1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
83 PP/01 Solaris build fix for Oracle's LDAP libraries.
84 Bugzilla 1109, patch from Stephen Usher.
86 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
88 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
91 TF/02 Fix a couple more cases where we did not log the error message
92 when unlink() failed. See also change 4.74-TF/03.
94 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
95 lock up or crash if it happened to be inside a call to libc when it
96 got a SIGUSR1 from exiwhat.
98 The SIGUSR1 handler appends the current process status to the process
99 log which is later printed by exiwhat. It used to use the general
100 purpose logging code to do this, but several functions it calls are
101 not safe for signals.
103 The new output code in the SIGUSR1 handler is specific to the process
104 log, and simple enough that it's easy to inspect for signal safety.
105 Removing some special cases also simplifies the general logging code.
106 Removing the spurious timestamps from the process log simplifies
109 TF/04 Improved ratelimit ACL condition.
111 The /noupdate option has been deprecated in favour of /readonly which
112 has clearer semantics. The /leaky, /strict, and /readonly update modes
113 are mutually exclusive. The update mode is no longer included in the
114 database key; it just determines when the database is updated. (This
115 means that when you upgrde Exim will forget old rate measurements.)
117 Exim now checks that the per_* options are used with an update mode that
118 makes sense for the current ACL. For example, when Exim is processing a
119 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
120 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
121 must specify per_mail/readonly. If you omit the update mode it defaults to
122 /leaky where that makes sense (as before) or /readonly where required.
124 The /noupdate option is now undocumented but still supported for
125 backwards compatibility. It is equivalent to /readonly except that in
126 ACLs where /readonly is required you may specify /leaky/noupdate or
127 /strict/noupdate which are treated the same as /readonly.
129 A useful new feature is the /count= option. This is a generalization
130 of the per_byte option, so that you can measure the throughput of other
131 aggregate values. For example, the per_byte option is now equivalent
132 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
134 The per_rcpt option has been generalized using the /count= mechanism
135 (though it's more complicated than the per_byte equivalence). When it is
136 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
137 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
138 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
139 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
140 Note that using per_rcpt with a non-readonly update mode in more than
141 one ACL will cause the recipients to be double-counted. (The per_mail
142 and per_byte options don't have this problem.)
144 The handling of very low rates has changed slightly. If the computed rate
145 is less than the event's count (usually one) then this event is the first
146 after a long gap. In this case the rate is set to the same as this event's
147 count, so that the first message of a spam run is counted properly.
149 The major new feature is a mechanism for counting the rate of unique
150 events. The new per_addr option counts the number of different
151 recipients that someone has sent messages to in the last time period. It
152 behaves like per_rcpt if all the recipient addresses are different, but
153 duplicate recipient addresses do not increase the measured rate. Like
154 the /count= option this is a general mechanism, so the per_addr option
155 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
156 example, measure the rate that a client uses different sender addresses
157 with the options per_mail/unique=$sender_address. There are further
158 details in the main documentation.
160 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
162 TF/06 Removed a few PCRE remnants.
164 TF/07 Automatically extract Exim's version number from tags in the git
165 repository when doing development or release builds.
167 PP/02 Raise smtp_cmd_buffer_size to 16kB.
168 Bugzilla 879. Patch from Paul Fisher.
170 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
171 Heavily based on revision 40f9a89a from Simon Arlott's tree.
174 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
176 PP/05 Variable $av_failed, true if the AV scanner deferred.
177 Bugzilla 1078. Patch from John Horne.
179 PP/06 Stop make process more reliably on build failure.
180 Bugzilla 1087. Patch from Heiko Schlittermann.
182 PP/07 Make maildir_use_size_file an _expandable_ boolean.
183 Bugzilla 1089. Patch from Heiko Schlittermann.
185 PP/08 Handle ${run} returning more data than OS pipe buffer size.
186 Bugzilla 1131. Patch from Holger Weiß.
188 PP/09 Handle IPv6 addresses with SPF.
189 Bugzilla 860. Patch from Wolfgang Breyha.
191 PP/10 GnuTLS: support TLS 1.2 & 1.1.
193 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
196 PP/11 match_* no longer expand right-hand-side by default.
197 New compile-time build option, EXPAND_LISTMATCH_RHS.
198 New expansion conditions, "inlist", "inlisti".
200 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
202 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
204 PP/14 fix log_write() format string regression from TF/03.
205 Bugzilla 1152. Patch from Dmitry Isaikin.
211 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
213 PP/02 Harmonised TLS library version reporting; only show if debugging.
214 Layout now matches that introduced for other libraries in 4.74 PP/03.
216 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
218 PP/04 New "dns_use_edns0" global option.
220 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
223 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
224 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
226 TK/01 Updated PolarSSL code to 0.14.2.
227 Bugzilla 1097. Patch from Andreas Metzler.
229 PP/07 Catch divide-by-zero in ${eval:...}.
232 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
235 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
236 format-string attack -- SECURITY: remote arbitrary code execution.
238 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
239 time unintentionally subject to list matching rules, letting the header
240 cause arbitrary Exim lookups (of items which can occur in lists, *not*
241 arbitrary string expansion). This allowed for information disclosure.
243 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
244 INT_MIN/-1 -- value coerced to INT_MAX.
250 NM/01 Workround for PCRE version dependency in version reporting
253 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
254 This fixes portability to compilers other than gcc, notably
255 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
257 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
258 makefiles for portability to HP-UX and POSIX correctness.
260 PP/01 Permit LOOKUP_foo enabling on the make command-line.
261 Also via indented variable definition in the Makefile.
262 (Debugging by Oliver Heesakkers).
264 PP/02 Restore caching of spamd results with expanded spamd_address.
265 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
267 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
268 Improves build reliability. Fix from: Frank Elsner
270 NM/02 Fix wide character breakage in the rfc2047 coding
271 Fixes bug 1064. Patch from Andrey N. Oktyabrski
273 NM/03 Allow underscore in dnslist lookups
274 Fixes bug 1026. Patch from Graeme Fowler
276 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
277 Code patches from Adam Ciarcinski of NetBSD.
279 NM/04 Fixed exiqgrep to cope with mailq missing size issue
282 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
283 is logged, to avoid truncation. Patch from John Horne.
285 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
286 Patch from Jakob Hirsch.
288 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
289 SQL string expansion failure details.
290 Patch from Andrey Oktyabrski.
292 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
293 Patch from Simon Arlott.
295 PP/09 New lookups functionality failed to compile on old gcc which rejects
296 extern declarations in function scope.
297 Patch from Oliver Fleischmann
299 PP/10 Use sig_atomic_t for flags set from signal handlers.
300 Check getgroups() return and improve debugging.
301 Fixed developed for diagnosis in bug 927 (which turned out to be
304 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
305 Patch from Mark Zealey.
307 PP/12 Bugzilla 1056: Improved spamd server selection.
308 Patch from Mark Zealey.
310 PP/13 Bugzilla 1086: Deal with maildir quota file races.
311 Based on patch from Heiko Schlittermann.
313 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
314 Patch from Uwe Doering, sign-off by Michael Haardt.
316 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
317 variable declaration deep within a block. Bug and patch from
320 PP/15 lookups-Makefile IRIX compatibilty coercion.
322 PP/16 Make DISABLE_DKIM build knob functional.
324 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
325 Patch by Simon Arlott
327 TF/03 Fix valgrind.h portability to C89 compilers that do not support
328 variable argument macros. Our copy now differs from upstream.
334 TF/01 Failure to get a lock on a hints database can have serious
335 consequences so log it to the panic log.
337 TF/02 Log LMTP confirmation messages in the same way as SMTP,
338 controlled using the smtp_confirmation log selector.
340 TF/03 Include the error message when we fail to unlink a spool file.
342 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
343 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
344 for maintaining out-of-tree patches for some time.
346 PP/01 Bugzilla 139: Documentation and portability issues.
347 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
348 Handle per-OS dynamic-module compilation flags.
350 PP/02 Let /dev/null have normal permissions.
351 The 4.73 fixes were a little too stringent and complained about the
352 permissions on /dev/null. Exempt it from some checks.
353 Reported by Andreas M. Kirchwitz.
355 PP/03 Report version information for many libraries, including
356 Exim version information for dynamically loaded libraries. Created
357 version.h, now support a version extension string for distributors
358 who patch heavily. Dynamic module ABI change.
360 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
361 privilege escalation vulnerability whereby the Exim run-time user
362 can cause root to append content of the attacker's choosing to
365 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
368 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
369 If dropping privileges for untrusted macros, we disabled normal logging
370 on the basis that it would fail; for the Exim run-time user, this is not
371 the case, and it resulted in successful deliveries going unlogged.
372 Fixed. Reported by Andreas Metzler.
378 PP/01 Date: & Message-Id: revert to normally being appended to a message,
379 only prepend for the Resent-* case. Fixes regression introduced in
380 Exim 4.70 by NM/22 for Bugzilla 607.
382 PP/02 Include check_rfc2047_length in configure.default because we're seeing
383 increasing numbers of administrators be bitten by this.
385 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
387 PP/03 Bugzilla 994: added openssl_options main configuration option.
389 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
391 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
393 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
395 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
396 without a peer certificate, leading to a segfault because of an
397 assumption that peers always have certificates. Be a little more
398 paranoid. Problem reported by Martin Tscholak.
400 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
401 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
402 NB: ClamAV planning to remove STREAM in "middle of 2010".
403 CL also introduces -bmalware, various -d+acl logging additions and
404 more caution in buffer sizes.
406 PP/09 Implemented reverse_ip expansion operator.
408 PP/10 Bugzilla 937: provide a "debug" ACL control.
410 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
412 PP/12 Bugzilla 973: Implement --version.
414 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
416 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
418 PP/15 Bugzilla 816: support multiple condition rules on Routers.
420 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
421 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
422 ignore trailing whitespace.
424 JJ/02 prevent non-panic DKIM error from being sent to paniclog
426 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
429 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
430 Notification from Dr Andrew Aitchison.
432 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
433 ExtendedDetectionInfo response format.
434 Notification from John Horne.
436 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
439 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
440 XSL and documented dependency on system catalogs, with examples of how
443 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
446 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
447 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
448 configuration file which is writeable by the Exim user or group.
450 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
451 of configuration files to cover files specified with the -C option if
452 they are going to be used with root privileges, not just the default
455 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
456 option (effectively making it always true).
458 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
459 files to be used while preserving root privileges.
461 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
462 that rogue child processes cannot use them.
464 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
465 run-time user, instead of root.
467 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
468 Exim run-time user without dropping privileges.
470 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
471 result string, instead of calling string_vformat() twice with the same
474 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
475 for other users. Others should always drop root privileges if they use
476 -C on the command line, even for a whitelisted configure file.
478 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
480 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
486 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
487 $data_path, and $header_path variables; fixed documentation bugs and
490 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
491 exipick to access non-standard spools, including the "frozen" queue
494 NM/01 Bugzilla 965: Support mysql stored procedures.
495 Patch from Alain Williams
497 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
499 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
500 Patch from Andreas Metzler
502 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
503 Patch from Kirill Miazine
505 NM/05 Bugzilla 671: Added umask to procmail example.
507 JJ/03 installed exipick 20100323.0, fixing doc bug
509 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
510 directory. Notification and patch from Dan Rosenberg.
512 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
514 TK/02 Improve log output when DKIM signing operation fails.
516 MH/01 Treat the transport option dkim_domain as a colon separated
517 list, not as a single string, and sign the message with each element,
518 omitting multiple occurences of the same signer.
520 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
521 Bugzilla 985, 986. Patch by Simon Arlott
523 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
524 Patch by Simon Arlott
526 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
527 MBX locking. Notification from Dan Rosenberg.
533 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
535 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
537 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
539 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
541 NM/04 Bugzilla 915: Flag broken perl installation during build.
547 TK/01 Added patch by Johannes Berg that expands the main option
548 "spamd_address" if it starts with a dollar sign.
550 TK/02 Write list of recipients to X-Envelope-Sender header when building
551 the mbox-format spool file for content scanning (suggested by Jakob
554 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
555 (http://www.dcc-servers.net/) support via dccifd. Activated by
556 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
558 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
559 by Mark Daniel Reidel <mr@df.eu>.
561 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
562 When building exim an external PCRE library is now needed -
563 PCRE is a system library on the majority of modern systems.
564 See entry on PCRE_LIBS in EDITME file.
566 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
567 conversation. Added nologin parameter to request.
568 Patch contributed by Kirill Miazine.
570 TF/01 Do not log submission mode rewrites if they do not change the address.
572 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
574 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
575 log files in place. Contributed by Roberto Lima.
577 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
579 TF/03 Bugzilla 615: When checking the local_parts router precondition
580 after a local_part_suffix or local_part_prefix option, Exim now
581 does not use the address's named list lookup cache, since this
582 contains cached lookups for the whole local part.
584 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
585 Robert Millan. Documentation is in experimental-spec.txt.
587 TF/04 Bugzilla 668: Fix parallel build (make -j).
589 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
591 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
592 Patch provided by Jan Srzednicki.
594 TF/05 Leading white space used to be stripped from $spam_report which
595 wrecked the formatting. Now it is preserved.
597 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
598 that they are available at delivery time.
600 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
602 TF/08 TLS error reporting now respects the incoming_interface and
603 incoming_port log selectors.
605 TF/09 Produce a more useful error message if an SMTP transport's hosts
606 setting expands to an empty string.
608 NM/06 Bugzilla 744: EXPN did not work under TLS.
609 Patch provided by Phil Pennock.
611 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
612 Patch provided by Richard Godbee.
614 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
615 acl_smtp_notquit, added index entry.
617 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
618 Patch provided by Eugene Bujak.
620 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
621 accept(). Patch provided by Maxim Dounin.
623 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
624 Patch provided by Phil Pennock.
626 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
628 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
629 Patch provided by Brad "anomie" Jorsch.
631 NM/14 Bugzilla 622: Added timeout setting to transport filter.
632 Patch provided by Dean Brooks.
634 TK/05 Add native DKIM support (does not depend on external libraries).
636 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
637 Patch provided by Graeme Fowler.
639 NM/16 Bugzilla 851: Documentation example syntax fix.
641 NM/17 Changed NOTICE file to remove references to embedded PCRE.
643 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
646 NM/19 Bugzilla 745: TLS version reporting.
647 Patch provided by Phil Pennock.
649 NM/20 Bugzilla 167: bool: condition support.
650 Patch provided by Phil Pennock.
652 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
653 clients. Patch provided by Phil Pennock.
655 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
656 Patch provided by Brad "anomie" Jorsch.
658 NM/23 Bugzilla 687: Fix misparses in eximstats.
659 Patch provided by Heiko Schlittermann.
661 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
662 Patch provided by Heiko Schlittermann.
664 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
665 plus update to original patch.
667 NM/26 Bugzilla 799: Documentation correction for ratelimit.
669 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
670 Patch provided by David Brownlee.
672 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
674 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
676 NM/30 Bugzilla 888: TLS documentation bugfixes.
678 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
680 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
681 Unlike the original bugzilla I have changed all shell scripts in src tree.
683 NM/33 Bugzilla 898: Transport filter timeout fix.
684 Patch by Todd Rinaldo.
686 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
687 Patch by Serge Demonchaux.
689 NM/35 Bugzilla 39: Base64 decode bug fixes.
690 Patch by Jakob Hirsch.
692 NM/36 Bugzilla 909: Correct connect() call in dcc code.
694 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
696 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
698 NM/39 Bugzilla 911: Fixed MakeLinks build script.
704 TK/01 Add preliminary DKIM support. Currently requires a forked version of
705 ALT-N's libdkim that I have put here:
706 http://duncanthrax.net/exim-experimental/
708 Note to Michael Haardt: I had to rename some vars in sieve.c. They
709 were called 'true' and it seems that C99 defines that as a reserved
710 keyword to be used with 'bool' variable types. That means you could
711 not include C99-style headers which use bools without triggering
712 build errors in sieve.c.
714 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
715 as mailq or other aliases. Changed the --help handling significantly
716 to do whats expected. exim_usage() emits usage/help information.
718 SC/01 Added the -bylocaldomain option to eximstats.
720 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
722 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
724 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
730 PH/01 Another patch from the Sieve maintainer.
732 PH/02 When an IPv6 address is converted to a string for single-key lookup
733 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
734 dots are used instead of colons so that keys in lsearch files need not
735 contain colons. This was done some time before quoting was made available
736 in lsearch files. However, iplsearch files do require colons in IPv6 keys
737 (notated using the quote facility) so as to distinguish them from IPv4
738 keys. This meant that lookups for IP addresses in host lists did not work
739 for iplsearch lookups.
741 This has been fixed by arranging for IPv6 addresses to be expressed with
742 colons if the lookup type is iplsearch. This is not incompatible, because
743 previously such lookups could never work.
745 The situation is now rather anomolous, since one *can* have colons in
746 ordinary lsearch keys. However, making the change in all cases is
747 incompatible and would probably break a number of configurations.
749 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
752 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
753 conversion specification without a maximum field width, thereby enabling
754 a rogue spamd server to cause a buffer overflow. While nobody in their
755 right mind would setup Exim to query an untrusted spamd server, an
756 attacker that gains access to a server running spamd could potentially
757 exploit this vulnerability to run arbitrary code as the Exim user.
759 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
760 $primary_hostname instead of what libspf2 thinks the hosts name is.
762 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
763 a directory entry by the name of the lookup key. Previously, if a
764 symlink pointed to a non-existing file or a file in a directory that
765 Exim lacked permissions to read, a lookup for a key matching that
766 symlink would fail. Now it is enough that a matching directory entry
767 exists, symlink or not. (Bugzilla 503.)
769 PH/03 The body_linecount and body_zerocount variables are now exported in the
772 PH/04 Added the $dnslist_matched variable.
774 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
775 This means they are set thereafter only if the connection becomes
778 PH/06 Added the client_condition to authenticators so that some can be skipped
779 by clients under certain conditions.
781 PH/07 The error message for a badly-placed control=no_multiline_responses left
782 "_responses" off the end of the name.
784 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
786 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
787 (without spaces) instead of just copying the configuration text.
789 PH/10 Added the /noupdate option to the ratelimit ACL condition.
791 PH/11 Added $max_received_linelength.
793 PH/12 Added +ignore_defer and +include_defer to host lists.
795 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
796 way in which PCRE > 7.0 is built.
798 PH/14 Implemented queue_only_load_latch.
800 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
801 MAIL command. The effect was to mangle the value on 64-bit systems.
803 PH/16 Another patch from the Sieve maintainer.
805 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
807 PH/18 If a system quota error occurred while trying to create the file for
808 a maildir delivery, the message "Mailbox is full" was not appended to the
809 bounce if the delivery eventually timed out. Change 4.67/27 below applied
810 only to a quota excession during the actual writing of the file.
812 PH/19 It seems that peer DN values may contain newlines (and other non-printing
813 characters?) which causes problems in log lines. The DN values are now
814 passed through string_printing() before being added to log lines.
816 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
817 and InterBase are left for another time.)
819 PH/21 Added message_body_newlines option.
821 PH/22 Guard against possible overflow in moan_check_errorcopy().
823 PH/23 POSIX allows open() to be a macro; guard against that.
825 PH/24 If the recipient of an error message contained an @ in the local part
826 (suitably quoted, of course), incorrect values were put in $domain and
827 $local_part during the evaluation of errors_copy.
833 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
834 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
837 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
838 issue a MAIL command.
840 PH/02 In an ACL statement such as
842 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
844 if a client was not listed at all, or was listed with a value other than
845 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
846 the condition was not true (as it should be), so access was not denied.
847 The bug was that the ! inversion was incorrectly passed on to the second
848 item. This has been fixed.
850 PH/03 Added additional dnslists conditions == and =& which are different from
851 = and & when the dns lookup returns more than one IP address.
853 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
854 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
856 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
857 FSYNC, which compiles an option called disable_fsync that allows for
858 bypassing fsync(). The documentation is heavily laced with warnings.
860 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
862 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
863 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
864 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
865 including adding "make clean"; (3) Added -fPIC when compiling the test
866 dynamically loaded module, to get rid of a warning.
868 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
869 message fails, move_frozen_messages = true and ignore_bounce_errors_after
870 = 0s. The bug is otherwise harmless.
872 PH/07 There was a bug in the dovecot authenticator such that the value of
873 $auth1 could be overwritten, and so not correctly preserved, after a
874 successful authentication. This usually meant that the value preserved by
875 the server_setid option was incorrect.
877 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
879 PH/09 Installed PCRE release 7.0.
881 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
882 run for batched SMTP input. It is now run at the start of every message
883 in the batch. While fixing this I discovered that the process information
884 (output by running exiwhat) was not always getting set for -bs and -bS
885 input. This is fixed, and it now also says "batched" for BSMTP.
887 PH/11 Added control=no_pipelining.
889 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
890 patch, slightly modified), and move the expansion of helo_data till after
891 the connection is made in the smtp transport (so it can use these
894 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
896 PH/14 Added log_selector = +pid.
898 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
900 PH/16 Add ${if forany and ${if forall.
902 PH/17 Added dsn_from option to vary the From: line in DSNs.
904 PH/18 Flush SMTP output before performing a callout, unless control =
905 no_callout_flush is set.
907 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
908 was true (the default) a successful delivery failed to delete the retry
909 item, thus causing premature timeout of the address. The bug is now
912 PH/20 Added hosts_avoid_pipelining to the smtp transport.
914 PH/21 Long custom messages for fakedefer and fakereject are now split up
915 into multiline reponses in the same way that messages for "deny" and
916 other ACL rejections are.
918 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
919 with slight modification.
921 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
922 draft, changing the syntax and factoring some duplicate code".
924 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
925 for deliveries of the second and subsequent messages over the same SMTP
928 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
929 ${reduce, with only minor "tidies".
931 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
933 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
934 expansion side effects.
936 PH/27 When a message times out after an over-quota error from an Exim-imposed
937 quota, the bounce message says "mailbox is full". This message was not
938 being given when it was a system quota that was exceeded. It now should
941 MH/03 Made $recipients available in local_scan(). local_scan() already has
942 better access to the recipient list through recipients_list[], but
943 $recipients can be useful in postmaster-provided expansion strings.
945 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
946 in the case of a MAIL command with additional options following the
947 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
948 were accidentally chopped off.
950 PH/29 SMTP synchronization checks are implemented when a command is read -
951 there is a check that no more input is waiting when there shouldn't be
952 any. However, for some commands, a delay in an ACL can mean that it is
953 some time before the response is written. In this time, more input might
954 arrive, invalidly. So now there are extra checks after an ACL has run for
955 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
956 pipelining has not been advertised.
958 PH/30 MH's patch to allow iscntrl() characters to be list separators.
960 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
961 returned in the SMTP response when smtp_return_error_details was false.
964 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
965 instead of the C I/O that was originally supplied, because problems were
968 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
969 Exim which did not show up earlier: it was assuming that a call to
970 SSL_CTX_set_info_callback() might give an error value. In fact, there is
971 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
972 was a macro that became an assignment, so it seemed to work. This has
973 changed to a proper function call with a void return, hence the compile
974 error. Exim's code has been fixed.
976 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
979 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
981 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
983 PH/37 If a message is not accepted after it has had an id assigned (e.g.
984 because it turns out to be too big or there is a timeout) there is no
985 "Completed" line in the log. When some messages of this type were
986 selected by exigrep, they were listed as "not completed". Others were
987 picked up by some special patterns. I have improved the selection
988 criteria to be more general.
990 PH/38 The host_find_failed option in the manualroute router can now be set
991 to "ignore", to completely ignore a host whose IP address cannot be
992 found. If all hosts are ignored, the behaviour is controlled by the new
993 host_all_ignored option.
995 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
996 homing or because of multiple MX records with /mx) generated more than
997 one IP address, and the following item turned out to be the local host,
998 all the secondary addresses of the first item were incorrectly removed
999 from the list, along with the local host and any following hosts (which
1000 is what is supposed to happen).
1002 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1003 whoever called Exim into the -H file. In the case of the daemon it was
1004 behaving confusingly. When first started, it used values for whoever
1005 started the daemon, but after a SIGHUP it used the Exim user (because it
1006 calls itself on a restart). I have changed the code so that it now always
1009 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1010 message are rejected with the same error (e.g. no authentication or bad
1011 sender address), and a DATA command is nevertheless sent (as can happen
1012 with PIPELINING or a stupid MUA), the error message that was given to the
1013 RCPT commands is included in the rejection of the DATA command. This is
1014 intended to be helpful for MUAs that show only the final error to their
1017 PH/42 Another patch from the Sieve maintainer.
1019 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1020 Eximstats - Fixed some broken HTML links and added missing column headers
1022 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1023 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1025 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1027 PH/43 Yet another patch from the Sieve maintainer.
1029 PH/44 I found a way to check for a TCP/IP connection going away before sending
1030 the response to the final '.' that terminates a message, but only in the
1031 case where the client has not sent further data following the '.'
1032 (unfortunately, this is allowed). However, in many cases there won't be
1033 any further data because there won't be any more messages to send. A call
1034 to select() can be used: if it shows that the input is "ready", there is
1035 either input waiting, or the socket has been closed. An attempt to read
1036 the next input character can distinguish the two cases. Previously, Exim
1037 would have sent an OK response which the client would never have see.
1038 This could lead to message repetition. This fix should cure that, at
1039 least in a lot of common cases.
1041 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1042 advertised in response to EHLO.
1048 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1049 fixed by 4.65/MH/01 (is this a record?) are fixed:
1051 (i) An empty string was always treated as zero by the numeric comparison
1052 operators. This behaviour has been restored.
1054 (ii) It is documented that the numeric comparison operators always treat
1055 their arguments as decimal numbers. This was broken in that numbers
1056 starting with 0 were being interpreted as octal.
1058 While fixing these problems I realized that there was another issue that
1059 hadn't been noticed. Values of message_size_limit (both the global option
1060 and the transport option) were treated as octal if they started with 0.
1061 The documentation was vague. These values are now always treated as
1062 decimal, and I will make that clear in the documentation.
1068 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1069 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1072 MH/01 Don't check that the operands of numeric comparison operators are
1073 integers when their expansion is in "skipping" mode (fixes bug
1074 introduced by 4.64-PH/07).
1076 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1077 child addresses, Exim now panics and dies. Previously, because the count
1078 is held in a short int, deliveries were likely to be lost. As such a
1079 large number of recipients for a single message is ridiculous
1080 (performance will be very, very poor), I have chosen to impose a limit
1081 rather than extend the field.
1087 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1088 leftover -K file (the existence of which was triggered by #402).
1089 While we were at it, introduced process PID as part of the -K
1090 filename. This should rule out race conditions when creating
1093 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1094 processing considerably. Previous code took too long for large mails,
1095 triggering a timeout which in turn triggers #401.
1097 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1098 in the DK code in transports.c. sendfile() is not really portable,
1099 hence the _LINUX specificness.
1101 TF/01 In the add_headers option to the mail command in an Exim filter,
1102 there was a bug that Exim would claim a syntax error in any
1103 header after the first one which had an odd number of characters
1106 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1107 callout verification, Exim cached a "reject" for the entire domain. This
1108 is correct for most verifications, but it is not correct for a recipient
1109 verification with use_sender or use_postmaster set, because in that case
1110 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1111 case of MAIL FROM:<> rejection from other early rejections (e.g.
1112 rejection of HELO). When verifying a recipient using a non-null MAIL
1113 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1114 Whatever the result of the callout, the value of the domain cache is
1115 left unchanged (for any other kind of callout, getting as far as trying
1116 RCPT means that the domain itself is ok).
1118 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1121 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1122 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1123 session, the daemon ignores ECONNECT errors and logs others; it now
1124 ignores EPIPE as well.
1126 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1127 (quoted-printable decoding).
1129 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1130 later the small subsequent patch to fix an introduced bug.
1132 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1134 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1136 PH/08 An error is now given if message_size_limit is specified negative.
1138 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1139 to be given (somewhat) arbitrary names.
1141 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1144 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1145 miscellaneous code fixes
1147 PH/10 Added the log_reject_target ACL modifier to specify where to log
1150 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1151 hostname. This is wrong, because it relates to the incoming message (and
1152 probably the interface on which it is arriving) and not to the outgoing
1153 callout (which could be using a different interface). This has been
1154 changed to use the value of the helo_data option from the smtp transport
1155 instead - this is what is used when a message is actually being sent. If
1156 there is no remote transport (possible with a router that sets up host
1157 addresses), $smtp_active_hostname is used.
1159 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1160 tweaks were necessary in order to get it to work (see also 21 below):
1161 (a) The code assumed that strncpy() returns a negative number on buffer
1162 overflow, which isn't the case. Replaced with Exim's string_format()
1164 (b) There were several signed/unsigned issues. I just did the minimum
1165 hacking in of casts. There is scope for a larger refactoring.
1166 (c) The code used strcasecmp() which is not a standard C function.
1167 Replaced with Exim's strcmpic() function.
1168 (d) The code set only $1; it now sets $auth1 as well.
1169 (e) A simple test gave the error "authentication client didn't specify
1170 service in request". It would seem that Dovecot has changed its
1171 interface. Fortunately there's a specification; I followed it and
1172 changed what the client sends and it appears to be working now.
1174 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1177 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1178 address A is aliased to B and C, where B exists and C does not. Without
1179 -v the output is "A verified" because verification stops after a
1180 successful redirection if more than one address is generated. However,
1181 with -v the child addresses are also verified. Exim was outputting "A
1182 failed to verify" and then showing the successful verification for C,
1183 with its parentage. It now outputs "B failed to verify", showing B's
1184 parentage before showing the successful verification of C.
1186 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1187 look up a TXT record in a specific list after matching in a combined
1190 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1191 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1192 they consult the DNS. I had assumed they would set it the way they
1193 wanted; and indeed my experiments on Linux seem to show that in some
1194 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1195 To be on the safe side, however, I have now made the interface to
1196 host_find_byname() similar to host_find_bydns(), with an argument
1197 containing the DNS resolver options. The host_find_byname() function now
1198 sets these options at its start, just as host_find_bydns() does. The smtp
1199 transport options dns_qualify_single and dns_search_parents are passed to
1200 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1201 of host_find_byname() use the default settings of RES_DEFNAMES
1202 (qualify_single) but not RES_DNSRCH (search_parents).
1204 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1205 spool_read_header() do less string testing, by means of a preliminary
1206 switch on the second character of optional "-foo" lines. (This is
1207 overdue, caused by the large number of possibilities that now exist.
1208 Originally there were few.) While I was there, I also converted the
1209 str(n)cmp tests so they don't re-test the leading "-" and the first
1210 character, in the hope this might squeeze out yet more improvement.
1212 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1213 flag allowing group syntax was set by the header_syntax check but not
1214 turned off, possible causing trouble later; (2) The flag was not being
1215 set at all for the header_verify test, causing "group"-style headers to
1216 be rejected. I have now set it in this case, and also caused header_
1217 verify to ignore an empty address taken from a group. While doing this, I
1218 came across some other cases where the code for allowing group syntax
1219 while scanning a header line wasn't quite right (mostly, not resetting
1220 the flag correctly in the right place). These bugs could have caused
1221 trouble for malformed header lines. I hope it is now all correct.
1223 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1224 with the "reply" argument non-NULL. The code, however (which originally
1225 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1226 but it didn't always do it. This confused somebody who was copying the
1227 code for some other use. I have removed all the tests.
1229 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1230 feature that was used to support insecure browsers during the U.S. crypto
1231 embargo. It requires special client support, and Exim is probably the
1232 only MTA that supported it -- and would never use it because real RSA is
1233 always available. This code has been removed, because it had the bad
1234 effect of slowing Exim down by computing (never used) parameters for the
1235 RSA_EXPORT functionality.
1237 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1238 authenticator to fail if there's a tab character in the incoming data
1239 (there should never be unless someone is messing about, as it's supposed
1240 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1241 if the connection is using TLS or if the remote IP is the same as the
1242 local IP, and the "valid-client-cert option" if a client certificate has
1245 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1246 authenticators. This can be used for authorization after authentication
1247 succeeds. (In the case of plaintext, it servers for both authentication
1250 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1251 if any retry times were supplied.
1253 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1254 connection, where there is no client IP address to check. In this
1255 situation, the verify now always succeeds.
1257 PH/25 Applied John Jetmore's -Mset patch.
1259 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1261 PH/27 In a string expansion for a processed (not raw) header when multiple
1262 headers of the same name were present, leading whitespace was being
1263 removed from all of them, but trailing whitespace was being removed only
1264 from the last one. Now trailing whitespace is removed from each header
1265 before concatenation. Completely empty headers in a concatenation (as
1266 before) are ignored.
1268 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1269 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1271 PH/29 [Removed. This was a change that I later backed out, and forgot to
1272 correct the ChangeLog entry (that I had efficiently created) before
1273 committing the later change.]
1275 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1276 address errors (4xx response to RCPT) over the same connection as other
1277 messages routed to the same hosts. Such deliveries are always "forced",
1278 so retry times are not inspected. This resulted in far too many retries
1279 for the affected addresses. The effect occurred only when there were more
1280 hosts than the hosts_max_try setting in the smtp transport when it had
1281 the 4xx errors. Those hosts that it had tried were not added to the list
1282 of hosts for which the message was waiting, so if all were tried, there
1283 was no problem. Two fixes have been applied:
1285 (i) If there are any address or message errors in an SMTP delivery, none
1286 of the hosts (tried or untried) are now added to the list of hosts
1287 for which the message is waiting, so the message should not be a
1288 candidate for sending over the same connection that was used for a
1289 successful delivery of some other message. This seems entirely
1290 reasonable: after all the message is NOT "waiting for some host".
1291 This is so "obvious" that I'm not sure why it wasn't done
1292 previously. Hope I haven't missed anything, but it can't do any
1293 harm, as the worst effect is to miss an optimization.
1295 (ii) If, despite (i), such a delivery is accidentally attempted, the
1296 routing retry time is respected, so at least it doesn't keep
1297 hammering the server.
1299 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1300 in ${readsocket because some servers need this prod.
1302 PH/32 Added some extra debug output when updating a wait-xxx database.
1304 PH/33 The hint "could be header name not terminated by colon", which has been
1305 given for certain expansion errors for a long time, was not being given
1306 for the ${if def:h_colon_omitted{... case.
1308 PH/34 The spec says: "With one important exception, whenever a domain list is
1309 being scanned, $domain contains the subject domain." There was at least
1310 one case where this was not true.
1312 PH/35 The error "getsockname() failed: connection reset by peer" was being
1313 written to the panic log as well as the main log, but it isn't really
1314 panic-worthy as it just means the connection died rather early on. I have
1315 removed the panic log writing for the ECONNRESET error when getsockname()
1318 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1319 runs only) independently of the message's sender address. This meant
1320 that, if the 4xx error was in fact related to the sender, a different
1321 message to the same recipient with a different sender could confuse
1322 things. In particualar, this can happen when sending to a greylisting
1323 server, but other circumstances could also provoke similar problems.
1324 I have changed the default so that the retry time for these errors is now
1325 based a combination of the sender and recipient addresses. This change
1326 can be overridden by setting address_retry_include_sender=false in the
1329 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1330 remote server are returned as part of bounce messages. This was not
1331 happening for LMTP over a pipe (the lmtp transport), but now it is the
1332 same for both kinds of LMTP.
1334 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1335 in header lines that were in fact CNAMEs. This is no longer the case.
1337 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1338 and queue runs started by the daemon processed all messages. This has
1339 been fixed so that -R and -S can now usefully be given with -q<time>.
1341 PH/40 Import PCRE release 6.7 (fixes some bugs).
1343 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1345 PH/42 Give an error if -q is specified more than once.
1347 PH/43 Renamed the variables $interface_address and $interface_port as
1348 $received_ip_address and $received_port, to make it clear that these
1349 values apply to message reception, and not to the outgoing interface when
1350 a message is delivered. (The old names remain recognized, of course.)
1352 PH/44 There was no timeout on the connect() call when using a Unix domain
1353 socket in the ${readsocket expansion. There now is.
1355 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1356 be meaningful with "accept".
1358 SC/01 Eximstats V1.43
1359 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1361 SC/02 Eximstats V1.44
1362 Use a glob alias rather than an array ref in the generated
1363 parser. This improves both readability and performance.
1365 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1366 Collect SpamAssassin and rejection statistics.
1367 Don't display local sender or destination tables unless
1368 there is data to show.
1369 Added average volumes into the top table text output.
1371 SC/04 Eximstats V1.46
1372 Collect data on the number of addresses (recipients)
1373 as well as the number of messages.
1375 SC/05 Eximstats V1.47
1376 Added 'Message too big' to the list of mail rejection
1377 reasons (thanks to Marco Gaiarin).
1379 SC/06 Eximstats V1.48
1380 Mainlog lines which have GMT offsets and are too short to
1381 have a flag are now skipped.
1383 SC/07 Eximstats V1.49 (Alain Williams)
1384 Added the -emptyok flag.
1386 SC/08 Eximstats V1.50
1387 Fixes for obtaining the IP address from reject messages.
1389 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1390 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1391 whitesspace changes from 4.64-PH/27
1393 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1396 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1397 are found, allow negative numbers in numeric criteria)
1399 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1401 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1402 to match changes made in 4.64-PH/43
1404 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1406 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1407 "domain" is actually an IP address, and force a failure. This locks out
1408 those revolvers/nameservers that support "A-for-A" lookups, in
1409 contravention of the specifications.
1411 PH/48 When a host name was looked up from an IP address, and the subsequent
1412 forward lookup of the name timed out, the host name was left in
1413 $sender_host_name, contrary to the specification.
1415 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1416 restricted to single-key lookups, Exim was not diagnosing an error if
1417 * or *@ was used with a query-style lookup.
1419 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1421 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1422 long ago, but noone interested enough thought of it. Let's just say that
1423 the "1.1" means that there are some new functions that weren't there at
1424 some point in the past.
1426 PH/51 Error processing for expansion failure of helo_data from an smtp
1427 transport during callout processing was broken.
1429 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1430 tested/used via the -bh/-bhc/-bs options.
1432 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1433 bug, fixed in subsequent PCRE releases).
1435 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1436 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1438 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1444 SC/01 Use a glob alias rather than an array ref in eximstats generated
1445 parser. This improves both readability and performance.
1447 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1448 Don't display local sender or destination tables in eximstats unless
1449 there is data to show.
1450 Added average volumes into the eximstats top table text output.
1452 SC/03 Collect data on the number of addresses (recipients) as well
1453 as the number of messages in eximstats.
1455 TF/01 Correct an error in the documentation for the redirect router. Exim
1456 does (usually) call initgroups() when daemonizing.
1458 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1459 with consistent privilege compared to when running as a daemon.
1461 TF/03 Note in the spec that $authenticated_id is not set for local
1462 submissions from trusted users.
1464 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1465 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1467 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1468 by adding some example configuration directives to the default
1469 configuration file. A little bit of work is required to uncomment the
1470 directives and define how usernames and passwords are checked, but
1471 there is now a framework to start from.
1473 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1474 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1475 without this. I don't know how relevant this is to other LDAP libraries.
1477 PH/02 Add the verb name to the "unknown ACL verb" error.
1479 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1481 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1483 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1484 directory not expanded when it should be if an expanded home directory
1485 was set for the address (which is overridden by the transport).
1487 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1490 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1491 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1492 because it is too late at that time, and has no effect.
1494 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1495 security issue with \' (bugzilla #107). I could not use the
1496 PQescapeStringConn() function, because it needs a PGconn value as one of
1499 PH/08 When testing addresses using -bt, indicate those final addresses that
1500 are duplicates that would not cause an additional delivery. At least one
1501 person was confused, thinking that -bt output corresponded to deliveries.
1502 (Suppressing duplicates isn't a good idea as you lose the information
1503 about possibly different redirections that led to the duplicates.)
1505 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1506 systems where poll() doesn't work, in particular OS X.
1508 PH/10 Added more information to debugging output for retry time not reached.
1510 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1511 operations in malware.c.
1513 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1516 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1517 syslog_duplication set false, log lines that would normally be written
1518 both the the main log and to the reject log were not written to syslog at
1521 PH/14 In the default configuration, change the use of "message" in ACL warn
1522 statements to "add_header".
1524 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1525 not followed by a command (e.g. "seen endif").
1527 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1528 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1531 PH/17 Added extra conditions to the default value of delay_warning_condition
1535 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1536 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1537 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1540 The Auto-Submitted: and various List- headers are standardised, whereas I
1541 don't think Precedence: ever was.
1543 PH/18 Refactored debugging code in route_finduser() to show more information,
1544 in particular, the error code if getpwnam() issues one.
1546 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1547 This is apparently needed in addition to the PH/07 change above to avoid
1548 any possible encoding problems.
1550 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1551 but not after initializing Perl.
1553 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1554 output them only if debugging. By default they are written stderr,
1555 apparently, which is not desirable.
1557 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1560 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1563 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1565 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1566 authenticated or an ident call has been made. Suppress the default
1567 values for $authenticated_id and $authenticated_sender (but permit -oMai
1568 and -oMas) when testing with -bh.
1570 PH/24 Re-jigged the order of the tests in the default configuration so that the
1571 tests for valid domains and recipients precede the DNS black list and CSA
1572 tests, on the grounds that those ones are more expensive.
1574 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1575 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1576 This bug exists in every version of Exim that I still have, right back to
1579 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1580 However, an attempt to turn on case-sensitivity in a regex key by
1581 including (?-i) didn't work because the subject string was already
1582 lowercased, and the effects were non-intuitive. It turns out that a
1583 one-line patch can be used to allow (?-i) to work as expected.
1589 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1590 other effects) broke the use of negated acl sub-conditions.
1592 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1595 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1596 "Deny" causes Exim to reject the incoming connection with a 554 error.
1597 Unfortunately, if there is a major crisis, such as a disk failure,
1598 tcp-wrappers gives "deny", whereas what one would like would be some
1599 kind of temporary error. A kludge has been added to help with this.
1600 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1601 554 error is used if errno is still zero or contains ENOENT (which occurs
1602 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1605 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1607 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1608 errors. Otherwise a message that provokes a temporary error (when other
1609 messages do not) can cause a whole host to time out.
1611 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1612 addresses were routed directly to files or pipes from a redirect router.
1613 File deliveries just didn't batch; pipe deliveries might have suffered
1616 PH/06 A failure to get a lock for a hints database would erroneously always say
1617 "Failed to get write lock", even when it was really a read lock.
1619 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1620 of 0600. This has been changed to use the value of the lockfile_mode
1621 option (which defaults to 0600).
1623 PH/08 Applied small patch from the Sieve maintainer.
1625 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1626 folder from quota calculations, a direct delivery into this folder messed
1627 up the contents of the maildirsize file. This was because the regex was
1628 used only to exclude .Trash (or whatever) when the size of the mailbox
1629 was calculated. There was no check that a delivery was happening into an
1630 excluded directory. This bug has been fixed by ignoring all quota
1631 processing for deliveries into excluded directories.
1633 PH/10 Added the maildirfolder_create_regex option to appendfile.
1639 PH/01 The code for finding all the local interface addresses on a FreeBSD
1640 system running IPv6 was broken. This may well have applied to all BSD
1641 systems, as well as to others that have similar system calls. The broken
1642 code found IPv4 interfaces correctly, but gave incorrect values for the
1643 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1644 that it would not match correctly against @[] and not recognize the IPv6
1647 PH/02 The ipliteral router was not recognizing addresses of the form user@
1648 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1650 PH/03 Added disable_ipv6.
1652 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1653 decoded form, because it is most often used to construct To: headers
1654 lines in autoreplies, and the decoded form may well be syntactically
1655 invalid. However, $reply_address has leading white space removed, and all
1656 newlines turned into spaces so that the autoreply transport does not
1659 PH/05 If group was specified without a user on a router, and no group or user
1660 was specified on a transport, the group from the router was ignored.
1662 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1663 for visible compile-time settings that can be used to change these
1664 numbers, for those that want even more. Backwards compatibility with old
1665 spool files has been maintained. However, going back to a previous Exim
1666 release will lost any variables that are in spool files.
1668 PH/07 Two small changes when running in the test harness: increase delay when
1669 passing a TCP/IP connection to a new process, in case the original
1670 process has to generate a bounce, and remove special handling of
1671 127.0.0.2 (sic), which is no longer necessary.
1673 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1674 be the same on different OS.
1676 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1679 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1680 whether --show-vars was specified or not
1682 JJ/02 exipick: Added support for new ACL variable spool format introduced
1685 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1686 syntactically invalid From: or Reply-to: line, and a filter used this to
1687 generate an autoreply, and therefore failed to obtain an address for the
1688 autoreply, Exim could try to deliver to a non-existent relative file
1689 name, causing unrelated and misleading errors. What now happens is that
1690 it logs this as a hard delivery error, but does not attempt to create a
1693 PH/11 The exinext utility has a -C option for testing purposes, but although
1694 the given file was scanned by exinext itself; it wasn't being passed on
1695 when Exim was called.
1697 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1698 an end-of-file indication when reading a command response.
1700 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1701 compiled. In many other places in Exim, IPv6 addresses are always
1702 recognized, so I have changed this. It also means that IPv4 domain
1703 literals of the form [IPV4:n.n.n.n] are now always recognized.
1705 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1706 used if the router is not running as root, for example, when verifying at
1707 ACL time, or when using -bh. The debugging output from this situation was
1708 non-existent - all you got was a failure to exec. I have made two
1711 (a) Failures to set uid/gid, the current directory, or a process leader
1712 in a subprocess such as that created by queryprogram now generate
1713 suitable debugging ouput when -d is set.
1715 (b) The queryprogram router detects when it is not running as root,
1716 outputs suitable debugging information if -d is set, and then runs
1717 the subprocess without attempting to change uid/gid.
1719 PH/15 Minor change to Makefile for building test_host (undocumented testing
1722 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1723 additional section of a DNS packet that returns MX or SRV records.
1724 Instead, it always explicitly searches for A/AAAA records. This avoids
1725 major problems that occur when a DNS server includes only records of one
1726 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1727 fixed another bug: if SRV records were looked up and the corresponding
1728 address records were *not* found in the additional section, the port
1729 values from the SRV records were lost.
1731 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1732 using the correct key (the original address) when searching the retry
1733 rules in order to find which one to use for generating the retry hint.
1735 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1736 adding the default one. Similarly, if it contains a Reply-To: header, the
1737 errors_reply_to option, if set, is not used.
1739 PH/19 When calculating a retry time, Exim used to measure the "time since
1740 failure" by looking at the "first failed" field in the retry record. Now
1741 it does not use this if it is later than than the arrival time of the
1742 message. Instead it uses the arrival time. This makes for better
1743 behaviour in cases where some deliveries succeed, thus re-setting the
1744 "first failed" field. An example is a quota failure for a huge message
1745 when small messages continue to be delivered. Without this change, the
1746 "time since failure" will always be short, possible causing more frequent
1747 delivery attempts for the huge message than are intended.
1748 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1750 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1751 $1, $2, $3) because the numerical variables can be reset during some
1752 expansion items (e.g. "match"), thereby losing the authentication data.
1754 PH/21 Make -bV show the size of off_t variables so that the test suite can
1755 decide whether to run tests for quotas > 2G.
1757 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1758 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1759 filecount value is greater than 2G or if a quota value is greater than 2G
1760 on a system where the size of off_t is not greater than 4, a panic error
1763 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1764 never match. The debug and -bh output now contains an explicit error
1765 message indicating a malformed IPv4 address or mask.
1767 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1768 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1769 PH/23 above applies.
1771 PH/25 Do not write to syslog when running in the test harness. The only
1772 occasion when this arises is a failure to open the main or panic logs
1773 (for which there is an explicit test).
1775 PH/26 Added the /no_tell option to "control=freeze".
1777 PH/27 If a host name lookup failed very early in a connection, for example, if
1778 the IP address matched host_lookup and the reverse lookup yielded a name
1779 that did not have a forward lookup, an error message of the form "no IP
1780 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1781 could be logged. Now it outputs the IP address instead of "NULL".
1783 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1784 allows the sender and the authenticated sender to be set when
1785 submitting a message from within Exim. Since child_open_exim() is
1786 documented for local_scan(), the new function should be too.
1788 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1789 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1790 results in an empty string is now treated as unset.
1792 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1794 PH/31 Added sender_verify_fail logging option.
1796 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1797 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1798 tidied the source and removed it altogether.
1800 PH/33 When a queue run was abandoned because the load average was too high, a
1801 log line was always written; now it is written only if the queue_run log
1802 selector is set. In addition, the log line for abandonment now contains
1803 information about the queue run such as the pid. This is always present
1804 in "start" and "stop" lines but was omitted from the "abandon" line.
1806 PH/34 Omit spaces between a header name and the colon in the error message that
1807 is given when verify = headers_syntax fails (if there are lots of them,
1808 the message gets confusing).
1810 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1811 names, as there are now some PTR records that contain slashes. This check
1812 is only to protect against broken name servers that fall over on strange
1813 characters, so the fact that it applies to all lookups doesn't matter.
1815 PH/36 Now that the new test suite is complete, we can remove some of the
1816 special code in Exim that was needed for the old test suite. For example,
1817 sorting DNS records because real resolvers return them in an arbitrary
1818 order. The new test suite's fake resolver always returns records in the
1821 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1822 bounces) except when queue_only is set, to avoid logging races between
1823 the different processes.
1825 PH/38 Panic-die if .include specifies a non-absolute path.
1827 PH/39 A tweak to the "H" retry rule from its user.
1829 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1830 a label. They prevented compilation on older perls.
1832 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1833 a warning to be raised on newish perls.
1835 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1836 on queue. Changes to match documented behaviour of showing count of
1837 messages matching specified criteria.
1839 PH/40 Changed the default ident timeout from 30s to 5s.
1841 PH/41 Added support for the use of login_cap features, on those BSD systems
1842 that have them, for controlling the resources used by pipe deliveries.
1844 PH/42 The content-scanning code uses fopen() to create files in which to put
1845 message data. Previously it was not paying any attention to the mode of
1846 the files. Exim runs with umask(0) because the rest of the code creates
1847 files with open(), and sets the required mode explicitly. Thus, these
1848 files were ending up world-writeable. This was not a big issue, because,
1849 being within the spool directory, they were not world-accessible. I have
1850 created a function called modefopen, which takes an additional mode
1851 argument. It sets umask(777), creates the file, chmods it to the required
1852 mode, then resets the umask. All the relevant calls to fopen() in the
1853 content scanning code have been changed to use this function.
1855 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1856 to 24 hours. This avoids potential overflow problems when processing G
1857 and H retry rules. I suspect nobody ever tinkers with this value.
1859 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1861 PH/45 When the plaintext authenticator is running as a client, the server's
1862 challenges are checked to ensure they are valid base64 strings. By
1863 default, the authentication attempt is cancelled if an invalid string is
1864 received. Setting client_ignore_invalid_base64 true ignores these errors.
1865 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1866 they are received. Thus, the responses can be made to depend on the
1867 challenges. If an invalid string is ignored, an empty string is placed in
1870 PH/46 Messages that are created by the autoreply transport now contains a
1871 References: header, in accordance with RFCs 2822 and 3834.
1873 PH/47 Added authenticated_sender_force to the smtp transport.
1875 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1877 PH/49 Installed latest patch from the Sieve maintainer.
1879 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1880 was also set, the appendfile transport was unnecessarily scanning a
1881 directory of message files (e.g. for maildir delivery) to find the count
1882 of files (along with the size), even though it did not need this
1883 information. It now does the scan only if it needs to find either the
1884 size of the count of files.
1886 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1888 PH/52 Two bugs concerned with error handling when the smtp transport is
1891 (i) Exim was not creating retry information for temporary errors given
1892 for individual recipients after the DATA command when the smtp transport
1893 was used in LMTP mode. This meant that they could be retried too
1894 frequently, and not timed out correctly.
1896 (ii) Exim was setting the flag that allows error details to be returned
1897 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1898 recipients that were returned after the DATA command.
1900 PH/53 This is related to PH/52, but is more general: for any failing address,
1901 when detailed error information was permitted to be returned to the
1902 sender, but the error was temporary, then after the final timeout, only
1903 "retry timeout exceeded" was returned. Now it returns the full error as
1904 well as "retry timeout exceeded".
1906 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1907 do this, and (what is worse) MTAs that accept it.
1909 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1910 will now be deprecated.
1912 PH/56 New os.c-cygwin from the Cygwin maintainer.
1914 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1915 formats (previously only available in exim formats via -bpr, -bpru,
1916 and -bpra. Now also available in native and exiqgrep formats)
1918 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1919 with very large, slow to parse queues
1921 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1923 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1925 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1926 responses to authentication challenges, though it was showing the
1927 challenges; (ii) I've removed the CR characters from the debug output for
1930 PH/58 Allow for the insertion of a newline as well as a space when a string
1931 is turned into more than one encoded-word during RFC 2047 encoding. The
1932 Sieve code now uses this.
1934 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1935 data_4xx, lost_connection, tls_required.
1937 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1938 message was being sent as an SMTP response.
1940 PH/61 Add -l and -k options to exicyclog.
1942 PH/62 When verifying, if an address was redirected to one new address, so that
1943 verification continued, and the new address failed or deferred after
1944 having set something in $address_data, the value of $address_data was not
1945 passed back to the ACL. This was different to the case when no
1946 redirection occurred. The value is now passed back in both cases.
1948 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1949 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1950 use login_cap.h, so on its own it isn't the distinguishing feature. The
1951 new name refers directly to the setclassresources() function.
1953 PH/65 Added configuration files for NetBSD3.
1955 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1957 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1958 is preferred over IPv4.
1960 PH/68 The bounce_return_message and bounce_return_body options were not being
1961 honoured for bounces generated during the reception of non-SMTP messages.
1962 In particular, this applied to messages rejected by the ACL. This bug has
1963 been fixed. However, if bounce_return_message is true and bounce_return_
1964 body is false, the headers that are returned for a non-SMTP message
1965 include only those that have been read before the error was detected.
1966 (In the case of an ACL rejection, they have all been read.)
1968 PH/69 The HTML version of the specification is now built in a directory called
1969 spec_html instead of spec.html, because the latter looks like a path with
1970 a MIME-type, and this confuses some software.
1972 PH/70 Catch two compiler warnings in sieve.c.
1974 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1975 function verify_get_ident() calls ip_connect() to connect a socket, but
1976 if the "connect()" function timed out, ip_connect() used to close the
1977 socket. However, verify_get_ident() also closes the socket later, and in
1978 between Exim writes to the log, which may get opened at this point. When
1979 the socket was closed in ip_connect(), the log could get the same file
1980 descriptor number as the socket. This naturally causes chaos. The fix is
1981 not to close the socket in ip_connect(); the socket should be closed by
1982 the function that creates it. There was only one place in the code where
1983 this was missing, in the iplookup router, which I don't think anybody now
1984 uses, but I've fixed it anyway.
1986 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1987 well as to direct DNS lookups. Otherwise the handling of names in host
1988 lists is inconsistent and therefore confusing.
1994 PH/01 Two changes to the default runtime configuration:
1996 (1) Move the checks for relay_from_hosts and authenticated clients from
1997 after to before the (commented out) DNS black list checks.
1999 (2) Add control=submission to the relay_from_hosts and authenticated
2000 clients checks, on the grounds that messages accepted by these
2001 statements are most likely to be submissions.
2003 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2005 (1) Generate an error if the third argument for the ${prvs expansion is
2008 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2011 (3) Reset the variables that are obtained from the first argument of
2012 ${prvscheck and used in the second argument before leaving the code,
2013 because their memory is reclaimed, so using them afterwards may do
2016 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2017 one (it's much easier than Tom thought :-).
2019 (5) Because of (4), we can now allow for the use of $prvscheck_result
2020 inside the third argument.
2022 PH/03 For some reason, the default setting of PATH when running a command from
2023 a pipe transport was just "/usr/bin". I have changed it to
2026 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2027 anything to be listed in the output from -bV.
2029 PH/05 When a filter generated an autoreply, the entire To: header line was
2030 quoted in the delivery log line, like this:
2032 => >A.N.Other <ano@some.domain> <original@ddress> ...
2034 This has been changed so that it extracts the operative address. There
2035 may be more than one such address. If so, they are comma-separated, like
2038 => >ano@some.domain,ona@other.domain <original@ddress> ...
2040 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2041 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2042 not being looked up in the rDNS to get a host name, Exim was showing the
2043 IP address twice in Received: lines, even though the IP addresses were
2044 identical. For example:
2046 Received: from [1.2.3.4] (helo=[1.2.3.4])
2048 However, if the real host name was known, it was omitting the HELO data
2049 if it matched the actual IP address. This has been tidied up so that it
2050 doesn't show the same IP address twice.
2052 PH/07 When both +timestamp and +memory debugging was on, the value given by
2053 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2054 called by the debug printing, thereby overwriting the timestamp buffer.
2055 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2057 PH/08 When the original message was included in an autoreply transport, it
2058 always said "this is a copy of the message, including all the headers",
2059 even if body_only or headers_only was set. It now gives an appropriate
2062 PH/09 Applied a patch from the Sieve maintainer which:
2064 o fixes some comments
2065 o adds the (disabled) notify extension core
2066 o adds some debug output for the result of if/elsif tests
2067 o points to the current vacation draft in the documentation
2068 and documents the missing references header update
2072 o fixes a bug in processing the envelope test (when testing
2073 multiple envelope elements, the last element determinted the
2076 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2077 Electronic Mail") by including:
2079 Auto-submitted: auto-generated
2081 in the messages that it generates (bounce messages and others, such as
2082 warnings). In the case of bounce messages for non-SMTP mesages, there was
2083 also a typo: it was using "Auto_submitted" (underscore instead of
2084 hyphen). Since every message generated by Exim is necessarily in response
2085 to another message, thes have all been changed to:
2087 Auto-Submitted: auto-replied
2089 in accordance with these statements in the RFC:
2091 The auto-replied keyword:
2093 - SHOULD be used on messages sent in direct response to another
2094 message by an automatic process,
2096 - MUST NOT be used on manually-generated messages,
2098 - MAY be used on Delivery Status Notifications (DSNs) and Message
2099 Disposition Notifications (MDNs),
2101 - MUST NOT be used on messages generated by automatic or periodic
2102 processes, except for messages which are automatic responses to
2105 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2106 to the default Received: header definition.
2108 PH/12 Added log selector acl_warn_skipped (default on).
2110 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2111 variables because (a) they are in the wrong storage pool and (b) even if
2112 they were copied, it wouldn't work properly because of the caching.
2114 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2115 checking when decoding. Apparently there are clients that generate
2116 overlong encoded strings. Why am I not surprised?
2118 PH/15 If the first argument of "${if match_address" was not empty, but did not
2119 contain an "@" character, Exim crashed. Now it writes a panic log message
2120 and treats the condition as false.
2122 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2124 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2125 extension "envelope-auth". The code is finished and in agreement with
2126 other implementations, but there is no documentation so far and in fact,
2127 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2128 not changing the active code.
2130 Print executed "if" and "elsif" statements when debugging is used. This
2131 helps a great deal to understand what a filter does.
2133 Document more things not specified clearly in RFC3028. I had all this
2134 sorted out, when out of a sudden new issues came to my mind. Oops."
2136 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2139 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2140 canonical form (as documented). However, after a host name lookup from
2141 the IP address, check_host() was doing a simple string comparison with
2142 addresses acquired from the DNS when checking that the found name did
2143 have the original IP as one of its addresses. Since any found IPv6
2144 addresses are likely to be in abbreviated form, the comparison could
2145 fail. Luckily, there already exists a function for doing the comparison
2146 by converting both addresses to binary, so now that is used instead of
2147 the text comparison.
2149 PH/20 There was another similar case to PH/19, when a complete host name was
2150 given in a host list; looking up its IP address could give an abbreviated
2151 form, whereas the current host's name might or might not be abbreviated.
2152 The same fix has been applied.
2158 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2159 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2162 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2163 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2165 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2167 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2168 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2169 Also, for "Auto-Submitted", treat anything other than "no" as
2170 non-personal, in accordance with RFC 3834. (Previously it treated
2171 anything starting "auto-" as non-personal.)
2173 TF/01 The control=submission/name=... option had a problem with syntax
2174 errors if the name included a slash character. The /name= option
2175 now slurps the rest of the string, so it can include any characters
2176 but it must come last in the list of options (after /sender_retain
2179 PH/05 Some modifications to the interface to the fake nameserver for the new
2187 TK/01 Added the "success_on_redirect" address verification option. See
2188 NewStuff for rationale and an example.
2190 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2192 PH/02 Patch to exigrep to allow it to work on syslog lines.
2194 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2195 fread() to skip over the body file's header line, because in Cygwin the
2196 header line is locked and is inaccessible.
2198 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2199 co-exist for some time) to make it clear that it is the Exim ID that is
2200 referenced, not the Message-ID: header line.
2202 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2203 string_format() function, because snprintf() does not exist on all
2206 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2207 ${stat: expansion item.
2209 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2210 protocol synchronization error", to keep the pedants happy.
2212 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2213 well as for IRIX systems, when gcc is being used. See the host.c source
2216 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2218 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2221 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2222 transport and to the smtp transport in LMTP mode.
2224 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2226 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2227 run a filter in a subprocess. This could lead to confusion in subsequent
2228 lookups in the parent process. There should also be a search_tidyup() at
2229 the end of the subprocess.
2231 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2232 only if the host matched helo_try_verify_hosts, which caused the
2233 verification to occur when the EHLO/HELO command was issued. The ACL just
2234 tested the remembered result. Now, if a previous verification attempt has
2235 not happened, "verify = helo" does it there and then.
2237 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2239 TK/03 Fix log output including CR from clamd.
2241 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2242 exist provoked a memory error which could cause a segfault.
2244 PH/15 Installed PCRE 6.2
2246 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2248 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2249 of the problem. Specifically, suggested +O2 rather than +O1 for the
2252 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2254 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2255 record, the port was not correctly logged when the outgoing_port log
2256 selector was set (it logged the transort's default port).
2258 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2259 fallback_hosts, and "hosts" in the smtp transport.
2261 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2262 host errors such as "Connection refused".
2264 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2265 authentication with radiusclient 0.4.9:
2267 - Error returned from rc_read_config was caught wrongly
2268 - Username/password not passed on to radius server due to wrong length.
2270 The presumption is that some radiusclient API changes for 4.51/PH/17
2271 were not taken care of correctly. The code is still untested by me (my
2272 Linux distribution still has 0.3.2 of radiusclient), but it was
2273 contributed by a Radius user.
2275 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2276 expanding the "port" option of the smtp transport.
2278 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2279 while reading a MIME header. Thanks to Tom Hughes for a patch.
2281 PH/24 Include config.h inside local_scan.h so that configuration settings are
2284 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2285 that in an ACL for RCPT (for example), you can examine exactly what was
2288 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2289 commands, but it was not correctly comparing the address with the actual
2290 client host address. Thus, it would show the EHLO address in Received:
2291 header lines when this was not necessary.
2293 PH/27 Added the % operator to ${eval:}.
2295 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2296 it should be ignoring failures (because with -C, for example, it has lost
2297 privilege). It wasn't ignoring creation failures other than "already
2300 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2303 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2304 because an input line was too long, either on its own, or by virtue of
2305 too many continuations, the temporary file was not being removed, and the
2306 return code was incorrect.
2308 PH/31 Missing "BOOL" in function definition in filtertest.c.
2310 PH/32 Applied Sieve patches from the maintainer.
2312 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2314 PH/33 Added "verify = not_blind".
2316 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2317 Local/Makefile (with some defaults set). These are used in built scripts
2318 such as exicyclog, but they have never been used in the exim_install
2319 script (though there are many overriding facilities there). I have
2320 arranged that the exim_install script now takes note of these two
2323 PH/35 Installed configuration files for Dragonfly.
2325 PH/36 When a locally submitted message by a trusted user did not contain a
2326 From: header, and the sender address was obtained from -f or from an SMTP
2327 MAIL command, and the trusted user did not use -F to supply a sender
2328 name, $originator_name was incorrectly used when constructing a From:
2329 header. Furthermore, $originator_name was used for submission mode
2330 messages from external hosts without From: headers in a similar way,
2331 which is clearly wrong.
2333 PH/37 Added control=suppress_local_fixups.
2335 PH/38 When log_selector = +received_sender was set, and the addition of the
2336 sender made the log line's construction buffer exactly full, or one byte
2337 less than full, an overflow happened when the terminating "\n" was
2340 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2341 when the result of a list match is failure because a DNS lookup failed.
2343 PH/40 RM_COMMAND is now used in the building process.
2345 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2346 the "build-* directories that it finds.
2348 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2349 address in a domain literal was a prefix of an interface address.
2351 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2352 when verifying a sender address, unless rewrite_headers is false.
2354 PH/44 Wrote a long comment about why errors_to addresses are verified as
2355 recipients, not senders.
2357 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2358 the ratelimit ACL was added.
2360 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2362 PH/46 Added extra information about PostgreSQL errors to the error string.
2364 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2365 avoiding the need to install special zones in a real server. This is
2366 backwards compatible; if it can't find the fake resolver, it drops back.
2367 Thus, both old and new test suites can be run.
2369 TF/02 Added util/ratelimit.pl
2371 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2372 clock is set back in time.
2374 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2375 Candler <B.Candler@pobox.com>.
2377 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2378 OK for addresses that are the result of redirections.
2380 PH/48 A number of further additions for the benefit of the new test suite,
2381 including a fake gethostbyname() that interfaces to the fake DNS resolver
2384 TF/06 The fix for widen_domains has also been applied to qualify_single and
2385 search_parents which are the other dnslookup options that can cause
2388 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2391 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2393 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2394 service" if it encountered an MX record with an empty target hostname.
2395 The message is now "an MX or SRV record indicated no SMTP service".
2397 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2398 if the DNS of the sending site is misconfigured. This is quite a
2399 common situation. This change restores the behaviour of treating a
2400 helo verification defer as a failure.
2402 PH/51 If self=fail was set on a router, the bounce message did not include the
2403 actual error message.
2409 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2411 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2412 command itself ended in error, the underlying message about the transport
2413 filter timeout was being overwritten with the pipe command error. Now the
2414 underlying error message should be appended to the second error message.
2416 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2418 PH/02 Reduce the amount of output that "make" produces by default. Full output
2419 can still be requested.
2421 PH/03 The warning log line about a condition test deferring for a "warn" verb
2422 was being output only once per connection, rather than after each
2423 occurrence (because it was using the same function as for successful
2424 "warn" verbs). This seems wrong, so I have changed it.
2426 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2427 it should not have, which might have caused a crash in the right
2428 circumstances, but probably never did.
2430 PH/04 Installed a modified version of Tony Finch's patch to make submission
2431 mode fix the return path as well as the Sender: header line, and to
2432 add a /name= option so that you can make the user's friendly name appear
2435 TF/03 Added the control = fakedefer ACL modifier.
2437 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2438 Mark Lowes for thorough testing.
2440 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2442 TK/03 Merged latest SRS patch from Miles Wilton.
2444 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2445 with the definition in sysexits.h (which is #included earlier).
2446 Fortunately, Exim does not actually use EX_OK. The code used to try to
2447 preserve the sysexits.h value, by assumimg that macro definitions were
2448 scanned for macro replacements. I have been disabused of this notion,
2449 so now the code just undefines EX_OK before #including unistd.h.
2451 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2452 in the smtp transport. When a block could not be written in a single
2453 write() function, the timeout was being re-applied to each part-write.
2454 This seems wrong - if the receiver was accepting one byte at a time it
2455 would take for ever. The timeout is now adjusted when this happens. It
2456 doesn't have to be particularly precise.
2458 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2459 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2461 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2462 without a domain if the check to <postmaster@domain> fails.
2464 SC/01 Eximstats: added -xls and the ability to specify output files
2465 (patch written by Frank Heydlauf).
2467 SC/02 Eximstats: use FileHandles for outputing results.
2469 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2471 SC/04 Eximstats: fixed display of large numbers with -nvr option
2473 SC/05 Eximstats: fixed merging of reports with empty tables.
2475 SC/06 Eximstats: added the -include_original_destination flag
2477 SC/07 Eximstats: removed tabs and trailing whitespace.
2479 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2481 TK/06 MBOX spool code: Add real "From " MBOX separator line
2482 so the .eml file is really in mbox format (even though
2483 most programs do not really care). Patch from Alex Miller.
2485 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2486 The latter is generated from $received_to and is only set if the
2487 message has one envelope recipient. SA can use these headers,
2488 obviously out-of-the-box. Patch from Alex Miller.
2490 PH/08 The ${def test on a variable was returning false if the variable's
2491 value was "0", contrary to what the specification has always said!
2492 The result should be true unless the variable is empty.
2494 PH/09 The syntax error of a character other than { following "${if
2495 def:variable_name" (after optional whitespace) was not being diagnosed.
2496 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2497 accidental colon was present, for example, could give incorrect results.
2499 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2500 result is used (not including appendfile, where other changes are about
2503 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2504 This involved changing a lot of size variables from int to off_t. It
2505 should work with maildirs and everything.
2507 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2508 spamd dying while we are connected to it.
2510 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2513 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2514 function for BDB 4.3.
2516 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2518 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2519 the sources for PCRE have been re-arranged and more files are now
2522 PH/15 The code I had for printing potentially long long variables in PH/11
2523 above was not the best (it lost precision). The length of off_t variables
2524 is now inspected at build time, and an appropriate printing format (%ld
2525 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2526 to be "long long int" or "long int". This is needed for the internal
2527 formatting function string_vformat().
2529 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2530 the configuration file to be ":syslog", then the script "guesses" where
2531 the logs files are, rather than using the compiled in default. In our
2532 case the guess is not the same as the compiled default, so the script
2533 suddenly stopped working when I started to use syslog. The patch checks
2534 to see if log_file_path is "". If so, it attempts to read it from exim
2535 with no configuration file to get the compiled in version, before it
2536 falls back to the previous guessing code."
2538 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2539 implementing BATV in an Exim configuration. See NewStuff for the gory
2542 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2543 Makefile that are specific to HP-UX.
2545 PH/18 If the "use_postmaster" option was set for a recipient callout together
2546 with the "random" option, the postmaster address was used as the MAIL
2547 FROM address for the random test, but not for the subsequent recipient
2548 test. It is now used for both.
2550 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2551 patch removes a few documentation additions to RFC 3028, because the
2552 latest draft now contains them. It adds the new en;ascii-case comparator
2553 and a new error check for 8bit text in MIME parts. Comparator and
2554 require names are now matched exactly. I enabled the subaddress
2555 extension, but it is not well tested yet (read: it works for me)."
2557 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2558 rework some of the code of TK/09 above to avoid the hardwired use of
2559 "%lld" and "long long". Replaced the call to snprintf() with a call to
2562 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2563 records point to non-existent hosts", "retry timeout exceeded", and
2564 "retry time not reached for any host after a long failure period".
2566 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2567 experimental DomainKeys support:
2569 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2570 (2) On an error such as an illegally used "control", the wrong name for
2571 the control was given.
2573 These problems did NOT occur unless DomainKeys support was compiled.
2575 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2577 PH/24 Added ${if match_ip condition.
2579 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2580 databases so that it will be absolutely obvious if a crash occurs in the
2581 DB library. This is a regular occurrence (often caused by mis-matched
2584 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2585 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2586 picked up on a user's system that detects such things. There doesn't seem
2587 to be a gcc warning option for this - only an attribute that has to be
2588 put on the function's prototype. It seems that in Fedora Core 4 they have
2589 set this on a number of new functions. No doubt there will be more in due
2592 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2593 specify a transport. However, if an address that was verified by such a
2594 router was the subject of a callout, Exim crashed because it tried to
2595 read the rcpt_include_affixes from the non-existent transport. Now it
2596 just assumes that the setting of that option is false. This bug was
2597 introduced by 4.51/PH/31.
2599 PH/28 Changed -d+all to exclude +memory, because that information is very
2600 rarely of interest, but it makes the output a lot bigger. People tend to
2601 do -d+all out of habit.
2603 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2604 code in os-type was giving problems when libc.so lives in lib64, like on
2607 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2608 aren't the modern standard, and it seems that some systems' include files
2609 don't always have them. Exim was already checking for some of the newer
2610 ones like T_AAAA, and defining it itself. I've added checks for all the
2611 record types that Exim uses.
2613 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2614 not automatically generating a new one, as it is supposed to. This
2615 prevented TLS from working. If the file did exist, but contained invalid
2616 data, a new version was generated, as expected. It was only the case of a
2617 non-existent file that was broken.
2619 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2620 with a change in libdomainkeys > 0.64.
2622 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2623 from DNS. If the selector record carries the flag, it now has
2624 precedence over the domain-wide flag.
2626 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2628 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2629 the use of a port name that isn't defined in /etc/services) occurred, the
2630 message was deferred as in a normal delivery, and thus remained on the
2631 spool, instead of being failed because of the mua_wrapper setting. This
2632 is now fixed, and I tidied up some of the mua_wrapper messages at the
2635 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2636 the messages in a hash of arrays rather than using individual hashes.
2637 This is a bit cleaner and results in dramatic memory savings, albeit
2638 at a slight CPU cost.
2640 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2641 as requested by Marc Sherman.
2643 SC/10 Eximstats: added histograms for user specified patterns as requested
2646 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2648 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2649 fopen() in the content-scanning modules that did not already have it.
2655 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2656 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2658 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2660 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2662 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2663 to test Sieve filters that use "vacation".
2665 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2666 that changes the way the GnuTLS parameters are stored in the cache file.
2667 The new format can be generated externally. For backward compatibility,
2668 if the data in the cache doesn't make sense, Exim assumes it has read an
2669 old-format file, and it generates new data and writes a new file. This
2670 means that you can't go back to an older release without removing the
2673 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2674 work if there are any delivery delays because "one_time" forces the
2675 parent to be marked "delivered", so its unseen clone is never tried
2676 again. For this reason, Exim now forbids the simultaneous setting of
2679 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2680 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2681 entry if you want to know the details. The fix, however, neglected to
2682 consider the case when local delivery batching is involved. The test for
2683 "previously delivered" was not happening when checking to see if an
2684 address could be batched with a previous (undelivered) one; under
2685 certain circumstances this could lead to multiple deliveries to the same
2688 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2689 in its include files, and this causes problems building Exim.
2691 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2692 header_syntax) but Exim was just ignoring anything given after a slash.
2693 In particular, this caused confusion with an attempt to use "verify =
2694 reverse_host_lookup/defer_ok". An error is now given when options are
2695 supplied for verify items that do not have them. (Maybe reverse_host_
2696 lookup should have a defer_ok option, but that's a different point.)
2698 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2699 defined by RFC 821) to 2048, because there were problems with some AUTH
2700 commands, and RFC 1869 says the size should be increased for extended
2701 SMTP commands that take arguments.
2703 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2706 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2707 "unknown" error; now it says that the functionality isn't in the binary.
2709 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2710 an address' error message when a string expansion fails (syntax or
2711 whatever). Otherwise the password may appear in the log. Following change
2712 PH/42 below, there is no longer a chance of it appearing in a bounce
2715 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2717 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2718 of its addresses was ever tried. (Bugzilla bug #2.)
2720 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2721 the result incorrectly in the debug output. (It correctly added a newline
2722 to what was transported.)
2724 TF/01 Added $received_time.
2726 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2727 commented out examples of how to interface to a virus scanner and to
2728 SpamAssassin. Also added commented examples of av_scanner and
2729 spamd_address settings.
2731 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2732 and controls are allowed in which ACLs. There were a couple of minor
2733 errors. Some of the entries in the conditions table (which is a table of
2734 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2735 negation of where the condition IS allowed.
2737 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2739 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2740 header file does not have a version number, so I've had to invent a new
2741 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2742 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2743 radiusclient), but it was contributed by a Radius user.
2745 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2746 files or directories, for OpenSSL.
2748 PH/19 When an Exim process that is running as root has to create an Exim log
2749 file, it does so in a subprocess that runs as exim:exim so as to get the
2750 ownership right at creation (otherwise, other Exim processes might see
2751 the file with the wrong ownership). There was no test for failure of this
2752 fork() call, which would lead to the process getting stuck as it waited
2753 for a non-existent subprocess. Forks do occasionally fail when resources
2754 run out. I reviewed all the other calls to fork(); they all seem to check
2757 PH/20 When checking for unexpected SMTP input at connect time (before writing
2758 the banner), Exim was not dealing correctly with a non-positive return
2759 from the read() function. If the client had disconnected by this time,
2760 the result was a log entry for a synchronization error with an empty
2761 string after "input=" when read() returned zero. If read() returned -1
2762 (an event I could not check), uninitialized data bytes were printed.
2763 There were reports of junk text (parts of files, etc) appearing after
2766 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2768 PH/22 Added support for macro redefinition, and (re)definition in between
2769 driver and ACL definitions.
2771 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2772 forgetting to use the resulting value; it was using the unexpanded value.
2774 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2775 hadn't been configured. The fix is from Juergen Kreileder, who
2776 understands it better than I do:
2778 "Here's what I see happening with three configured cyrus_sasl
2779 authenticators configured (plain, login, cram-md5):
2781 On startup auth_cyrus_sasl_init() gets called for each of these.
2782 This means three calls to sasl_listmech() without a specified mech_list.
2783 => SASL tests which mechs of all available mechs actually work
2784 => three warnings about OTP not working
2785 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2787 With the patch, sasl_listmech() also gets called three times. But now
2788 SASL's mech_list option is set to the server_mech specified in the the
2789 authenticator. Or in other words, the answer from sasl_listmech()
2790 gets limited to just the mech you're testing for (which is different
2792 => the return list contains just 'plain' or 'login', 'cram-md5' or
2793 nothing depending on the value of ob->server_mech.
2795 I've just tested the patch: Authentication still works fine,
2796 unavailable mechs specified in the exim configuration are still
2797 caught, and the auth.log warnings about OTP are gone."
2799 PH/25 When debugging is enabled, the contents of the command line are added
2800 to the debugging output, even when log_selector=+arguments is not
2803 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2804 answer is "GNU", and only if the return is "GNU/something" is the answer
2807 PH/27 $acl_verify_message is now set immediately after the failure of a
2808 verification in an ACL, and so is available in subsequent modifiers. In
2809 particular, the message can be preserved by coding like this:
2811 warn !verify = sender
2812 set acl_m0 = $acl_verify_message
2814 Previously, $acl_verify_message was set only while expanding "message"
2815 and "log_message" when a very denied access.
2817 PH/28 Modified OS/os.c-Linux with
2819 -#ifndef OS_LOAD_AVERAGE
2820 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2822 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2823 nomenclature these days.)
2825 PH/29 Installed patch from the Sieve maintainer that adds the options
2826 sieve_useraddress and sieve_subaddress to the redirect router.
2828 PH/30 In these circumstances:
2829 . Two addresses routed to the same list of hosts;
2830 . First host does not offer TLS;
2831 . First host accepts first address;
2832 . First host gives temporary error to second address;
2833 . Second host offers TLS and a TLS session is established;
2834 . Second host accepts second address.
2835 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2836 and peerdn, if requested) that were in fact used only for the second
2839 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2840 attention to any local part prefix or suffix that was matched by the
2841 router that accepted the address. It now behaves in the same way as it
2842 does for delivery: the affixes are removed from the local part unless
2843 rcpt_include_affixes is set on the transport.
2845 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2846 timeout during the DATA phase of an incoming message.
2848 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2849 applied a patch sanctioned by the Sieve maintainer.
2851 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2852 the uid or gid is negative. A case of a negative gid caused this to be
2853 noticed. The fix allows for either to be negative.
2855 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2856 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2859 PH/36 The change PH/12 above was broken. Fixed it.
2861 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2862 the grounds that routing the same address twice would always produce the
2863 same answer. This might have been true once, but it is certainly no
2864 longer true now. Routing a child address may depend on the previous
2865 routing that produced that child. Some complicated redirection strategies
2866 went wrong when messages had multiple recipients, and made Exim's
2867 behaviour dependent on the order in which the addresses were given.
2869 I have moved the duplicate checking until after the routing is complete.
2870 Exim scans the addresses that are assigned to local and remote
2871 transports, and removes any duplicates. This means that more work will be
2872 done, as duplicates will always all be routed, but duplicates are
2873 presumably rare, so I don't expect this is of any significance.
2875 For deliveries to pipes, files, and autoreplies, the duplicate checking
2876 still happens during the routing process, since they are not going to be
2879 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2880 It corrects a timeout issue with spamd. This is Ian's comment: "The
2881 background is that sometimes spamd either never reads data from a
2882 connection it has accepted, or it never writes response data. The exiscan
2883 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2884 blindly assumes that writes won't block so it may never time out."
2886 PH/39 Allow G after quota size as well as K and M.
2888 PH/40 The value set for $authenticated_id in an authenticator may not contain
2889 binary zeroes or newlines because the value is written to log lines and
2890 to spool files. There was no check on this. Now the value is run through
2891 the string_printing() function so that such characters are converted to
2892 printable escape sequences.
2894 PH/41 $message_linecount is a new variable that contains the total number of
2895 lines in the message. Compare $body_linecount, which is the count for the
2898 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2899 bounce and delay warning messages, except in certain special cases, which
2902 (a) An SMTP error message from a remote host;
2903 (b) A message specified in a :fail: redirection;
2904 (c) A message specified in a "fail" command in a system filter;
2905 (d) A message specified in a FAIL return from the queryprogram router;
2906 (e) A message specified by the cannot_route_message router option.
2908 In these cases only, Exim does include the error details in bounce and
2909 warning messages. There are also a few cases where bland messages such
2910 as "unrouteable address" or "local delivery error" are given.
2912 PH/43 $value is now also set for the "else" part of a ${run expansion.
2914 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2915 being worked on, but at least Exim now implements the latest version to
2918 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2919 process to complete was treated as a delivery failure, a timeout while
2920 writing the message to the pipe was logged, but erroneously treated as a
2921 successful delivery. Such timeouts include transport filter timeouts. For
2922 consistency with the overall process timeout, these timeouts are now
2923 treated as errors, giving rise to delivery failures by default. However,
2924 there is now a new Boolean option for the pipe transport called
2925 timeout_defer, which, if set TRUE, converts the failures into defers for
2926 both kinds of timeout. A transport filter timeout is now identified in
2929 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2930 systems where "make" and "gmake" are different, calling "gmake" at top
2931 level broke things. I've arranged for the value of $(MAKE) to be passed
2932 from the Makefile to this script so that it can call the same version of
2936 A note about Exim versions 4.44 and 4.50
2937 ----------------------------------------
2939 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2940 changes of various kinds. As a consequence, a big documentation update was
2941 needed. This delayed the release for rather longer than seemed good, especially
2942 in the light of a couple of (minor) security issues. Therefore, the changes
2943 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2944 release. So 4.44 and 4.50 are in effect two different branches that both start
2947 I have left the 4.50 change log unchanged; it contains all the changes since
2948 4.43. The change log for 4.44 is below; many of its items are identical to
2949 those for 4.50. This seems to be the most sensible way to preserve the
2950 historical information.
2956 1. Minor wording change to the doc/README.SIEVE file.
2958 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2959 computation of the current number of files was incorrect.
2961 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2962 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2965 4. Give more explanation in the error message when the command for a transport
2966 filter fails to execute.
2968 5. There are several places where Exim runs a non-Exim command in a
2969 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2970 was being done only for the command run by the queryprogram router. It is
2971 now done for all such subprocesses. The other cases are: ${run, transport
2972 filters, and the commands run by the lmtp and pipe transports.
2974 6. Added CONFIGURE_GROUP build-time option.
2976 7. Some older OS have a limit of 256 on the maximum number of file
2977 descriptors. Exim was using setrlimit() to set 1000 as a large value
2978 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2979 systems. I've change it so that if it can't get 1000, it tries for 256.
2981 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2982 was an oversight, and furthermore, ever since the addition of extra
2983 controls (e.g. 4.43/32), the checks on when to allow different forms of
2984 "control" were broken. There should now be diagnostics for all cases when a
2985 control that does not make sense is encountered.
2987 9. Added the /retain_sender option to "control=submission".
2989 10. $recipients is now available in the predata ACL (oversight).
2991 11. Tidy the search cache before the fork to do a delivery from a message
2992 received from the command line. Otherwise the child will trigger a lookup
2993 failure and thereby defer the delivery if it tries to use (for example) a
2994 cached ldap connection that the parent has called unbind on.
2996 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2997 of $address_data from the recipient verification was clobbered by the
2998 sender verification.
3000 13. The value of address_data from a sender verification is now available in
3001 $sender_address_data in subsequent conditions in the ACL statement.
3003 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3005 15. Added a new option "connect=<time>" to callout options, to set a different
3008 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3009 was its contents. (It was OK if the option was not defined at all.)
3011 17. A "Completed" log line is now written for messages that are removed from
3012 the spool by the -Mrm option.
3014 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3015 information about exactly what failed.
3017 19. Added -dd to debug only the daemon process.
3019 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3020 handles timeouts, both on the server side and network timeouts. Renamed the
3021 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3023 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3024 It is now set to "smtps".
3026 22. $host_address is now set to the target address during the checking of
3027 ignore_target_hosts.
3029 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3030 being passed; this would have caused $sender_host_name to have been used if
3031 matching the list had actually called for a host name (not very likely,
3032 since this list is usually IP addresses). A host name is now passed as
3035 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3036 code that specifies a non-restarting handler (typically sigaction() in
3037 modern systems) in an attempt to fix a rare and obscure crash bug.
3039 25. Narrowed the window for a race in the daemon that could cause it to ignore
3040 SIGCHLD signals. This is not a major problem, because they are used only to
3041 wake it up if nothing else does.
3043 26. A malformed maildirsize file could cause Exim to calculate negative values
3044 for the mailbox size or file count. Odd effects could occur as a result.
3045 The maildirsize information is now recalculated if the size or filecount
3048 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3049 support for a long time. Removed HAVE_SYS_VFS_H.
3051 28. Installed the latest version of exipick from John Jetmore.
3053 29. In an address list, if the pattern was not a regular expression, an empty
3054 subject address (from a bounce message) matched only if the pattern was an
3055 empty string. Non-empty patterns were not even tested. This was the wrong
3056 because it is perfectly reasonable to use an empty address as part of a
3057 database query. An empty address is now tested by patterns that are
3058 lookups. However, all the other forms of pattern expect the subject to
3059 contain a local part and a domain, and therefore, for them, an empty
3060 address still always fails if the pattern is not itself empty.
3062 30. Exim went into a mad DNS loop when attempting to do a callout where the
3063 host was specified on an smtp transport, and looking it up yielded more
3064 than one IP address.
3066 31. Re-factored the code for checking spool and log partition space into a
3067 function that finds that data and another that does the check. The former
3068 is then used to implement four new variables: $spool_space, $log_space,
3069 $spool_inodes, and $log_inodes.
3071 32. The RFC2047 encoding function was originally intended for short strings
3072 such as real names; it was not keeping to the 75-character limit for
3073 encoded words that the RFC imposes. It now respects the limit, and
3074 generates multiple encoded words if necessary. To be on the safe side, I
3075 have increased the buffer size for the ${rfc2047: expansion operator from
3078 33. It is now permitted to omit both strings after an "if" condition; if the
3079 condition is true, the result is "true". As before, when the second string
3080 is omitted, a false condition yields an empty string. This makes it less
3081 cumbersome to write custom ACL and router conditions.
3083 34. Failure to deliver a bounce message always caused it to be frozen, even if
3084 there was an errors_to setting on the router. The errors_to setting is now
3087 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3088 canonical form (fully expanded) before being placed in
3089 $sender_host_address.
3091 36. The table in the code that translates DNS record types into text (T_A to
3092 "A" for instance) was missing entries for NS and CNAME. It is just possible
3093 that this could have caused confusion if both these types were looked up
3094 for the same domain, because the text type is used as part of Exim's
3095 per-process caching. But the chance of anyone hitting this buglet seems
3098 37. The dnsdb lookup has been extended in a number of ways.
3100 (1) There is a new type, "zns", which walks up the domain tree until it
3101 finds some nameserver records. It should be used with care.
3103 (2) There is a new type, "mxh", which is like "mx" except that it returns
3104 just the host names, not the priorities.
3106 (3) It is now possible to give a list of domains (or IP addresses) to be
3107 looked up. The behaviour when one of the lookups defers can be
3108 controlled by a keyword.
3110 (4) It is now possible to specify the separator character for use when
3111 multiple records are returned.
3113 38. The dnslists ACL condition has been extended: it is now possible to supply
3114 a list of IP addresses and/or domains to be looked up in a particular DNS
3117 39. Added log_selector=+queue_time_overall.
3119 40. When running the queue in the test harness, wait just a tad after forking a
3120 delivery process, to get repeatability of debugging output.
3122 41. Include certificate and key file names in error message when GnuTLS fails
3123 to set them up, because the GnuTLS error message doesn't include the name
3124 of the failing file when there is a problem reading it.
3126 42. Allow both -bf and -bF in the same test run.
3128 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3130 44. The "Exiscan patch" is now merged into the mainline Exim source.
3132 45. Sometimes the final signoff response after QUIT could fail to get
3133 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3134 < 0 before doing a fflush(). This bug looks as though it goes back to the
3135 introduction of TLS in release 3.20, but "sometimes" must have been rare
3136 because the tests only now provoked it.
3138 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3139 (this can affect the format of dates).
3141 47. exim_tidydb, when checking for the continued existence of a message for
3142 which it has found a message-specific retry record, was not finding
3143 messages that were in split spool directories. Consequently, it was
3144 deleting retry records that should have stayed in existence.
3146 48. Steve fixed some bugs in eximstats.
3148 49. The SPA authentication driver was not abandoning authentication and moving
3149 on to the next authenticator when an expansion was forced to fail,
3150 contradicting the general specification for all authenticators. Instead it
3151 was generating a temporary error. It now behaves as specified.
3153 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3154 (the order specifies the preference for clients). The order is now AES256,
3155 AES128, 3DES, ARCFOUR128.
3157 51. Small patch to Sieve code - explicitly set From: when generating an
3160 52. Exim crashed if a remote delivery caused a very long error message to be
3161 recorded - for instance if somebody sent an entire SpamAssassin report back
3162 as a large number of 550 error lines. This bug was coincidentally fixed by
3163 increasing the size of one of Exim's internal buffers (big_buffer) that
3164 happened as part of the Exiscan merge. However, to be on the safe side, I
3165 have made the code more robust (and fixed the comments that describe what
3168 53. Now that there can be additional text after "Completed" in log lines (if
3169 the queue_time_overall log selector is set), a one-byte patch to exigrep
3170 was needed to allow it to recognize "Completed" as not the last thing in
3173 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3174 patch that reportedly fixes this has been added. I am not expert enough to
3175 create a test for it. This is what the patch creator wrote:
3177 "I found a little strange behaviour of ldap code when working with
3178 Windows 2003 AD Domain, where users was placed in more than one
3179 Organization Units. When I tried to give exim partial DN, the exit code
3180 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3181 But simultaneously result of request was absolutely normal ldap result,
3182 so I produce this patch..."
3184 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3185 REFERENCE, so I have modified the code to exclude the patch when that macro
3188 55. Some experimental protocols are using DNS PTR records for new purposes. The
3189 keys for these records are domain names, not reversed IP addresses. The
3190 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3191 leaves it alone. Component reversal etc. now happens only for IP addresses.
3194 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3196 57. Double the size of the debug message buffer (to 2048) so that more of very
3197 long debug lines gets shown.
3199 58. The exicyclog utility now does better if the number of log files to keep
3200 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3202 59. Two changes related to the smtp_active_hostname option:
3204 (1) $smtp_active_hostname is now available as a variable.
3205 (2) The default for smtp_banner uses $smtp_active_hostname instead
3206 of $primary_hostname.
3208 60. The host_aton() function is supposed to be passed a string that is known
3209 to be a valid IP address. However, in the case of IPv6 addresses, it was
3210 not checking this. This is a hostage to fortune. Exim now panics and dies
3211 if the condition is not met. A case was found where this could be provoked
3212 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3213 components; fortuitously, this particular loophole had already been fixed
3214 by change 4.50/55 above.
3216 If there are any other similar loopholes, the new check in host_aton()
3217 itself should stop them being exploited. The report I received stated that
3218 data on the command line could provoke the exploit when Exim was running as
3219 exim, but did not say which command line option was involved. All I could
3220 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3221 running as the user.
3224 61. There was a buffer overflow vulnerability in the SPA authentication code
3225 (which came originally from the Samba project). I have added a test to the
3226 spa_base64_to_bits() function which I hope fixes it.
3229 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3230 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3232 63. The daemon start-up calls getloadavg() while still root for those OS that
3233 need the first call to be done as root, but it missed one case: when
3234 deliver_queue_load_max is set with deliver_drop_privilege. This is
3235 necessary for the benefit of the queue runner, because there is no re-exec
3236 when deliver_drop_privilege is set.
3238 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3239 This has been fixed.
3241 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3242 was in use, was not putting the data itself into the right store pool;
3243 consequently, it could be overwritten for a subsequent message in the same
3244 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3247 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3249 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3250 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3251 the function were treating the return as a boolean value, which happened to
3252 work because 0=false and not-0=true, but is not correct code.
3254 68. The host_aton() function was not handling scoped IPv6 addresses (those
3255 with, for example, "%eth0" on the end) correctly.
3257 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3258 negated items (that is, ~something) in unsigned ints. Some compilers
3259 apparently mutter when there is no cast.
3261 70. If an address verification called from an ACL failed, and did not produce a
3262 user-specific message (i.e. there was only a "system" message), nothing was
3263 put in $acl_verify_message. In this situation, it now puts the system
3266 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3267 session; change 4.31/43 added the unwanted input to the log line - except
3268 that it did not do this in the start of session case. It now does.
3270 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3271 This is wrong and can cause the other end to generate a synchronization
3272 error if it is another Exim or anything else that does the synchronization
3273 check. A QUIT command is no longer sent after a timeout.
3275 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3276 during host lookups.
3278 74. The defer_ok option of callout verification was not working if it was used
3279 when verifying addresses in header lines, that is, for this case:
3281 verify = header_sender/callout=defer_ok
3283 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3284 those file descriptors could be used for SMTP connections. If anything
3285 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3286 could be sent to the SMTP client, causing chaos. The daemon now opens
3287 stdin, stdout, and stderr to /dev/null when it puts itself into the
3290 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3291 log by default. The user can override this with suitable Perl magic.
3293 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3294 the log message when discard triggers, was not working for the DATA ACL or
3295 for the non-SMTP ACL.
3297 78. Error message wording change in sieve.c.
3299 79. If smtp_accept_max_per_host was set, the number of connections could be
3300 restricted to fewer than expected, because the daemon was trying to set up
3301 a new connection before checking whether the processes handling previous
3302 connections had finished. The check for completed processes is now done
3303 earlier. On busy systems, this bug wouldn't be noticed because something
3304 else would have woken the daemon, and it would have reaped the completed
3307 80. If a message was submitted locally by a user whose login name contained one
3308 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3309 It caused a spool format error. I have fixed the spool reading code. A
3310 related problem was that the "from" clause in the Received: line became
3311 illegal because of the space(s). It is now covered by ${quote_local_part.
3313 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3316 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3318 83. Patch from Sieve maintainer to fix unterminated string problem in
3319 "vacation" handling.
3321 84. Some minor changes to the Linux configuration files to help with other
3322 OS variants using glibc.
3324 85. One more patch for Sieve to update vacation handling to latest spec.
3327 ----------------------------------------------------
3328 See the note above about the 4.44 and 4.50 releases.
3329 ----------------------------------------------------
3335 1. Change 4.43/35 introduced a bug that caused file counts to be
3336 incorrectly computed when quota_filecount was set in an appendfile
3339 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3340 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3343 3. Give more explanation in the error message when the command for a transport
3344 filter fails to execute.
3346 4. There are several places where Exim runs a non-Exim command in a
3347 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3348 was being done only for the command run by the queryprogram router. It is
3349 now done for all such subprocesses. The other cases are: ${run, transport
3350 filters, and the commands run by the lmtp and pipe transports.
3352 5. Some older OS have a limit of 256 on the maximum number of file
3353 descriptors. Exim was using setrlimit() to set 1000 as a large value
3354 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3355 systems. I've change it so that if it can't get 1000, it tries for 256.
3357 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3358 was an oversight, and furthermore, ever since the addition of extra
3359 controls (e.g. 4.43/32), the checks on when to allow different forms of
3360 "control" were broken. There should now be diagnostics for all cases when a
3361 control that does not make sense is encountered.
3363 7. $recipients is now available in the predata ACL (oversight).
3365 8. Tidy the search cache before the fork to do a delivery from a message
3366 received from the command line. Otherwise the child will trigger a lookup
3367 failure and thereby defer the delivery if it tries to use (for example) a
3368 cached ldap connection that the parent has called unbind on.
3370 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3371 of $address_data from the recipient verification was clobbered by the
3372 sender verification.
3374 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3375 was its contents. (It was OK if the option was not defined at all.)
3377 11. A "Completed" log line is now written for messages that are removed from
3378 the spool by the -Mrm option.
3380 12. $host_address is now set to the target address during the checking of
3381 ignore_target_hosts.
3383 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3384 being passed; this would have caused $sender_host_name to have been used if
3385 matching the list had actually called for a host name (not very likely,
3386 since this list is usually IP addresses). A host name is now passed as
3389 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3390 code that specifies a non-restarting handler (typically sigaction() in
3391 modern systems) in an attempt to fix a rare and obscure crash bug.
3393 15. Narrowed the window for a race in the daemon that could cause it to ignore
3394 SIGCHLD signals. This is not a major problem, because they are used only to
3395 wake it up if nothing else does.
3397 16. A malformed maildirsize file could cause Exim to calculate negative values
3398 for the mailbox size or file count. Odd effects could occur as a result.
3399 The maildirsize information is now recalculated if the size or filecount
3402 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3403 support for a long time. Removed HAVE_SYS_VFS_H.
3405 18. Updated exipick to current release from John Jetmore.
3407 19. Allow an empty sender to be matched against a lookup in an address list.
3408 Previously the only cases considered were a regular expression, or an
3411 20. Exim went into a mad DNS lookup loop when doing a callout where the
3412 host was specified on the transport, if the DNS lookup yielded more than
3415 21. The RFC2047 encoding function was originally intended for short strings
3416 such as real names; it was not keeping to the 75-character limit for
3417 encoded words that the RFC imposes. It now respects the limit, and
3418 generates multiple encoded words if necessary. To be on the safe side, I
3419 have increased the buffer size for the ${rfc2047: expansion operator from
3422 22. Failure to deliver a bounce message always caused it to be frozen, even if
3423 there was an errors_to setting on the router. The errors_to setting is now
3426 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3427 canonical form (fully expanded) before being placed in
3428 $sender_host_address.
3430 24. Updated eximstats to version 1.33
3432 25. Include certificate and key file names in error message when GnuTLS fails
3433 to set them up, because the GnuTLS error message doesn't include the name
3434 of the failing file when there is a problem reading it.
3436 26. Expand error message when OpenSSL has problems setting up cert/key files.
3439 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3440 (this can affect the format of dates).
3442 28. exim_tidydb, when checking for the continued existence of a message for
3443 which it has found a message-specific retry record, was not finding
3444 messages that were in split spool directories. Consequently, it was
3445 deleting retry records that should have stayed in existence.
3447 29. eximstats updated to version 1.35
3448 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3449 1.35 - bugfix such that pie charts by volume are generated correctly
3451 30. The SPA authentication driver was not abandoning authentication and moving
3452 on to the next authenticator when an expansion was forced to fail,
3453 contradicting the general specification for all authenticators. Instead it
3454 was generating a temporary error. It now behaves as specified.
3456 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3457 (the order specifies the preference for clients). The order is now AES256,
3458 AES128, 3DES, ARCFOUR128.
3460 31. Small patch to Sieve code - explicitly set From: when generating an
3463 32. Exim crashed if a remote delivery caused a very long error message to be
3464 recorded - for instance if somebody sent an entire SpamAssassin report back
3465 as a large number of 550 error lines. This bug was coincidentally fixed by
3466 increasing the size of one of Exim's internal buffers (big_buffer) that
3467 happened as part of the Exiscan merge. However, to be on the safe side, I
3468 have made the code more robust (and fixed the comments that describe what
3471 33. Some experimental protocols are using DNS PTR records for new purposes. The
3472 keys for these records are domain names, not reversed IP addresses. The
3473 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3474 leaves it alone. Component reversal etc. now happens only for IP addresses.
3477 34. The host_aton() function is supposed to be passed a string that is known
3478 to be a valid IP address. However, in the case of IPv6 addresses, it was
3479 not checking this. This is a hostage to fortune. Exim now panics and dies
3480 if the condition is not met. A case was found where this could be provoked
3481 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3482 components; fortuitously, this particular loophole had already been fixed
3483 by change 4.50/55 or 4.44/33 above.
3485 If there are any other similar loopholes, the new check in host_aton()
3486 itself should stop them being exploited. The report I received stated that
3487 data on the command line could provoke the exploit when Exim was running as
3488 exim, but did not say which command line option was involved. All I could
3489 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3490 running as the user.
3493 35. There was a buffer overflow vulnerability in the SPA authentication code
3494 (which came originally from the Samba project). I have added a test to the
3495 spa_base64_to_bits() function which I hope fixes it.
3498 36. The daemon start-up calls getloadavg() while still root for those OS that
3499 need the first call to be done as root, but it missed one case: when
3500 deliver_queue_load_max is set with deliver_drop_privilege. This is
3501 necessary for the benefit of the queue runner, because there is no re-exec
3502 when deliver_drop_privilege is set.
3504 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3505 was in use, was not putting the data itself into the right store pool;
3506 consequently, it could be overwritten for a subsequent message in the same
3507 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3510 38. Sometimes the final signoff response after QUIT could fail to get
3511 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3512 < 0 before doing a fflush(). This bug looks as though it goes back to the
3513 introduction of TLS in release 3.20, but "sometimes" must have been rare
3514 because the tests only now provoked it.
3520 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3521 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3522 Now it returns an integer. A number of calls were still expecting a T/F
3523 return. Fortuitously, in all cases, the tests worked in OK situations,
3524 which is the norm. However, things would have gone wrong on any write
3525 failures on the smtp file descriptor. This function is used when sending
3526 messages over SMTP and also when doing verify callouts.
3528 2. When Exim is called to do synchronous delivery of a locally submitted
3529 message (the -odf or -odi options), it no longer closes stderr before doing
3532 3. Implemented the mua_wrapper option.
3534 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3536 5. Implemented the functions header_remove(), header_testname(),
3537 header_add_at_position(), and receive_remove_recipient(), and exported them
3540 6. If an ACL "warn" statement specified the addition of headers, Exim already
3541 inserted X-ACL-Warn: at the start if there was no header name. However, it
3542 was not making this test for the second and subsequent header lines if
3543 there were newlines in the string. This meant that an invalid header could
3544 be inserted if Exim was badly configured.
3546 7. Allow an ACL "warn" statement to add header lines at the start or after all
3547 the Received: headers, as well as at the end.
3549 8. Added the rcpt_4xx retry error code.
3551 9. Added postmaster_mailfrom=xxx to callout verification option.
3553 10. Added mailfrom=xxxx to the callout verification option, for verify=
3556 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3557 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3559 12. Inserted some casts to stop certain compilers warning when using pointer
3560 differences as field lengths or precisions in printf-type calls (mostly
3561 affecting debugging statements).
3563 13. Added optional readline() support for -be (dynamically loaded).
3565 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3566 same clock tick as a message's arrival, so that its received time was the
3567 same as the "first fail" time on the retry record, and that message
3568 remained on the queue past the ultimate address timeout, every queue runner
3569 would try a delivery (because it was past the ultimate address timeout) but
3570 after another failure, the ultimate address timeout, which should have then
3571 bounced the address, did not kick in. This was a "< instead of <=" error;
3572 in most cases the first failure would have been in the next clock tick
3573 after the received time, and all would be well.
3575 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3576 being recognized when the domain list was tested by the match_domain
3577 condition in an expansion string.
3579 16. Added the ${str2b64: operator.
3581 17. Exim was always calling setrlimit() to set a large limit for the number of
3582 processes, without checking whether the existing limit was already
3583 adequate. (It did check for the limit on file descriptors.) Furthermore,
3584 errors from getrlimit() and setrlimit() were being ignored. Now they are
3585 logged to the main and panic logs, but Exim does carry on, to try to do its
3586 job under whatever limits there are.
3588 18. Imported PCRE 5.0.
3590 19. Trivial typo in log message " temporarily refused connection" (the leading
3593 20. If the log selector return_path_on_delivery was set and an address was
3594 redirected to /dev/null, the delivery process crashed because it assumed
3595 that a return path would always be set for a "successful" delivery. In this
3596 case, the whole delivery is bypassed as an optimization, and therefore no
3599 21. Internal re-arrangement: the function for sending a challenge and reading
3600 a response while authentication was assuming a zero-terminated challenge
3601 string. It's now changed to take a pointer and a length, to allow for
3602 binary data in such strings.
3604 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3606 23. Exim was not respecting finduser_retries when seeking the login of the
3607 uid under which it was called; it was always trying 10 times. (The default
3608 setting of finduser_retries is zero.) Also, it was sleeping after the final
3609 failure, which is pointless.
3611 24. Implemented tls_on_connect_ports.
3613 25. Implemented acl_smtp_predata.
3615 26. If the domain in control=submission is set empty, Exim assumes that the
3616 authenticated id is a complete email address when it generates From: or
3617 Sender: header lines.
3619 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3620 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3621 chown and chgrp in /bin and hostname in /usr/bin.
3623 28. Exim was keeping the "process log" file open after each use, just as it
3624 does for the main log. This opens the possibility of it remaining open for
3625 long periods when the USR1 signal hits a daemon. Occasional processlog
3626 errors were reported, that could have been caused by this. Anyway, it seems
3627 much more sensible not to leave this file open at all, so that is what now
3630 29. The long-running daemon process does not normally write to the log once it
3631 has entered its main loop, and it closes the log before doing so. This is
3632 so that log files can straightforwardly be renamed and moved. However,
3633 there are a couple of unusual error situations where the daemon does write
3634 log entries, and I had neglected to close the log afterwards.
3636 30. The text of an SMTP error response that was received during a remote
3637 delivery was being truncated at 512 bytes. This is too short for some of
3638 the long messages that one sometimes sees. I've increased the limit to
3641 31. It is now possible to make retry rules that apply only when a message has a
3642 specific sender, in particular, an empty sender.
3644 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3645 it possible to be selective about when SMTP synchronization is enforced.
3647 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3649 32. Implemented hosts_connection_nolog.
3651 33. Added an ACL for QUIT.
3653 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3656 35. Added mailbox_size and mailbox_filecount to appendfile.
3658 36. Added control = no_multiline_responses to ACLs.
3660 37. There was a bug in the logic of the code that waits for the clock to tick
3661 in the case where the clock went backwards by a substantial amount such
3662 that the microsecond fraction of "now" was more than the microsecond
3663 fraction of "then" (but the whole seconds number was less).
3665 38. Added support for the libradius Radius client library this is found on
3666 FreeBSD (previously only the radiusclient library was supported).
3672 1. When certain lookups returned multiple values in the form name=value, the
3673 quoting of the values was not always being done properly. Specifically:
3674 (a) If the value started with a double quote, but contained no whitespace,
3676 (b) If the value contained whitespace other than a space character (i.e.
3677 tabs or newlines or carriage returns) it was not quoted.
3678 This fix has been applied to the mysql and pgsql lookups by writing a
3679 separate quoting function and calling it from the lookup code. The fix
3680 should probably also be applied to nisplus, ibase and oracle lookups, but
3681 since I cannot test any of those, I have not disturbed their existing code.
3683 2. A hit in the callout cache for a specific address caused a log line with no
3684 reason for rejecting RCPT. Now it says "Previous (cached) callout
3685 verification failure".
3687 3. There was an off-by-one bug in the queryprogram router. An over-long
3688 return line was truncated at 256 instead of 255 characters, thereby
3689 overflowing its buffer with the terminating zero. As well as fixing this, I
3690 have increased the buffer size to 1024 (and made a note to document this).
3692 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3693 when Exim is waiting for an SMTP response from a remote server, Exim
3694 restarts its select() call on the socket, thereby resetting its timeout.
3695 This is not a problem when such interrupts are rare. Somebody set up a cron
3696 job to run exiwhat every 2 minutes, which is less than the normal select()
3697 timeout (5 or 10 minutes). This meant that the select() timeout never
3698 kicked in because it was always reset. I have fixed this by comparing the
3699 time when an interrupt arrives with the time at the start of the first call
3700 to select(). If more time than the timeout has elapsed, the interrupt is
3701 treated as a timeout.
3703 5. Some internal re-factoring in preparation for the addition of Sieve
3704 extensions (by MH). In particular, the "personal" test is moved to a
3705 separate function, and given an option for scanning Cc: and Bcc: (which is
3706 not set for Exim filters).
3708 6. When Exim created an email address using the login of the caller as the
3709 local part (e.g. when creating a From: or Sender: header line), it was not
3710 quoting the local part when it contained special characters such as @.
3712 7. Installed new OpenBSD configuration files.
3714 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3715 try to make them clearer.
3717 9. Callout options, other than the timeout value, were being ignored when
3718 verifying sender addresses in header lines. For example, when using
3720 verify = header_sender/callout=no_cache
3722 the cache was (incorrectly) being used.
3724 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3725 only the Cygwin environment.
3727 11. When return_path_on_delivery was set as a log selector, if different remote
3728 addresses in the same message used different return paths and parallel
3729 remote delivery occurred, the wrong values would sometimes be logged.
3730 (Whenever a remote delivery process finished, the return path value from
3731 the most recently started remote delivery process was logged.)
3733 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3734 lines when AUTH and/or TLS are in use. This is the "received protocol"
3735 field. Exim used to use "asmtp" for authenticated SMTP, without any
3736 indication (in the protocol name) for TLS use. Now it follows the RFC and
3737 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3738 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3739 names appear in log lines as well as in Received: header lines.
3741 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3742 extensions, and comparison tests, and to fix some bugs.
3744 14. Changes to the "personal" filter test:
3746 (1) The test was buggy in that it was just doing the equivalent of
3747 "contains" tests on header lines. For example, if a user's address was
3748 anne@some.where, the "personal" test would incorrectly be true for
3750 To: susanne@some.where
3752 This test is now done by extracting each address from the header in turn,
3753 and checking the entire address. Other tests that are part of "personal"
3754 are now done using regular expressions (for example, to check local parts
3755 of addresses in From: header lines).
3757 (2) The list of non-personal local parts in From: addresses has been
3758 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3759 taken from the Sieve specification recommendations.
3761 (3) If the message contains any header line starting with "List-" it is
3762 treated as non-personal.
3764 (4) The test for "circular" in the Subject: header line has been removed
3765 because it now seems ill-conceived.
3767 15. Minor typos in src/EDITME comments corrected.
3769 16. Installed latest exipick from John Jetmore.
3771 17. If headers_add on a router specified a text string that was too long for
3772 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3773 of string_sprintf() is now avoided.
3775 18. $message_body_size was not set (it was always zero) when running the DATA
3776 ACL and the local_scan() function.
3778 19. For the "mail" command in an Exim filter, no default was being set for
3779 the once_repeat time, causing a random time value to be used if "once" was
3780 specified. (If the value happened to be <= 0, no repeat happened.) The
3781 default is now 0s, meaning "never repeat". The "vacation" command was OK
3782 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3783 (I found it when inspecting the code).
3785 20. There is now an overall timeout for performing a callout verification. It
3786 defaults to 4 times the callout timeout, which applies to individual SMTP
3787 commands during the callout. The overall timeout applies when there is more
3788 than one host that can be tried. The timeout is checked before trying the
3789 next host. This prevents very long delays if there are a large number of
3790 hosts and all are timing out (e.g. when the network connections are timing
3791 out). The value of the overall timeout can be changed by specifying an
3792 additional sub-option for "callout", called "maxwait". For example:
3794 verify = sender/callout=5s,maxwait=20s
3796 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3797 to the end before writing, but this should make it even safer).
3799 22. Exim was forgetting that it had advertised PIPELINING for the second and
3800 subsequent messages on an SMTP connection. It was also not resetting its
3801 memory on STARTTLS and an internal HELO.
3803 23. When Exim logs an SMTP synchronization error within a session, it now
3804 records whether PIPELINING has been advertised or not.
3806 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3807 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3808 rather than long int.
3810 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3812 26. Added the never_mail option to autoreply.
3818 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3819 crash if the getsockname() call failed; this can happen if a connection is
3820 closed very soon after it is established. The problem was simply in the
3821 order in which certain operations were done, causing Exim to try to write
3822 to the SMTP stream before it had set up the file descriptor. The bug has
3823 been fixed by making things happen in the correct order.
3829 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3830 before the connection was closed, thus losing the rejection response.
3832 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3833 some early Solaris releases, but causes trouble in current releases where
3834 socklen_t is defined.
3836 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3839 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3842 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3843 that releasing the top of it at the end releases what was used for sub-
3844 expansions (unless the block got too big). However, discard this block if
3845 the first thing is a variable or header, so that we can use its block when
3846 it is dynamic (useful for very large $message_headers, for example).
3848 6. Lookups now cache *every* query, not just the most recent. A new, separate
3849 store pool is used for this. It can be recovered when all lookup caches are
3850 flushed. Lookups now release memory at the end of their result strings.
3851 This has involved some general refactoring of the lookup sources.
3853 7. Some code has been added to the store_xxx() functions to reduce the amount
3854 of flapping under certain conditions.
3856 8. log_incoming_interface used to affect only the <= reception log lines. Now
3857 it causes the local interface and port to be added to several more SMTP log
3858 lines, for example "SMTP connection from", and rejection lines.
3860 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3862 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3864 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3865 could be overwritten at the end of the current message (or the start of a
3866 new message if it was set in a HELO ACL). The value is now preserved for
3867 the duration of the SMTP connection.
3869 12. If a transport had a headers_rewrite setting, and a matching header line
3870 contained an unqualified address, that address was qualified, even if it
3871 did not match any rewriting rules. The underlying bug was that the values
3872 of the flags that permit the existence of unqualified sender and recipient
3873 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3874 non-local messages, and by -bnq for local messages) were not being
3875 preserved with the message after it was received.
3877 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3878 "next input=" as part of the text comprising the host identity instead of
3879 the correct text. The code was using the same buffer for two different
3880 strings. However, depending on which order the printing function evaluated
3881 its arguments, the bug did not always show up. Under Linux, for example, my
3882 test suite worked just fine.
3884 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3885 doesn't work with some older versions of Perl. It has been changed to "my",
3886 which in any case is probably the better facility to use.
3888 15. A really picky compiler found some instances of statements for creating
3889 error messages that either had too many or two few arguments for the format
3892 16. The size of the buffer for calls to the DNS resolver has been increased
3893 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3894 for addresses that have a lot of PTR records. This alleviates a problem; it
3895 does not fully solve it.
3897 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3898 buffer now truncates the list and logs the incident, which is the same
3899 action as happens when Exim is looking up a host name and its aliases.
3900 Previously in this situation something unpredictable would happen;
3901 sometimes it was "internal error: store_reset failed".
3903 18. If a server dropped the connection unexpectedly when an Exim client was
3904 using GnuTLS and trying to read a response, the client delivery process
3905 crashed while trying to generate an error log message.
3907 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3908 string, for example:
3910 warn message = H1: something\nH2: something
3912 the text was added as a single header line from Exim's point of view
3913 though it ended up OK in the delivered message. However, searching for the
3914 second and subsequent header lines using $h_h2: did not work. This has been
3915 fixed. Similarly, if a system filter added multiple headers in this way,
3916 the routers could not see them.
3918 20. Expanded the error message when iplsearch is called with an invalid key to
3919 suggest using net-iplsearch in a host list.
3921 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3922 ACLs are no longer actually imposed (and a message to that effect is
3925 22. If a "gecos" field in a passwd entry contained escaped characters, in
3926 particular, if it contained a \" sequence, Exim got it wrong when building
3927 a From: or a Sender: header from that name. A second bug also caused
3928 incorrect handling when an unquoted " was present following a character
3929 that needed quoting.
3931 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3932 was not being matched caselessly.
3934 24. Arranged for all hyphens in the exim.8 source to be escaped with
3937 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3938 properly. Recipient callout cache records were still being keyed to include
3939 the sender, even when use_sender was set false. This led to far more
3940 callouts that were necessary. The sender is no longer included in the key
3941 when use_sender is false.
3943 26. Added "control = submission" modifier to ACLs.
3945 27. Added the ${base62d: operator to decode base 62 numbers.
3947 28. dnsdb lookups can now access SRV records.
3949 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3950 the configuration file.
3952 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3953 (-v) mode. This makes the output for a verbose queue run more intelligible.
3955 31. Added a use_postmaster feature to recipient callouts.
3957 32. Added the $body_zerocount variable, containing the number of binary zero
3958 bytes in the message body.
3960 33. The time of last modification of the "new" subdirectory is now used as the
3961 "mailbox time last read" when there is a quota error for a maildir
3964 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3966 35. Added +ignore_unknown as a special item in host lists.
3968 36. Code for decoding IPv6 addresses in host lists is now included, even if
3969 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3970 address was recognized as an IP address, but was then not correctly decoded
3971 into binary, causing unexpected and incorrect effects when compared with
3978 1. Very minor rewording of debugging text in manualroute to say "list of
3979 hosts" instead of "hostlist".
3981 2. If verify=header_syntax was set, and a header line with an unqualified
3982 address (no domain) and a large number of spaces between the end of the
3983 name and the colon was received, the reception process suffered a buffer
3984 overflow, and (when I tested it) crashed. This was caused by some obsolete
3985 code that should have been removed. The fix is to remove it!
3987 3. When running in the test harness, delay a bit after writing a bounce
3988 message to get a bit more predictability in the log output.
3990 4. Added a call to search_tidyup() just before forking a reception process. In
3991 theory, someone could use a lookup in the expansion of smtp_accept_max_
3992 per_host which, without the tidyup, could leave open a database connection.
3994 5. Added the variables $recipient_data and $sender_data which get set from a
3995 lookup success in an ACL "recipients" or "senders" condition, or a router
3996 "senders" option, similar to $domain_data and $local_part_data.
3998 6. Moved the writing of debug_print from before to after the "senders" test
4001 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4002 problems for message scanning, either using a data ACL, or using
4003 local_scan() because the Received: header was not generated till after they
4004 were called (in order to set the time as the time of reception completion).
4005 I have revised the way this works. The header is now generated after the
4006 body is received, but before the ACL or local_scan() are called. After they
4007 are run, the timestamp in the header is updated.
4013 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4014 before starting a queue runner without re-exec. This happened only when
4015 deliver_drop_privilege was set or when the Exim user was set to root. The
4016 effect of the bug was that timeouts during subsequent deliveries caused
4017 crashes instead of being properly handled. The handler is now left at its
4018 default (and expected) setting.
4020 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4021 message, again when deliver_drop_privilege is set or Exim is run as root.
4022 The bug described in (1) was not present in this case, but the tidying up
4023 of the other signals was missing. I have made the two cases consistent.
4025 3. The ignore_target_hosts setting on a manualroute router was being ignored
4026 for hosts that were looked up using the /MX notation.
4028 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4031 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4032 operated on the sender address. After changing the $sender_address to <>
4033 for the sender address verify, Exim was re-instated it as the original
4034 (before rewriting) address, but remembering that it had rewritten it, so it
4035 wasn't rewriting it again. This bug also had the effect of breaking the
4036 sender address verification caching when the sender address was rewritten.
4038 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4039 This has been changed so that if the ip literal address matches
4040 ignore_target_hosts, the router declines.
4042 7. Added expansion conditions match_domain, match_address, and match_local_
4043 part (NOT match_host).
4045 8. The placeholder for the Received: header didn't have a length field set.
4047 9. Added code to Exim itself and to exim_lock to test for a specific race
4048 condition that could lead to file corruption when using MBX delivery. The
4049 issue is with the lockfile that is created in /tmp. If this file is removed
4050 after a process has opened it but before that process has acquired a lock,
4051 there is the potential for a second process to recreate the file and also
4052 acquire a lock. This could lead to two Exim processes writing to the file
4053 at the same time. The added code performs the same test as UW imapd; it
4054 checks after acquiring the lock that its file descriptor still refers to
4055 the same named file.
4057 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4058 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4061 11. Added the smtp_active_hostname option. If used, this will typically be made
4062 to depend on the incoming interface address. Because $interface_address is
4063 not set up until the daemon has forked a reception process, error responses
4064 that can happen earlier (such as "too many connections") no longer contain
4067 12. If an expansion in a condition on a "warn" statement fails because a lookup
4068 defers, the "warn" statement is abandoned, and the next ACL statement is
4069 processed. Previously this caused the whole ACL to be aborted.
4071 13. Added the iplsearch lookup type.
4073 14. Added ident_timeout as a log selector.
4075 15. Added tls_certificate_verified as a log selector.
4077 16. Added a global option tls_require_ciphers (compare the smtp transport
4078 option of the same name). This controls incoming TLS connections.
4080 17. I finally figured out how to make tls_require_ciphers do a similar thing
4081 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4082 before starting the TLS session.
4084 18. Tabs are now shown as \t in -bP output.
4086 19. If the log selector return_path_on_delivery was set, Exim crashed when
4087 bouncing a message because it had too many Received: header lines.
4089 20. If two routers both had headers_remove settings, and the first one included
4090 a superfluous trailing colon, the final name in the first list and the
4091 first name in the second list were incorrectly joined into one item (with a
4092 colon in the middle).
4098 1. Added -C and -D options to the exinext utility, mainly to make it easier
4099 to include in the automated testing, but these could be helpful when
4100 multiple configurations are in use.
4102 2. The exinext utility was not formatting the output nicely when there was
4103 an alternate port involved in the retry record key, nor when there was a
4104 message id as well (for retries that were specific to a specific message
4105 and a specific host). It was also confused by IPv6 addresses, because of
4106 the additional colons they contain. I have fixed the IPv4 problem, and
4107 patched it up to do a reasonable job for IPv6.
4109 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4110 delivery, the log line now contains "pipelined" if PIPELINING was used.
4112 4. An SMTP transport process used to panic and die if the bind() call to set
4113 an explicit outgoing interface failed. This has been changed; it is now
4114 treated in the same way as a connect() failure.
4116 5. A reference to $sender_host_name in the part of a conditional expansion
4117 that was being skipped was still causing a DNS lookup. This no longer
4120 6. The def: expansion condition was not recognizing references to header lines
4121 that used bh_ and bheader_.
4123 7. Added the _cache feature to named lists.
4125 8. The code for checking quota_filecount in the appendfile transport was
4126 allowing one more file than it should have been.
4128 9. For compatibility with Sendmail, the command line option
4136 and sets the incoming protocol and host name (for trusted callers). The
4137 host name and its colon can be omitted when only the protocol is to be set.
4138 Note the Exim already has two private options, -pd and -ps, that refer to
4139 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4140 "s", but I don't think that's a major issue.
4142 10. A number of refactoring changes to the code, none of which should affect
4145 (a) The number of logging options was getting close to filling up the
4146 32-bit word that was used as a bit map. I have split them into two classes:
4147 those that are passed in the argument to log_write(), and those that are
4148 only ever tested independently outside of that function. These are now in
4149 separate 32-bit words, so there is plenty of room for expansion again.
4150 There is no change in the user interface or the logging behaviour.
4152 (b) When building, for example, log lines, the code previously used a
4153 macro that called string_cat() twice, in order to add two strings. This is
4154 not really sufficiently general. Furthermore, there was one instance where
4155 it was actually wrong because one of the argument was used twice, and in
4156 one call a function was used. (As it happened, calling the function twice
4157 did not affect the overall behaviour.) The macro has been replaced by a
4158 function that can join an arbitrary number of extra strings onto a growing
4161 (c) The code for expansion conditions now uses a table and a binary chop
4162 instead of a serial search (which was left over from when there were very
4163 few conditions). Also, it now recognizes conditions like "pam" even when
4164 the relevant support is not compiled in: a suitably worded error message is
4165 given if an attempt is made to use such a condition.
4167 11. Added ${time_interval:xxxxx}.
4169 12. A bug was causing one of the ddress fields not to be passed back correctly
4170 from remote delivery subprocesses. The field in question was not being
4171 subsequently used, so this caused to problems in practice.
4173 13. Added new log selectors queue_time and deliver_time.
4175 14. Might have fixed a bug in maildirsizefile handling that threw up
4176 "unexpected character" debug warnings, and recalculated the data
4177 unnecessarily. In any case, I expanded the warning message to give more
4180 15. Added the message "Restricted characters in address" to the statements in
4181 the default ACL that block characters like @ and % in local parts.
4183 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4184 Three changes have been made:
4186 (a) There was a serious bug; a negative response to MAIL caused the whole
4187 recipient domain to be cached as invalid, thereby blocking all messages
4188 to all local parts at the same domain, from all senders. This bug has
4189 been fixed. The domain is no longer cached after a negative response to
4190 MAIL if the sender used is not empty.
4192 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4195 (c) A new callout option, "use_sender" has been added for people who want
4196 the modified behaviour.
4202 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4205 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4206 indeed breaks things for older releases.
4208 3. Added additional logging to the case where there is a problem reading data
4209 from a filter that is running in a subprocess using a pipe, in order to
4210 try to track down a specific problem.
4212 4. Testing facility fudge: when running in the test harness and attempting
4213 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4214 getting "No route to host". Convert this to a timeout.
4216 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4219 6. Some OS don't have socklen_t but use size_t instead. This affects the
4220 fifth argument of getsockopt() amongst other things. This is now
4221 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4222 can be set for individual OS. I have set it for SunOS5, OSF1, and
4223 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4224 some earlier ones do not.
4226 7. Change 4.30/15 was not doing the test caselessly.
4228 8. The standard form for an IPv6 address literal was being rejected by address
4229 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4230 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4231 this, as well as the form without the "IPv6" on the front (but only when
4232 address literals are enabled, of course).
4234 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4236 10. Exim crashed if a message with an empty sender address specified by -f
4237 encountered a router with an errors_to setting. This could be provoked only
4238 by a command such as
4242 where an empty string was supplied; "<>" did not hit this bug.
4244 11. Installed PCRE release 4.5.
4246 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4247 remained set. It is now erased.
4249 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4250 times from message ids (which are base 36 rather than the normal 62).
4252 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4253 were being counted as actual protocol errors, and logged if the log
4254 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4255 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4256 and DATA following a set of rejected RCPTs do not count as protocol errors.
4257 In other words, Exim assumes they were pipelined, though this may not
4258 actually be the case. Of course, in all cases the client gets an
4259 appropriate error code.
4261 15. If a lookup fails in an ACL condition, a message about the failure may
4262 be available; it is used if testing the ACL cannot continue, because most
4263 such messages specify what the cause of the deferral is. However, some
4264 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4265 that caused an old message to be retained and used if a later statement
4266 caused a defer, replacing the real cause of the deferral.
4268 16. If an IP address had so many PTR records that the DNS lookup buffer
4269 was not large enough to hold them, Exim could crash while trying to process
4270 the truncated data. It now detects and logs this case.
4272 17. Further to 4.21/58, another change has been made: if (and only if) the
4273 first line of a message (the first header line) ends with CRLF, a bare LF
4274 in a subsequent header line has a space inserted after it, so as not to
4275 terminate the header.
4277 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4278 unnecessarily, used atoi() instead of strtol(), and didn't check the
4279 termination when getting file sizes from file names by regex.
4281 19. Completely re-implemented the support for maildirsize files, in the light
4282 of a number of problems with the previous contributed implementation
4283 (4.30/29). In particular:
4285 . If the quota is zero, the maildirsize file is maintained, but no quota is
4288 . If the maildir directory does not exist, it is created before any attempt
4289 to write a maildirsize file.
4291 . The quota value in the file is just a cache; if the quota is changed in
4292 the transport, the new value overrides.
4294 . A regular expression is available for excluding directories from the
4297 20. The autoreply transport checks the characters in options that define the
4298 message's headers; it allows continued headers, but it was checking with
4299 isspace() after an embedded newline instead of explicitly looking for a
4302 21. If all the "regular" hosts to which an address was routed had passed their
4303 expiry times, and had not reached their retry times, the address was
4304 bounced, even if fallback hosts were defined. Now Exim should go on to try
4307 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4308 equivalent code in the SMTP transport. Some hosts send humungous responses
4309 to HELO/EHLO, more than 1024 it seems.
4311 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4312 gives compiler warnings in some environments. I've now done it "properly",
4315 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4316 (because of problems with the built-in one) was declared to return uschar *
4317 instead of char *, causing compiler failure.
4319 25. Fixed a file descriptor leak when processing alias/forward files.
4321 26. Fixed a minor format string issue in dbfn.c.
4323 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4325 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4326 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4329 29. When a sender address is verified, it is cached, to save repeating the test
4330 when there is more than one recipient in a message. However, when the
4331 verification involves a callout, it is possible for different callout
4332 options to be set for different recipients. It is too complicated to keep
4333 track of this in the cache, so now Exim always runs a verification when a
4334 callout is required, relying on the callout cache for the optimization.
4335 The overhead is duplication of the address routing, but this should not be
4338 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4339 to be verified with callout=postmaster, and the main callout worked but the
4340 postmaster check failed, the verification correctly failed. However, if a
4341 subsequent RCPT command asked for sender verification *without* the
4342 postmaster check, incorrect caching caused this verification also to fail,
4345 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4346 it was not caching the DNS options (qualify_single, search_parents) that
4347 were used when the lookup failed. A subsequent lookup with different
4348 options therefore always gave the same answer, though there were cases
4349 where it should not have. (Example: a "domains = !$mx_any" option on a
4350 dnslookup router: the "domains" option is always processed without any
4351 widening, but the router might have qualify_single set.) Now Exim uses the
4352 cached value only when the same options are set.
4354 32. Added John Jetmore's "exipick" utility to the distribution.
4356 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4357 than a timeout (e.g. a certificate is required, and is not provided), an
4358 Exim server now closes the connection immediately. Previously it waited for
4359 the client to close - but if the client is SSL, it seems that they each
4360 wait for each other, leading to a delay before one of them times out.
4362 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4363 maintained 0.8.x compatibility because I don't think many are using it, and
4364 it is clearly obsolete.
4366 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4369 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4370 client certificate was expired. A simple patch fixes this, though I don't
4371 understand the full logic of why the verify callback is called multiple
4374 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4375 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4376 which causes problems with some clients (such as the Certicom SSL Plus
4377 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4378 disables the coutermeasure allowing Eudora to connect."
4380 38. Exim was not checking that a write() to a log file succeeded. This could
4381 lead to Bad Things if a log got too big, in particular if it hit a file
4382 size limit. Exim now panics and dies if it cannot write to a log file, just
4383 as it does if it cannot open a log file.
4385 39. Modified OS/Makefile-Linux so that it now contains
4387 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4389 The two -D definitions ensure that Exim is compiled with large file
4390 support, which makes it possible to handle log files that are bigger than
4393 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4394 instance) a domain was checked against a named list that involved a lookup,
4395 causing $domain_data to be set, then another domain was checked against the
4396 same list, then the first domain was re-checked, the value of $domain_data
4397 after the final check could be wrong. In particular, if the second check
4398 failed, it could be set empty. This bug probably also applied to
4401 41. The strip_trailing_dot option was not being applied to the address given
4402 with the -f command-line option.
4404 42. The code for reading a message's header from the spool was incrementing
4405 $received_count, but never initializing it. This meant that the value was
4406 incorrect (doubled) while delivering a message in the same process in which
4407 it was received. In the most common configuration of Exim, this never
4408 happens - a fresh exec is done - but it can happen when
4409 deliver_drop_privilege is set.
4411 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4412 it now includes up to 150 characters of the unexpected data in the log
4415 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4416 and building data strings. The size of both of these buffers was 10 000
4417 bytes - far larger than anybody would *ever* want, thought I. Needless to
4418 say, somebody hit the limit. I have increased the maximum line length to
4419 20 000 and the maximum data length of concatenated lines to 100 000. I have
4420 also fixed two bugs, because there was no checking on these buffers. Tsk,
4421 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4422 buffer is too small.
4424 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4425 lsearch lookups. Now it does.
4427 46. When parsing a route_list item in a manualroute router, a fixed-length
4428 buffer was used for the list of hosts. I made this 1024 bytes long,
4429 thinking that nobody would ever have a list of hosts that long. Wrong.
4430 Somebody had a whole pile of complicated expansion conditions, and the
4431 string was silently truncated, leading to an expansion error. It turns out
4432 that it is easier to change to an unlimited length (owing to other changes
4433 that have happened since this code was originally written) than to build
4434 structure for giving a limitation error. The length of the item that
4435 expands into the list of hosts is now unlimited.
4437 47. The lsearch lookup could not handle data where the length of text line was
4438 more than 4095 characters. Such lines were truncated, leading to shortened
4439 data being returned. It should now handle lines of any length.
4441 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4442 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4445 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4446 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4447 no longer happens in comment lines. A list of replacements is now placed
4448 at the head of all of the source files, except those whose only change is
4449 to replace PERL_COMMAND in the very first #! line.
4451 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4452 messages on the queue, with a bottom-up merge sort, using code contributed
4453 by Michael Haardt. This should make operations like -bp somewhat faster on
4454 large queues. It won't affect queue runners, except when queue_run_in_order
4457 51. Installed eximstats 1.31 in the distribution.
4459 52. Added support for SRV lookups to the dnslookup router.
4461 53. If an ACL referred to $message_body or $message_body_end, the value was not
4462 reset for any messages that followed in the same SMTP session.
4464 54. The store-handling optimization for building very long strings was not
4465 differentiating between the different store pools. I don't think this
4466 actually made any difference in practice, but I've tidied it.
4468 55. While running the routers to verify a sender address, $sender_address
4469 was still set to the sender address. This is wrong, because when routing to
4470 send a bounce to the sender, it would be empty. Therefore, I have changed
4471 it so that, while verifying a sender address, $sender_address is set to <>.
4472 (There is no change to what happens when verifying a recipient address.)
4474 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4475 target A or AAAA records (if not already returned) without resetting the
4476 qualify_single or search_parents options of the DNS resolver. These are
4477 inappropriate in this case because the targets of MX and SRV records must
4478 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4479 target that, when qualified, matched something in the local domain. These
4480 two options are now turned off when doing these lookups.
4482 57. It seems that at least some releases of Reiserfs (which does not have the
4483 concept of a fixed number of inodes) returns zero and not -1 for the
4484 number of available inodes. This interacted badly with check_spool_inodes,
4485 which assumed that -1 was the "no such thing" setting. What I have done is
4486 to check that the total number of inodes is greater than zero before doing
4487 the test of how many are available.
4489 58. When a "warn" ACL statement has a log_message modifier, the message is
4490 remembered, and not repeated. This is to avoid a lot of repetition when a
4491 message has many recipients that cause the same warning to be written.
4492 Howewer, Exim was preserving the list of already written lines for an
4493 entire SMTP session, which doesn't seem right. The memory is now reset if a
4494 new message is started.
4496 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4497 debugging output unless log_rewrite was also set.
4499 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4500 of (int)(handle) when we know that handle contains (void *)(-1).
4502 61. The Exim daemon panic-logs an error return when it closes the incoming
4503 connection. However "connection reset by peer" seems to be common, and
4504 isn't really an error worthy of noting specially, so that particular error
4507 62. When Exim is trying to find all the local interfaces, it used to panic and
4508 die if the ioctl to get the interface flags failed. However, it seems that
4509 on at least one OS (Solaris 9) it is possible to have an interface that is
4510 included in the list of interfaces, but for which you get a failure error
4511 for this call. This happens when the interface is not "plumbed" into a
4512 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4513 failure of the "get flags" call assumes that the interface is down.
4515 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4516 makes life easier for people who are doing arithmetic on fields extracted
4517 from dates, where you often get leading zeros that should not be
4518 interpreted as octal.
4520 64. Added qualify_domain to the redirect router, to override the global
4523 65. If a pathologically long header line contained very many addresses (the
4524 report of this problem mentioned 10 000) and each of them was rewritten,
4525 Exim could use up a very large amount of memory. (It kept on making new
4526 copies of the header line as it rewrote, and never released the old ones.)
4527 At the expense of a bit more processing, the header rewriting function has
4528 been changed so that it no longer eats memory in this way.
4530 66. The generation of the Received: header has been moved from the time that a
4531 message starts to be received, to the time that it finishes. The timestamp
4532 in the Received: header should now be very close to that of the <= log
4533 line. There are two side-effects of this change:
4535 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4536 logged header lines no longer include the local Received: line, because
4537 it has not yet been created. The same applies to a copy of the message
4538 that is returned to a non-SMTP sender when a message is rejected.
4540 (b) When a filter file is tested using -bf, no additional Received: header
4541 is added to the test message. After some thought, I decided that this
4544 This change does not affect the value of $received_for. It is still set
4545 after address rewriting, but before local_scan() is called.
4547 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4549 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4550 gave an unhelpful panic error message, and a defer error. I have managed to
4551 change this behaviour so that it now rejects any supplied certificate,
4552 which seems right, as the list of acceptable certificates is empty.
4554 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4555 gave an unhelpful defer error. I have not managed to make this reject any
4556 supplied certificates, but the error message it gives is "no certificate
4557 supplied", which is not helpful.
4559 70. exigrep's output now also includes lines that are not associated with any
4560 message, but which match the given pattern. Implemented by a patch from
4561 Martin Sluka, which also tidied up the Perl a bit.
4563 71. Recipient callout verification, like sender verification, was using <> in
4564 the MAIL FROM command. This isn't really the right thing, since the actual
4565 sender may affect whether the remote host accepts the recipient or not. I
4566 have changed it to use the actual sender in the callout; this means that
4567 the cache record is now keyed on a recipient/sender pair, not just the
4568 recipient address. There doesn't seem to be a real danger of callout loops,
4569 since a callout by the remote host to check the sender would use <>.
4570 [SEE ABOVE: changed after hitting problems.]
4572 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4573 temporary errors. However, in the case of such a code being given after
4574 the end of a data transmission (i.e. after ".") Exim was failing to write
4575 a retry record for the message. (Yes, there was some broken host that was
4576 actually sending 8xx at this point.)
4578 73. An unknown lookup type in a host list could cause Exim to panic-die when
4579 the list was checked. (An example that provoked this was putting <; in the
4580 middle of a list instead of at the start.) If this happened during a DATA
4581 ACL check, a -D file could be left lying around. This kind of configuration
4582 error no longer causes Exim to die; instead it causes a defer errror. The
4583 incident is still logged to the main and panic logs.
4585 74. Buglet left over from Exim 3 conversion. The message "too many messages
4586 in one connection" was written to the rejectlog but not the mainlog, except
4587 when address rewriting (yes!) was being logged.
4589 75. Added write_rejectlog option.
4591 76. When a system filter was run not as root (that is, when system_filter_user
4592 was set), the values of the $n variables were not being returned to the
4593 main process; thus, they were not subsequently available in the $sn
4596 77. Added +return_path_on_delivery log selector.
4598 78. A connection timeout was being treated differently from recipients deferred
4599 when testing hosts_max_try with a message that was older than the host's
4600 retry timeout. (The host should not be counted, thus allowing all hosts to
4601 be tried at least once before bouncing.) This may have been the cause of an
4602 occasionally reported bug whereby a message would remain on the queue
4603 longer than the retry timeout, but would be bounced if a delivery was
4604 forced. I say "may" because I never totally pinned down the problem;
4605 setting up timeout/retry tests is difficult. See also the next item.
4607 79. The ultimate address timeout was not being applied to errors that involved
4608 a combination of host plus message (for example, a timeout on a MAIL
4609 command). When an address resolved to a number of possible hosts, and they
4610 were not all tried for each delivery (e.g. because of hosts_max_try), a
4611 message could remain on the queue longer than the retry timeout.
4613 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4616 81. Fixed an obscure SMTP outgoing bug which required at least the following
4617 conditions: (a) there was another message waiting for the same server;
4618 (b) the server returned 5xx to all RCPT commands in the first message so
4619 that the message was not completed; (c) the server dropped the connection
4620 or gave a negative response to the RSET that Exim sends to abort the
4621 transaction. The observed case was a dropped connection after DATA that had
4622 been sent in pipelining mode. That is, the server had advertised PIPELINING
4623 but was not implementing it correctly. The effect of the bug was incorrect
4624 behaviour, such as trying another host, and this could lead to a crash.
4630 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4631 and daemon.c were passed as pointers to ints; they should have been
4632 pointers to socklen_t variables (which are typically unsigned ints).
4634 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4637 3. Fixed a really odd bug that affected only the testing scheme; patching a
4638 certain fixed string in the binary changed the value of another string that
4639 happened to be identical to the end of the original first string.
4641 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4642 name", it returns that address as the IP address. On some operating
4643 systems (e.g. Solaris), it also passes back the IP address string as the
4644 "host name". However, on others (e.g. Linux), it passes back an empty
4645 string. Exim wasn't checking for this, and was changing the host name to an
4646 empty string, assuming it had been canonicized.
4648 5. Although rare, it is permitted to have more than one PTR record for a given
4649 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4650 all the names associated with an address, because they do in Solaris.
4651 However, it seems that they do not in Linux for data that comes from the
4652 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4653 I found this out when I moved to a new Linux workstation and tried to run
4654 the Exim test suite.
4656 To get round this problem I have changed the code so that it now does its
4657 own call to the DNS to look up PTR records when searching for a host name.
4658 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4659 addresses that are only in /etc/hosts are still found.
4661 This behaviour is, however, controlled by an option called host_lookup_
4662 order, which defaults to "bydns:byaddr". If people want to use the other
4663 order, or indeed, just use one or the other means of lookup, they can
4664 specify it in this variable.
4666 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4667 some operating systems, this comes back from gethostbyaddr() as an empty
4668 string, and this is what Exim used to test for. However, it seems that in
4669 other systems, "." is yielded. Exim now tests for this case too.
4671 7. The values of check_spool_space and check_log_space are now held internally
4672 as a number of kilobytes instead of an absolute number of bytes. If a
4673 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4674 kilobyte. This means that much larger values can be stored.
4676 8. Exim monitor: an attempt to get the action menu when not actually pointing
4677 at a message produces an empty menu entitled "No message selected". This
4678 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4679 no entries in it ("Shell widget menu has zero width and/or height"). So I
4680 have added a single, blank menu entry in this case.
4682 9. Added ${quote_local_part.
4684 10. MIME decoding is now applied to the contents of Subject: header lines when
4687 11. Now that a reference to $sender_host_address automatically causes a reverse
4688 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4689 host lookup before query-style lookups in lists that might use this
4690 variable. This has therefore been abolished, and the "net-" prefix is no
4691 longer necessary for query-style lookups.
4693 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4694 have been a typo for LFLAGS, so it has been changed.
4696 13. The install script calls Exim with "-C /dev/null" in order to find the
4697 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4698 to be output. Howeve, since Exim outputs its version number before the
4699 error, it didn't break the script. It just looked ugly. I fixed this by
4700 always allowing "-C /dev/null" if the caller is root.
4702 14. Ignore overlarge ACL variable number when reading spool file - insurance
4703 against a later release with more variables having written the file.
4705 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4706 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4707 this, as well as the form without the "IPv6" on the front.
4709 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4710 OS/Makefile-Darwin file.
4712 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4713 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4715 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4716 variable. However, some OS use ints for inodes. I've added cast to long int
4717 to get rid of the compiler warning.
4719 19. I had forgotten to lock out "/../" in configuration file names when
4720 ALT_CONFIG_PREFIX was set.
4722 20. Routers used for verification do not need to specify transports. However,
4723 if such a router generated a host list, and callout was configured, Exim
4724 crashed, because it could not find a port number from the (non-existent)
4725 transport. It now assumes port 25 in this circumstance.
4727 21. Added the -t option to exigrep.
4729 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4730 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4731 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4732 them is set, LOOKUP_LSEARCH is forced.
4734 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4736 24. Added sender and host information to the "rejected by local_scan()" log
4737 line; previously there was no indication of these.
4739 25. Added .include_if_exists.
4741 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4742 Linux. It turns out that not all file systems support this. Apparently some
4743 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4744 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4745 not supported on the file descriptor, is now ignored when Exim is trying to
4746 sync a directory. This applies only to Linux.
4748 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4750 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4751 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4754 29. Added support for maildirsize files from supplied patch (modified a bit).
4756 30. The use of :fail: followed by an empty string could lead Exim to respond to
4757 sender verification failures with (e.g.):
4759 550 Verification failed for <xxx>
4760 550 Sender verify failed
4762 where the first response line was missing the '-' that indicates it is not
4763 the final line of the response.
4765 31. The loop for finding the name of the user that called Exim had a hardwired
4766 limit of 10; it now uses the value of finduser_retries, which is used for
4767 all other user lookups.
4769 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4772 33. Exim was neglecting to zero errno before one call of strtol() when
4773 expanding a string and expecting an integer value. On some systems this
4774 resulted in spurious "integer overflow" errors. Also, it was casting the
4775 result into an int without checking.
4777 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4778 did not work. The code looks as if it has *never* worked, though it appears
4779 to have been documented since at least releast 1.62. I have made it work.
4781 35. The "timeout_DNS" error in retry rules, also documented since at least
4782 1.62, also never worked. As it isn't clear exactly what this means, and
4783 clearly it isn't a major issue, I have abolished the feature by treating it
4784 as "timeout", and writing a warning to the main and panic logs.
4786 36. The display of retry rules for -brt wasn't always showing the error code
4789 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4790 timeout_connect_A, timeout_connect_MX.
4792 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4793 to the empty sender.
4795 39. The daemon was not analysing the content of -oX till after it had closed
4796 stderr and disconnected from the controlling terminal. This meant that any
4797 syntax errors were only noted on the panic log, and the return code from
4798 the command was 0. By re-arranging the code a little, I've made the
4799 decoding happen first, so such errors now appear on stderr, and the return
4800 code is 1. However, the actual setting up of the sockets still happens in
4801 the disconnected process, so errors there are still only recorded on the
4804 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4805 connections (as happens on some IP stacks) was logged at start up time as
4806 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4807 it from "IPv6 and IPv4", which means that two separate sockets are being
4810 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4811 says whether AF_INET or AF_INET6 was passed as an argument.
4813 42. Exiwhat output was messed up when time zones were included in log
4816 43. Exiwhat now gives more information about the daemon's listening ports,
4817 and whether -tls-on-connect was used.
4819 44. The "port" option of the smtp transport is now expanded.
4821 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4822 silently ignored. Now an error message is written to the main and panic
4825 46. There's a new ACL modifier called "logwrite" which writes to a log file
4826 as soon as it is encountered.
4828 47. Added $local_user_uid and $local_user_gid at routing time.
4830 48. Exim crashed when trying to verify a sender address that was being
4833 49. Exim was recognizing only a space character after ".include". It now also
4834 recognizes a tab character.
4836 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4837 extracting the relevant information from the specification. The man page no
4838 longer contains scrambled data for the -d option, and I've added a section
4839 at the front about calling Exim under different names.
4841 51. Added "extra_headers" argument to the "mail" command in filter files.
4843 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4846 53. Installed eximstats 1.29.
4848 54. Added transport_filter_timeout as a generic transport option.
4850 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4851 Cc: header lines. This was required by RFC 822, but it not required by RFC
4854 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4855 incoming messages that did not have them. Now it does so only if the
4856 message originates locally, that is, if there is no associated remote host
4857 address. When Resent- header lines are present, this applies to the Resent-
4858 lines rather than the non-Resent- lines.
4860 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4861 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4863 58. Messages for configuration errors now include the name of the main
4864 configuration file - useful now that there may be more than one file in a
4865 list (.included file names were always shown).
4867 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4868 for those rare installations that do not start the daemon as root or run it
4869 setuid root. I've cut out the call to initgroups() if the daemon is not
4872 60. The Exim user and group can now be bound into the binary as text strings
4873 that are looked up at the start of Exim's processing.
4875 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4877 62. Added $mailstore_basename variable.
4879 63. Installed patch to sieve.c from Michael Haardt.
4881 64. When Exim failed to open the panic log after failing to open the main log,
4882 the original message it was trying to log was written to stderr and debug
4883 output, but if they were not available (the usual case in production), it
4884 was lost. Now it is written to syslog before the two lines that record the
4885 failures to open the logs.
4887 65. Users' Exim filters run in subprocesses under the user's uid. It is
4888 possible for a "deliver" command or an alias in a "personal" command to
4889 provoke an address rewrite. If logging of address rewriting is configured,
4890 this fails because the process is not running as root or exim. There may be
4891 a better way of dealing with this, but for the moment (because 4.30 needs
4892 to be released), I have disabled address rewrite logging when running a
4893 filter in a non-root, non-exim process.
4899 1. The buildconfig auxiliary program wasn't quoting the value set for
4900 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4901 not defined. This bug was masked in 4.22 by the effect that was fixed in
4904 2. Some messages that were rejected after a message id was allocated were
4905 shown as "incomplete" by exigrep. It no longer does this for messages that
4906 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4908 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4909 have allow_domain_literals set, the ID did not get logged in the <= line.
4910 Domain literals are now always recognized in Message-ID: header lines.
4912 4. The first argument for a ${extract expansion item is the key name or field
4913 number. Leading and trailing spaces in this item were not being ignored,
4914 causing some misleading effects.
4916 5. When deliver_drop_privilege was set, single queue runner processes started
4917 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4918 same command in the process it spins off) were not dropping privilege.
4920 6. When the daemon running as "exim" started a queue runner, it always
4921 re-executed Exim in the spun-off process. This is a waste of effort when
4922 deliver_drop_privilege is set. The new process now just calls the
4923 queue-runner function directly.
4929 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4932 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4933 ignored. Though the use of -oP was forcing the writing of a pid file, it
4934 was always written to the default place.
4936 3. If the message "no IP address found for host xxxx" is generated during
4937 incoming verification, it is now followed by identification of the incoming
4938 connection (so you can more easily find what provoked it).
4940 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4942 5. Added some features to "harden" Exim a bit more against certain attacks:
4944 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4945 be put in Local/Makefile. This is like the never_users runtime option,
4946 but it cannot be overridden. The default setting is "root".
4948 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4949 prefix string with which any file named in a -C command line option
4952 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4953 is retained for -C and -D only if the caller of Exim is root. Without
4954 it, the exim user may also use -C and -D and retain privilege.
4956 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4957 command line option is disabled.
4959 6. Macro names set by the -D option must start with an upper case letter, just
4960 like macro names defined in the configuration file.
4962 7. Added "dereference=" facility to LDAP.
4964 8. Two instances of the typo "uknown" in the source files are fixed.
4966 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4967 the Configure-Makefile script screwed up while processing it.
4969 10. Incorporated PCRE 4.4.
4971 11. The SMTP synchronization check was not operating right at the start of an
4972 SMTP session. For example, it could not catch a HELO sent before the client
4973 waited for the greeting. There is now a check for outstanding input at the
4974 point when the greeting is written. Because of the duplex, asynchronous
4975 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4976 way, but not yet received, when the check is performed.
4978 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4979 on TCP/IP sockets, because this apparently causes some broken clients to
4982 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4983 unchanged) from the Cygwin maintainer.
4985 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4986 was supported instead of testing for mbx. Effectively a typo.
4988 15. The spa authenticator server code was not checking that the input it
4989 received was valid base64.
4991 16. The debug output line for the "set" modifier in ACLs was not showing the
4992 name of the variable that was being set.
4994 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4996 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4997 lookup on its own. Something else was needed to trigger the lookup. For
4998 example, a match in host_lookup or the need for a host name in a host list.
4999 Now, if $sender_host_name is referenced and the host name has not yet been
5000 looked up, a lookup is performed. If the lookup fails, the variable remains
5001 empty, and $host_lookup_failed is set to "1".
5003 19. Added "eqi" as a case-independent comparison operator.
5005 20. The saslauthd authentication condition could segfault if neither service
5006 nor realm was specified.
5008 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5009 error message that was logged was misleading, and incoming SMTP
5010 connections were dropped. The message is now more accurate, and temporary
5011 errors are given to SMTP connections.
5013 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5014 (except RSET) with a 421 error, until QUIT is received. However, it was
5015 failing to send a response to QUIT.
5017 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5018 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5019 ACL is now run after the helo_try_verify_hosts code.
5021 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5022 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5023 case-independent, so other case variants are also recognized). Apparently
5024 some systems use these upper case variants.
5026 25. If more than two messages were waiting for the same host, and a transport
5027 filter was specified for the transport, Exim sent two messages over the
5028 same TCP/IP connection, and then failed with "socket operation on non-
5029 socket" when it tried to send the third.
5031 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5033 27. The extern definition of crypt16() in expand.c was not being excluded when
5034 the OS had its own crypt16() function.
5036 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5037 as a preferred synonym for return_size_limit, both as an option and as an
5040 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5042 30. Changed the default configuration ACL to relax the local part checking rule
5043 for addresses that are not in any local domains. For these addresses,
5044 slashes and pipe symbols are allowed within local parts, but the sequence
5045 /../ is explicitly forbidden.
5047 31. SPA server authentication was not clearing the challenge buffer before
5050 32. log_message in a "warn" ACL statement was writing to the reject log as
5051 well as to the main log, which contradicts the documentation and doesn't
5052 seem right (because no rejection is happening). So I have stopped it.
5054 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5055 However, I am unable to do any testing of this.
5057 34. Fixed an infelicity in the appendfile transport. When checking directories
5058 for a mailbox, to see if any needed to be created, it was accidentally
5059 using path names with one or more superfluous leading slashes; tracing
5060 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5062 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5063 contents are added to the log line that is written for every discarded
5064 recipient. (Previously a log_message setting was ignored.)
5066 36. The ${quote: operator now quotes the string if it is empty.
5068 37. The install script runs exim in order to find its version number. If for
5069 some reason other than non-existence or emptiness, which it checks, it
5070 could not run './exim', it was installing it with an empty version number,
5071 i.e. as "exim-". This error state is now caught, and the installation is
5074 38. An argument was missing from the function that creates an error message
5075 when Exim fails to connect to the socket for saslauthd authentication.
5076 This could cause Exim to crash, or give a corrupted message.
5078 39. Added isip, isip4, and isip6 to ${if conditions.
5080 40. The ACL variables $acl_xx are now saved with the message, and can be
5081 accessed later in routers, transports, and filters.
5083 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5084 strings in the file are not string-expanded.
5086 42. If a MAIL command specified a SIZE value that was too large to fit into an
5087 int variable, the check against message_size_limit failed. Such values are
5088 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5089 day this will have to be increased, but I don't think I want to be around
5090 when emails are that large.
5097 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5098 iconv() is not standard in FreeBSD.
5100 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5101 IPv6 enabled. The observed symptom was a segmentation fault on return from
5102 the function os_common_find_running_interfaces() in src/os.c.
5104 3. In the check_special_case() function in daemon.c I had used "errno" as an
5105 argument name, which causes warnings on some systems. This was basically a
5106 typo, since it was named "eno" in the comments!
5108 4. The code that waits for the clock to tick (at a resolution of some fraction
5109 of a second) so as to ensure message-id uniqueness was always waiting for
5110 at least one whole tick, when it could have waited for less. [This is
5111 almost certainly not relevant at current processor speeds, where it is
5112 unlikely to ever wait at all. But we try to future-proof.]
5114 5. The function that sleeps for a time interval that includes fractions of a
5115 second contained a race. It did not block SIGALRM between setting the
5116 timer, and suspending (a couple of lines later). If the interval was short
5117 and the sigsuspend() was delayed until after it had expired, the suspension
5118 never ended. On busy systems this could lead to processes getting stuck for
5121 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5122 process, before it forks any delivery processes. The open lookup caching
5123 mechanism meant that the open file or database connection was passed into
5124 the delivery process. The problem was that delivery processes always tidy
5125 up cached lookup data. This could cause a problem for the next delivery
5126 process started by the queue runner, because the external queue runner
5127 process does not know about the closure. So the next delivery process
5128 still has data in the lookup cache. In the case of a file lookup, there was
5129 no problem because closing a file descriptor in a subprocess doesn't affect
5130 the parent. However, if the lookup was caching a connection to a database,
5131 the connection was closed, and the second delivery process was likely to
5132 see errors such as "PGSQL: query failed: server closed the connection
5133 unexpectedly". The problem has been fixed by closing all cached lookups
5134 in a queue runner before running a delivery process.
5136 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5137 seem to have the "const" qualifier which it has on other OS. I've
5140 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5141 *of the same type* (client or server) with the same public name that an
5142 error should be diagnosed.
5144 9. When Exim looked up a host name for an IP address, but failed to find the
5145 original IP address when looking up the host name (a safety check), it
5146 output the message "<ip address> does not match any IP for NULL", which was
5147 confusing, to say the least. The bug was that the host name should have
5148 appeared instead of "NULL".
5150 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5151 user that is built into the binary, and the -C or -D options is used, root
5152 privilege is dropped before the configuration file is read. In addition,
5153 logging is switched to stderr instead of the normal log files. If the
5154 configuration then re-defines the Exim user, the unprivileged environment
5155 is probably not what is expected, so Exim logs a panic warning message (but
5158 However, if deliver_drop_privilege is set, the unprivileged state may well
5159 be exactly what is intended, so the warning has been cut out in that case,
5160 and Exim is allowed to try to write to its normal log files.
5166 1. smtp_return_error_details was not giving details for temporary sender
5167 or receiver verification errors.
5169 2. Diagnose a configuration error if two authenticators have the same public
5172 3. Exim used not to create the message log file for a message until the first
5173 delivery attempt. This could be confusing when incoming messages were held
5174 for policy or load reasons. The message log file is now created at the time
5175 the message is received, and an initial "Received" line is written to it.
5177 4. The automatically generated man page for command line options had a minor
5178 bug that caused no ill effects; however, a more serious problem was that
5179 the procedure for building the man page automatically didn't always
5180 operate. Consequently, release 4.20 contains an out-of-date version. This
5181 shouldn't happen again.
5183 5. When building Exim with embedded Perl support, the script that builds the
5184 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5185 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5187 6. The freeze_tell option was not being used for messages that were frozen on
5188 arrival, either by an ACL or by local_scan().
5190 7. Added the smtp_incomplete_transaction log selector.
5192 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5193 was accepting AUTH without a new EHLO.
5195 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5196 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5199 10. Logging of TCP/IP connections (when configured) now happens in the main
5200 daemon process instead of the child process, so that the TCP/IP connection
5201 count is more accurate (but it can never be perfect).
5203 11. The use of "drop" in a nested ACL was not being handled correctly in the
5204 outer ACL. Now, if condition failure induced by the nested "drop" causes
5205 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5206 or "require"), the connection is dropped.
5208 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5209 that yield "discard" can now be used with an "accept" or a "discard" verb,
5210 but an error is generated for any others (because I can't see a useful way
5211 to define what should happen).
5213 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5214 are now processed in the same way as lines in the Exim configuration file.
5215 In particular, continuation lines are supported.
5217 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5219 15. Added -ti meaning -t -i.
5221 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5222 domains, and warn by logging if others are found.
5224 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5225 returned from GIFCONF when Exim is trying to find the list of interfaces on
5226 a host. The code in os.c has been modified to copy each ifreq to an aligned
5227 structure in all cases.
5229 Also, in some cases, the returned ifreq's were being copied to a 'struct
5230 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5231 means the last couple of bytes of an IPv6 address could be chopped if the
5232 ifreq contained only a normal sockaddr (14 bytes storage).
5234 18. Named domain lists were not supported in the hosts_treat_as_local option.
5235 An entry such as +xxxx was not recognized, and was treated as a literal
5238 19. Ensure that header lines added by a DATA ACL are included in the reject log
5239 if the ACL subsequently rejects the message.
5241 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5242 MD5 (which is deprecated).
5244 21. When testing a filter file using -bf, Exim was writing a message when it
5245 took the sender from a "From " line in the message, but it was not doing so
5246 when it took $return_path from a Return-Path: header line. It now does.
5248 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5249 with a valid header line field name (a series of printing characters
5250 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5252 23. Changed "disc" in the source to "disk" to conform to the documentation and
5253 the book and for uniformity.
5255 24. Ignore Sendmail's -Ooption=value command line item.
5257 25. When execve() failed while trying to run a command in a pipe transport,
5258 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5259 could be confused with a return value of 69 from the command itself. This
5260 has been changed to 127, the value the shell returns if it is asked to run
5261 a non-existent command. The wording for the related log line suggests a
5262 non-existent command as the problem.
5264 26. If received_header_text expands to an empty string, do not add a Received:
5265 header line to the message. (Well, it adds a token one on the spool, but
5266 marks it "old" so that it doesn't get used or transmitted.)
5268 27. Installed eximstats 1.28 (addition of -nt option).
5270 28. There was no check for failure on the call to getsockname() in the daemon
5271 code. This can fail if there is a shortage of resources on the system, with
5272 ENOMEM, for example. A temporary error is now given on failure.
5274 29. Contrary to the C standard, it seems that in some environments, the
5275 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5276 program. Exim now does this explicitly; it affects the formatting of
5277 timestamps using strftime().
5279 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5280 complaints. I've now initialized all the variables, to avoid this.
5282 32. Header lines added by a system filter were not being "seen" during
5283 transport-time rewrites.
5285 33. The info_callback() function passed to OpenSSL is set up with type void
5286 (*)(SSL *, int, int), as described somewhere. However, when calling the
5287 function (actually a macro) that sets it up, the type void(*)() is
5288 expected. I've put in a cast to prevent warnings from picky compilers.
5290 34. If a DNS black list lookup found a CNAME record, but there were no A
5291 records associated with the domain it pointed at, Exim crashed.
5293 35. If a DNS black list lookup returned more than one A record, Exim ignored
5294 all but the first. It now scans all returned addresses if a particular IP
5295 value is being sought. In this situation, the contents of the
5296 $dnslist_value variable are a list of all the addresses, separated by a
5299 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5300 to accept a host name and all its aliases if the forward lookup for any of
5301 them yielded the IP address of the incoming connection. Now it accepts only
5302 those names whose forward lookup yields the correct IP address. Any other
5303 names are discarded. This closes a loophole whereby a rogue DNS
5304 administrator could create reverse DNS records to break through a
5305 wildcarded host restriction in an ACL.
5307 37. If a user filter or a system filter that ran in a subprocess used any of
5308 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5309 the wrong values were passed to the pipe command ($thisaddress had the
5310 value of $0, $0 had the value of $1, etc). This bug was introduced by
5311 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5313 38. Improved the line breaking for long SMTP error messages from ACLs.
5314 Previously, if there was no break point between 40 and 75 characters, Exim
5315 left the rest of the message alone. Two changes have been made: (a) I've
5316 reduced the minimum length to 35 characters; (b) if it can't find a break
5317 point between 35 and 75 characters, it looks ahead and uses the first one
5318 that it finds. This may give the occasional overlong line, but at least the
5319 remaining text gets split now.
5321 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5322 file descriptors might be low, and that setting 1000 would always raise it.
5323 It turns out that in some environments, the limit is already over 1000 and
5324 that lowering it causes trouble. So now Exim takes care not to decrease it.
5326 40. When delivering a message, the value of $return_path is set to $sender_
5327 address at the start of routing (routers may change the value). By an
5328 oversight, this default was not being set up when an address was tested by
5329 -bt or -bv, which affected the outcome if any router or filter referred to
5332 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5333 log only when "message" or "log_message" are set. However, if one of the
5334 conditions was an address verification, or a call to a nested ACL, the
5335 messages generated by the underlying test were being passed through. This
5336 no longer happens. The underlying message is available in $acl_verify_
5337 message for both "message" and "log_message" expansions, so it can be
5338 passed through if needed.
5340 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5341 new expansion $bh_ to give the encoded byte string without charset
5342 translation. Translation happens only if iconv() is available; HAVE_ICONV
5343 indicates this at build time. HEADERS_CHARSET gives the charset to
5344 translate to; headers_charset can change it in the configuration, and
5345 "headers charset" can change it in an individual filter file.
5347 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5348 that creates RFC 2047 encoded "words" labels them as that charset instead
5349 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5350 expansion operator; (ii) when Exim creates a From: line for a local
5351 message; (iii) when a header line is rewritten to include a "phrase" part.
5353 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5354 buggy, causing it to skip the first lines of messages whose message ID
5355 ended in 'D'. This would not have bitten before Exim release 4.14, because
5356 message IDs were unlikely to end in 'D' before then. The effect was to have
5357 incorrect size information for certain domains.
5359 45. #include "config.h" was missing at the start of the crypt16.c module. This
5360 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5363 46. If there was a timeout during a "random" callout check, Exim treated it as
5364 a failure of the random address, and carried on sending RSET and the real
5365 address. If the delay was just some slowness somewhere, the response to the
5366 original RCPT would be taken as a response to RSET and so on, causing
5367 mayhem of various kinds.
5369 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5370 when I implemented it. It didn't allow for the fact that some option values
5371 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5372 the right test for positive values.
5374 48. Domain names in DNS records are case-independent. Exim always looks them up
5375 in lower case. Some resolvers return domain names in exactly the case they
5376 appear in the zone file, that is, they may contain uppercase letters. Not
5377 all resolvers do this - some return always lower case. Exim was treating a
5378 change of case by a resolver as a change of domain, similar to a widening
5379 of a domain abbreviation. This triggered its re-routing code and so it was
5380 trying to route what was effectively the same domain again. This normally
5381 caused routing to fail (because the router wouldn't handle the domain
5382 twice). Now Exim checks for this case specially, and just changes the
5383 casing of the domain that it ultimately uses when it transmits the message
5386 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5389 50. If a filter generated a file delivery with a non-absolute name (possible if
5390 no home directory exists for the router), the forbid_file option was not
5393 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5394 the existing equality matching.
5396 52. Exim was using ints instead of ino_t variables in some places where it was
5397 dealing with inode numbers.
5399 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5400 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5401 called TMPDIR, and if it finds it is different, it changes its value.
5403 54. The smtp_printf() function is now made available to local_scan() so
5404 additional output lines can be written before returning. There is also an
5405 smtp_fflush() function to enable the detection of a dropped connection.
5406 The variables smtp_input and smtp_batched_input are exported to
5409 55. Changed the default runtime configuration: the message "Unknown user"
5410 has been removed from the ACL, and instead placed on the localuser router,
5411 using the cannot_route_message feature. This means that any verification
5412 failures that generate their own messages won't get overridden. Similarly,
5413 the "Unrouteable address" message that was in the ACL for unverifiable
5414 relay addresses has also been removed.
5416 56. Added hosts_avoid_esmtp to the smtp transport.
5418 57. The exicyclog script was not checking for the esoteric option
5419 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5420 will work only if exicyclog is run under the appropriate euid.
5422 58. Following a discussion on the list, the rules by which Exim recognises line
5423 endings on incoming messages have been changed. The -dropcr and drop_cr
5424 options are now no-ops, retained only for backwards compatibility. The
5425 following line terminators are recognized: LF CRLF CR. However, special
5426 processing applies to CR:
5428 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5429 nor a local message in the state where . is a terminator.
5431 (ii) If a bare CR is encountered in a header line, an extra space is added
5432 after the line terminator so as not to end the header. The reasoning
5433 behind this is that bare CRs in header lines are most likely either
5434 to be mistakes, or people trying to play silly games.
5436 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5437 was being incorrectly given as 18 bytes larger than it should have been.
5438 This is a VOB (very old bug).
5440 60. This may never have affected anything current, but just in case it has:
5441 When the local host is found other than at the start of a list of hosts,
5442 the local host, those with the same MX, and any that follow, are discarded.
5443 When the list in question was part of a longer list of hosts, the following
5444 hosts (not currently being processed) were also being discarded. This no
5445 longer happens. I'm not sure if this situation could ever has previously
5448 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5451 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5452 header. This is something that is recommended in a new Internet Draft, and
5453 is something that is documented as being done by Sendmail. There are two
5454 possible values. For messages generated by the autoreply transport, Exim
5457 Auto-Submitted: auto-replied
5459 whereas for all other generated messages (e.g. bounces) it adds
5461 Auto-Submitted: auto-generated
5463 63. The "personal" condition in filters now includes a test for the
5464 Auto-Submitted: header. If it contains the string "auto-" the message it
5465 not considered personal.
5467 64. Added rcpt_include_affixes as a generic transport option.
5469 65. Added queue_only_override (default true).
5471 66. Added the syslog_duplication option.
5473 67. If what should have been the first header line of a message consisted of
5474 a space followed by a colon, Exim was mis-interpreting it as a header line.
5475 It isn't of course - it is syntactically invalid and should therefore be
5476 treated as the start of the message body. The misbehaviour could have
5477 caused a number of strange effects, including loss of data in subsequent
5478 header lines, and spool format errors.
5480 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5481 client host had authenticated. This control can now be exercised by an ACL
5482 for more flexibility.
5484 69. By default, callouts do not happen when testing with -bh. There is now a
5485 variant, -bhc, which does actually run the callout code, including
5486 consulting and updating the callout cache.
5488 70. Added support for saslauthd authentication, courtesy of Alexander
5491 71. If statvfs() failed on the spool or log directories while checking their
5492 size for availability, Exim confusingly gave the error "space shortage".
5493 Furthermore, in debugging mode it crashed with a floating point exception.
5494 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5495 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5496 serious problem, Exim now writes to the main and panic logs when this
5497 happens, with details of the failure. It then refuses to accept the
5498 incoming message, giving the message "spool directory problem" or "log
5499 directory problem" with a 421 code for SMTP messages.
5501 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5502 0, 1, and 2 exist, because some OS complain for execs without them (see
5503 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5504 descriptors. However, the code omitted to check that the open succeeded,
5505 causing mysterious errors if for some reason the permissions on /dev/null
5506 got screwed. Now Exim writes a message to the main and panic logs, and
5507 bombs out if it can't open /dev/null.
5509 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5510 interact so that it is all more flexible. It is supposed to remain
5511 backwards compatible. Also added extra_local_interfaces.
5513 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5514 to bomb out with an assertion failure - to the client this appears as a
5515 connection drop. This problem occurs in the part of the code that was taken
5516 from the Samba project. Fortunately, the assertion is in a very simple
5517 function, so I have fixed this by reproducing the function inline in the
5518 one place where it is called, and arranging for authentication to fail
5519 instead of killing the process with assert().
5521 75. The SPA client code was not working when the server requested OEM rather
5522 than Unicode encoding.
5524 76. Added code to make require_files with a specific uid setting more usable in
5525 the case where statting the file as root fails - usually a non-root-mounted
5526 NFS file system. When this happens and the failure is EACCES, Exim now
5527 forks a subprocess and does the per-uid checking as the relevant uid.
5529 77. Added process_log_path.
5531 78. If log_file_path was not explicitly set, a setting of check_log_space or
5532 check_log_inodes was ignored.
5534 79. If a space check for the spool or log partitions fails, the incident is now
5535 logged. Of course, in the latter case the data may get lost...
5537 80. Added the %p formatting code to string_format() so that it can be used to
5538 print addresses in debug_print(). Adjusted all the address printing in the
5539 debugging in store.c to use %p rather than %d.
5541 81. There was a concern that a line of code in smtp_in.c could overflow a
5542 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5543 initially expressed, the concern was not well-founded, because trailing
5544 spaces are removed early. However, if the trailing spaces were followed by
5545 a NULL, they did not get removed, so the overflow was possible. Two fixes
5548 (a) I re-wrote the offending code in a cleaner fashion.
5549 (b) If an incoming SMTP command contains a NULL character, it is rejected
5552 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5553 runs initgroups(), so that if the Exim user is in any additional groups,
5554 they will be used during message reception.
5560 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob