recipients is now available in the predata ACL (oversight).
3362 8. Tidy the search cache before the fork to do a delivery from a message
3363 received from the command line. Otherwise the child will trigger a lookup
3364 failure and thereby defer the delivery if it tries to use (for example) a
3365 cached ldap connection that the parent has called unbind on.
3367 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3368 of $address_data from the recipient verification was clobbered by the
3369 sender verification.
3371 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3372 was its contents. (It was OK if the option was not defined at all.)
3374 11. A "Completed" log line is now written for messages that are removed from
3375 the spool by the -Mrm option.
3377 12. $host_address is now set to the target address during the checking of
3378 ignore_target_hosts.
3380 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3381 being passed; this would have caused $sender_host_name to have been used if
3382 matching the list had actually called for a host name (not very likely,
3383 since this list is usually IP addresses). A host name is now passed as
3386 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3387 code that specifies a non-restarting handler (typically sigaction() in
3388 modern systems) in an attempt to fix a rare and obscure crash bug.
3390 15. Narrowed the window for a race in the daemon that could cause it to ignore
3391 SIGCHLD signals. This is not a major problem, because they are used only to
3392 wake it up if nothing else does.
3394 16. A malformed maildirsize file could cause Exim to calculate negative values
3395 for the mailbox size or file count. Odd effects could occur as a result.
3396 The maildirsize information is now recalculated if the size or filecount
3399 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3400 support for a long time. Removed HAVE_SYS_VFS_H.
3402 18. Updated exipick to current release from John Jetmore.
3404 19. Allow an empty sender to be matched against a lookup in an address list.
3405 Previously the only cases considered were a regular expression, or an
3408 20. Exim went into a mad DNS lookup loop when doing a callout where the
3409 host was specified on the transport, if the DNS lookup yielded more than
3412 21. The RFC2047 encoding function was originally intended for short strings
3413 such as real names; it was not keeping to the 75-character limit for
3414 encoded words that the RFC imposes. It now respects the limit, and
3415 generates multiple encoded words if necessary. To be on the safe side, I
3416 have increased the buffer size for the ${rfc2047: expansion operator from
3419 22. Failure to deliver a bounce message always caused it to be frozen, even if
3420 there was an errors_to setting on the router. The errors_to setting is now
3423 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3424 canonical form (fully expanded) before being placed in
3425 $sender_host_address.
3427 24. Updated eximstats to version 1.33
3429 25. Include certificate and key file names in error message when GnuTLS fails
3430 to set them up, because the GnuTLS error message doesn't include the name
3431 of the failing file when there is a problem reading it.
3433 26. Expand error message when OpenSSL has problems setting up cert/key files.
3436 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3437 (this can affect the format of dates).
3439 28. exim_tidydb, when checking for the continued existence of a message for
3440 which it has found a message-specific retry record, was not finding
3441 messages that were in split spool directories. Consequently, it was
3442 deleting retry records that should have stayed in existence.
3444 29. eximstats updated to version 1.35
3445 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3446 1.35 - bugfix such that pie charts by volume are generated correctly
3448 30. The SPA authentication driver was not abandoning authentication and moving
3449 on to the next authenticator when an expansion was forced to fail,
3450 contradicting the general specification for all authenticators. Instead it
3451 was generating a temporary error. It now behaves as specified.
3453 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3454 (the order specifies the preference for clients). The order is now AES256,
3455 AES128, 3DES, ARCFOUR128.
3457 31. Small patch to Sieve code - explicitly set From: when generating an
3460 32. Exim crashed if a remote delivery caused a very long error message to be
3461 recorded - for instance if somebody sent an entire SpamAssassin report back
3462 as a large number of 550 error lines. This bug was coincidentally fixed by
3463 increasing the size of one of Exim's internal buffers (big_buffer) that
3464 happened as part of the Exiscan merge. However, to be on the safe side, I
3465 have made the code more robust (and fixed the comments that describe what
3468 33. Some experimental protocols are using DNS PTR records for new purposes. The
3469 keys for these records are domain names, not reversed IP addresses. The
3470 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3471 leaves it alone. Component reversal etc. now happens only for IP addresses.
3474 34. The host_aton() function is supposed to be passed a string that is known
3475 to be a valid IP address. However, in the case of IPv6 addresses, it was
3476 not checking this. This is a hostage to fortune. Exim now panics and dies
3477 if the condition is not met. A case was found where this could be provoked
3478 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3479 components; fortuitously, this particular loophole had already been fixed
3480 by change 4.50/55 or 4.44/33 above.
3482 If there are any other similar loopholes, the new check in host_aton()
3483 itself should stop them being exploited. The report I received stated that
3484 data on the command line could provoke the exploit when Exim was running as
3485 exim, but did not say which command line option was involved. All I could
3486 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3487 running as the user.
3490 35. There was a buffer overflow vulnerability in the SPA authentication code
3491 (which came originally from the Samba project). I have added a test to the
3492 spa_base64_to_bits() function which I hope fixes it.
3495 36. The daemon start-up calls getloadavg() while still root for those OS that
3496 need the first call to be done as root, but it missed one case: when
3497 deliver_queue_load_max is set with deliver_drop_privilege. This is
3498 necessary for the benefit of the queue runner, because there is no re-exec
3499 when deliver_drop_privilege is set.
3501 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3502 was in use, was not putting the data itself into the right store pool;
3503 consequently, it could be overwritten for a subsequent message in the same
3504 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3507 38. Sometimes the final signoff response after QUIT could fail to get
3508 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3509 < 0 before doing a fflush(). This bug looks as though it goes back to the
3510 introduction of TLS in release 3.20, but "sometimes" must have been rare
3511 because the tests only now provoked it.
3517 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3518 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3519 Now it returns an integer. A number of calls were still expecting a T/F
3520 return. Fortuitously, in all cases, the tests worked in OK situations,
3521 which is the norm. However, things would have gone wrong on any write
3522 failures on the smtp file descriptor. This function is used when sending
3523 messages over SMTP and also when doing verify callouts.
3525 2. When Exim is called to do synchronous delivery of a locally submitted
3526 message (the -odf or -odi options), it no longer closes stderr before doing
3529 3. Implemented the mua_wrapper option.
3531 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3533 5. Implemented the functions header_remove(), header_testname(),
3534 header_add_at_position(), and receive_remove_recipient(), and exported them
3537 6. If an ACL "warn" statement specified the addition of headers, Exim already
3538 inserted X-ACL-Warn: at the start if there was no header name. However, it
3539 was not making this test for the second and subsequent header lines if
3540 there were newlines in the string. This meant that an invalid header could
3541 be inserted if Exim was badly configured.
3543 7. Allow an ACL "warn" statement to add header lines at the start or after all
3544 the Received: headers, as well as at the end.
3546 8. Added the rcpt_4xx retry error code.
3548 9. Added postmaster_mailfrom=xxx to callout verification option.
3550 10. Added mailfrom=xxxx to the callout verification option, for verify=
3553 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3554 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3556 12. Inserted some casts to stop certain compilers warning when using pointer
3557 differences as field lengths or precisions in printf-type calls (mostly
3558 affecting debugging statements).
3560 13. Added optional readline() support for -be (dynamically loaded).
3562 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3563 same clock tick as a message's arrival, so that its received time was the
3564 same as the "first fail" time on the retry record, and that message
3565 remained on the queue past the ultimate address timeout, every queue runner
3566 would try a delivery (because it was past the ultimate address timeout) but
3567 after another failure, the ultimate address timeout, which should have then
3568 bounced the address, did not kick in. This was a "< instead of <=" error;
3569 in most cases the first failure would have been in the next clock tick
3570 after the received time, and all would be well.
3572 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3573 being recognized when the domain list was tested by the match_domain
3574 condition in an expansion string.
3576 16. Added the ${str2b64: operator.
3578 17. Exim was always calling setrlimit() to set a large limit for the number of
3579 processes, without checking whether the existing limit was already
3580 adequate. (It did check for the limit on file descriptors.) Furthermore,
3581 errors from getrlimit() and setrlimit() were being ignored. Now they are
3582 logged to the main and panic logs, but Exim does carry on, to try to do its
3583 job under whatever limits there are.
3585 18. Imported PCRE 5.0.
3587 19. Trivial typo in log message " temporarily refused connection" (the leading
3590 20. If the log selector return_path_on_delivery was set and an address was
3591 redirected to /dev/null, the delivery process crashed because it assumed
3592 that a return path would always be set for a "successful" delivery. In this
3593 case, the whole delivery is bypassed as an optimization, and therefore no
3596 21. Internal re-arrangement: the function for sending a challenge and reading
3597 a response while authentication was assuming a zero-terminated challenge
3598 string. It's now changed to take a pointer and a length, to allow for
3599 binary data in such strings.
3601 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3603 23. Exim was not respecting finduser_retries when seeking the login of the
3604 uid under which it was called; it was always trying 10 times. (The default
3605 setting of finduser_retries is zero.) Also, it was sleeping after the final
3606 failure, which is pointless.
3608 24. Implemented tls_on_connect_ports.
3610 25. Implemented acl_smtp_predata.
3612 26. If the domain in control=submission is set empty, Exim assumes that the
3613 authenticated id is a complete email address when it generates From: or
3614 Sender: header lines.
3616 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3617 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3618 chown and chgrp in /bin and hostname in /usr/bin.
3620 28. Exim was keeping the "process log" file open after each use, just as it
3621 does for the main log. This opens the possibility of it remaining open for
3622 long periods when the USR1 signal hits a daemon. Occasional processlog
3623 errors were reported, that could have been caused by this. Anyway, it seems
3624 much more sensible not to leave this file open at all, so that is what now
3627 29. The long-running daemon process does not normally write to the log once it
3628 has entered its main loop, and it closes the log before doing so. This is
3629 so that log files can straightforwardly be renamed and moved. However,
3630 there are a couple of unusual error situations where the daemon does write
3631 log entries, and I had neglected to close the log afterwards.
3633 30. The text of an SMTP error response that was received during a remote
3634 delivery was being truncated at 512 bytes. This is too short for some of
3635 the long messages that one sometimes sees. I've increased the limit to
3638 31. It is now possible to make retry rules that apply only when a message has a
3639 specific sender, in particular, an empty sender.
3641 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3642 it possible to be selective about when SMTP synchronization is enforced.
3644 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3646 32. Implemented hosts_connection_nolog.
3648 33. Added an ACL for QUIT.
3650 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3653 35. Added mailbox_size and mailbox_filecount to appendfile.
3655 36. Added control = no_multiline_responses to ACLs.
3657 37. There was a bug in the logic of the code that waits for the clock to tick
3658 in the case where the clock went backwards by a substantial amount such
3659 that the microsecond fraction of "now" was more than the microsecond
3660 fraction of "then" (but the whole seconds number was less).
3662 38. Added support for the libradius Radius client library this is found on
3663 FreeBSD (previously only the radiusclient library was supported).
3669 1. When certain lookups returned multiple values in the form name=value, the
3670 quoting of the values was not always being done properly. Specifically:
3671 (a) If the value started with a double quote, but contained no whitespace,
3673 (b) If the value contained whitespace other than a space character (i.e.
3674 tabs or newlines or carriage returns) it was not quoted.
3675 This fix has been applied to the mysql and pgsql lookups by writing a
3676 separate quoting function and calling it from the lookup code. The fix
3677 should probably also be applied to nisplus, ibase and oracle lookups, but
3678 since I cannot test any of those, I have not disturbed their existing code.
3680 2. A hit in the callout cache for a specific address caused a log line with no
3681 reason for rejecting RCPT. Now it says "Previous (cached) callout
3682 verification failure".
3684 3. There was an off-by-one bug in the queryprogram router. An over-long
3685 return line was truncated at 256 instead of 255 characters, thereby
3686 overflowing its buffer with the terminating zero. As well as fixing this, I
3687 have increased the buffer size to 1024 (and made a note to document this).
3689 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3690 when Exim is waiting for an SMTP response from a remote server, Exim
3691 restarts its select() call on the socket, thereby resetting its timeout.
3692 This is not a problem when such interrupts are rare. Somebody set up a cron
3693 job to run exiwhat every 2 minutes, which is less than the normal select()
3694 timeout (5 or 10 minutes). This meant that the select() timeout never
3695 kicked in because it was always reset. I have fixed this by comparing the
3696 time when an interrupt arrives with the time at the start of the first call
3697 to select(). If more time than the timeout has elapsed, the interrupt is
3698 treated as a timeout.
3700 5. Some internal re-factoring in preparation for the addition of Sieve
3701 extensions (by MH). In particular, the "personal" test is moved to a
3702 separate function, and given an option for scanning Cc: and Bcc: (which is
3703 not set for Exim filters).
3705 6. When Exim created an email address using the login of the caller as the
3706 local part (e.g. when creating a From: or Sender: header line), it was not
3707 quoting the local part when it contained special characters such as @.
3709 7. Installed new OpenBSD configuration files.
3711 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3712 try to make them clearer.
3714 9. Callout options, other than the timeout value, were being ignored when
3715 verifying sender addresses in header lines. For example, when using
3717 verify = header_sender/callout=no_cache
3719 the cache was (incorrectly) being used.
3721 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3722 only the Cygwin environment.
3724 11. When return_path_on_delivery was set as a log selector, if different remote
3725 addresses in the same message used different return paths and parallel
3726 remote delivery occurred, the wrong values would sometimes be logged.
3727 (Whenever a remote delivery process finished, the return path value from
3728 the most recently started remote delivery process was logged.)
3730 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3731 lines when AUTH and/or TLS are in use. This is the "received protocol"
3732 field. Exim used to use "asmtp" for authenticated SMTP, without any
3733 indication (in the protocol name) for TLS use. Now it follows the RFC and
3734 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3735 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3736 names appear in log lines as well as in Received: header lines.
3738 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3739 extensions, and comparison tests, and to fix some bugs.
3741 14. Changes to the "personal" filter test:
3743 (1) The test was buggy in that it was just doing the equivalent of
3744 "contains" tests on header lines. For example, if a user's address was
3745 anne@some.where, the "personal" test would incorrectly be true for
3747 To: susanne@some.where
3749 This test is now done by extracting each address from the header in turn,
3750 and checking the entire address. Other tests that are part of "personal"
3751 are now done using regular expressions (for example, to check local parts
3752 of addresses in From: header lines).
3754 (2) The list of non-personal local parts in From: addresses has been
3755 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3756 taken from the Sieve specification recommendations.
3758 (3) If the message contains any header line starting with "List-" it is
3759 treated as non-personal.
3761 (4) The test for "circular" in the Subject: header line has been removed
3762 because it now seems ill-conceived.
3764 15. Minor typos in src/EDITME comments corrected.
3766 16. Installed latest exipick from John Jetmore.
3768 17. If headers_add on a router specified a text string that was too long for
3769 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3770 of string_sprintf() is now avoided.
3772 18. $message_body_size was not set (it was always zero) when running the DATA
3773 ACL and the local_scan() function.
3775 19. For the "mail" command in an Exim filter, no default was being set for
3776 the once_repeat time, causing a random time value to be used if "once" was
3777 specified. (If the value happened to be <= 0, no repeat happened.) The
3778 default is now 0s, meaning "never repeat". The "vacation" command was OK
3779 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3780 (I found it when inspecting the code).
3782 20. There is now an overall timeout for performing a callout verification. It
3783 defaults to 4 times the callout timeout, which applies to individual SMTP
3784 commands during the callout. The overall timeout applies when there is more
3785 than one host that can be tried. The timeout is checked before trying the
3786 next host. This prevents very long delays if there are a large number of
3787 hosts and all are timing out (e.g. when the network connections are timing
3788 out). The value of the overall timeout can be changed by specifying an
3789 additional sub-option for "callout", called "maxwait". For example:
3791 verify = sender/callout=5s,maxwait=20s
3793 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3794 to the end before writing, but this should make it even safer).
3796 22. Exim was forgetting that it had advertised PIPELINING for the second and
3797 subsequent messages on an SMTP connection. It was also not resetting its
3798 memory on STARTTLS and an internal HELO.
3800 23. When Exim logs an SMTP synchronization error within a session, it now
3801 records whether PIPELINING has been advertised or not.
3803 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3804 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3805 rather than long int.
3807 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3809 26. Added the never_mail option to autoreply.
3815 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3816 crash if the getsockname() call failed; this can happen if a connection is
3817 closed very soon after it is established. The problem was simply in the
3818 order in which certain operations were done, causing Exim to try to write
3819 to the SMTP stream before it had set up the file descriptor. The bug has
3820 been fixed by making things happen in the correct order.
3826 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3827 before the connection was closed, thus losing the rejection response.
3829 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3830 some early Solaris releases, but causes trouble in current releases where
3831 socklen_t is defined.
3833 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3836 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3839 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3840 that releasing the top of it at the end releases what was used for sub-
3841 expansions (unless the block got too big). However, discard this block if
3842 the first thing is a variable or header, so that we can use its block when
3843 it is dynamic (useful for very large $message_headers, for example).
3845 6. Lookups now cache *every* query, not just the most recent. A new, separate
3846 store pool is used for this. It can be recovered when all lookup caches are
3847 flushed. Lookups now release memory at the end of their result strings.
3848 This has involved some general refactoring of the lookup sources.
3850 7. Some code has been added to the store_xxx() functions to reduce the amount
3851 of flapping under certain conditions.
3853 8. log_incoming_interface used to affect only the <= reception log lines. Now
3854 it causes the local interface and port to be added to several more SMTP log
3855 lines, for example "SMTP connection from", and rejection lines.
3857 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3859 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3861 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3862 could be overwritten at the end of the current message (or the start of a
3863 new message if it was set in a HELO ACL). The value is now preserved for
3864 the duration of the SMTP connection.
3866 12. If a transport had a headers_rewrite setting, and a matching header line
3867 contained an unqualified address, that address was qualified, even if it
3868 did not match any rewriting rules. The underlying bug was that the values
3869 of the flags that permit the existence of unqualified sender and recipient
3870 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3871 non-local messages, and by -bnq for local messages) were not being
3872 preserved with the message after it was received.
3874 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3875 "next input=" as part of the text comprising the host identity instead of
3876 the correct text. The code was using the same buffer for two different
3877 strings. However, depending on which order the printing function evaluated
3878 its arguments, the bug did not always show up. Under Linux, for example, my
3879 test suite worked just fine.
3881 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3882 doesn't work with some older versions of Perl. It has been changed to "my",
3883 which in any case is probably the better facility to use.
3885 15. A really picky compiler found some instances of statements for creating
3886 error messages that either had too many or two few arguments for the format
3889 16. The size of the buffer for calls to the DNS resolver has been increased
3890 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3891 for addresses that have a lot of PTR records. This alleviates a problem; it
3892 does not fully solve it.
3894 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3895 buffer now truncates the list and logs the incident, which is the same
3896 action as happens when Exim is looking up a host name and its aliases.
3897 Previously in this situation something unpredictable would happen;
3898 sometimes it was "internal error: store_reset failed".
3900 18. If a server dropped the connection unexpectedly when an Exim client was
3901 using GnuTLS and trying to read a response, the client delivery process
3902 crashed while trying to generate an error log message.
3904 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3905 string, for example:
3907 warn message = H1: something\nH2: something
3909 the text was added as a single header line from Exim's point of view
3910 though it ended up OK in the delivered message. However, searching for the
3911 second and subsequent header lines using $h_h2: did not work. This has been
3912 fixed. Similarly, if a system filter added multiple headers in this way,
3913 the routers could not see them.
3915 20. Expanded the error message when iplsearch is called with an invalid key to
3916 suggest using net-iplsearch in a host list.
3918 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3919 ACLs are no longer actually imposed (and a message to that effect is
3922 22. If a "gecos" field in a passwd entry contained escaped characters, in
3923 particular, if it contained a \" sequence, Exim got it wrong when building
3924 a From: or a Sender: header from that name. A second bug also caused
3925 incorrect handling when an unquoted " was present following a character
3926 that needed quoting.
3928 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3929 was not being matched caselessly.
3931 24. Arranged for all hyphens in the exim.8 source to be escaped with
3934 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3935 properly. Recipient callout cache records were still being keyed to include
3936 the sender, even when use_sender was set false. This led to far more
3937 callouts that were necessary. The sender is no longer included in the key
3938 when use_sender is false.
3940 26. Added "control = submission" modifier to ACLs.
3942 27. Added the ${base62d: operator to decode base 62 numbers.
3944 28. dnsdb lookups can now access SRV records.
3946 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3947 the configuration file.
3949 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3950 (-v) mode. This makes the output for a verbose queue run more intelligible.
3952 31. Added a use_postmaster feature to recipient callouts.
3954 32. Added the $body_zerocount variable, containing the number of binary zero
3955 bytes in the message body.
3957 33. The time of last modification of the "new" subdirectory is now used as the
3958 "mailbox time last read" when there is a quota error for a maildir
3961 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3963 35. Added +ignore_unknown as a special item in host lists.
3965 36. Code for decoding IPv6 addresses in host lists is now included, even if
3966 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3967 address was recognized as an IP address, but was then not correctly decoded
3968 into binary, causing unexpected and incorrect effects when compared with
3975 1. Very minor rewording of debugging text in manualroute to say "list of
3976 hosts" instead of "hostlist".
3978 2. If verify=header_syntax was set, and a header line with an unqualified
3979 address (no domain) and a large number of spaces between the end of the
3980 name and the colon was received, the reception process suffered a buffer
3981 overflow, and (when I tested it) crashed. This was caused by some obsolete
3982 code that should have been removed. The fix is to remove it!
3984 3. When running in the test harness, delay a bit after writing a bounce
3985 message to get a bit more predictability in the log output.
3987 4. Added a call to search_tidyup() just before forking a reception process. In
3988 theory, someone could use a lookup in the expansion of smtp_accept_max_
3989 per_host which, without the tidyup, could leave open a database connection.
3991 5. Added the variables $recipient_data and $sender_data which get set from a
3992 lookup success in an ACL "recipients" or "senders" condition, or a router
3993 "senders" option, similar to $domain_data and $local_part_data.
3995 6. Moved the writing of debug_print from before to after the "senders" test
3998 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3999 problems for message scanning, either using a data ACL, or using
4000 local_scan() because the Received: header was not generated till after they
4001 were called (in order to set the time as the time of reception completion).
4002 I have revised the way this works. The header is now generated after the
4003 body is received, but before the ACL or local_scan() are called. After they
4004 are run, the timestamp in the header is updated.
4010 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4011 before starting a queue runner without re-exec. This happened only when
4012 deliver_drop_privilege was set or when the Exim user was set to root. The
4013 effect of the bug was that timeouts during subsequent deliveries caused
4014 crashes instead of being properly handled. The handler is now left at its
4015 default (and expected) setting.
4017 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4018 message, again when deliver_drop_privilege is set or Exim is run as root.
4019 The bug described in (1) was not present in this case, but the tidying up
4020 of the other signals was missing. I have made the two cases consistent.
4022 3. The ignore_target_hosts setting on a manualroute router was being ignored
4023 for hosts that were looked up using the /MX notation.
4025 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4028 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4029 operated on the sender address. After changing the $sender_address to <>
4030 for the sender address verify, Exim was re-instated it as the original
4031 (before rewriting) address, but remembering that it had rewritten it, so it
4032 wasn't rewriting it again. This bug also had the effect of breaking the
4033 sender address verification caching when the sender address was rewritten.
4035 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4036 This has been changed so that if the ip literal address matches
4037 ignore_target_hosts, the router declines.
4039 7. Added expansion conditions match_domain, match_address, and match_local_
4040 part (NOT match_host).
4042 8. The placeholder for the Received: header didn't have a length field set.
4044 9. Added code to Exim itself and to exim_lock to test for a specific race
4045 condition that could lead to file corruption when using MBX delivery. The
4046 issue is with the lockfile that is created in /tmp. If this file is removed
4047 after a process has opened it but before that process has acquired a lock,
4048 there is the potential for a second process to recreate the file and also
4049 acquire a lock. This could lead to two Exim processes writing to the file
4050 at the same time. The added code performs the same test as UW imapd; it
4051 checks after acquiring the lock that its file descriptor still refers to
4052 the same named file.
4054 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4055 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4058 11. Added the smtp_active_hostname option. If used, this will typically be made
4059 to depend on the incoming interface address. Because $interface_address is
4060 not set up until the daemon has forked a reception process, error responses
4061 that can happen earlier (such as "too many connections") no longer contain
4064 12. If an expansion in a condition on a "warn" statement fails because a lookup
4065 defers, the "warn" statement is abandoned, and the next ACL statement is
4066 processed. Previously this caused the whole ACL to be aborted.
4068 13. Added the iplsearch lookup type.
4070 14. Added ident_timeout as a log selector.
4072 15. Added tls_certificate_verified as a log selector.
4074 16. Added a global option tls_require_ciphers (compare the smtp transport
4075 option of the same name). This controls incoming TLS connections.
4077 17. I finally figured out how to make tls_require_ciphers do a similar thing
4078 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4079 before starting the TLS session.
4081 18. Tabs are now shown as \t in -bP output.
4083 19. If the log selector return_path_on_delivery was set, Exim crashed when
4084 bouncing a message because it had too many Received: header lines.
4086 20. If two routers both had headers_remove settings, and the first one included
4087 a superfluous trailing colon, the final name in the first list and the
4088 first name in the second list were incorrectly joined into one item (with a
4089 colon in the middle).
4095 1. Added -C and -D options to the exinext utility, mainly to make it easier
4096 to include in the automated testing, but these could be helpful when
4097 multiple configurations are in use.
4099 2. The exinext utility was not formatting the output nicely when there was
4100 an alternate port involved in the retry record key, nor when there was a
4101 message id as well (for retries that were specific to a specific message
4102 and a specific host). It was also confused by IPv6 addresses, because of
4103 the additional colons they contain. I have fixed the IPv4 problem, and
4104 patched it up to do a reasonable job for IPv6.
4106 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4107 delivery, the log line now contains "pipelined" if PIPELINING was used.
4109 4. An SMTP transport process used to panic and die if the bind() call to set
4110 an explicit outgoing interface failed. This has been changed; it is now
4111 treated in the same way as a connect() failure.
4113 5. A reference to $sender_host_name in the part of a conditional expansion
4114 that was being skipped was still causing a DNS lookup. This no longer
4117 6. The def: expansion condition was not recognizing references to header lines
4118 that used bh_ and bheader_.
4120 7. Added the _cache feature to named lists.
4122 8. The code for checking quota_filecount in the appendfile transport was
4123 allowing one more file than it should have been.
4125 9. For compatibility with Sendmail, the command line option
4133 and sets the incoming protocol and host name (for trusted callers). The
4134 host name and its colon can be omitted when only the protocol is to be set.
4135 Note the Exim already has two private options, -pd and -ps, that refer to
4136 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4137 "s", but I don't think that's a major issue.
4139 10. A number of refactoring changes to the code, none of which should affect
4142 (a) The number of logging options was getting close to filling up the
4143 32-bit word that was used as a bit map. I have split them into two classes:
4144 those that are passed in the argument to log_write(), and those that are
4145 only ever tested independently outside of that function. These are now in
4146 separate 32-bit words, so there is plenty of room for expansion again.
4147 There is no change in the user interface or the logging behaviour.
4149 (b) When building, for example, log lines, the code previously used a
4150 macro that called string_cat() twice, in order to add two strings. This is
4151 not really sufficiently general. Furthermore, there was one instance where
4152 it was actually wrong because one of the argument was used twice, and in
4153 one call a function was used. (As it happened, calling the function twice
4154 did not affect the overall behaviour.) The macro has been replaced by a
4155 function that can join an arbitrary number of extra strings onto a growing
4158 (c) The code for expansion conditions now uses a table and a binary chop
4159 instead of a serial search (which was left over from when there were very
4160 few conditions). Also, it now recognizes conditions like "pam" even when
4161 the relevant support is not compiled in: a suitably worded error message is
4162 given if an attempt is made to use such a condition.
4164 11. Added ${time_interval:xxxxx}.
4166 12. A bug was causing one of the ddress fields not to be passed back correctly
4167 from remote delivery subprocesses. The field in question was not being
4168 subsequently used, so this caused to problems in practice.
4170 13. Added new log selectors queue_time and deliver_time.
4172 14. Might have fixed a bug in maildirsizefile handling that threw up
4173 "unexpected character" debug warnings, and recalculated the data
4174 unnecessarily. In any case, I expanded the warning message to give more
4177 15. Added the message "Restricted characters in address" to the statements in
4178 the default ACL that block characters like @ and % in local parts.
4180 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4181 Three changes have been made:
4183 (a) There was a serious bug; a negative response to MAIL caused the whole
4184 recipient domain to be cached as invalid, thereby blocking all messages
4185 to all local parts at the same domain, from all senders. This bug has
4186 been fixed. The domain is no longer cached after a negative response to
4187 MAIL if the sender used is not empty.
4189 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4192 (c) A new callout option, "use_sender" has been added for people who want
4193 the modified behaviour.
4199 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4202 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4203 indeed breaks things for older releases.
4205 3. Added additional logging to the case where there is a problem reading data
4206 from a filter that is running in a subprocess using a pipe, in order to
4207 try to track down a specific problem.
4209 4. Testing facility fudge: when running in the test harness and attempting
4210 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4211 getting "No route to host". Convert this to a timeout.
4213 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4216 6. Some OS don't have socklen_t but use size_t instead. This affects the
4217 fifth argument of getsockopt() amongst other things. This is now
4218 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4219 can be set for individual OS. I have set it for SunOS5, OSF1, and
4220 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4221 some earlier ones do not.
4223 7. Change 4.30/15 was not doing the test caselessly.
4225 8. The standard form for an IPv6 address literal was being rejected by address
4226 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4227 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4228 this, as well as the form without the "IPv6" on the front (but only when
4229 address literals are enabled, of course).
4231 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4233 10. Exim crashed if a message with an empty sender address specified by -f
4234 encountered a router with an errors_to setting. This could be provoked only
4235 by a command such as
4239 where an empty string was supplied; "<>" did not hit this bug.
4241 11. Installed PCRE release 4.5.
4243 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4244 remained set. It is now erased.
4246 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4247 times from message ids (which are base 36 rather than the normal 62).
4249 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4250 were being counted as actual protocol errors, and logged if the log
4251 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4252 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4253 and DATA following a set of rejected RCPTs do not count as protocol errors.
4254 In other words, Exim assumes they were pipelined, though this may not
4255 actually be the case. Of course, in all cases the client gets an
4256 appropriate error code.
4258 15. If a lookup fails in an ACL condition, a message about the failure may
4259 be available; it is used if testing the ACL cannot continue, because most
4260 such messages specify what the cause of the deferral is. However, some
4261 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4262 that caused an old message to be retained and used if a later statement
4263 caused a defer, replacing the real cause of the deferral.
4265 16. If an IP address had so many PTR records that the DNS lookup buffer
4266 was not large enough to hold them, Exim could crash while trying to process
4267 the truncated data. It now detects and logs this case.
4269 17. Further to 4.21/58, another change has been made: if (and only if) the
4270 first line of a message (the first header line) ends with CRLF, a bare LF
4271 in a subsequent header line has a space inserted after it, so as not to
4272 terminate the header.
4274 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4275 unnecessarily, used atoi() instead of strtol(), and didn't check the
4276 termination when getting file sizes from file names by regex.
4278 19. Completely re-implemented the support for maildirsize files, in the light
4279 of a number of problems with the previous contributed implementation
4280 (4.30/29). In particular:
4282 . If the quota is zero, the maildirsize file is maintained, but no quota is
4285 . If the maildir directory does not exist, it is created before any attempt
4286 to write a maildirsize file.
4288 . The quota value in the file is just a cache; if the quota is changed in
4289 the transport, the new value overrides.
4291 . A regular expression is available for excluding directories from the
4294 20. The autoreply transport checks the characters in options that define the
4295 message's headers; it allows continued headers, but it was checking with
4296 isspace() after an embedded newline instead of explicitly looking for a
4299 21. If all the "regular" hosts to which an address was routed had passed their
4300 expiry times, and had not reached their retry times, the address was
4301 bounced, even if fallback hosts were defined. Now Exim should go on to try
4304 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4305 equivalent code in the SMTP transport. Some hosts send humungous responses
4306 to HELO/EHLO, more than 1024 it seems.
4308 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4309 gives compiler warnings in some environments. I've now done it "properly",
4312 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4313 (because of problems with the built-in one) was declared to return uschar *
4314 instead of char *, causing compiler failure.
4316 25. Fixed a file descriptor leak when processing alias/forward files.
4318 26. Fixed a minor format string issue in dbfn.c.
4320 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4322 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4323 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4326 29. When a sender address is verified, it is cached, to save repeating the test
4327 when there is more than one recipient in a message. However, when the
4328 verification involves a callout, it is possible for different callout
4329 options to be set for different recipients. It is too complicated to keep
4330 track of this in the cache, so now Exim always runs a verification when a
4331 callout is required, relying on the callout cache for the optimization.
4332 The overhead is duplication of the address routing, but this should not be
4335 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4336 to be verified with callout=postmaster, and the main callout worked but the
4337 postmaster check failed, the verification correctly failed. However, if a
4338 subsequent RCPT command asked for sender verification *without* the
4339 postmaster check, incorrect caching caused this verification also to fail,
4342 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4343 it was not caching the DNS options (qualify_single, search_parents) that
4344 were used when the lookup failed. A subsequent lookup with different
4345 options therefore always gave the same answer, though there were cases
4346 where it should not have. (Example: a "domains = !$mx_any" option on a
4347 dnslookup router: the "domains" option is always processed without any
4348 widening, but the router might have qualify_single set.) Now Exim uses the
4349 cached value only when the same options are set.
4351 32. Added John Jetmore's "exipick" utility to the distribution.
4353 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4354 than a timeout (e.g. a certificate is required, and is not provided), an
4355 Exim server now closes the connection immediately. Previously it waited for
4356 the client to close - but if the client is SSL, it seems that they each
4357 wait for each other, leading to a delay before one of them times out.
4359 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4360 maintained 0.8.x compatibility because I don't think many are using it, and
4361 it is clearly obsolete.
4363 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4366 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4367 client certificate was expired. A simple patch fixes this, though I don't
4368 understand the full logic of why the verify callback is called multiple
4371 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4372 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4373 which causes problems with some clients (such as the Certicom SSL Plus
4374 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4375 disables the coutermeasure allowing Eudora to connect."
4377 38. Exim was not checking that a write() to a log file succeeded. This could
4378 lead to Bad Things if a log got too big, in particular if it hit a file
4379 size limit. Exim now panics and dies if it cannot write to a log file, just
4380 as it does if it cannot open a log file.
4382 39. Modified OS/Makefile-Linux so that it now contains
4384 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4386 The two -D definitions ensure that Exim is compiled with large file
4387 support, which makes it possible to handle log files that are bigger than
4390 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4391 instance) a domain was checked against a named list that involved a lookup,
4392 causing $domain_data to be set, then another domain was checked against the
4393 same list, then the first domain was re-checked, the value of $domain_data
4394 after the final check could be wrong. In particular, if the second check
4395 failed, it could be set empty. This bug probably also applied to
4398 41. The strip_trailing_dot option was not being applied to the address given
4399 with the -f command-line option.
4401 42. The code for reading a message's header from the spool was incrementing
4402 $received_count, but never initializing it. This meant that the value was
4403 incorrect (doubled) while delivering a message in the same process in which
4404 it was received. In the most common configuration of Exim, this never
4405 happens - a fresh exec is done - but it can happen when
4406 deliver_drop_privilege is set.
4408 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4409 it now includes up to 150 characters of the unexpected data in the log
4412 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4413 and building data strings. The size of both of these buffers was 10 000
4414 bytes - far larger than anybody would *ever* want, thought I. Needless to
4415 say, somebody hit the limit. I have increased the maximum line length to
4416 20 000 and the maximum data length of concatenated lines to 100 000. I have
4417 also fixed two bugs, because there was no checking on these buffers. Tsk,
4418 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4419 buffer is too small.
4421 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4422 lsearch lookups. Now it does.
4424 46. When parsing a route_list item in a manualroute router, a fixed-length
4425 buffer was used for the list of hosts. I made this 1024 bytes long,
4426 thinking that nobody would ever have a list of hosts that long. Wrong.
4427 Somebody had a whole pile of complicated expansion conditions, and the
4428 string was silently truncated, leading to an expansion error. It turns out
4429 that it is easier to change to an unlimited length (owing to other changes
4430 that have happened since this code was originally written) than to build
4431 structure for giving a limitation error. The length of the item that
4432 expands into the list of hosts is now unlimited.
4434 47. The lsearch lookup could not handle data where the length of text line was
4435 more than 4095 characters. Such lines were truncated, leading to shortened
4436 data being returned. It should now handle lines of any length.
4438 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4439 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4442 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4443 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4444 no longer happens in comment lines. A list of replacements is now placed
4445 at the head of all of the source files, except those whose only change is
4446 to replace PERL_COMMAND in the very first #! line.
4448 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4449 messages on the queue, with a bottom-up merge sort, using code contributed
4450 by Michael Haardt. This should make operations like -bp somewhat faster on
4451 large queues. It won't affect queue runners, except when queue_run_in_order
4454 51. Installed eximstats 1.31 in the distribution.
4456 52. Added support for SRV lookups to the dnslookup router.
4458 53. If an ACL referred to $message_body or $message_body_end, the value was not
4459 reset for any messages that followed in the same SMTP session.
4461 54. The store-handling optimization for building very long strings was not
4462 differentiating between the different store pools. I don't think this
4463 actually made any difference in practice, but I've tidied it.
4465 55. While running the routers to verify a sender address, $sender_address
4466 was still set to the sender address. This is wrong, because when routing to
4467 send a bounce to the sender, it would be empty. Therefore, I have changed
4468 it so that, while verifying a sender address, $sender_address is set to <>.
4469 (There is no change to what happens when verifying a recipient address.)
4471 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4472 target A or AAAA records (if not already returned) without resetting the
4473 qualify_single or search_parents options of the DNS resolver. These are
4474 inappropriate in this case because the targets of MX and SRV records must
4475 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4476 target that, when qualified, matched something in the local domain. These
4477 two options are now turned off when doing these lookups.
4479 57. It seems that at least some releases of Reiserfs (which does not have the
4480 concept of a fixed number of inodes) returns zero and not -1 for the
4481 number of available inodes. This interacted badly with check_spool_inodes,
4482 which assumed that -1 was the "no such thing" setting. What I have done is
4483 to check that the total number of inodes is greater than zero before doing
4484 the test of how many are available.
4486 58. When a "warn" ACL statement has a log_message modifier, the message is
4487 remembered, and not repeated. This is to avoid a lot of repetition when a
4488 message has many recipients that cause the same warning to be written.
4489 Howewer, Exim was preserving the list of already written lines for an
4490 entire SMTP session, which doesn't seem right. The memory is now reset if a
4491 new message is started.
4493 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4494 debugging output unless log_rewrite was also set.
4496 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4497 of (int)(handle) when we know that handle contains (void *)(-1).
4499 61. The Exim daemon panic-logs an error return when it closes the incoming
4500 connection. However "connection reset by peer" seems to be common, and
4501 isn't really an error worthy of noting specially, so that particular error
4504 62. When Exim is trying to find all the local interfaces, it used to panic and
4505 die if the ioctl to get the interface flags failed. However, it seems that
4506 on at least one OS (Solaris 9) it is possible to have an interface that is
4507 included in the list of interfaces, but for which you get a failure error
4508 for this call. This happens when the interface is not "plumbed" into a
4509 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4510 failure of the "get flags" call assumes that the interface is down.
4512 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4513 makes life easier for people who are doing arithmetic on fields extracted
4514 from dates, where you often get leading zeros that should not be
4515 interpreted as octal.
4517 64. Added qualify_domain to the redirect router, to override the global
4520 65. If a pathologically long header line contained very many addresses (the
4521 report of this problem mentioned 10 000) and each of them was rewritten,
4522 Exim could use up a very large amount of memory. (It kept on making new
4523 copies of the header line as it rewrote, and never released the old ones.)
4524 At the expense of a bit more processing, the header rewriting function has
4525 been changed so that it no longer eats memory in this way.
4527 66. The generation of the Received: header has been moved from the time that a
4528 message starts to be received, to the time that it finishes. The timestamp
4529 in the Received: header should now be very close to that of the <= log
4530 line. There are two side-effects of this change:
4532 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4533 logged header lines no longer include the local Received: line, because
4534 it has not yet been created. The same applies to a copy of the message
4535 that is returned to a non-SMTP sender when a message is rejected.
4537 (b) When a filter file is tested using -bf, no additional Received: header
4538 is added to the test message. After some thought, I decided that this
4541 This change does not affect the value of $received_for. It is still set
4542 after address rewriting, but before local_scan() is called.
4544 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4546 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4547 gave an unhelpful panic error message, and a defer error. I have managed to
4548 change this behaviour so that it now rejects any supplied certificate,
4549 which seems right, as the list of acceptable certificates is empty.
4551 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4552 gave an unhelpful defer error. I have not managed to make this reject any
4553 supplied certificates, but the error message it gives is "no certificate
4554 supplied", which is not helpful.
4556 70. exigrep's output now also includes lines that are not associated with any
4557 message, but which match the given pattern. Implemented by a patch from
4558 Martin Sluka, which also tidied up the Perl a bit.
4560 71. Recipient callout verification, like sender verification, was using <> in
4561 the MAIL FROM command. This isn't really the right thing, since the actual
4562 sender may affect whether the remote host accepts the recipient or not. I
4563 have changed it to use the actual sender in the callout; this means that
4564 the cache record is now keyed on a recipient/sender pair, not just the
4565 recipient address. There doesn't seem to be a real danger of callout loops,
4566 since a callout by the remote host to check the sender would use <>.
4567 [SEE ABOVE: changed after hitting problems.]
4569 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4570 temporary errors. However, in the case of such a code being given after
4571 the end of a data transmission (i.e. after ".") Exim was failing to write
4572 a retry record for the message. (Yes, there was some broken host that was
4573 actually sending 8xx at this point.)
4575 73. An unknown lookup type in a host list could cause Exim to panic-die when
4576 the list was checked. (An example that provoked this was putting <; in the
4577 middle of a list instead of at the start.) If this happened during a DATA
4578 ACL check, a -D file could be left lying around. This kind of configuration
4579 error no longer causes Exim to die; instead it causes a defer errror. The
4580 incident is still logged to the main and panic logs.
4582 74. Buglet left over from Exim 3 conversion. The message "too many messages
4583 in one connection" was written to the rejectlog but not the mainlog, except
4584 when address rewriting (yes!) was being logged.
4586 75. Added write_rejectlog option.
4588 76. When a system filter was run not as root (that is, when system_filter_user
4589 was set), the values of the $n variables were not being returned to the
4590 main process; thus, they were not subsequently available in the $sn
4593 77. Added +return_path_on_delivery log selector.
4595 78. A connection timeout was being treated differently from recipients deferred
4596 when testing hosts_max_try with a message that was older than the host's
4597 retry timeout. (The host should not be counted, thus allowing all hosts to
4598 be tried at least once before bouncing.) This may have been the cause of an
4599 occasionally reported bug whereby a message would remain on the queue
4600 longer than the retry timeout, but would be bounced if a delivery was
4601 forced. I say "may" because I never totally pinned down the problem;
4602 setting up timeout/retry tests is difficult. See also the next item.
4604 79. The ultimate address timeout was not being applied to errors that involved
4605 a combination of host plus message (for example, a timeout on a MAIL
4606 command). When an address resolved to a number of possible hosts, and they
4607 were not all tried for each delivery (e.g. because of hosts_max_try), a
4608 message could remain on the queue longer than the retry timeout.
4610 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4613 81. Fixed an obscure SMTP outgoing bug which required at least the following
4614 conditions: (a) there was another message waiting for the same server;
4615 (b) the server returned 5xx to all RCPT commands in the first message so
4616 that the message was not completed; (c) the server dropped the connection
4617 or gave a negative response to the RSET that Exim sends to abort the
4618 transaction. The observed case was a dropped connection after DATA that had
4619 been sent in pipelining mode. That is, the server had advertised PIPELINING
4620 but was not implementing it correctly. The effect of the bug was incorrect
4621 behaviour, such as trying another host, and this could lead to a crash.
4627 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4628 and daemon.c were passed as pointers to ints; they should have been
4629 pointers to socklen_t variables (which are typically unsigned ints).
4631 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4634 3. Fixed a really odd bug that affected only the testing scheme; patching a
4635 certain fixed string in the binary changed the value of another string that
4636 happened to be identical to the end of the original first string.
4638 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4639 name", it returns that address as the IP address. On some operating
4640 systems (e.g. Solaris), it also passes back the IP address string as the
4641 "host name". However, on others (e.g. Linux), it passes back an empty
4642 string. Exim wasn't checking for this, and was changing the host name to an
4643 empty string, assuming it had been canonicized.
4645 5. Although rare, it is permitted to have more than one PTR record for a given
4646 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4647 all the names associated with an address, because they do in Solaris.
4648 However, it seems that they do not in Linux for data that comes from the
4649 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4650 I found this out when I moved to a new Linux workstation and tried to run
4651 the Exim test suite.
4653 To get round this problem I have changed the code so that it now does its
4654 own call to the DNS to look up PTR records when searching for a host name.
4655 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4656 addresses that are only in /etc/hosts are still found.
4658 This behaviour is, however, controlled by an option called host_lookup_
4659 order, which defaults to "bydns:byaddr". If people want to use the other
4660 order, or indeed, just use one or the other means of lookup, they can
4661 specify it in this variable.
4663 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4664 some operating systems, this comes back from gethostbyaddr() as an empty
4665 string, and this is what Exim used to test for. However, it seems that in
4666 other systems, "." is yielded. Exim now tests for this case too.
4668 7. The values of check_spool_space and check_log_space are now held internally
4669 as a number of kilobytes instead of an absolute number of bytes. If a
4670 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4671 kilobyte. This means that much larger values can be stored.
4673 8. Exim monitor: an attempt to get the action menu when not actually pointing
4674 at a message produces an empty menu entitled "No message selected". This
4675 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4676 no entries in it ("Shell widget menu has zero width and/or height"). So I
4677 have added a single, blank menu entry in this case.
4679 9. Added ${quote_local_part.
4681 10. MIME decoding is now applied to the contents of Subject: header lines when
4684 11. Now that a reference to $sender_host_address automatically causes a reverse
4685 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4686 host lookup before query-style lookups in lists that might use this
4687 variable. This has therefore been abolished, and the "net-" prefix is no
4688 longer necessary for query-style lookups.
4690 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4691 have been a typo for LFLAGS, so it has been changed.
4693 13. The install script calls Exim with "-C /dev/null" in order to find the
4694 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4695 to be output. Howeve, since Exim outputs its version number before the
4696 error, it didn't break the script. It just looked ugly. I fixed this by
4697 always allowing "-C /dev/null" if the caller is root.
4699 14. Ignore overlarge ACL variable number when reading spool file - insurance
4700 against a later release with more variables having written the file.
4702 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4703 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4704 this, as well as the form without the "IPv6" on the front.
4706 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4707 OS/Makefile-Darwin file.
4709 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4710 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4712 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4713 variable. However, some OS use ints for inodes. I've added cast to long int
4714 to get rid of the compiler warning.
4716 19. I had forgotten to lock out "/../" in configuration file names when
4717 ALT_CONFIG_PREFIX was set.
4719 20. Routers used for verification do not need to specify transports. However,
4720 if such a router generated a host list, and callout was configured, Exim
4721 crashed, because it could not find a port number from the (non-existent)
4722 transport. It now assumes port 25 in this circumstance.
4724 21. Added the -t option to exigrep.
4726 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4727 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4728 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4729 them is set, LOOKUP_LSEARCH is forced.
4731 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4733 24. Added sender and host information to the "rejected by local_scan()" log
4734 line; previously there was no indication of these.
4736 25. Added .include_if_exists.
4738 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4739 Linux. It turns out that not all file systems support this. Apparently some
4740 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4741 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4742 not supported on the file descriptor, is now ignored when Exim is trying to
4743 sync a directory. This applies only to Linux.
4745 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4747 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4748 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4751 29. Added support for maildirsize files from supplied patch (modified a bit).
4753 30. The use of :fail: followed by an empty string could lead Exim to respond to
4754 sender verification failures with (e.g.):
4756 550 Verification failed for <xxx>
4757 550 Sender verify failed
4759 where the first response line was missing the '-' that indicates it is not
4760 the final line of the response.
4762 31. The loop for finding the name of the user that called Exim had a hardwired
4763 limit of 10; it now uses the value of finduser_retries, which is used for
4764 all other user lookups.
4766 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4769 33. Exim was neglecting to zero errno before one call of strtol() when
4770 expanding a string and expecting an integer value. On some systems this
4771 resulted in spurious "integer overflow" errors. Also, it was casting the
4772 result into an int without checking.
4774 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4775 did not work. The code looks as if it has *never* worked, though it appears
4776 to have been documented since at least releast 1.62. I have made it work.
4778 35. The "timeout_DNS" error in retry rules, also documented since at least
4779 1.62, also never worked. As it isn't clear exactly what this means, and
4780 clearly it isn't a major issue, I have abolished the feature by treating it
4781 as "timeout", and writing a warning to the main and panic logs.
4783 36. The display of retry rules for -brt wasn't always showing the error code
4786 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4787 timeout_connect_A, timeout_connect_MX.
4789 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4790 to the empty sender.
4792 39. The daemon was not analysing the content of -oX till after it had closed
4793 stderr and disconnected from the controlling terminal. This meant that any
4794 syntax errors were only noted on the panic log, and the return code from
4795 the command was 0. By re-arranging the code a little, I've made the
4796 decoding happen first, so such errors now appear on stderr, and the return
4797 code is 1. However, the actual setting up of the sockets still happens in
4798 the disconnected process, so errors there are still only recorded on the
4801 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4802 connections (as happens on some IP stacks) was logged at start up time as
4803 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4804 it from "IPv6 and IPv4", which means that two separate sockets are being
4807 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4808 says whether AF_INET or AF_INET6 was passed as an argument.
4810 42. Exiwhat output was messed up when time zones were included in log
4813 43. Exiwhat now gives more information about the daemon's listening ports,
4814 and whether -tls-on-connect was used.
4816 44. The "port" option of the smtp transport is now expanded.
4818 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4819 silently ignored. Now an error message is written to the main and panic
4822 46. There's a new ACL modifier called "logwrite" which writes to a log file
4823 as soon as it is encountered.
4825 47. Added $local_user_uid and $local_user_gid at routing time.
4827 48. Exim crashed when trying to verify a sender address that was being
4830 49. Exim was recognizing only a space character after ".include". It now also
4831 recognizes a tab character.
4833 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4834 extracting the relevant information from the specification. The man page no
4835 longer contains scrambled data for the -d option, and I've added a section
4836 at the front about calling Exim under different names.
4838 51. Added "extra_headers" argument to the "mail" command in filter files.
4840 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4843 53. Installed eximstats 1.29.
4845 54. Added transport_filter_timeout as a generic transport option.
4847 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4848 Cc: header lines. This was required by RFC 822, but it not required by RFC
4851 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4852 incoming messages that did not have them. Now it does so only if the
4853 message originates locally, that is, if there is no associated remote host
4854 address. When Resent- header lines are present, this applies to the Resent-
4855 lines rather than the non-Resent- lines.
4857 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4858 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4860 58. Messages for configuration errors now include the name of the main
4861 configuration file - useful now that there may be more than one file in a
4862 list (.included file names were always shown).
4864 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4865 for those rare installations that do not start the daemon as root or run it
4866 setuid root. I've cut out the call to initgroups() if the daemon is not
4869 60. The Exim user and group can now be bound into the binary as text strings
4870 that are looked up at the start of Exim's processing.
4872 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4874 62. Added $mailstore_basename variable.
4876 63. Installed patch to sieve.c from Michael Haardt.
4878 64. When Exim failed to open the panic log after failing to open the main log,
4879 the original message it was trying to log was written to stderr and debug
4880 output, but if they were not available (the usual case in production), it
4881 was lost. Now it is written to syslog before the two lines that record the
4882 failures to open the logs.
4884 65. Users' Exim filters run in subprocesses under the user's uid. It is
4885 possible for a "deliver" command or an alias in a "personal" command to
4886 provoke an address rewrite. If logging of address rewriting is configured,
4887 this fails because the process is not running as root or exim. There may be
4888 a better way of dealing with this, but for the moment (because 4.30 needs
4889 to be released), I have disabled address rewrite logging when running a
4890 filter in a non-root, non-exim process.
4896 1. The buildconfig auxiliary program wasn't quoting the value set for
4897 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4898 not defined. This bug was masked in 4.22 by the effect that was fixed in
4901 2. Some messages that were rejected after a message id was allocated were
4902 shown as "incomplete" by exigrep. It no longer does this for messages that
4903 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4905 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4906 have allow_domain_literals set, the ID did not get logged in the <= line.
4907 Domain literals are now always recognized in Message-ID: header lines.
4909 4. The first argument for a ${extract expansion item is the key name or field
4910 number. Leading and trailing spaces in this item were not being ignored,
4911 causing some misleading effects.
4913 5. When deliver_drop_privilege was set, single queue runner processes started
4914 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4915 same command in the process it spins off) were not dropping privilege.
4917 6. When the daemon running as "exim" started a queue runner, it always
4918 re-executed Exim in the spun-off process. This is a waste of effort when
4919 deliver_drop_privilege is set. The new process now just calls the
4920 queue-runner function directly.
4926 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4929 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4930 ignored. Though the use of -oP was forcing the writing of a pid file, it
4931 was always written to the default place.
4933 3. If the message "no IP address found for host xxxx" is generated during
4934 incoming verification, it is now followed by identification of the incoming
4935 connection (so you can more easily find what provoked it).
4937 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4939 5. Added some features to "harden" Exim a bit more against certain attacks:
4941 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4942 be put in Local/Makefile. This is like the never_users runtime option,
4943 but it cannot be overridden. The default setting is "root".
4945 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4946 prefix string with which any file named in a -C command line option
4949 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4950 is retained for -C and -D only if the caller of Exim is root. Without
4951 it, the exim user may also use -C and -D and retain privilege.
4953 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4954 command line option is disabled.
4956 6. Macro names set by the -D option must start with an upper case letter, just
4957 like macro names defined in the configuration file.
4959 7. Added "dereference=" facility to LDAP.
4961 8. Two instances of the typo "uknown" in the source files are fixed.
4963 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4964 the Configure-Makefile script screwed up while processing it.
4966 10. Incorporated PCRE 4.4.
4968 11. The SMTP synchronization check was not operating right at the start of an
4969 SMTP session. For example, it could not catch a HELO sent before the client
4970 waited for the greeting. There is now a check for outstanding input at the
4971 point when the greeting is written. Because of the duplex, asynchronous
4972 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4973 way, but not yet received, when the check is performed.
4975 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4976 on TCP/IP sockets, because this apparently causes some broken clients to
4979 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4980 unchanged) from the Cygwin maintainer.
4982 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4983 was supported instead of testing for mbx. Effectively a typo.
4985 15. The spa authenticator server code was not checking that the input it
4986 received was valid base64.
4988 16. The debug output line for the "set" modifier in ACLs was not showing the
4989 name of the variable that was being set.
4991 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4993 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4994 lookup on its own. Something else was needed to trigger the lookup. For
4995 example, a match in host_lookup or the need for a host name in a host list.
4996 Now, if $sender_host_name is referenced and the host name has not yet been
4997 looked up, a lookup is performed. If the lookup fails, the variable remains
4998 empty, and $host_lookup_failed is set to "1".
5000 19. Added "eqi" as a case-independent comparison operator.
5002 20. The saslauthd authentication condition could segfault if neither service
5003 nor realm was specified.
5005 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5006 error message that was logged was misleading, and incoming SMTP
5007 connections were dropped. The message is now more accurate, and temporary
5008 errors are given to SMTP connections.
5010 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5011 (except RSET) with a 421 error, until QUIT is received. However, it was
5012 failing to send a response to QUIT.
5014 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5015 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5016 ACL is now run after the helo_try_verify_hosts code.
5018 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5019 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5020 case-independent, so other case variants are also recognized). Apparently
5021 some systems use these upper case variants.
5023 25. If more than two messages were waiting for the same host, and a transport
5024 filter was specified for the transport, Exim sent two messages over the
5025 same TCP/IP connection, and then failed with "socket operation on non-
5026 socket" when it tried to send the third.
5028 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5030 27. The extern definition of crypt16() in expand.c was not being excluded when
5031 the OS had its own crypt16() function.
5033 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5034 as a preferred synonym for return_size_limit, both as an option and as an
5037 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5039 30. Changed the default configuration ACL to relax the local part checking rule
5040 for addresses that are not in any local domains. For these addresses,
5041 slashes and pipe symbols are allowed within local parts, but the sequence
5042 /../ is explicitly forbidden.
5044 31. SPA server authentication was not clearing the challenge buffer before
5047 32. log_message in a "warn" ACL statement was writing to the reject log as
5048 well as to the main log, which contradicts the documentation and doesn't
5049 seem right (because no rejection is happening). So I have stopped it.
5051 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5052 However, I am unable to do any testing of this.
5054 34. Fixed an infelicity in the appendfile transport. When checking directories
5055 for a mailbox, to see if any needed to be created, it was accidentally
5056 using path names with one or more superfluous leading slashes; tracing
5057 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5059 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5060 contents are added to the log line that is written for every discarded
5061 recipient. (Previously a log_message setting was ignored.)
5063 36. The ${quote: operator now quotes the string if it is empty.
5065 37. The install script runs exim in order to find its version number. If for
5066 some reason other than non-existence or emptiness, which it checks, it
5067 could not run './exim', it was installing it with an empty version number,
5068 i.e. as "exim-". This error state is now caught, and the installation is
5071 38. An argument was missing from the function that creates an error message
5072 when Exim fails to connect to the socket for saslauthd authentication.
5073 This could cause Exim to crash, or give a corrupted message.
5075 39. Added isip, isip4, and isip6 to ${if conditions.
5077 40. The ACL variables $acl_xx are now saved with the message, and can be
5078 accessed later in routers, transports, and filters.
5080 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5081 strings in the file are not string-expanded.
5083 42. If a MAIL command specified a SIZE value that was too large to fit into an
5084 int variable, the check against message_size_limit failed. Such values are
5085 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5086 day this will have to be increased, but I don't think I want to be around
5087 when emails are that large.
5094 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5095 iconv() is not standard in FreeBSD.
5097 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5098 IPv6 enabled. The observed symptom was a segmentation fault on return from
5099 the function os_common_find_running_interfaces() in src/os.c.
5101 3. In the check_special_case() function in daemon.c I had used "errno" as an
5102 argument name, which causes warnings on some systems. This was basically a
5103 typo, since it was named "eno" in the comments!
5105 4. The code that waits for the clock to tick (at a resolution of some fraction
5106 of a second) so as to ensure message-id uniqueness was always waiting for
5107 at least one whole tick, when it could have waited for less. [This is
5108 almost certainly not relevant at current processor speeds, where it is
5109 unlikely to ever wait at all. But we try to future-proof.]
5111 5. The function that sleeps for a time interval that includes fractions of a
5112 second contained a race. It did not block SIGALRM between setting the
5113 timer, and suspending (a couple of lines later). If the interval was short
5114 and the sigsuspend() was delayed until after it had expired, the suspension
5115 never ended. On busy systems this could lead to processes getting stuck for
5118 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5119 process, before it forks any delivery processes. The open lookup caching
5120 mechanism meant that the open file or database connection was passed into
5121 the delivery process. The problem was that delivery processes always tidy
5122 up cached lookup data. This could cause a problem for the next delivery
5123 process started by the queue runner, because the external queue runner
5124 process does not know about the closure. So the next delivery process
5125 still has data in the lookup cache. In the case of a file lookup, there was
5126 no problem because closing a file descriptor in a subprocess doesn't affect
5127 the parent. However, if the lookup was caching a connection to a database,
5128 the connection was closed, and the second delivery process was likely to
5129 see errors such as "PGSQL: query failed: server closed the connection
5130 unexpectedly". The problem has been fixed by closing all cached lookups
5131 in a queue runner before running a delivery process.
5133 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5134 seem to have the "const" qualifier which it has on other OS. I've
5137 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5138 *of the same type* (client or server) with the same public name that an
5139 error should be diagnosed.
5141 9. When Exim looked up a host name for an IP address, but failed to find the
5142 original IP address when looking up the host name (a safety check), it
5143 output the message "<ip address> does not match any IP for NULL", which was
5144 confusing, to say the least. The bug was that the host name should have
5145 appeared instead of "NULL".
5147 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5148 user that is built into the binary, and the -C or -D options is used, root
5149 privilege is dropped before the configuration file is read. In addition,
5150 logging is switched to stderr instead of the normal log files. If the
5151 configuration then re-defines the Exim user, the unprivileged environment
5152 is probably not what is expected, so Exim logs a panic warning message (but
5155 However, if deliver_drop_privilege is set, the unprivileged state may well
5156 be exactly what is intended, so the warning has been cut out in that case,
5157 and Exim is allowed to try to write to its normal log files.
5163 1. smtp_return_error_details was not giving details for temporary sender
5164 or receiver verification errors.
5166 2. Diagnose a configuration error if two authenticators have the same public
5169 3. Exim used not to create the message log file for a message until the first
5170 delivery attempt. This could be confusing when incoming messages were held
5171 for policy or load reasons. The message log file is now created at the time
5172 the message is received, and an initial "Received" line is written to it.
5174 4. The automatically generated man page for command line options had a minor
5175 bug that caused no ill effects; however, a more serious problem was that
5176 the procedure for building the man page automatically didn't always
5177 operate. Consequently, release 4.20 contains an out-of-date version. This
5178 shouldn't happen again.
5180 5. When building Exim with embedded Perl support, the script that builds the
5181 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5182 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5184 6. The freeze_tell option was not being used for messages that were frozen on
5185 arrival, either by an ACL or by local_scan().
5187 7. Added the smtp_incomplete_transaction log selector.
5189 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5190 was accepting AUTH without a new EHLO.
5192 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5193 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5196 10. Logging of TCP/IP connections (when configured) now happens in the main
5197 daemon process instead of the child process, so that the TCP/IP connection
5198 count is more accurate (but it can never be perfect).
5200 11. The use of "drop" in a nested ACL was not being handled correctly in the
5201 outer ACL. Now, if condition failure induced by the nested "drop" causes
5202 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5203 or "require"), the connection is dropped.
5205 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5206 that yield "discard" can now be used with an "accept" or a "discard" verb,
5207 but an error is generated for any others (because I can't see a useful way
5208 to define what should happen).
5210 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5211 are now processed in the same way as lines in the Exim configuration file.
5212 In particular, continuation lines are supported.
5214 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5216 15. Added -ti meaning -t -i.
5218 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5219 domains, and warn by logging if others are found.
5221 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5222 returned from GIFCONF when Exim is trying to find the list of interfaces on
5223 a host. The code in os.c has been modified to copy each ifreq to an aligned
5224 structure in all cases.
5226 Also, in some cases, the returned ifreq's were being copied to a 'struct
5227 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5228 means the last couple of bytes of an IPv6 address could be chopped if the
5229 ifreq contained only a normal sockaddr (14 bytes storage).
5231 18. Named domain lists were not supported in the hosts_treat_as_local option.
5232 An entry such as +xxxx was not recognized, and was treated as a literal
5235 19. Ensure that header lines added by a DATA ACL are included in the reject log
5236 if the ACL subsequently rejects the message.
5238 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5239 MD5 (which is deprecated).
5241 21. When testing a filter file using -bf, Exim was writing a message when it
5242 took the sender from a "From " line in the message, but it was not doing so
5243 when it took $return_path from a Return-Path: header line. It now does.
5245 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5246 with a valid header line field name (a series of printing characters
5247 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5249 23. Changed "disc" in the source to "disk" to conform to the documentation and
5250 the book and for uniformity.
5252 24. Ignore Sendmail's -Ooption=value command line item.
5254 25. When execve() failed while trying to run a command in a pipe transport,
5255 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5256 could be confused with a return value of 69 from the command itself. This
5257 has been changed to 127, the value the shell returns if it is asked to run
5258 a non-existent command. The wording for the related log line suggests a
5259 non-existent command as the problem.
5261 26. If received_header_text expands to an empty string, do not add a Received:
5262 header line to the message. (Well, it adds a token one on the spool, but
5263 marks it "old" so that it doesn't get used or transmitted.)
5265 27. Installed eximstats 1.28 (addition of -nt option).
5267 28. There was no check for failure on the call to getsockname() in the daemon
5268 code. This can fail if there is a shortage of resources on the system, with
5269 ENOMEM, for example. A temporary error is now given on failure.
5271 29. Contrary to the C standard, it seems that in some environments, the
5272 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5273 program. Exim now does this explicitly; it affects the formatting of
5274 timestamps using strftime().
5276 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5277 complaints. I've now initialized all the variables, to avoid this.
5279 32. Header lines added by a system filter were not being "seen" during
5280 transport-time rewrites.
5282 33. The info_callback() function passed to OpenSSL is set up with type void
5283 (*)(SSL *, int, int), as described somewhere. However, when calling the
5284 function (actually a macro) that sets it up, the type void(*)() is
5285 expected. I've put in a cast to prevent warnings from picky compilers.
5287 34. If a DNS black list lookup found a CNAME record, but there were no A
5288 records associated with the domain it pointed at, Exim crashed.
5290 35. If a DNS black list lookup returned more than one A record, Exim ignored
5291 all but the first. It now scans all returned addresses if a particular IP
5292 value is being sought. In this situation, the contents of the
5293 $dnslist_value variable are a list of all the addresses, separated by a
5296 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5297 to accept a host name and all its aliases if the forward lookup for any of
5298 them yielded the IP address of the incoming connection. Now it accepts only
5299 those names whose forward lookup yields the correct IP address. Any other
5300 names are discarded. This closes a loophole whereby a rogue DNS
5301 administrator could create reverse DNS records to break through a
5302 wildcarded host restriction in an ACL.
5304 37. If a user filter or a system filter that ran in a subprocess used any of
5305 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5306 the wrong values were passed to the pipe command ($thisaddress had the
5307 value of $0, $0 had the value of $1, etc). This bug was introduced by
5308 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5310 38. Improved the line breaking for long SMTP error messages from ACLs.
5311 Previously, if there was no break point between 40 and 75 characters, Exim
5312 left the rest of the message alone. Two changes have been made: (a) I've
5313 reduced the minimum length to 35 characters; (b) if it can't find a break
5314 point between 35 and 75 characters, it looks ahead and uses the first one
5315 that it finds. This may give the occasional overlong line, but at least the
5316 remaining text gets split now.
5318 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5319 file descriptors might be low, and that setting 1000 would always raise it.
5320 It turns out that in some environments, the limit is already over 1000 and
5321 that lowering it causes trouble. So now Exim takes care not to decrease it.
5323 40. When delivering a message, the value of $return_path is set to $sender_
5324 address at the start of routing (routers may change the value). By an
5325 oversight, this default was not being set up when an address was tested by
5326 -bt or -bv, which affected the outcome if any router or filter referred to
5329 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5330 log only when "message" or "log_message" are set. However, if one of the
5331 conditions was an address verification, or a call to a nested ACL, the
5332 messages generated by the underlying test were being passed through. This
5333 no longer happens. The underlying message is available in $acl_verify_
5334 message for both "message" and "log_message" expansions, so it can be
5335 passed through if needed.
5337 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5338 new expansion $bh_ to give the encoded byte string without charset
5339 translation. Translation happens only if iconv() is available; HAVE_ICONV
5340 indicates this at build time. HEADERS_CHARSET gives the charset to
5341 translate to; headers_charset can change it in the configuration, and
5342 "headers charset" can change it in an individual filter file.
5344 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5345 that creates RFC 2047 encoded "words" labels them as that charset instead
5346 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5347 expansion operator; (ii) when Exim creates a From: line for a local
5348 message; (iii) when a header line is rewritten to include a "phrase" part.
5350 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5351 buggy, causing it to skip the first lines of messages whose message ID
5352 ended in 'D'. This would not have bitten before Exim release 4.14, because
5353 message IDs were unlikely to end in 'D' before then. The effect was to have
5354 incorrect size information for certain domains.
5356 45. #include "config.h" was missing at the start of the crypt16.c module. This
5357 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5360 46. If there was a timeout during a "random" callout check, Exim treated it as
5361 a failure of the random address, and carried on sending RSET and the real
5362 address. If the delay was just some slowness somewhere, the response to the
5363 original RCPT would be taken as a response to RSET and so on, causing
5364 mayhem of various kinds.
5366 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5367 when I implemented it. It didn't allow for the fact that some option values
5368 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5369 the right test for positive values.
5371 48. Domain names in DNS records are case-independent. Exim always looks them up
5372 in lower case. Some resolvers return domain names in exactly the case they
5373 appear in the zone file, that is, they may contain uppercase letters. Not
5374 all resolvers do this - some return always lower case. Exim was treating a
5375 change of case by a resolver as a change of domain, similar to a widening
5376 of a domain abbreviation. This triggered its re-routing code and so it was
5377 trying to route what was effectively the same domain again. This normally
5378 caused routing to fail (because the router wouldn't handle the domain
5379 twice). Now Exim checks for this case specially, and just changes the
5380 casing of the domain that it ultimately uses when it transmits the message
5383 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5386 50. If a filter generated a file delivery with a non-absolute name (possible if
5387 no home directory exists for the router), the forbid_file option was not
5390 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5391 the existing equality matching.
5393 52. Exim was using ints instead of ino_t variables in some places where it was
5394 dealing with inode numbers.
5396 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5397 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5398 called TMPDIR, and if it finds it is different, it changes its value.
5400 54. The smtp_printf() function is now made available to local_scan() so
5401 additional output lines can be written before returning. There is also an
5402 smtp_fflush() function to enable the detection of a dropped connection.
5403 The variables smtp_input and smtp_batched_input are exported to
5406 55. Changed the default runtime configuration: the message "Unknown user"
5407 has been removed from the ACL, and instead placed on the localuser router,
5408 using the cannot_route_message feature. This means that any verification
5409 failures that generate their own messages won't get overridden. Similarly,
5410 the "Unrouteable address" message that was in the ACL for unverifiable
5411 relay addresses has also been removed.
5413 56. Added hosts_avoid_esmtp to the smtp transport.
5415 57. The exicyclog script was not checking for the esoteric option
5416 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5417 will work only if exicyclog is run under the appropriate euid.
5419 58. Following a discussion on the list, the rules by which Exim recognises line
5420 endings on incoming messages have been changed. The -dropcr and drop_cr
5421 options are now no-ops, retained only for backwards compatibility. The
5422 following line terminators are recognized: LF CRLF CR. However, special
5423 processing applies to CR:
5425 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5426 nor a local message in the state where . is a terminator.
5428 (ii) If a bare CR is encountered in a header line, an extra space is added
5429 after the line terminator so as not to end the header. The reasoning
5430 behind this is that bare CRs in header lines are most likely either
5431 to be mistakes, or people trying to play silly games.
5433 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5434 was being incorrectly given as 18 bytes larger than it should have been.
5435 This is a VOB (very old bug).
5437 60. This may never have affected anything current, but just in case it has:
5438 When the local host is found other than at the start of a list of hosts,
5439 the local host, those with the same MX, and any that follow, are discarded.
5440 When the list in question was part of a longer list of hosts, the following
5441 hosts (not currently being processed) were also being discarded. This no
5442 longer happens. I'm not sure if this situation could ever has previously
5445 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5448 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5449 header. This is something that is recommended in a new Internet Draft, and
5450 is something that is documented as being done by Sendmail. There are two
5451 possible values. For messages generated by the autoreply transport, Exim
5454 Auto-Submitted: auto-replied
5456 whereas for all other generated messages (e.g. bounces) it adds
5458 Auto-Submitted: auto-generated
5460 63. The "personal" condition in filters now includes a test for the
5461 Auto-Submitted: header. If it contains the string "auto-" the message it
5462 not considered personal.
5464 64. Added rcpt_include_affixes as a generic transport option.
5466 65. Added queue_only_override (default true).
5468 66. Added the syslog_duplication option.
5470 67. If what should have been the first header line of a message consisted of
5471 a space followed by a colon, Exim was mis-interpreting it as a header line.
5472 It isn't of course - it is syntactically invalid and should therefore be
5473 treated as the start of the message body. The misbehaviour could have
5474 caused a number of strange effects, including loss of data in subsequent
5475 header lines, and spool format errors.
5477 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5478 client host had authenticated. This control can now be exercised by an ACL
5479 for more flexibility.
5481 69. By default, callouts do not happen when testing with -bh. There is now a
5482 variant, -bhc, which does actually run the callout code, including
5483 consulting and updating the callout cache.
5485 70. Added support for saslauthd authentication, courtesy of Alexander
5488 71. If statvfs() failed on the spool or log directories while checking their
5489 size for availability, Exim confusingly gave the error "space shortage".
5490 Furthermore, in debugging mode it crashed with a floating point exception.
5491 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5492 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5493 serious problem, Exim now writes to the main and panic logs when this
5494 happens, with details of the failure. It then refuses to accept the
5495 incoming message, giving the message "spool directory problem" or "log
5496 directory problem" with a 421 code for SMTP messages.
5498 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5499 0, 1, and 2 exist, because some OS complain for execs without them (see
5500 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5501 descriptors. However, the code omitted to check that the open succeeded,
5502 causing mysterious errors if for some reason the permissions on /dev/null
5503 got screwed. Now Exim writes a message to the main and panic logs, and
5504 bombs out if it can't open /dev/null.
5506 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5507 interact so that it is all more flexible. It is supposed to remain
5508 backwards compatible. Also added extra_local_interfaces.
5510 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5511 to bomb out with an assertion failure - to the client this appears as a
5512 connection drop. This problem occurs in the part of the code that was taken
5513 from the Samba project. Fortunately, the assertion is in a very simple
5514 function, so I have fixed this by reproducing the function inline in the
5515 one place where it is called, and arranging for authentication to fail
5516 instead of killing the process with assert().
5518 75. The SPA client code was not working when the server requested OEM rather
5519 than Unicode encoding.
5521 76. Added code to make require_files with a specific uid setting more usable in
5522 the case where statting the file as root fails - usually a non-root-mounted
5523 NFS file system. When this happens and the failure is EACCES, Exim now
5524 forks a subprocess and does the per-uid checking as the relevant uid.
5526 77. Added process_log_path.
5528 78. If log_file_path was not explicitly set, a setting of check_log_space or
5529 check_log_inodes was ignored.
5531 79. If a space check for the spool or log partitions fails, the incident is now
5532 logged. Of course, in the latter case the data may get lost...
5534 80. Added the %p formatting code to string_format() so that it can be used to
5535 print addresses in debug_print(). Adjusted all the address printing in the
5536 debugging in store.c to use %p rather than %d.
5538 81. There was a concern that a line of code in smtp_in.c could overflow a
5539 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5540 initially expressed, the concern was not well-founded, because trailing
5541 spaces are removed early. However, if the trailing spaces were followed by
5542 a NULL, they did not get removed, so the overflow was possible. Two fixes
5545 (a) I re-wrote the offending code in a cleaner fashion.
5546 (b) If an incoming SMTP command contains a NULL character, it is rejected
5549 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5550 runs initgroups(), so that if the Exim user is in any additional groups,
5551 they will be used during message reception.
5557 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob