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
65 PP/01 Solaris build fix for Oracle's LDAP libraries.
66 Bugzilla 1109, patch from Stephen Usher.
68 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
70 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
73 TF/02 Fix a couple more cases where we did not log the error message
74 when unlink() failed. See also change 4.74-TF/03.
76 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
77 lock up or crash if it happened to be inside a call to libc when it
78 got a SIGUSR1 from exiwhat.
80 The SIGUSR1 handler appends the current process status to the process
81 log which is later printed by exiwhat. It used to use the general
82 purpose logging code to do this, but several functions it calls are
85 The new output code in the SIGUSR1 handler is specific to the process
86 log, and simple enough that it's easy to inspect for signal safety.
87 Removing some special cases also simplifies the general logging code.
88 Removing the spurious timestamps from the process log simplifies
91 TF/04 Improved ratelimit ACL condition.
93 The /noupdate option has been deprecated in favour of /readonly which
94 has clearer semantics. The /leaky, /strict, and /readonly update modes
95 are mutually exclusive. The update mode is no longer included in the
96 database key; it just determines when the database is updated. (This
97 means that when you upgrde Exim will forget old rate measurements.)
99 Exim now checks that the per_* options are used with an update mode that
100 makes sense for the current ACL. For example, when Exim is processing a
101 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
102 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
103 must specify per_mail/readonly. If you omit the update mode it defaults to
104 /leaky where that makes sense (as before) or /readonly where required.
106 The /noupdate option is now undocumented but still supported for
107 backwards compatibility. It is equivalent to /readonly except that in
108 ACLs where /readonly is required you may specify /leaky/noupdate or
109 /strict/noupdate which are treated the same as /readonly.
111 A useful new feature is the /count= option. This is a generalization
112 of the per_byte option, so that you can measure the throughput of other
113 aggregate values. For example, the per_byte option is now equivalent
114 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
116 The per_rcpt option has been generalized using the /count= mechanism
117 (though it's more complicated than the per_byte equivalence). When it is
118 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
119 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
120 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
121 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
122 Note that using per_rcpt with a non-readonly update mode in more than
123 one ACL will cause the recipients to be double-counted. (The per_mail
124 and per_byte options don't have this problem.)
126 The handling of very low rates has changed slightly. If the computed rate
127 is less than the event's count (usually one) then this event is the first
128 after a long gap. In this case the rate is set to the same as this event's
129 count, so that the first message of a spam run is counted properly.
131 The major new feature is a mechanism for counting the rate of unique
132 events. The new per_addr option counts the number of different
133 recipients that someone has sent messages to in the last time period. It
134 behaves like per_rcpt if all the recipient addresses are different, but
135 duplicate recipient addresses do not increase the measured rate. Like
136 the /count= option this is a general mechanism, so the per_addr option
137 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
138 example, measure the rate that a client uses different sender addresses
139 with the options per_mail/unique=$sender_address. There are further
140 details in the main documentation.
142 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
144 TF/06 Removed a few PCRE remnants.
146 TF/07 Automatically extract Exim's version number from tags in the git
147 repository when doing development or release builds.
149 PP/02 Raise smtp_cmd_buffer_size to 16kB.
150 Bugzilla 879. Patch from Paul Fisher.
152 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
153 Heavily based on revision 40f9a89a from Simon Arlott's tree.
156 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
158 PP/05 Variable $av_failed, true if the AV scanner deferred.
159 Bugzilla 1078. Patch from John Horne.
161 PP/06 Stop make process more reliably on build failure.
162 Bugzilla 1087. Patch from Heiko Schlittermann.
164 PP/07 Make maildir_use_size_file an _expandable_ boolean.
165 Bugzilla 1089. Patch from Heiko Schlittermann.
167 PP/08 Handle ${run} returning more data than OS pipe buffer size.
168 Bugzilla 1131. Patch from Holger Weiß.
170 PP/09 Handle IPv6 addresses with SPF.
171 Bugzilla 860. Patch from Wolfgang Breyha.
173 PP/10 GnuTLS: support TLS 1.2 & 1.1.
175 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
178 PP/11 match_* no longer expand right-hand-side by default.
179 New compile-time build option, EXPAND_LISTMATCH_RHS.
180 New expansion conditions, "inlist", "inlisti".
182 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
184 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
186 PP/14 fix log_write() format string regression from TF/03.
187 Bugzilla 1152. Patch from Dmitry Isaikin.
193 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
195 PP/02 Harmonised TLS library version reporting; only show if debugging.
196 Layout now matches that introduced for other libraries in 4.74 PP/03.
198 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
200 PP/04 New "dns_use_edns0" global option.
202 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
205 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
206 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
208 TK/01 Updated PolarSSL code to 0.14.2.
209 Bugzilla 1097. Patch from Andreas Metzler.
211 PP/07 Catch divide-by-zero in ${eval:...}.
214 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
217 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
218 format-string attack -- SECURITY: remote arbitrary code execution.
220 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
221 time unintentionally subject to list matching rules, letting the header
222 cause arbitrary Exim lookups (of items which can occur in lists, *not*
223 arbitrary string expansion). This allowed for information disclosure.
225 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
226 INT_MIN/-1 -- value coerced to INT_MAX.
232 NM/01 Workround for PCRE version dependency in version reporting
235 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
236 This fixes portability to compilers other than gcc, notably
237 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
239 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
240 makefiles for portability to HP-UX and POSIX correctness.
242 PP/01 Permit LOOKUP_foo enabling on the make command-line.
243 Also via indented variable definition in the Makefile.
244 (Debugging by Oliver Heesakkers).
246 PP/02 Restore caching of spamd results with expanded spamd_address.
247 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
249 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
250 Improves build reliability. Fix from: Frank Elsner
252 NM/02 Fix wide character breakage in the rfc2047 coding
253 Fixes bug 1064. Patch from Andrey N. Oktyabrski
255 NM/03 Allow underscore in dnslist lookups
256 Fixes bug 1026. Patch from Graeme Fowler
258 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
259 Code patches from Adam Ciarcinski of NetBSD.
261 NM/04 Fixed exiqgrep to cope with mailq missing size issue
264 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
265 is logged, to avoid truncation. Patch from John Horne.
267 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
268 Patch from Jakob Hirsch.
270 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
271 SQL string expansion failure details.
272 Patch from Andrey Oktyabrski.
274 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
275 Patch from Simon Arlott.
277 PP/09 New lookups functionality failed to compile on old gcc which rejects
278 extern declarations in function scope.
279 Patch from Oliver Fleischmann
281 PP/10 Use sig_atomic_t for flags set from signal handlers.
282 Check getgroups() return and improve debugging.
283 Fixed developed for diagnosis in bug 927 (which turned out to be
286 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
287 Patch from Mark Zealey.
289 PP/12 Bugzilla 1056: Improved spamd server selection.
290 Patch from Mark Zealey.
292 PP/13 Bugzilla 1086: Deal with maildir quota file races.
293 Based on patch from Heiko Schlittermann.
295 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
296 Patch from Uwe Doering, sign-off by Michael Haardt.
298 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
299 variable declaration deep within a block. Bug and patch from
302 PP/15 lookups-Makefile IRIX compatibilty coercion.
304 PP/16 Make DISABLE_DKIM build knob functional.
306 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
307 Patch by Simon Arlott
309 TF/03 Fix valgrind.h portability to C89 compilers that do not support
310 variable argument macros. Our copy now differs from upstream.
316 TF/01 Failure to get a lock on a hints database can have serious
317 consequences so log it to the panic log.
319 TF/02 Log LMTP confirmation messages in the same way as SMTP,
320 controlled using the smtp_confirmation log selector.
322 TF/03 Include the error message when we fail to unlink a spool file.
324 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
325 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
326 for maintaining out-of-tree patches for some time.
328 PP/01 Bugzilla 139: Documentation and portability issues.
329 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
330 Handle per-OS dynamic-module compilation flags.
332 PP/02 Let /dev/null have normal permissions.
333 The 4.73 fixes were a little too stringent and complained about the
334 permissions on /dev/null. Exempt it from some checks.
335 Reported by Andreas M. Kirchwitz.
337 PP/03 Report version information for many libraries, including
338 Exim version information for dynamically loaded libraries. Created
339 version.h, now support a version extension string for distributors
340 who patch heavily. Dynamic module ABI change.
342 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
343 privilege escalation vulnerability whereby the Exim run-time user
344 can cause root to append content of the attacker's choosing to
347 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
350 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
351 If dropping privileges for untrusted macros, we disabled normal logging
352 on the basis that it would fail; for the Exim run-time user, this is not
353 the case, and it resulted in successful deliveries going unlogged.
354 Fixed. Reported by Andreas Metzler.
360 PP/01 Date: & Message-Id: revert to normally being appended to a message,
361 only prepend for the Resent-* case. Fixes regression introduced in
362 Exim 4.70 by NM/22 for Bugzilla 607.
364 PP/02 Include check_rfc2047_length in configure.default because we're seeing
365 increasing numbers of administrators be bitten by this.
367 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
369 PP/03 Bugzilla 994: added openssl_options main configuration option.
371 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
373 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
375 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
377 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
378 without a peer certificate, leading to a segfault because of an
379 assumption that peers always have certificates. Be a little more
380 paranoid. Problem reported by Martin Tscholak.
382 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
383 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
384 NB: ClamAV planning to remove STREAM in "middle of 2010".
385 CL also introduces -bmalware, various -d+acl logging additions and
386 more caution in buffer sizes.
388 PP/09 Implemented reverse_ip expansion operator.
390 PP/10 Bugzilla 937: provide a "debug" ACL control.
392 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
394 PP/12 Bugzilla 973: Implement --version.
396 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
398 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
400 PP/15 Bugzilla 816: support multiple condition rules on Routers.
402 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
403 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
404 ignore trailing whitespace.
406 JJ/02 prevent non-panic DKIM error from being sent to paniclog
408 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
411 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
412 Notification from Dr Andrew Aitchison.
414 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
415 ExtendedDetectionInfo response format.
416 Notification from John Horne.
418 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
421 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
422 XSL and documented dependency on system catalogs, with examples of how
425 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
428 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
429 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
430 configuration file which is writeable by the Exim user or group.
432 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
433 of configuration files to cover files specified with the -C option if
434 they are going to be used with root privileges, not just the default
437 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
438 option (effectively making it always true).
440 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
441 files to be used while preserving root privileges.
443 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
444 that rogue child processes cannot use them.
446 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
447 run-time user, instead of root.
449 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
450 Exim run-time user without dropping privileges.
452 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
453 result string, instead of calling string_vformat() twice with the same
456 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
457 for other users. Others should always drop root privileges if they use
458 -C on the command line, even for a whitelisted configure file.
460 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
462 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
468 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
469 $data_path, and $header_path variables; fixed documentation bugs and
472 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
473 exipick to access non-standard spools, including the "frozen" queue
476 NM/01 Bugzilla 965: Support mysql stored procedures.
477 Patch from Alain Williams
479 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
481 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
482 Patch from Andreas Metzler
484 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
485 Patch from Kirill Miazine
487 NM/05 Bugzilla 671: Added umask to procmail example.
489 JJ/03 installed exipick 20100323.0, fixing doc bug
491 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
492 directory. Notification and patch from Dan Rosenberg.
494 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
496 TK/02 Improve log output when DKIM signing operation fails.
498 MH/01 Treat the transport option dkim_domain as a colon separated
499 list, not as a single string, and sign the message with each element,
500 omitting multiple occurences of the same signer.
502 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
503 Bugzilla 985, 986. Patch by Simon Arlott
505 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
506 Patch by Simon Arlott
508 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
509 MBX locking. Notification from Dan Rosenberg.
515 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
517 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
519 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
521 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
523 NM/04 Bugzilla 915: Flag broken perl installation during build.
529 TK/01 Added patch by Johannes Berg that expands the main option
530 "spamd_address" if it starts with a dollar sign.
532 TK/02 Write list of recipients to X-Envelope-Sender header when building
533 the mbox-format spool file for content scanning (suggested by Jakob
536 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
537 (http://www.dcc-servers.net/) support via dccifd. Activated by
538 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
540 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
541 by Mark Daniel Reidel <mr@df.eu>.
543 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
544 When building exim an external PCRE library is now needed -
545 PCRE is a system library on the majority of modern systems.
546 See entry on PCRE_LIBS in EDITME file.
548 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
549 conversation. Added nologin parameter to request.
550 Patch contributed by Kirill Miazine.
552 TF/01 Do not log submission mode rewrites if they do not change the address.
554 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
556 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
557 log files in place. Contributed by Roberto Lima.
559 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
561 TF/03 Bugzilla 615: When checking the local_parts router precondition
562 after a local_part_suffix or local_part_prefix option, Exim now
563 does not use the address's named list lookup cache, since this
564 contains cached lookups for the whole local part.
566 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
567 Robert Millan. Documentation is in experimental-spec.txt.
569 TF/04 Bugzilla 668: Fix parallel build (make -j).
571 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
573 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
574 Patch provided by Jan Srzednicki.
576 TF/05 Leading white space used to be stripped from $spam_report which
577 wrecked the formatting. Now it is preserved.
579 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
580 that they are available at delivery time.
582 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
584 TF/08 TLS error reporting now respects the incoming_interface and
585 incoming_port log selectors.
587 TF/09 Produce a more useful error message if an SMTP transport's hosts
588 setting expands to an empty string.
590 NM/06 Bugzilla 744: EXPN did not work under TLS.
591 Patch provided by Phil Pennock.
593 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
594 Patch provided by Richard Godbee.
596 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
597 acl_smtp_notquit, added index entry.
599 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
600 Patch provided by Eugene Bujak.
602 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
603 accept(). Patch provided by Maxim Dounin.
605 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
606 Patch provided by Phil Pennock.
608 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
610 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
611 Patch provided by Brad "anomie" Jorsch.
613 NM/14 Bugzilla 622: Added timeout setting to transport filter.
614 Patch provided by Dean Brooks.
616 TK/05 Add native DKIM support (does not depend on external libraries).
618 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
619 Patch provided by Graeme Fowler.
621 NM/16 Bugzilla 851: Documentation example syntax fix.
623 NM/17 Changed NOTICE file to remove references to embedded PCRE.
625 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
628 NM/19 Bugzilla 745: TLS version reporting.
629 Patch provided by Phil Pennock.
631 NM/20 Bugzilla 167: bool: condition support.
632 Patch provided by Phil Pennock.
634 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
635 clients. Patch provided by Phil Pennock.
637 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
638 Patch provided by Brad "anomie" Jorsch.
640 NM/23 Bugzilla 687: Fix misparses in eximstats.
641 Patch provided by Heiko Schlittermann.
643 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
644 Patch provided by Heiko Schlittermann.
646 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
647 plus update to original patch.
649 NM/26 Bugzilla 799: Documentation correction for ratelimit.
651 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
652 Patch provided by David Brownlee.
654 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
656 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
658 NM/30 Bugzilla 888: TLS documentation bugfixes.
660 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
662 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
663 Unlike the original bugzilla I have changed all shell scripts in src tree.
665 NM/33 Bugzilla 898: Transport filter timeout fix.
666 Patch by Todd Rinaldo.
668 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
669 Patch by Serge Demonchaux.
671 NM/35 Bugzilla 39: Base64 decode bug fixes.
672 Patch by Jakob Hirsch.
674 NM/36 Bugzilla 909: Correct connect() call in dcc code.
676 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
678 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
680 NM/39 Bugzilla 911: Fixed MakeLinks build script.
686 TK/01 Add preliminary DKIM support. Currently requires a forked version of
687 ALT-N's libdkim that I have put here:
688 http://duncanthrax.net/exim-experimental/
690 Note to Michael Haardt: I had to rename some vars in sieve.c. They
691 were called 'true' and it seems that C99 defines that as a reserved
692 keyword to be used with 'bool' variable types. That means you could
693 not include C99-style headers which use bools without triggering
694 build errors in sieve.c.
696 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
697 as mailq or other aliases. Changed the --help handling significantly
698 to do whats expected. exim_usage() emits usage/help information.
700 SC/01 Added the -bylocaldomain option to eximstats.
702 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
704 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
706 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
712 PH/01 Another patch from the Sieve maintainer.
714 PH/02 When an IPv6 address is converted to a string for single-key lookup
715 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
716 dots are used instead of colons so that keys in lsearch files need not
717 contain colons. This was done some time before quoting was made available
718 in lsearch files. However, iplsearch files do require colons in IPv6 keys
719 (notated using the quote facility) so as to distinguish them from IPv4
720 keys. This meant that lookups for IP addresses in host lists did not work
721 for iplsearch lookups.
723 This has been fixed by arranging for IPv6 addresses to be expressed with
724 colons if the lookup type is iplsearch. This is not incompatible, because
725 previously such lookups could never work.
727 The situation is now rather anomolous, since one *can* have colons in
728 ordinary lsearch keys. However, making the change in all cases is
729 incompatible and would probably break a number of configurations.
731 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
734 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
735 conversion specification without a maximum field width, thereby enabling
736 a rogue spamd server to cause a buffer overflow. While nobody in their
737 right mind would setup Exim to query an untrusted spamd server, an
738 attacker that gains access to a server running spamd could potentially
739 exploit this vulnerability to run arbitrary code as the Exim user.
741 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
742 $primary_hostname instead of what libspf2 thinks the hosts name is.
744 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
745 a directory entry by the name of the lookup key. Previously, if a
746 symlink pointed to a non-existing file or a file in a directory that
747 Exim lacked permissions to read, a lookup for a key matching that
748 symlink would fail. Now it is enough that a matching directory entry
749 exists, symlink or not. (Bugzilla 503.)
751 PH/03 The body_linecount and body_zerocount variables are now exported in the
754 PH/04 Added the $dnslist_matched variable.
756 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
757 This means they are set thereafter only if the connection becomes
760 PH/06 Added the client_condition to authenticators so that some can be skipped
761 by clients under certain conditions.
763 PH/07 The error message for a badly-placed control=no_multiline_responses left
764 "_responses" off the end of the name.
766 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
768 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
769 (without spaces) instead of just copying the configuration text.
771 PH/10 Added the /noupdate option to the ratelimit ACL condition.
773 PH/11 Added $max_received_linelength.
775 PH/12 Added +ignore_defer and +include_defer to host lists.
777 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
778 way in which PCRE > 7.0 is built.
780 PH/14 Implemented queue_only_load_latch.
782 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
783 MAIL command. The effect was to mangle the value on 64-bit systems.
785 PH/16 Another patch from the Sieve maintainer.
787 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
789 PH/18 If a system quota error occurred while trying to create the file for
790 a maildir delivery, the message "Mailbox is full" was not appended to the
791 bounce if the delivery eventually timed out. Change 4.67/27 below applied
792 only to a quota excession during the actual writing of the file.
794 PH/19 It seems that peer DN values may contain newlines (and other non-printing
795 characters?) which causes problems in log lines. The DN values are now
796 passed through string_printing() before being added to log lines.
798 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
799 and InterBase are left for another time.)
801 PH/21 Added message_body_newlines option.
803 PH/22 Guard against possible overflow in moan_check_errorcopy().
805 PH/23 POSIX allows open() to be a macro; guard against that.
807 PH/24 If the recipient of an error message contained an @ in the local part
808 (suitably quoted, of course), incorrect values were put in $domain and
809 $local_part during the evaluation of errors_copy.
815 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
816 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
819 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
820 issue a MAIL command.
822 PH/02 In an ACL statement such as
824 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
826 if a client was not listed at all, or was listed with a value other than
827 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
828 the condition was not true (as it should be), so access was not denied.
829 The bug was that the ! inversion was incorrectly passed on to the second
830 item. This has been fixed.
832 PH/03 Added additional dnslists conditions == and =& which are different from
833 = and & when the dns lookup returns more than one IP address.
835 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
836 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
838 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
839 FSYNC, which compiles an option called disable_fsync that allows for
840 bypassing fsync(). The documentation is heavily laced with warnings.
842 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
844 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
845 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
846 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
847 including adding "make clean"; (3) Added -fPIC when compiling the test
848 dynamically loaded module, to get rid of a warning.
850 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
851 message fails, move_frozen_messages = true and ignore_bounce_errors_after
852 = 0s. The bug is otherwise harmless.
854 PH/07 There was a bug in the dovecot authenticator such that the value of
855 $auth1 could be overwritten, and so not correctly preserved, after a
856 successful authentication. This usually meant that the value preserved by
857 the server_setid option was incorrect.
859 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
861 PH/09 Installed PCRE release 7.0.
863 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
864 run for batched SMTP input. It is now run at the start of every message
865 in the batch. While fixing this I discovered that the process information
866 (output by running exiwhat) was not always getting set for -bs and -bS
867 input. This is fixed, and it now also says "batched" for BSMTP.
869 PH/11 Added control=no_pipelining.
871 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
872 patch, slightly modified), and move the expansion of helo_data till after
873 the connection is made in the smtp transport (so it can use these
876 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
878 PH/14 Added log_selector = +pid.
880 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
882 PH/16 Add ${if forany and ${if forall.
884 PH/17 Added dsn_from option to vary the From: line in DSNs.
886 PH/18 Flush SMTP output before performing a callout, unless control =
887 no_callout_flush is set.
889 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
890 was true (the default) a successful delivery failed to delete the retry
891 item, thus causing premature timeout of the address. The bug is now
894 PH/20 Added hosts_avoid_pipelining to the smtp transport.
896 PH/21 Long custom messages for fakedefer and fakereject are now split up
897 into multiline reponses in the same way that messages for "deny" and
898 other ACL rejections are.
900 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
901 with slight modification.
903 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
904 draft, changing the syntax and factoring some duplicate code".
906 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
907 for deliveries of the second and subsequent messages over the same SMTP
910 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
911 ${reduce, with only minor "tidies".
913 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
915 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
916 expansion side effects.
918 PH/27 When a message times out after an over-quota error from an Exim-imposed
919 quota, the bounce message says "mailbox is full". This message was not
920 being given when it was a system quota that was exceeded. It now should
923 MH/03 Made $recipients available in local_scan(). local_scan() already has
924 better access to the recipient list through recipients_list[], but
925 $recipients can be useful in postmaster-provided expansion strings.
927 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
928 in the case of a MAIL command with additional options following the
929 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
930 were accidentally chopped off.
932 PH/29 SMTP synchronization checks are implemented when a command is read -
933 there is a check that no more input is waiting when there shouldn't be
934 any. However, for some commands, a delay in an ACL can mean that it is
935 some time before the response is written. In this time, more input might
936 arrive, invalidly. So now there are extra checks after an ACL has run for
937 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
938 pipelining has not been advertised.
940 PH/30 MH's patch to allow iscntrl() characters to be list separators.
942 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
943 returned in the SMTP response when smtp_return_error_details was false.
946 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
947 instead of the C I/O that was originally supplied, because problems were
950 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
951 Exim which did not show up earlier: it was assuming that a call to
952 SSL_CTX_set_info_callback() might give an error value. In fact, there is
953 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
954 was a macro that became an assignment, so it seemed to work. This has
955 changed to a proper function call with a void return, hence the compile
956 error. Exim's code has been fixed.
958 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
961 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
963 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
965 PH/37 If a message is not accepted after it has had an id assigned (e.g.
966 because it turns out to be too big or there is a timeout) there is no
967 "Completed" line in the log. When some messages of this type were
968 selected by exigrep, they were listed as "not completed". Others were
969 picked up by some special patterns. I have improved the selection
970 criteria to be more general.
972 PH/38 The host_find_failed option in the manualroute router can now be set
973 to "ignore", to completely ignore a host whose IP address cannot be
974 found. If all hosts are ignored, the behaviour is controlled by the new
975 host_all_ignored option.
977 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
978 homing or because of multiple MX records with /mx) generated more than
979 one IP address, and the following item turned out to be the local host,
980 all the secondary addresses of the first item were incorrectly removed
981 from the list, along with the local host and any following hosts (which
982 is what is supposed to happen).
984 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
985 whoever called Exim into the -H file. In the case of the daemon it was
986 behaving confusingly. When first started, it used values for whoever
987 started the daemon, but after a SIGHUP it used the Exim user (because it
988 calls itself on a restart). I have changed the code so that it now always
991 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
992 message are rejected with the same error (e.g. no authentication or bad
993 sender address), and a DATA command is nevertheless sent (as can happen
994 with PIPELINING or a stupid MUA), the error message that was given to the
995 RCPT commands is included in the rejection of the DATA command. This is
996 intended to be helpful for MUAs that show only the final error to their
999 PH/42 Another patch from the Sieve maintainer.
1001 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1002 Eximstats - Fixed some broken HTML links and added missing column headers
1004 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1005 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1007 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1009 PH/43 Yet another patch from the Sieve maintainer.
1011 PH/44 I found a way to check for a TCP/IP connection going away before sending
1012 the response to the final '.' that terminates a message, but only in the
1013 case where the client has not sent further data following the '.'
1014 (unfortunately, this is allowed). However, in many cases there won't be
1015 any further data because there won't be any more messages to send. A call
1016 to select() can be used: if it shows that the input is "ready", there is
1017 either input waiting, or the socket has been closed. An attempt to read
1018 the next input character can distinguish the two cases. Previously, Exim
1019 would have sent an OK response which the client would never have see.
1020 This could lead to message repetition. This fix should cure that, at
1021 least in a lot of common cases.
1023 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1024 advertised in response to EHLO.
1030 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1031 fixed by 4.65/MH/01 (is this a record?) are fixed:
1033 (i) An empty string was always treated as zero by the numeric comparison
1034 operators. This behaviour has been restored.
1036 (ii) It is documented that the numeric comparison operators always treat
1037 their arguments as decimal numbers. This was broken in that numbers
1038 starting with 0 were being interpreted as octal.
1040 While fixing these problems I realized that there was another issue that
1041 hadn't been noticed. Values of message_size_limit (both the global option
1042 and the transport option) were treated as octal if they started with 0.
1043 The documentation was vague. These values are now always treated as
1044 decimal, and I will make that clear in the documentation.
1050 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1051 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1054 MH/01 Don't check that the operands of numeric comparison operators are
1055 integers when their expansion is in "skipping" mode (fixes bug
1056 introduced by 4.64-PH/07).
1058 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1059 child addresses, Exim now panics and dies. Previously, because the count
1060 is held in a short int, deliveries were likely to be lost. As such a
1061 large number of recipients for a single message is ridiculous
1062 (performance will be very, very poor), I have chosen to impose a limit
1063 rather than extend the field.
1069 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1070 leftover -K file (the existence of which was triggered by #402).
1071 While we were at it, introduced process PID as part of the -K
1072 filename. This should rule out race conditions when creating
1075 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1076 processing considerably. Previous code took too long for large mails,
1077 triggering a timeout which in turn triggers #401.
1079 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1080 in the DK code in transports.c. sendfile() is not really portable,
1081 hence the _LINUX specificness.
1083 TF/01 In the add_headers option to the mail command in an Exim filter,
1084 there was a bug that Exim would claim a syntax error in any
1085 header after the first one which had an odd number of characters
1088 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1089 callout verification, Exim cached a "reject" for the entire domain. This
1090 is correct for most verifications, but it is not correct for a recipient
1091 verification with use_sender or use_postmaster set, because in that case
1092 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1093 case of MAIL FROM:<> rejection from other early rejections (e.g.
1094 rejection of HELO). When verifying a recipient using a non-null MAIL
1095 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1096 Whatever the result of the callout, the value of the domain cache is
1097 left unchanged (for any other kind of callout, getting as far as trying
1098 RCPT means that the domain itself is ok).
1100 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1103 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1104 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1105 session, the daemon ignores ECONNECT errors and logs others; it now
1106 ignores EPIPE as well.
1108 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1109 (quoted-printable decoding).
1111 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1112 later the small subsequent patch to fix an introduced bug.
1114 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1116 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1118 PH/08 An error is now given if message_size_limit is specified negative.
1120 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1121 to be given (somewhat) arbitrary names.
1123 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1126 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1127 miscellaneous code fixes
1129 PH/10 Added the log_reject_target ACL modifier to specify where to log
1132 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1133 hostname. This is wrong, because it relates to the incoming message (and
1134 probably the interface on which it is arriving) and not to the outgoing
1135 callout (which could be using a different interface). This has been
1136 changed to use the value of the helo_data option from the smtp transport
1137 instead - this is what is used when a message is actually being sent. If
1138 there is no remote transport (possible with a router that sets up host
1139 addresses), $smtp_active_hostname is used.
1141 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1142 tweaks were necessary in order to get it to work (see also 21 below):
1143 (a) The code assumed that strncpy() returns a negative number on buffer
1144 overflow, which isn't the case. Replaced with Exim's string_format()
1146 (b) There were several signed/unsigned issues. I just did the minimum
1147 hacking in of casts. There is scope for a larger refactoring.
1148 (c) The code used strcasecmp() which is not a standard C function.
1149 Replaced with Exim's strcmpic() function.
1150 (d) The code set only $1; it now sets $auth1 as well.
1151 (e) A simple test gave the error "authentication client didn't specify
1152 service in request". It would seem that Dovecot has changed its
1153 interface. Fortunately there's a specification; I followed it and
1154 changed what the client sends and it appears to be working now.
1156 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1159 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1160 address A is aliased to B and C, where B exists and C does not. Without
1161 -v the output is "A verified" because verification stops after a
1162 successful redirection if more than one address is generated. However,
1163 with -v the child addresses are also verified. Exim was outputting "A
1164 failed to verify" and then showing the successful verification for C,
1165 with its parentage. It now outputs "B failed to verify", showing B's
1166 parentage before showing the successful verification of C.
1168 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1169 look up a TXT record in a specific list after matching in a combined
1172 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1173 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1174 they consult the DNS. I had assumed they would set it the way they
1175 wanted; and indeed my experiments on Linux seem to show that in some
1176 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1177 To be on the safe side, however, I have now made the interface to
1178 host_find_byname() similar to host_find_bydns(), with an argument
1179 containing the DNS resolver options. The host_find_byname() function now
1180 sets these options at its start, just as host_find_bydns() does. The smtp
1181 transport options dns_qualify_single and dns_search_parents are passed to
1182 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1183 of host_find_byname() use the default settings of RES_DEFNAMES
1184 (qualify_single) but not RES_DNSRCH (search_parents).
1186 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1187 spool_read_header() do less string testing, by means of a preliminary
1188 switch on the second character of optional "-foo" lines. (This is
1189 overdue, caused by the large number of possibilities that now exist.
1190 Originally there were few.) While I was there, I also converted the
1191 str(n)cmp tests so they don't re-test the leading "-" and the first
1192 character, in the hope this might squeeze out yet more improvement.
1194 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1195 flag allowing group syntax was set by the header_syntax check but not
1196 turned off, possible causing trouble later; (2) The flag was not being
1197 set at all for the header_verify test, causing "group"-style headers to
1198 be rejected. I have now set it in this case, and also caused header_
1199 verify to ignore an empty address taken from a group. While doing this, I
1200 came across some other cases where the code for allowing group syntax
1201 while scanning a header line wasn't quite right (mostly, not resetting
1202 the flag correctly in the right place). These bugs could have caused
1203 trouble for malformed header lines. I hope it is now all correct.
1205 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1206 with the "reply" argument non-NULL. The code, however (which originally
1207 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1208 but it didn't always do it. This confused somebody who was copying the
1209 code for some other use. I have removed all the tests.
1211 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1212 feature that was used to support insecure browsers during the U.S. crypto
1213 embargo. It requires special client support, and Exim is probably the
1214 only MTA that supported it -- and would never use it because real RSA is
1215 always available. This code has been removed, because it had the bad
1216 effect of slowing Exim down by computing (never used) parameters for the
1217 RSA_EXPORT functionality.
1219 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1220 authenticator to fail if there's a tab character in the incoming data
1221 (there should never be unless someone is messing about, as it's supposed
1222 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1223 if the connection is using TLS or if the remote IP is the same as the
1224 local IP, and the "valid-client-cert option" if a client certificate has
1227 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1228 authenticators. This can be used for authorization after authentication
1229 succeeds. (In the case of plaintext, it servers for both authentication
1232 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1233 if any retry times were supplied.
1235 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1236 connection, where there is no client IP address to check. In this
1237 situation, the verify now always succeeds.
1239 PH/25 Applied John Jetmore's -Mset patch.
1241 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1243 PH/27 In a string expansion for a processed (not raw) header when multiple
1244 headers of the same name were present, leading whitespace was being
1245 removed from all of them, but trailing whitespace was being removed only
1246 from the last one. Now trailing whitespace is removed from each header
1247 before concatenation. Completely empty headers in a concatenation (as
1248 before) are ignored.
1250 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1251 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1253 PH/29 [Removed. This was a change that I later backed out, and forgot to
1254 correct the ChangeLog entry (that I had efficiently created) before
1255 committing the later change.]
1257 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1258 address errors (4xx response to RCPT) over the same connection as other
1259 messages routed to the same hosts. Such deliveries are always "forced",
1260 so retry times are not inspected. This resulted in far too many retries
1261 for the affected addresses. The effect occurred only when there were more
1262 hosts than the hosts_max_try setting in the smtp transport when it had
1263 the 4xx errors. Those hosts that it had tried were not added to the list
1264 of hosts for which the message was waiting, so if all were tried, there
1265 was no problem. Two fixes have been applied:
1267 (i) If there are any address or message errors in an SMTP delivery, none
1268 of the hosts (tried or untried) are now added to the list of hosts
1269 for which the message is waiting, so the message should not be a
1270 candidate for sending over the same connection that was used for a
1271 successful delivery of some other message. This seems entirely
1272 reasonable: after all the message is NOT "waiting for some host".
1273 This is so "obvious" that I'm not sure why it wasn't done
1274 previously. Hope I haven't missed anything, but it can't do any
1275 harm, as the worst effect is to miss an optimization.
1277 (ii) If, despite (i), such a delivery is accidentally attempted, the
1278 routing retry time is respected, so at least it doesn't keep
1279 hammering the server.
1281 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1282 in ${readsocket because some servers need this prod.
1284 PH/32 Added some extra debug output when updating a wait-xxx database.
1286 PH/33 The hint "could be header name not terminated by colon", which has been
1287 given for certain expansion errors for a long time, was not being given
1288 for the ${if def:h_colon_omitted{... case.
1290 PH/34 The spec says: "With one important exception, whenever a domain list is
1291 being scanned, $domain contains the subject domain." There was at least
1292 one case where this was not true.
1294 PH/35 The error "getsockname() failed: connection reset by peer" was being
1295 written to the panic log as well as the main log, but it isn't really
1296 panic-worthy as it just means the connection died rather early on. I have
1297 removed the panic log writing for the ECONNRESET error when getsockname()
1300 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1301 runs only) independently of the message's sender address. This meant
1302 that, if the 4xx error was in fact related to the sender, a different
1303 message to the same recipient with a different sender could confuse
1304 things. In particualar, this can happen when sending to a greylisting
1305 server, but other circumstances could also provoke similar problems.
1306 I have changed the default so that the retry time for these errors is now
1307 based a combination of the sender and recipient addresses. This change
1308 can be overridden by setting address_retry_include_sender=false in the
1311 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1312 remote server are returned as part of bounce messages. This was not
1313 happening for LMTP over a pipe (the lmtp transport), but now it is the
1314 same for both kinds of LMTP.
1316 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1317 in header lines that were in fact CNAMEs. This is no longer the case.
1319 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1320 and queue runs started by the daemon processed all messages. This has
1321 been fixed so that -R and -S can now usefully be given with -q<time>.
1323 PH/40 Import PCRE release 6.7 (fixes some bugs).
1325 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1327 PH/42 Give an error if -q is specified more than once.
1329 PH/43 Renamed the variables $interface_address and $interface_port as
1330 $received_ip_address and $received_port, to make it clear that these
1331 values apply to message reception, and not to the outgoing interface when
1332 a message is delivered. (The old names remain recognized, of course.)
1334 PH/44 There was no timeout on the connect() call when using a Unix domain
1335 socket in the ${readsocket expansion. There now is.
1337 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1338 be meaningful with "accept".
1340 SC/01 Eximstats V1.43
1341 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1343 SC/02 Eximstats V1.44
1344 Use a glob alias rather than an array ref in the generated
1345 parser. This improves both readability and performance.
1347 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1348 Collect SpamAssassin and rejection statistics.
1349 Don't display local sender or destination tables unless
1350 there is data to show.
1351 Added average volumes into the top table text output.
1353 SC/04 Eximstats V1.46
1354 Collect data on the number of addresses (recipients)
1355 as well as the number of messages.
1357 SC/05 Eximstats V1.47
1358 Added 'Message too big' to the list of mail rejection
1359 reasons (thanks to Marco Gaiarin).
1361 SC/06 Eximstats V1.48
1362 Mainlog lines which have GMT offsets and are too short to
1363 have a flag are now skipped.
1365 SC/07 Eximstats V1.49 (Alain Williams)
1366 Added the -emptyok flag.
1368 SC/08 Eximstats V1.50
1369 Fixes for obtaining the IP address from reject messages.
1371 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1372 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1373 whitesspace changes from 4.64-PH/27
1375 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1378 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1379 are found, allow negative numbers in numeric criteria)
1381 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1383 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1384 to match changes made in 4.64-PH/43
1386 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1388 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1389 "domain" is actually an IP address, and force a failure. This locks out
1390 those revolvers/nameservers that support "A-for-A" lookups, in
1391 contravention of the specifications.
1393 PH/48 When a host name was looked up from an IP address, and the subsequent
1394 forward lookup of the name timed out, the host name was left in
1395 $sender_host_name, contrary to the specification.
1397 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1398 restricted to single-key lookups, Exim was not diagnosing an error if
1399 * or *@ was used with a query-style lookup.
1401 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1403 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1404 long ago, but noone interested enough thought of it. Let's just say that
1405 the "1.1" means that there are some new functions that weren't there at
1406 some point in the past.
1408 PH/51 Error processing for expansion failure of helo_data from an smtp
1409 transport during callout processing was broken.
1411 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1412 tested/used via the -bh/-bhc/-bs options.
1414 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1415 bug, fixed in subsequent PCRE releases).
1417 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1418 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1420 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1426 SC/01 Use a glob alias rather than an array ref in eximstats generated
1427 parser. This improves both readability and performance.
1429 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1430 Don't display local sender or destination tables in eximstats unless
1431 there is data to show.
1432 Added average volumes into the eximstats top table text output.
1434 SC/03 Collect data on the number of addresses (recipients) as well
1435 as the number of messages in eximstats.
1437 TF/01 Correct an error in the documentation for the redirect router. Exim
1438 does (usually) call initgroups() when daemonizing.
1440 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1441 with consistent privilege compared to when running as a daemon.
1443 TF/03 Note in the spec that $authenticated_id is not set for local
1444 submissions from trusted users.
1446 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1447 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1449 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1450 by adding some example configuration directives to the default
1451 configuration file. A little bit of work is required to uncomment the
1452 directives and define how usernames and passwords are checked, but
1453 there is now a framework to start from.
1455 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1456 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1457 without this. I don't know how relevant this is to other LDAP libraries.
1459 PH/02 Add the verb name to the "unknown ACL verb" error.
1461 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1463 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1465 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1466 directory not expanded when it should be if an expanded home directory
1467 was set for the address (which is overridden by the transport).
1469 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1472 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1473 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1474 because it is too late at that time, and has no effect.
1476 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1477 security issue with \' (bugzilla #107). I could not use the
1478 PQescapeStringConn() function, because it needs a PGconn value as one of
1481 PH/08 When testing addresses using -bt, indicate those final addresses that
1482 are duplicates that would not cause an additional delivery. At least one
1483 person was confused, thinking that -bt output corresponded to deliveries.
1484 (Suppressing duplicates isn't a good idea as you lose the information
1485 about possibly different redirections that led to the duplicates.)
1487 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1488 systems where poll() doesn't work, in particular OS X.
1490 PH/10 Added more information to debugging output for retry time not reached.
1492 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1493 operations in malware.c.
1495 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1498 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1499 syslog_duplication set false, log lines that would normally be written
1500 both the the main log and to the reject log were not written to syslog at
1503 PH/14 In the default configuration, change the use of "message" in ACL warn
1504 statements to "add_header".
1506 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1507 not followed by a command (e.g. "seen endif").
1509 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1510 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1513 PH/17 Added extra conditions to the default value of delay_warning_condition
1517 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1518 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1519 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1522 The Auto-Submitted: and various List- headers are standardised, whereas I
1523 don't think Precedence: ever was.
1525 PH/18 Refactored debugging code in route_finduser() to show more information,
1526 in particular, the error code if getpwnam() issues one.
1528 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1529 This is apparently needed in addition to the PH/07 change above to avoid
1530 any possible encoding problems.
1532 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1533 but not after initializing Perl.
1535 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1536 output them only if debugging. By default they are written stderr,
1537 apparently, which is not desirable.
1539 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1542 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1545 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1547 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1548 authenticated or an ident call has been made. Suppress the default
1549 values for $authenticated_id and $authenticated_sender (but permit -oMai
1550 and -oMas) when testing with -bh.
1552 PH/24 Re-jigged the order of the tests in the default configuration so that the
1553 tests for valid domains and recipients precede the DNS black list and CSA
1554 tests, on the grounds that those ones are more expensive.
1556 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1557 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1558 This bug exists in every version of Exim that I still have, right back to
1561 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1562 However, an attempt to turn on case-sensitivity in a regex key by
1563 including (?-i) didn't work because the subject string was already
1564 lowercased, and the effects were non-intuitive. It turns out that a
1565 one-line patch can be used to allow (?-i) to work as expected.
1571 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1572 other effects) broke the use of negated acl sub-conditions.
1574 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1577 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1578 "Deny" causes Exim to reject the incoming connection with a 554 error.
1579 Unfortunately, if there is a major crisis, such as a disk failure,
1580 tcp-wrappers gives "deny", whereas what one would like would be some
1581 kind of temporary error. A kludge has been added to help with this.
1582 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1583 554 error is used if errno is still zero or contains ENOENT (which occurs
1584 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1587 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1589 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1590 errors. Otherwise a message that provokes a temporary error (when other
1591 messages do not) can cause a whole host to time out.
1593 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1594 addresses were routed directly to files or pipes from a redirect router.
1595 File deliveries just didn't batch; pipe deliveries might have suffered
1598 PH/06 A failure to get a lock for a hints database would erroneously always say
1599 "Failed to get write lock", even when it was really a read lock.
1601 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1602 of 0600. This has been changed to use the value of the lockfile_mode
1603 option (which defaults to 0600).
1605 PH/08 Applied small patch from the Sieve maintainer.
1607 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1608 folder from quota calculations, a direct delivery into this folder messed
1609 up the contents of the maildirsize file. This was because the regex was
1610 used only to exclude .Trash (or whatever) when the size of the mailbox
1611 was calculated. There was no check that a delivery was happening into an
1612 excluded directory. This bug has been fixed by ignoring all quota
1613 processing for deliveries into excluded directories.
1615 PH/10 Added the maildirfolder_create_regex option to appendfile.
1621 PH/01 The code for finding all the local interface addresses on a FreeBSD
1622 system running IPv6 was broken. This may well have applied to all BSD
1623 systems, as well as to others that have similar system calls. The broken
1624 code found IPv4 interfaces correctly, but gave incorrect values for the
1625 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1626 that it would not match correctly against @[] and not recognize the IPv6
1629 PH/02 The ipliteral router was not recognizing addresses of the form user@
1630 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1632 PH/03 Added disable_ipv6.
1634 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1635 decoded form, because it is most often used to construct To: headers
1636 lines in autoreplies, and the decoded form may well be syntactically
1637 invalid. However, $reply_address has leading white space removed, and all
1638 newlines turned into spaces so that the autoreply transport does not
1641 PH/05 If group was specified without a user on a router, and no group or user
1642 was specified on a transport, the group from the router was ignored.
1644 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1645 for visible compile-time settings that can be used to change these
1646 numbers, for those that want even more. Backwards compatibility with old
1647 spool files has been maintained. However, going back to a previous Exim
1648 release will lost any variables that are in spool files.
1650 PH/07 Two small changes when running in the test harness: increase delay when
1651 passing a TCP/IP connection to a new process, in case the original
1652 process has to generate a bounce, and remove special handling of
1653 127.0.0.2 (sic), which is no longer necessary.
1655 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1656 be the same on different OS.
1658 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1661 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1662 whether --show-vars was specified or not
1664 JJ/02 exipick: Added support for new ACL variable spool format introduced
1667 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1668 syntactically invalid From: or Reply-to: line, and a filter used this to
1669 generate an autoreply, and therefore failed to obtain an address for the
1670 autoreply, Exim could try to deliver to a non-existent relative file
1671 name, causing unrelated and misleading errors. What now happens is that
1672 it logs this as a hard delivery error, but does not attempt to create a
1675 PH/11 The exinext utility has a -C option for testing purposes, but although
1676 the given file was scanned by exinext itself; it wasn't being passed on
1677 when Exim was called.
1679 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1680 an end-of-file indication when reading a command response.
1682 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1683 compiled. In many other places in Exim, IPv6 addresses are always
1684 recognized, so I have changed this. It also means that IPv4 domain
1685 literals of the form [IPV4:n.n.n.n] are now always recognized.
1687 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1688 used if the router is not running as root, for example, when verifying at
1689 ACL time, or when using -bh. The debugging output from this situation was
1690 non-existent - all you got was a failure to exec. I have made two
1693 (a) Failures to set uid/gid, the current directory, or a process leader
1694 in a subprocess such as that created by queryprogram now generate
1695 suitable debugging ouput when -d is set.
1697 (b) The queryprogram router detects when it is not running as root,
1698 outputs suitable debugging information if -d is set, and then runs
1699 the subprocess without attempting to change uid/gid.
1701 PH/15 Minor change to Makefile for building test_host (undocumented testing
1704 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1705 additional section of a DNS packet that returns MX or SRV records.
1706 Instead, it always explicitly searches for A/AAAA records. This avoids
1707 major problems that occur when a DNS server includes only records of one
1708 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1709 fixed another bug: if SRV records were looked up and the corresponding
1710 address records were *not* found in the additional section, the port
1711 values from the SRV records were lost.
1713 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1714 using the correct key (the original address) when searching the retry
1715 rules in order to find which one to use for generating the retry hint.
1717 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1718 adding the default one. Similarly, if it contains a Reply-To: header, the
1719 errors_reply_to option, if set, is not used.
1721 PH/19 When calculating a retry time, Exim used to measure the "time since
1722 failure" by looking at the "first failed" field in the retry record. Now
1723 it does not use this if it is later than than the arrival time of the
1724 message. Instead it uses the arrival time. This makes for better
1725 behaviour in cases where some deliveries succeed, thus re-setting the
1726 "first failed" field. An example is a quota failure for a huge message
1727 when small messages continue to be delivered. Without this change, the
1728 "time since failure" will always be short, possible causing more frequent
1729 delivery attempts for the huge message than are intended.
1730 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1732 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1733 $1, $2, $3) because the numerical variables can be reset during some
1734 expansion items (e.g. "match"), thereby losing the authentication data.
1736 PH/21 Make -bV show the size of off_t variables so that the test suite can
1737 decide whether to run tests for quotas > 2G.
1739 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1740 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1741 filecount value is greater than 2G or if a quota value is greater than 2G
1742 on a system where the size of off_t is not greater than 4, a panic error
1745 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1746 never match. The debug and -bh output now contains an explicit error
1747 message indicating a malformed IPv4 address or mask.
1749 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1750 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1751 PH/23 above applies.
1753 PH/25 Do not write to syslog when running in the test harness. The only
1754 occasion when this arises is a failure to open the main or panic logs
1755 (for which there is an explicit test).
1757 PH/26 Added the /no_tell option to "control=freeze".
1759 PH/27 If a host name lookup failed very early in a connection, for example, if
1760 the IP address matched host_lookup and the reverse lookup yielded a name
1761 that did not have a forward lookup, an error message of the form "no IP
1762 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1763 could be logged. Now it outputs the IP address instead of "NULL".
1765 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1766 allows the sender and the authenticated sender to be set when
1767 submitting a message from within Exim. Since child_open_exim() is
1768 documented for local_scan(), the new function should be too.
1770 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1771 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1772 results in an empty string is now treated as unset.
1774 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1776 PH/31 Added sender_verify_fail logging option.
1778 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1779 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1780 tidied the source and removed it altogether.
1782 PH/33 When a queue run was abandoned because the load average was too high, a
1783 log line was always written; now it is written only if the queue_run log
1784 selector is set. In addition, the log line for abandonment now contains
1785 information about the queue run such as the pid. This is always present
1786 in "start" and "stop" lines but was omitted from the "abandon" line.
1788 PH/34 Omit spaces between a header name and the colon in the error message that
1789 is given when verify = headers_syntax fails (if there are lots of them,
1790 the message gets confusing).
1792 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1793 names, as there are now some PTR records that contain slashes. This check
1794 is only to protect against broken name servers that fall over on strange
1795 characters, so the fact that it applies to all lookups doesn't matter.
1797 PH/36 Now that the new test suite is complete, we can remove some of the
1798 special code in Exim that was needed for the old test suite. For example,
1799 sorting DNS records because real resolvers return them in an arbitrary
1800 order. The new test suite's fake resolver always returns records in the
1803 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1804 bounces) except when queue_only is set, to avoid logging races between
1805 the different processes.
1807 PH/38 Panic-die if .include specifies a non-absolute path.
1809 PH/39 A tweak to the "H" retry rule from its user.
1811 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1812 a label. They prevented compilation on older perls.
1814 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1815 a warning to be raised on newish perls.
1817 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1818 on queue. Changes to match documented behaviour of showing count of
1819 messages matching specified criteria.
1821 PH/40 Changed the default ident timeout from 30s to 5s.
1823 PH/41 Added support for the use of login_cap features, on those BSD systems
1824 that have them, for controlling the resources used by pipe deliveries.
1826 PH/42 The content-scanning code uses fopen() to create files in which to put
1827 message data. Previously it was not paying any attention to the mode of
1828 the files. Exim runs with umask(0) because the rest of the code creates
1829 files with open(), and sets the required mode explicitly. Thus, these
1830 files were ending up world-writeable. This was not a big issue, because,
1831 being within the spool directory, they were not world-accessible. I have
1832 created a function called modefopen, which takes an additional mode
1833 argument. It sets umask(777), creates the file, chmods it to the required
1834 mode, then resets the umask. All the relevant calls to fopen() in the
1835 content scanning code have been changed to use this function.
1837 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1838 to 24 hours. This avoids potential overflow problems when processing G
1839 and H retry rules. I suspect nobody ever tinkers with this value.
1841 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1843 PH/45 When the plaintext authenticator is running as a client, the server's
1844 challenges are checked to ensure they are valid base64 strings. By
1845 default, the authentication attempt is cancelled if an invalid string is
1846 received. Setting client_ignore_invalid_base64 true ignores these errors.
1847 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1848 they are received. Thus, the responses can be made to depend on the
1849 challenges. If an invalid string is ignored, an empty string is placed in
1852 PH/46 Messages that are created by the autoreply transport now contains a
1853 References: header, in accordance with RFCs 2822 and 3834.
1855 PH/47 Added authenticated_sender_force to the smtp transport.
1857 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1859 PH/49 Installed latest patch from the Sieve maintainer.
1861 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1862 was also set, the appendfile transport was unnecessarily scanning a
1863 directory of message files (e.g. for maildir delivery) to find the count
1864 of files (along with the size), even though it did not need this
1865 information. It now does the scan only if it needs to find either the
1866 size of the count of files.
1868 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1870 PH/52 Two bugs concerned with error handling when the smtp transport is
1873 (i) Exim was not creating retry information for temporary errors given
1874 for individual recipients after the DATA command when the smtp transport
1875 was used in LMTP mode. This meant that they could be retried too
1876 frequently, and not timed out correctly.
1878 (ii) Exim was setting the flag that allows error details to be returned
1879 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1880 recipients that were returned after the DATA command.
1882 PH/53 This is related to PH/52, but is more general: for any failing address,
1883 when detailed error information was permitted to be returned to the
1884 sender, but the error was temporary, then after the final timeout, only
1885 "retry timeout exceeded" was returned. Now it returns the full error as
1886 well as "retry timeout exceeded".
1888 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1889 do this, and (what is worse) MTAs that accept it.
1891 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1892 will now be deprecated.
1894 PH/56 New os.c-cygwin from the Cygwin maintainer.
1896 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1897 formats (previously only available in exim formats via -bpr, -bpru,
1898 and -bpra. Now also available in native and exiqgrep formats)
1900 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1901 with very large, slow to parse queues
1903 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1905 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1907 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1908 responses to authentication challenges, though it was showing the
1909 challenges; (ii) I've removed the CR characters from the debug output for
1912 PH/58 Allow for the insertion of a newline as well as a space when a string
1913 is turned into more than one encoded-word during RFC 2047 encoding. The
1914 Sieve code now uses this.
1916 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1917 data_4xx, lost_connection, tls_required.
1919 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1920 message was being sent as an SMTP response.
1922 PH/61 Add -l and -k options to exicyclog.
1924 PH/62 When verifying, if an address was redirected to one new address, so that
1925 verification continued, and the new address failed or deferred after
1926 having set something in $address_data, the value of $address_data was not
1927 passed back to the ACL. This was different to the case when no
1928 redirection occurred. The value is now passed back in both cases.
1930 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1931 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1932 use login_cap.h, so on its own it isn't the distinguishing feature. The
1933 new name refers directly to the setclassresources() function.
1935 PH/65 Added configuration files for NetBSD3.
1937 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1939 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1940 is preferred over IPv4.
1942 PH/68 The bounce_return_message and bounce_return_body options were not being
1943 honoured for bounces generated during the reception of non-SMTP messages.
1944 In particular, this applied to messages rejected by the ACL. This bug has
1945 been fixed. However, if bounce_return_message is true and bounce_return_
1946 body is false, the headers that are returned for a non-SMTP message
1947 include only those that have been read before the error was detected.
1948 (In the case of an ACL rejection, they have all been read.)
1950 PH/69 The HTML version of the specification is now built in a directory called
1951 spec_html instead of spec.html, because the latter looks like a path with
1952 a MIME-type, and this confuses some software.
1954 PH/70 Catch two compiler warnings in sieve.c.
1956 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1957 function verify_get_ident() calls ip_connect() to connect a socket, but
1958 if the "connect()" function timed out, ip_connect() used to close the
1959 socket. However, verify_get_ident() also closes the socket later, and in
1960 between Exim writes to the log, which may get opened at this point. When
1961 the socket was closed in ip_connect(), the log could get the same file
1962 descriptor number as the socket. This naturally causes chaos. The fix is
1963 not to close the socket in ip_connect(); the socket should be closed by
1964 the function that creates it. There was only one place in the code where
1965 this was missing, in the iplookup router, which I don't think anybody now
1966 uses, but I've fixed it anyway.
1968 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1969 well as to direct DNS lookups. Otherwise the handling of names in host
1970 lists is inconsistent and therefore confusing.
1976 PH/01 Two changes to the default runtime configuration:
1978 (1) Move the checks for relay_from_hosts and authenticated clients from
1979 after to before the (commented out) DNS black list checks.
1981 (2) Add control=submission to the relay_from_hosts and authenticated
1982 clients checks, on the grounds that messages accepted by these
1983 statements are most likely to be submissions.
1985 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1987 (1) Generate an error if the third argument for the ${prvs expansion is
1990 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1993 (3) Reset the variables that are obtained from the first argument of
1994 ${prvscheck and used in the second argument before leaving the code,
1995 because their memory is reclaimed, so using them afterwards may do
1998 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1999 one (it's much easier than Tom thought :-).
2001 (5) Because of (4), we can now allow for the use of $prvscheck_result
2002 inside the third argument.
2004 PH/03 For some reason, the default setting of PATH when running a command from
2005 a pipe transport was just "/usr/bin". I have changed it to
2008 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2009 anything to be listed in the output from -bV.
2011 PH/05 When a filter generated an autoreply, the entire To: header line was
2012 quoted in the delivery log line, like this:
2014 => >A.N.Other <ano@some.domain> <original@ddress> ...
2016 This has been changed so that it extracts the operative address. There
2017 may be more than one such address. If so, they are comma-separated, like
2020 => >ano@some.domain,ona@other.domain <original@ddress> ...
2022 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2023 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2024 not being looked up in the rDNS to get a host name, Exim was showing the
2025 IP address twice in Received: lines, even though the IP addresses were
2026 identical. For example:
2028 Received: from [1.2.3.4] (helo=[1.2.3.4])
2030 However, if the real host name was known, it was omitting the HELO data
2031 if it matched the actual IP address. This has been tidied up so that it
2032 doesn't show the same IP address twice.
2034 PH/07 When both +timestamp and +memory debugging was on, the value given by
2035 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2036 called by the debug printing, thereby overwriting the timestamp buffer.
2037 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2039 PH/08 When the original message was included in an autoreply transport, it
2040 always said "this is a copy of the message, including all the headers",
2041 even if body_only or headers_only was set. It now gives an appropriate
2044 PH/09 Applied a patch from the Sieve maintainer which:
2046 o fixes some comments
2047 o adds the (disabled) notify extension core
2048 o adds some debug output for the result of if/elsif tests
2049 o points to the current vacation draft in the documentation
2050 and documents the missing references header update
2054 o fixes a bug in processing the envelope test (when testing
2055 multiple envelope elements, the last element determinted the
2058 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2059 Electronic Mail") by including:
2061 Auto-submitted: auto-generated
2063 in the messages that it generates (bounce messages and others, such as
2064 warnings). In the case of bounce messages for non-SMTP mesages, there was
2065 also a typo: it was using "Auto_submitted" (underscore instead of
2066 hyphen). Since every message generated by Exim is necessarily in response
2067 to another message, thes have all been changed to:
2069 Auto-Submitted: auto-replied
2071 in accordance with these statements in the RFC:
2073 The auto-replied keyword:
2075 - SHOULD be used on messages sent in direct response to another
2076 message by an automatic process,
2078 - MUST NOT be used on manually-generated messages,
2080 - MAY be used on Delivery Status Notifications (DSNs) and Message
2081 Disposition Notifications (MDNs),
2083 - MUST NOT be used on messages generated by automatic or periodic
2084 processes, except for messages which are automatic responses to
2087 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2088 to the default Received: header definition.
2090 PH/12 Added log selector acl_warn_skipped (default on).
2092 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2093 variables because (a) they are in the wrong storage pool and (b) even if
2094 they were copied, it wouldn't work properly because of the caching.
2096 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2097 checking when decoding. Apparently there are clients that generate
2098 overlong encoded strings. Why am I not surprised?
2100 PH/15 If the first argument of "${if match_address" was not empty, but did not
2101 contain an "@" character, Exim crashed. Now it writes a panic log message
2102 and treats the condition as false.
2104 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2106 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2107 extension "envelope-auth". The code is finished and in agreement with
2108 other implementations, but there is no documentation so far and in fact,
2109 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2110 not changing the active code.
2112 Print executed "if" and "elsif" statements when debugging is used. This
2113 helps a great deal to understand what a filter does.
2115 Document more things not specified clearly in RFC3028. I had all this
2116 sorted out, when out of a sudden new issues came to my mind. Oops."
2118 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2121 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2122 canonical form (as documented). However, after a host name lookup from
2123 the IP address, check_host() was doing a simple string comparison with
2124 addresses acquired from the DNS when checking that the found name did
2125 have the original IP as one of its addresses. Since any found IPv6
2126 addresses are likely to be in abbreviated form, the comparison could
2127 fail. Luckily, there already exists a function for doing the comparison
2128 by converting both addresses to binary, so now that is used instead of
2129 the text comparison.
2131 PH/20 There was another similar case to PH/19, when a complete host name was
2132 given in a host list; looking up its IP address could give an abbreviated
2133 form, whereas the current host's name might or might not be abbreviated.
2134 The same fix has been applied.
2140 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2141 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2144 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2145 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2147 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2149 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2150 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2151 Also, for "Auto-Submitted", treat anything other than "no" as
2152 non-personal, in accordance with RFC 3834. (Previously it treated
2153 anything starting "auto-" as non-personal.)
2155 TF/01 The control=submission/name=... option had a problem with syntax
2156 errors if the name included a slash character. The /name= option
2157 now slurps the rest of the string, so it can include any characters
2158 but it must come last in the list of options (after /sender_retain
2161 PH/05 Some modifications to the interface to the fake nameserver for the new
2169 TK/01 Added the "success_on_redirect" address verification option. See
2170 NewStuff for rationale and an example.
2172 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2174 PH/02 Patch to exigrep to allow it to work on syslog lines.
2176 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2177 fread() to skip over the body file's header line, because in Cygwin the
2178 header line is locked and is inaccessible.
2180 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2181 co-exist for some time) to make it clear that it is the Exim ID that is
2182 referenced, not the Message-ID: header line.
2184 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2185 string_format() function, because snprintf() does not exist on all
2188 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2189 ${stat: expansion item.
2191 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2192 protocol synchronization error", to keep the pedants happy.
2194 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2195 well as for IRIX systems, when gcc is being used. See the host.c source
2198 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2200 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2203 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2204 transport and to the smtp transport in LMTP mode.
2206 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2208 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2209 run a filter in a subprocess. This could lead to confusion in subsequent
2210 lookups in the parent process. There should also be a search_tidyup() at
2211 the end of the subprocess.
2213 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2214 only if the host matched helo_try_verify_hosts, which caused the
2215 verification to occur when the EHLO/HELO command was issued. The ACL just
2216 tested the remembered result. Now, if a previous verification attempt has
2217 not happened, "verify = helo" does it there and then.
2219 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2221 TK/03 Fix log output including CR from clamd.
2223 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2224 exist provoked a memory error which could cause a segfault.
2226 PH/15 Installed PCRE 6.2
2228 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2230 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2231 of the problem. Specifically, suggested +O2 rather than +O1 for the
2234 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2236 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2237 record, the port was not correctly logged when the outgoing_port log
2238 selector was set (it logged the transort's default port).
2240 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2241 fallback_hosts, and "hosts" in the smtp transport.
2243 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2244 host errors such as "Connection refused".
2246 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2247 authentication with radiusclient 0.4.9:
2249 - Error returned from rc_read_config was caught wrongly
2250 - Username/password not passed on to radius server due to wrong length.
2252 The presumption is that some radiusclient API changes for 4.51/PH/17
2253 were not taken care of correctly. The code is still untested by me (my
2254 Linux distribution still has 0.3.2 of radiusclient), but it was
2255 contributed by a Radius user.
2257 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2258 expanding the "port" option of the smtp transport.
2260 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2261 while reading a MIME header. Thanks to Tom Hughes for a patch.
2263 PH/24 Include config.h inside local_scan.h so that configuration settings are
2266 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2267 that in an ACL for RCPT (for example), you can examine exactly what was
2270 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2271 commands, but it was not correctly comparing the address with the actual
2272 client host address. Thus, it would show the EHLO address in Received:
2273 header lines when this was not necessary.
2275 PH/27 Added the % operator to ${eval:}.
2277 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2278 it should be ignoring failures (because with -C, for example, it has lost
2279 privilege). It wasn't ignoring creation failures other than "already
2282 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2285 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2286 because an input line was too long, either on its own, or by virtue of
2287 too many continuations, the temporary file was not being removed, and the
2288 return code was incorrect.
2290 PH/31 Missing "BOOL" in function definition in filtertest.c.
2292 PH/32 Applied Sieve patches from the maintainer.
2294 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2296 PH/33 Added "verify = not_blind".
2298 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2299 Local/Makefile (with some defaults set). These are used in built scripts
2300 such as exicyclog, but they have never been used in the exim_install
2301 script (though there are many overriding facilities there). I have
2302 arranged that the exim_install script now takes note of these two
2305 PH/35 Installed configuration files for Dragonfly.
2307 PH/36 When a locally submitted message by a trusted user did not contain a
2308 From: header, and the sender address was obtained from -f or from an SMTP
2309 MAIL command, and the trusted user did not use -F to supply a sender
2310 name, $originator_name was incorrectly used when constructing a From:
2311 header. Furthermore, $originator_name was used for submission mode
2312 messages from external hosts without From: headers in a similar way,
2313 which is clearly wrong.
2315 PH/37 Added control=suppress_local_fixups.
2317 PH/38 When log_selector = +received_sender was set, and the addition of the
2318 sender made the log line's construction buffer exactly full, or one byte
2319 less than full, an overflow happened when the terminating "\n" was
2322 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2323 when the result of a list match is failure because a DNS lookup failed.
2325 PH/40 RM_COMMAND is now used in the building process.
2327 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2328 the "build-* directories that it finds.
2330 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2331 address in a domain literal was a prefix of an interface address.
2333 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2334 when verifying a sender address, unless rewrite_headers is false.
2336 PH/44 Wrote a long comment about why errors_to addresses are verified as
2337 recipients, not senders.
2339 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2340 the ratelimit ACL was added.
2342 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2344 PH/46 Added extra information about PostgreSQL errors to the error string.
2346 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2347 avoiding the need to install special zones in a real server. This is
2348 backwards compatible; if it can't find the fake resolver, it drops back.
2349 Thus, both old and new test suites can be run.
2351 TF/02 Added util/ratelimit.pl
2353 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2354 clock is set back in time.
2356 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2357 Candler <B.Candler@pobox.com>.
2359 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2360 OK for addresses that are the result of redirections.
2362 PH/48 A number of further additions for the benefit of the new test suite,
2363 including a fake gethostbyname() that interfaces to the fake DNS resolver
2366 TF/06 The fix for widen_domains has also been applied to qualify_single and
2367 search_parents which are the other dnslookup options that can cause
2370 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2373 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2375 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2376 service" if it encountered an MX record with an empty target hostname.
2377 The message is now "an MX or SRV record indicated no SMTP service".
2379 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2380 if the DNS of the sending site is misconfigured. This is quite a
2381 common situation. This change restores the behaviour of treating a
2382 helo verification defer as a failure.
2384 PH/51 If self=fail was set on a router, the bounce message did not include the
2385 actual error message.
2391 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2393 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2394 command itself ended in error, the underlying message about the transport
2395 filter timeout was being overwritten with the pipe command error. Now the
2396 underlying error message should be appended to the second error message.
2398 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2400 PH/02 Reduce the amount of output that "make" produces by default. Full output
2401 can still be requested.
2403 PH/03 The warning log line about a condition test deferring for a "warn" verb
2404 was being output only once per connection, rather than after each
2405 occurrence (because it was using the same function as for successful
2406 "warn" verbs). This seems wrong, so I have changed it.
2408 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2409 it should not have, which might have caused a crash in the right
2410 circumstances, but probably never did.
2412 PH/04 Installed a modified version of Tony Finch's patch to make submission
2413 mode fix the return path as well as the Sender: header line, and to
2414 add a /name= option so that you can make the user's friendly name appear
2417 TF/03 Added the control = fakedefer ACL modifier.
2419 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2420 Mark Lowes for thorough testing.
2422 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2424 TK/03 Merged latest SRS patch from Miles Wilton.
2426 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2427 with the definition in sysexits.h (which is #included earlier).
2428 Fortunately, Exim does not actually use EX_OK. The code used to try to
2429 preserve the sysexits.h value, by assumimg that macro definitions were
2430 scanned for macro replacements. I have been disabused of this notion,
2431 so now the code just undefines EX_OK before #including unistd.h.
2433 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2434 in the smtp transport. When a block could not be written in a single
2435 write() function, the timeout was being re-applied to each part-write.
2436 This seems wrong - if the receiver was accepting one byte at a time it
2437 would take for ever. The timeout is now adjusted when this happens. It
2438 doesn't have to be particularly precise.
2440 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2441 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2443 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2444 without a domain if the check to <postmaster@domain> fails.
2446 SC/01 Eximstats: added -xls and the ability to specify output files
2447 (patch written by Frank Heydlauf).
2449 SC/02 Eximstats: use FileHandles for outputing results.
2451 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2453 SC/04 Eximstats: fixed display of large numbers with -nvr option
2455 SC/05 Eximstats: fixed merging of reports with empty tables.
2457 SC/06 Eximstats: added the -include_original_destination flag
2459 SC/07 Eximstats: removed tabs and trailing whitespace.
2461 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2463 TK/06 MBOX spool code: Add real "From " MBOX separator line
2464 so the .eml file is really in mbox format (even though
2465 most programs do not really care). Patch from Alex Miller.
2467 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2468 The latter is generated from $received_to and is only set if the
2469 message has one envelope recipient. SA can use these headers,
2470 obviously out-of-the-box. Patch from Alex Miller.
2472 PH/08 The ${def test on a variable was returning false if the variable's
2473 value was "0", contrary to what the specification has always said!
2474 The result should be true unless the variable is empty.
2476 PH/09 The syntax error of a character other than { following "${if
2477 def:variable_name" (after optional whitespace) was not being diagnosed.
2478 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2479 accidental colon was present, for example, could give incorrect results.
2481 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2482 result is used (not including appendfile, where other changes are about
2485 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2486 This involved changing a lot of size variables from int to off_t. It
2487 should work with maildirs and everything.
2489 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2490 spamd dying while we are connected to it.
2492 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2495 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2496 function for BDB 4.3.
2498 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2500 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2501 the sources for PCRE have been re-arranged and more files are now
2504 PH/15 The code I had for printing potentially long long variables in PH/11
2505 above was not the best (it lost precision). The length of off_t variables
2506 is now inspected at build time, and an appropriate printing format (%ld
2507 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2508 to be "long long int" or "long int". This is needed for the internal
2509 formatting function string_vformat().
2511 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2512 the configuration file to be ":syslog", then the script "guesses" where
2513 the logs files are, rather than using the compiled in default. In our
2514 case the guess is not the same as the compiled default, so the script
2515 suddenly stopped working when I started to use syslog. The patch checks
2516 to see if log_file_path is "". If so, it attempts to read it from exim
2517 with no configuration file to get the compiled in version, before it
2518 falls back to the previous guessing code."
2520 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2521 implementing BATV in an Exim configuration. See NewStuff for the gory
2524 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2525 Makefile that are specific to HP-UX.
2527 PH/18 If the "use_postmaster" option was set for a recipient callout together
2528 with the "random" option, the postmaster address was used as the MAIL
2529 FROM address for the random test, but not for the subsequent recipient
2530 test. It is now used for both.
2532 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2533 patch removes a few documentation additions to RFC 3028, because the
2534 latest draft now contains them. It adds the new en;ascii-case comparator
2535 and a new error check for 8bit text in MIME parts. Comparator and
2536 require names are now matched exactly. I enabled the subaddress
2537 extension, but it is not well tested yet (read: it works for me)."
2539 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2540 rework some of the code of TK/09 above to avoid the hardwired use of
2541 "%lld" and "long long". Replaced the call to snprintf() with a call to
2544 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2545 records point to non-existent hosts", "retry timeout exceeded", and
2546 "retry time not reached for any host after a long failure period".
2548 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2549 experimental DomainKeys support:
2551 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2552 (2) On an error such as an illegally used "control", the wrong name for
2553 the control was given.
2555 These problems did NOT occur unless DomainKeys support was compiled.
2557 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2559 PH/24 Added ${if match_ip condition.
2561 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2562 databases so that it will be absolutely obvious if a crash occurs in the
2563 DB library. This is a regular occurrence (often caused by mis-matched
2566 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2567 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2568 picked up on a user's system that detects such things. There doesn't seem
2569 to be a gcc warning option for this - only an attribute that has to be
2570 put on the function's prototype. It seems that in Fedora Core 4 they have
2571 set this on a number of new functions. No doubt there will be more in due
2574 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2575 specify a transport. However, if an address that was verified by such a
2576 router was the subject of a callout, Exim crashed because it tried to
2577 read the rcpt_include_affixes from the non-existent transport. Now it
2578 just assumes that the setting of that option is false. This bug was
2579 introduced by 4.51/PH/31.
2581 PH/28 Changed -d+all to exclude +memory, because that information is very
2582 rarely of interest, but it makes the output a lot bigger. People tend to
2583 do -d+all out of habit.
2585 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2586 code in os-type was giving problems when libc.so lives in lib64, like on
2589 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2590 aren't the modern standard, and it seems that some systems' include files
2591 don't always have them. Exim was already checking for some of the newer
2592 ones like T_AAAA, and defining it itself. I've added checks for all the
2593 record types that Exim uses.
2595 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2596 not automatically generating a new one, as it is supposed to. This
2597 prevented TLS from working. If the file did exist, but contained invalid
2598 data, a new version was generated, as expected. It was only the case of a
2599 non-existent file that was broken.
2601 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2602 with a change in libdomainkeys > 0.64.
2604 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2605 from DNS. If the selector record carries the flag, it now has
2606 precedence over the domain-wide flag.
2608 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2610 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2611 the use of a port name that isn't defined in /etc/services) occurred, the
2612 message was deferred as in a normal delivery, and thus remained on the
2613 spool, instead of being failed because of the mua_wrapper setting. This
2614 is now fixed, and I tidied up some of the mua_wrapper messages at the
2617 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2618 the messages in a hash of arrays rather than using individual hashes.
2619 This is a bit cleaner and results in dramatic memory savings, albeit
2620 at a slight CPU cost.
2622 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2623 as requested by Marc Sherman.
2625 SC/10 Eximstats: added histograms for user specified patterns as requested
2628 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2630 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2631 fopen() in the content-scanning modules that did not already have it.
2637 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2638 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2640 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2642 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2644 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2645 to test Sieve filters that use "vacation".
2647 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2648 that changes the way the GnuTLS parameters are stored in the cache file.
2649 The new format can be generated externally. For backward compatibility,
2650 if the data in the cache doesn't make sense, Exim assumes it has read an
2651 old-format file, and it generates new data and writes a new file. This
2652 means that you can't go back to an older release without removing the
2655 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2656 work if there are any delivery delays because "one_time" forces the
2657 parent to be marked "delivered", so its unseen clone is never tried
2658 again. For this reason, Exim now forbids the simultaneous setting of
2661 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2662 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2663 entry if you want to know the details. The fix, however, neglected to
2664 consider the case when local delivery batching is involved. The test for
2665 "previously delivered" was not happening when checking to see if an
2666 address could be batched with a previous (undelivered) one; under
2667 certain circumstances this could lead to multiple deliveries to the same
2670 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2671 in its include files, and this causes problems building Exim.
2673 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2674 header_syntax) but Exim was just ignoring anything given after a slash.
2675 In particular, this caused confusion with an attempt to use "verify =
2676 reverse_host_lookup/defer_ok". An error is now given when options are
2677 supplied for verify items that do not have them. (Maybe reverse_host_
2678 lookup should have a defer_ok option, but that's a different point.)
2680 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2681 defined by RFC 821) to 2048, because there were problems with some AUTH
2682 commands, and RFC 1869 says the size should be increased for extended
2683 SMTP commands that take arguments.
2685 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2688 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2689 "unknown" error; now it says that the functionality isn't in the binary.
2691 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2692 an address' error message when a string expansion fails (syntax or
2693 whatever). Otherwise the password may appear in the log. Following change
2694 PH/42 below, there is no longer a chance of it appearing in a bounce
2697 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2699 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2700 of its addresses was ever tried. (Bugzilla bug #2.)
2702 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2703 the result incorrectly in the debug output. (It correctly added a newline
2704 to what was transported.)
2706 TF/01 Added $received_time.
2708 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2709 commented out examples of how to interface to a virus scanner and to
2710 SpamAssassin. Also added commented examples of av_scanner and
2711 spamd_address settings.
2713 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2714 and controls are allowed in which ACLs. There were a couple of minor
2715 errors. Some of the entries in the conditions table (which is a table of
2716 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2717 negation of where the condition IS allowed.
2719 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2721 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2722 header file does not have a version number, so I've had to invent a new
2723 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2724 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2725 radiusclient), but it was contributed by a Radius user.
2727 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2728 files or directories, for OpenSSL.
2730 PH/19 When an Exim process that is running as root has to create an Exim log
2731 file, it does so in a subprocess that runs as exim:exim so as to get the
2732 ownership right at creation (otherwise, other Exim processes might see
2733 the file with the wrong ownership). There was no test for failure of this
2734 fork() call, which would lead to the process getting stuck as it waited
2735 for a non-existent subprocess. Forks do occasionally fail when resources
2736 run out. I reviewed all the other calls to fork(); they all seem to check
2739 PH/20 When checking for unexpected SMTP input at connect time (before writing
2740 the banner), Exim was not dealing correctly with a non-positive return
2741 from the read() function. If the client had disconnected by this time,
2742 the result was a log entry for a synchronization error with an empty
2743 string after "input=" when read() returned zero. If read() returned -1
2744 (an event I could not check), uninitialized data bytes were printed.
2745 There were reports of junk text (parts of files, etc) appearing after
2748 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2750 PH/22 Added support for macro redefinition, and (re)definition in between
2751 driver and ACL definitions.
2753 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2754 forgetting to use the resulting value; it was using the unexpanded value.
2756 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2757 hadn't been configured. The fix is from Juergen Kreileder, who
2758 understands it better than I do:
2760 "Here's what I see happening with three configured cyrus_sasl
2761 authenticators configured (plain, login, cram-md5):
2763 On startup auth_cyrus_sasl_init() gets called for each of these.
2764 This means three calls to sasl_listmech() without a specified mech_list.
2765 => SASL tests which mechs of all available mechs actually work
2766 => three warnings about OTP not working
2767 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2769 With the patch, sasl_listmech() also gets called three times. But now
2770 SASL's mech_list option is set to the server_mech specified in the the
2771 authenticator. Or in other words, the answer from sasl_listmech()
2772 gets limited to just the mech you're testing for (which is different
2774 => the return list contains just 'plain' or 'login', 'cram-md5' or
2775 nothing depending on the value of ob->server_mech.
2777 I've just tested the patch: Authentication still works fine,
2778 unavailable mechs specified in the exim configuration are still
2779 caught, and the auth.log warnings about OTP are gone."
2781 PH/25 When debugging is enabled, the contents of the command line are added
2782 to the debugging output, even when log_selector=+arguments is not
2785 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2786 answer is "GNU", and only if the return is "GNU/something" is the answer
2789 PH/27 $acl_verify_message is now set immediately after the failure of a
2790 verification in an ACL, and so is available in subsequent modifiers. In
2791 particular, the message can be preserved by coding like this:
2793 warn !verify = sender
2794 set acl_m0 = $acl_verify_message
2796 Previously, $acl_verify_message was set only while expanding "message"
2797 and "log_message" when a very denied access.
2799 PH/28 Modified OS/os.c-Linux with
2801 -#ifndef OS_LOAD_AVERAGE
2802 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2804 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2805 nomenclature these days.)
2807 PH/29 Installed patch from the Sieve maintainer that adds the options
2808 sieve_useraddress and sieve_subaddress to the redirect router.
2810 PH/30 In these circumstances:
2811 . Two addresses routed to the same list of hosts;
2812 . First host does not offer TLS;
2813 . First host accepts first address;
2814 . First host gives temporary error to second address;
2815 . Second host offers TLS and a TLS session is established;
2816 . Second host accepts second address.
2817 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2818 and peerdn, if requested) that were in fact used only for the second
2821 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2822 attention to any local part prefix or suffix that was matched by the
2823 router that accepted the address. It now behaves in the same way as it
2824 does for delivery: the affixes are removed from the local part unless
2825 rcpt_include_affixes is set on the transport.
2827 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2828 timeout during the DATA phase of an incoming message.
2830 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2831 applied a patch sanctioned by the Sieve maintainer.
2833 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2834 the uid or gid is negative. A case of a negative gid caused this to be
2835 noticed. The fix allows for either to be negative.
2837 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2838 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2841 PH/36 The change PH/12 above was broken. Fixed it.
2843 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2844 the grounds that routing the same address twice would always produce the
2845 same answer. This might have been true once, but it is certainly no
2846 longer true now. Routing a child address may depend on the previous
2847 routing that produced that child. Some complicated redirection strategies
2848 went wrong when messages had multiple recipients, and made Exim's
2849 behaviour dependent on the order in which the addresses were given.
2851 I have moved the duplicate checking until after the routing is complete.
2852 Exim scans the addresses that are assigned to local and remote
2853 transports, and removes any duplicates. This means that more work will be
2854 done, as duplicates will always all be routed, but duplicates are
2855 presumably rare, so I don't expect this is of any significance.
2857 For deliveries to pipes, files, and autoreplies, the duplicate checking
2858 still happens during the routing process, since they are not going to be
2861 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2862 It corrects a timeout issue with spamd. This is Ian's comment: "The
2863 background is that sometimes spamd either never reads data from a
2864 connection it has accepted, or it never writes response data. The exiscan
2865 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2866 blindly assumes that writes won't block so it may never time out."
2868 PH/39 Allow G after quota size as well as K and M.
2870 PH/40 The value set for $authenticated_id in an authenticator may not contain
2871 binary zeroes or newlines because the value is written to log lines and
2872 to spool files. There was no check on this. Now the value is run through
2873 the string_printing() function so that such characters are converted to
2874 printable escape sequences.
2876 PH/41 $message_linecount is a new variable that contains the total number of
2877 lines in the message. Compare $body_linecount, which is the count for the
2880 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2881 bounce and delay warning messages, except in certain special cases, which
2884 (a) An SMTP error message from a remote host;
2885 (b) A message specified in a :fail: redirection;
2886 (c) A message specified in a "fail" command in a system filter;
2887 (d) A message specified in a FAIL return from the queryprogram router;
2888 (e) A message specified by the cannot_route_message router option.
2890 In these cases only, Exim does include the error details in bounce and
2891 warning messages. There are also a few cases where bland messages such
2892 as "unrouteable address" or "local delivery error" are given.
2894 PH/43 $value is now also set for the "else" part of a ${run expansion.
2896 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2897 being worked on, but at least Exim now implements the latest version to
2900 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2901 process to complete was treated as a delivery failure, a timeout while
2902 writing the message to the pipe was logged, but erroneously treated as a
2903 successful delivery. Such timeouts include transport filter timeouts. For
2904 consistency with the overall process timeout, these timeouts are now
2905 treated as errors, giving rise to delivery failures by default. However,
2906 there is now a new Boolean option for the pipe transport called
2907 timeout_defer, which, if set TRUE, converts the failures into defers for
2908 both kinds of timeout. A transport filter timeout is now identified in
2911 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2912 systems where "make" and "gmake" are different, calling "gmake" at top
2913 level broke things. I've arranged for the value of $(MAKE) to be passed
2914 from the Makefile to this script so that it can call the same version of
2918 A note about Exim versions 4.44 and 4.50
2919 ----------------------------------------
2921 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2922 changes of various kinds. As a consequence, a big documentation update was
2923 needed. This delayed the release for rather longer than seemed good, especially
2924 in the light of a couple of (minor) security issues. Therefore, the changes
2925 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2926 release. So 4.44 and 4.50 are in effect two different branches that both start
2929 I have left the 4.50 change log unchanged; it contains all the changes since
2930 4.43. The change log for 4.44 is below; many of its items are identical to
2931 those for 4.50. This seems to be the most sensible way to preserve the
2932 historical information.
2938 1. Minor wording change to the doc/README.SIEVE file.
2940 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2941 computation of the current number of files was incorrect.
2943 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2944 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2947 4. Give more explanation in the error message when the command for a transport
2948 filter fails to execute.
2950 5. There are several places where Exim runs a non-Exim command in a
2951 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2952 was being done only for the command run by the queryprogram router. It is
2953 now done for all such subprocesses. The other cases are: ${run, transport
2954 filters, and the commands run by the lmtp and pipe transports.
2956 6. Added CONFIGURE_GROUP build-time option.
2958 7. Some older OS have a limit of 256 on the maximum number of file
2959 descriptors. Exim was using setrlimit() to set 1000 as a large value
2960 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2961 systems. I've change it so that if it can't get 1000, it tries for 256.
2963 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2964 was an oversight, and furthermore, ever since the addition of extra
2965 controls (e.g. 4.43/32), the checks on when to allow different forms of
2966 "control" were broken. There should now be diagnostics for all cases when a
2967 control that does not make sense is encountered.
2969 9. Added the /retain_sender option to "control=submission".
2971 10. $recipients is now available in the predata ACL (oversight).
2973 11. Tidy the search cache before the fork to do a delivery from a message
2974 received from the command line. Otherwise the child will trigger a lookup
2975 failure and thereby defer the delivery if it tries to use (for example) a
2976 cached ldap connection that the parent has called unbind on.
2978 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2979 of $address_data from the recipient verification was clobbered by the
2980 sender verification.
2982 13. The value of address_data from a sender verification is now available in
2983 $sender_address_data in subsequent conditions in the ACL statement.
2985 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2987 15. Added a new option "connect=<time>" to callout options, to set a different
2990 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2991 was its contents. (It was OK if the option was not defined at all.)
2993 17. A "Completed" log line is now written for messages that are removed from
2994 the spool by the -Mrm option.
2996 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2997 information about exactly what failed.
2999 19. Added -dd to debug only the daemon process.
3001 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3002 handles timeouts, both on the server side and network timeouts. Renamed the
3003 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3005 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3006 It is now set to "smtps".
3008 22. $host_address is now set to the target address during the checking of
3009 ignore_target_hosts.
3011 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3012 being passed; this would have caused $sender_host_name to have been used if
3013 matching the list had actually called for a host name (not very likely,
3014 since this list is usually IP addresses). A host name is now passed as
3017 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3018 code that specifies a non-restarting handler (typically sigaction() in
3019 modern systems) in an attempt to fix a rare and obscure crash bug.
3021 25. Narrowed the window for a race in the daemon that could cause it to ignore
3022 SIGCHLD signals. This is not a major problem, because they are used only to
3023 wake it up if nothing else does.
3025 26. A malformed maildirsize file could cause Exim to calculate negative values
3026 for the mailbox size or file count. Odd effects could occur as a result.
3027 The maildirsize information is now recalculated if the size or filecount
3030 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3031 support for a long time. Removed HAVE_SYS_VFS_H.
3033 28. Installed the latest version of exipick from John Jetmore.
3035 29. In an address list, if the pattern was not a regular expression, an empty
3036 subject address (from a bounce message) matched only if the pattern was an
3037 empty string. Non-empty patterns were not even tested. This was the wrong
3038 because it is perfectly reasonable to use an empty address as part of a
3039 database query. An empty address is now tested by patterns that are
3040 lookups. However, all the other forms of pattern expect the subject to
3041 contain a local part and a domain, and therefore, for them, an empty
3042 address still always fails if the pattern is not itself empty.
3044 30. Exim went into a mad DNS loop when attempting to do a callout where the
3045 host was specified on an smtp transport, and looking it up yielded more
3046 than one IP address.
3048 31. Re-factored the code for checking spool and log partition space into a
3049 function that finds that data and another that does the check. The former
3050 is then used to implement four new variables: $spool_space, $log_space,
3051 $spool_inodes, and $log_inodes.
3053 32. The RFC2047 encoding function was originally intended for short strings
3054 such as real names; it was not keeping to the 75-character limit for
3055 encoded words that the RFC imposes. It now respects the limit, and
3056 generates multiple encoded words if necessary. To be on the safe side, I
3057 have increased the buffer size for the ${rfc2047: expansion operator from
3060 33. It is now permitted to omit both strings after an "if" condition; if the
3061 condition is true, the result is "true". As before, when the second string
3062 is omitted, a false condition yields an empty string. This makes it less
3063 cumbersome to write custom ACL and router conditions.
3065 34. Failure to deliver a bounce message always caused it to be frozen, even if
3066 there was an errors_to setting on the router. The errors_to setting is now
3069 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3070 canonical form (fully expanded) before being placed in
3071 $sender_host_address.
3073 36. The table in the code that translates DNS record types into text (T_A to
3074 "A" for instance) was missing entries for NS and CNAME. It is just possible
3075 that this could have caused confusion if both these types were looked up
3076 for the same domain, because the text type is used as part of Exim's
3077 per-process caching. But the chance of anyone hitting this buglet seems
3080 37. The dnsdb lookup has been extended in a number of ways.
3082 (1) There is a new type, "zns", which walks up the domain tree until it
3083 finds some nameserver records. It should be used with care.
3085 (2) There is a new type, "mxh", which is like "mx" except that it returns
3086 just the host names, not the priorities.
3088 (3) It is now possible to give a list of domains (or IP addresses) to be
3089 looked up. The behaviour when one of the lookups defers can be
3090 controlled by a keyword.
3092 (4) It is now possible to specify the separator character for use when
3093 multiple records are returned.
3095 38. The dnslists ACL condition has been extended: it is now possible to supply
3096 a list of IP addresses and/or domains to be looked up in a particular DNS
3099 39. Added log_selector=+queue_time_overall.
3101 40. When running the queue in the test harness, wait just a tad after forking a
3102 delivery process, to get repeatability of debugging output.
3104 41. Include certificate and key file names in error message when GnuTLS fails
3105 to set them up, because the GnuTLS error message doesn't include the name
3106 of the failing file when there is a problem reading it.
3108 42. Allow both -bf and -bF in the same test run.
3110 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3112 44. The "Exiscan patch" is now merged into the mainline Exim source.
3114 45. Sometimes the final signoff response after QUIT could fail to get
3115 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3116 < 0 before doing a fflush(). This bug looks as though it goes back to the
3117 introduction of TLS in release 3.20, but "sometimes" must have been rare
3118 because the tests only now provoked it.
3120 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3121 (this can affect the format of dates).
3123 47. exim_tidydb, when checking for the continued existence of a message for
3124 which it has found a message-specific retry record, was not finding
3125 messages that were in split spool directories. Consequently, it was
3126 deleting retry records that should have stayed in existence.
3128 48. Steve fixed some bugs in eximstats.
3130 49. The SPA authentication driver was not abandoning authentication and moving
3131 on to the next authenticator when an expansion was forced to fail,
3132 contradicting the general specification for all authenticators. Instead it
3133 was generating a temporary error. It now behaves as specified.
3135 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3136 (the order specifies the preference for clients). The order is now AES256,
3137 AES128, 3DES, ARCFOUR128.
3139 51. Small patch to Sieve code - explicitly set From: when generating an
3142 52. Exim crashed if a remote delivery caused a very long error message to be
3143 recorded - for instance if somebody sent an entire SpamAssassin report back
3144 as a large number of 550 error lines. This bug was coincidentally fixed by
3145 increasing the size of one of Exim's internal buffers (big_buffer) that
3146 happened as part of the Exiscan merge. However, to be on the safe side, I
3147 have made the code more robust (and fixed the comments that describe what
3150 53. Now that there can be additional text after "Completed" in log lines (if
3151 the queue_time_overall log selector is set), a one-byte patch to exigrep
3152 was needed to allow it to recognize "Completed" as not the last thing in
3155 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3156 patch that reportedly fixes this has been added. I am not expert enough to
3157 create a test for it. This is what the patch creator wrote:
3159 "I found a little strange behaviour of ldap code when working with
3160 Windows 2003 AD Domain, where users was placed in more than one
3161 Organization Units. When I tried to give exim partial DN, the exit code
3162 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3163 But simultaneously result of request was absolutely normal ldap result,
3164 so I produce this patch..."
3166 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3167 REFERENCE, so I have modified the code to exclude the patch when that macro
3170 55. Some experimental protocols are using DNS PTR records for new purposes. The
3171 keys for these records are domain names, not reversed IP addresses. The
3172 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3173 leaves it alone. Component reversal etc. now happens only for IP addresses.
3176 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3178 57. Double the size of the debug message buffer (to 2048) so that more of very
3179 long debug lines gets shown.
3181 58. The exicyclog utility now does better if the number of log files to keep
3182 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3184 59. Two changes related to the smtp_active_hostname option:
3186 (1) $smtp_active_hostname is now available as a variable.
3187 (2) The default for smtp_banner uses $smtp_active_hostname instead
3188 of $primary_hostname.
3190 60. The host_aton() function is supposed to be passed a string that is known
3191 to be a valid IP address. However, in the case of IPv6 addresses, it was
3192 not checking this. This is a hostage to fortune. Exim now panics and dies
3193 if the condition is not met. A case was found where this could be provoked
3194 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3195 components; fortuitously, this particular loophole had already been fixed
3196 by change 4.50/55 above.
3198 If there are any other similar loopholes, the new check in host_aton()
3199 itself should stop them being exploited. The report I received stated that
3200 data on the command line could provoke the exploit when Exim was running as
3201 exim, but did not say which command line option was involved. All I could
3202 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3203 running as the user.
3206 61. There was a buffer overflow vulnerability in the SPA authentication code
3207 (which came originally from the Samba project). I have added a test to the
3208 spa_base64_to_bits() function which I hope fixes it.
3211 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3212 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3214 63. The daemon start-up calls getloadavg() while still root for those OS that
3215 need the first call to be done as root, but it missed one case: when
3216 deliver_queue_load_max is set with deliver_drop_privilege. This is
3217 necessary for the benefit of the queue runner, because there is no re-exec
3218 when deliver_drop_privilege is set.
3220 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3221 This has been fixed.
3223 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3224 was in use, was not putting the data itself into the right store pool;
3225 consequently, it could be overwritten for a subsequent message in the same
3226 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3229 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3231 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3232 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3233 the function were treating the return as a boolean value, which happened to
3234 work because 0=false and not-0=true, but is not correct code.
3236 68. The host_aton() function was not handling scoped IPv6 addresses (those
3237 with, for example, "%eth0" on the end) correctly.
3239 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3240 negated items (that is, ~something) in unsigned ints. Some compilers
3241 apparently mutter when there is no cast.
3243 70. If an address verification called from an ACL failed, and did not produce a
3244 user-specific message (i.e. there was only a "system" message), nothing was
3245 put in $acl_verify_message. In this situation, it now puts the system
3248 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3249 session; change 4.31/43 added the unwanted input to the log line - except
3250 that it did not do this in the start of session case. It now does.
3252 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3253 This is wrong and can cause the other end to generate a synchronization
3254 error if it is another Exim or anything else that does the synchronization
3255 check. A QUIT command is no longer sent after a timeout.
3257 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3258 during host lookups.
3260 74. The defer_ok option of callout verification was not working if it was used
3261 when verifying addresses in header lines, that is, for this case:
3263 verify = header_sender/callout=defer_ok
3265 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3266 those file descriptors could be used for SMTP connections. If anything
3267 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3268 could be sent to the SMTP client, causing chaos. The daemon now opens
3269 stdin, stdout, and stderr to /dev/null when it puts itself into the
3272 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3273 log by default. The user can override this with suitable Perl magic.
3275 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3276 the log message when discard triggers, was not working for the DATA ACL or
3277 for the non-SMTP ACL.
3279 78. Error message wording change in sieve.c.
3281 79. If smtp_accept_max_per_host was set, the number of connections could be
3282 restricted to fewer than expected, because the daemon was trying to set up
3283 a new connection before checking whether the processes handling previous
3284 connections had finished. The check for completed processes is now done
3285 earlier. On busy systems, this bug wouldn't be noticed because something
3286 else would have woken the daemon, and it would have reaped the completed
3289 80. If a message was submitted locally by a user whose login name contained one
3290 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3291 It caused a spool format error. I have fixed the spool reading code. A
3292 related problem was that the "from" clause in the Received: line became
3293 illegal because of the space(s). It is now covered by ${quote_local_part.
3295 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3298 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3300 83. Patch from Sieve maintainer to fix unterminated string problem in
3301 "vacation" handling.
3303 84. Some minor changes to the Linux configuration files to help with other
3304 OS variants using glibc.
3306 85. One more patch for Sieve to update vacation handling to latest spec.
3309 ----------------------------------------------------
3310 See the note above about the 4.44 and 4.50 releases.
3311 ----------------------------------------------------
3317 1. Change 4.43/35 introduced a bug that caused file counts to be
3318 incorrectly computed when quota_filecount was set in an appendfile
3321 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3322 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3325 3. Give more explanation in the error message when the command for a transport
3326 filter fails to execute.
3328 4. There are several places where Exim runs a non-Exim command in a
3329 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3330 was being done only for the command run by the queryprogram router. It is
3331 now done for all such subprocesses. The other cases are: ${run, transport
3332 filters, and the commands run by the lmtp and pipe transports.
3334 5. Some older OS have a limit of 256 on the maximum number of file
3335 descriptors. Exim was using setrlimit() to set 1000 as a large value
3336 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3337 systems. I've change it so that if it can't get 1000, it tries for 256.
3339 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3340 was an oversight, and furthermore, ever since the addition of extra
3341 controls (e.g. 4.43/32), the checks on when to allow different forms of
3342 "control" were broken. There should now be diagnostics for all cases when a
3343 control that does not make sense is encountered.
3345 7. $recipients is now available in the predata ACL (oversight).
3347 8. Tidy the search cache before the fork to do a delivery from a message
3348 received from the command line. Otherwise the child will trigger a lookup
3349 failure and thereby defer the delivery if it tries to use (for example) a
3350 cached ldap connection that the parent has called unbind on.
3352 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3353 of $address_data from the recipient verification was clobbered by the
3354 sender verification.
3356 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3357 was its contents. (It was OK if the option was not defined at all.)
3359 11. A "Completed" log line is now written for messages that are removed from
3360 the spool by the -Mrm option.
3362 12. $host_address is now set to the target address during the checking of
3363 ignore_target_hosts.
3365 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3366 being passed; this would have caused $sender_host_name to have been used if
3367 matching the list had actually called for a host name (not very likely,
3368 since this list is usually IP addresses). A host name is now passed as
3371 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3372 code that specifies a non-restarting handler (typically sigaction() in
3373 modern systems) in an attempt to fix a rare and obscure crash bug.
3375 15. Narrowed the window for a race in the daemon that could cause it to ignore
3376 SIGCHLD signals. This is not a major problem, because they are used only to
3377 wake it up if nothing else does.
3379 16. A malformed maildirsize file could cause Exim to calculate negative values
3380 for the mailbox size or file count. Odd effects could occur as a result.
3381 The maildirsize information is now recalculated if the size or filecount
3384 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3385 support for a long time. Removed HAVE_SYS_VFS_H.
3387 18. Updated exipick to current release from John Jetmore.
3389 19. Allow an empty sender to be matched against a lookup in an address list.
3390 Previously the only cases considered were a regular expression, or an
3393 20. Exim went into a mad DNS lookup loop when doing a callout where the
3394 host was specified on the transport, if the DNS lookup yielded more than
3397 21. The RFC2047 encoding function was originally intended for short strings
3398 such as real names; it was not keeping to the 75-character limit for
3399 encoded words that the RFC imposes. It now respects the limit, and
3400 generates multiple encoded words if necessary. To be on the safe side, I
3401 have increased the buffer size for the ${rfc2047: expansion operator from
3404 22. Failure to deliver a bounce message always caused it to be frozen, even if
3405 there was an errors_to setting on the router. The errors_to setting is now
3408 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3409 canonical form (fully expanded) before being placed in
3410 $sender_host_address.
3412 24. Updated eximstats to version 1.33
3414 25. Include certificate and key file names in error message when GnuTLS fails
3415 to set them up, because the GnuTLS error message doesn't include the name
3416 of the failing file when there is a problem reading it.
3418 26. Expand error message when OpenSSL has problems setting up cert/key files.
3421 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3422 (this can affect the format of dates).
3424 28. exim_tidydb, when checking for the continued existence of a message for
3425 which it has found a message-specific retry record, was not finding
3426 messages that were in split spool directories. Consequently, it was
3427 deleting retry records that should have stayed in existence.
3429 29. eximstats updated to version 1.35
3430 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3431 1.35 - bugfix such that pie charts by volume are generated correctly
3433 30. The SPA authentication driver was not abandoning authentication and moving
3434 on to the next authenticator when an expansion was forced to fail,
3435 contradicting the general specification for all authenticators. Instead it
3436 was generating a temporary error. It now behaves as specified.
3438 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3439 (the order specifies the preference for clients). The order is now AES256,
3440 AES128, 3DES, ARCFOUR128.
3442 31. Small patch to Sieve code - explicitly set From: when generating an
3445 32. Exim crashed if a remote delivery caused a very long error message to be
3446 recorded - for instance if somebody sent an entire SpamAssassin report back
3447 as a large number of 550 error lines. This bug was coincidentally fixed by
3448 increasing the size of one of Exim's internal buffers (big_buffer) that
3449 happened as part of the Exiscan merge. However, to be on the safe side, I
3450 have made the code more robust (and fixed the comments that describe what
3453 33. Some experimental protocols are using DNS PTR records for new purposes. The
3454 keys for these records are domain names, not reversed IP addresses. The
3455 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3456 leaves it alone. Component reversal etc. now happens only for IP addresses.
3459 34. The host_aton() function is supposed to be passed a string that is known
3460 to be a valid IP address. However, in the case of IPv6 addresses, it was
3461 not checking this. This is a hostage to fortune. Exim now panics and dies
3462 if the condition is not met. A case was found where this could be provoked
3463 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3464 components; fortuitously, this particular loophole had already been fixed
3465 by change 4.50/55 or 4.44/33 above.
3467 If there are any other similar loopholes, the new check in host_aton()
3468 itself should stop them being exploited. The report I received stated that
3469 data on the command line could provoke the exploit when Exim was running as
3470 exim, but did not say which command line option was involved. All I could
3471 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3472 running as the user.
3475 35. There was a buffer overflow vulnerability in the SPA authentication code
3476 (which came originally from the Samba project). I have added a test to the
3477 spa_base64_to_bits() function which I hope fixes it.
3480 36. The daemon start-up calls getloadavg() while still root for those OS that
3481 need the first call to be done as root, but it missed one case: when
3482 deliver_queue_load_max is set with deliver_drop_privilege. This is
3483 necessary for the benefit of the queue runner, because there is no re-exec
3484 when deliver_drop_privilege is set.
3486 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3487 was in use, was not putting the data itself into the right store pool;
3488 consequently, it could be overwritten for a subsequent message in the same
3489 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3492 38. Sometimes the final signoff response after QUIT could fail to get
3493 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3494 < 0 before doing a fflush(). This bug looks as though it goes back to the
3495 introduction of TLS in release 3.20, but "sometimes" must have been rare
3496 because the tests only now provoked it.
3502 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3503 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3504 Now it returns an integer. A number of calls were still expecting a T/F
3505 return. Fortuitously, in all cases, the tests worked in OK situations,
3506 which is the norm. However, things would have gone wrong on any write
3507 failures on the smtp file descriptor. This function is used when sending
3508 messages over SMTP and also when doing verify callouts.
3510 2. When Exim is called to do synchronous delivery of a locally submitted
3511 message (the -odf or -odi options), it no longer closes stderr before doing
3514 3. Implemented the mua_wrapper option.
3516 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3518 5. Implemented the functions header_remove(), header_testname(),
3519 header_add_at_position(), and receive_remove_recipient(), and exported them
3522 6. If an ACL "warn" statement specified the addition of headers, Exim already
3523 inserted X-ACL-Warn: at the start if there was no header name. However, it
3524 was not making this test for the second and subsequent header lines if
3525 there were newlines in the string. This meant that an invalid header could
3526 be inserted if Exim was badly configured.
3528 7. Allow an ACL "warn" statement to add header lines at the start or after all
3529 the Received: headers, as well as at the end.
3531 8. Added the rcpt_4xx retry error code.
3533 9. Added postmaster_mailfrom=xxx to callout verification option.
3535 10. Added mailfrom=xxxx to the callout verification option, for verify=
3538 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3539 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3541 12. Inserted some casts to stop certain compilers warning when using pointer
3542 differences as field lengths or precisions in printf-type calls (mostly
3543 affecting debugging statements).
3545 13. Added optional readline() support for -be (dynamically loaded).
3547 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3548 same clock tick as a message's arrival, so that its received time was the
3549 same as the "first fail" time on the retry record, and that message
3550 remained on the queue past the ultimate address timeout, every queue runner
3551 would try a delivery (because it was past the ultimate address timeout) but
3552 after another failure, the ultimate address timeout, which should have then
3553 bounced the address, did not kick in. This was a "< instead of <=" error;
3554 in most cases the first failure would have been in the next clock tick
3555 after the received time, and all would be well.
3557 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3558 being recognized when the domain list was tested by the match_domain
3559 condition in an expansion string.
3561 16. Added the ${str2b64: operator.
3563 17. Exim was always calling setrlimit() to set a large limit for the number of
3564 processes, without checking whether the existing limit was already
3565 adequate. (It did check for the limit on file descriptors.) Furthermore,
3566 errors from getrlimit() and setrlimit() were being ignored. Now they are
3567 logged to the main and panic logs, but Exim does carry on, to try to do its
3568 job under whatever limits there are.
3570 18. Imported PCRE 5.0.
3572 19. Trivial typo in log message " temporarily refused connection" (the leading
3575 20. If the log selector return_path_on_delivery was set and an address was
3576 redirected to /dev/null, the delivery process crashed because it assumed
3577 that a return path would always be set for a "successful" delivery. In this
3578 case, the whole delivery is bypassed as an optimization, and therefore no
3581 21. Internal re-arrangement: the function for sending a challenge and reading
3582 a response while authentication was assuming a zero-terminated challenge
3583 string. It's now changed to take a pointer and a length, to allow for
3584 binary data in such strings.
3586 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3588 23. Exim was not respecting finduser_retries when seeking the login of the
3589 uid under which it was called; it was always trying 10 times. (The default
3590 setting of finduser_retries is zero.) Also, it was sleeping after the final
3591 failure, which is pointless.
3593 24. Implemented tls_on_connect_ports.
3595 25. Implemented acl_smtp_predata.
3597 26. If the domain in control=submission is set empty, Exim assumes that the
3598 authenticated id is a complete email address when it generates From: or
3599 Sender: header lines.
3601 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3602 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3603 chown and chgrp in /bin and hostname in /usr/bin.
3605 28. Exim was keeping the "process log" file open after each use, just as it
3606 does for the main log. This opens the possibility of it remaining open for
3607 long periods when the USR1 signal hits a daemon. Occasional processlog
3608 errors were reported, that could have been caused by this. Anyway, it seems
3609 much more sensible not to leave this file open at all, so that is what now
3612 29. The long-running daemon process does not normally write to the log once it
3613 has entered its main loop, and it closes the log before doing so. This is
3614 so that log files can straightforwardly be renamed and moved. However,
3615 there are a couple of unusual error situations where the daemon does write
3616 log entries, and I had neglected to close the log afterwards.
3618 30. The text of an SMTP error response that was received during a remote
3619 delivery was being truncated at 512 bytes. This is too short for some of
3620 the long messages that one sometimes sees. I've increased the limit to
3623 31. It is now possible to make retry rules that apply only when a message has a
3624 specific sender, in particular, an empty sender.
3626 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3627 it possible to be selective about when SMTP synchronization is enforced.
3629 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3631 32. Implemented hosts_connection_nolog.
3633 33. Added an ACL for QUIT.
3635 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3638 35. Added mailbox_size and mailbox_filecount to appendfile.
3640 36. Added control = no_multiline_responses to ACLs.
3642 37. There was a bug in the logic of the code that waits for the clock to tick
3643 in the case where the clock went backwards by a substantial amount such
3644 that the microsecond fraction of "now" was more than the microsecond
3645 fraction of "then" (but the whole seconds number was less).
3647 38. Added support for the libradius Radius client library this is found on
3648 FreeBSD (previously only the radiusclient library was supported).
3654 1. When certain lookups returned multiple values in the form name=value, the
3655 quoting of the values was not always being done properly. Specifically:
3656 (a) If the value started with a double quote, but contained no whitespace,
3658 (b) If the value contained whitespace other than a space character (i.e.
3659 tabs or newlines or carriage returns) it was not quoted.
3660 This fix has been applied to the mysql and pgsql lookups by writing a
3661 separate quoting function and calling it from the lookup code. The fix
3662 should probably also be applied to nisplus, ibase and oracle lookups, but
3663 since I cannot test any of those, I have not disturbed their existing code.
3665 2. A hit in the callout cache for a specific address caused a log line with no
3666 reason for rejecting RCPT. Now it says "Previous (cached) callout
3667 verification failure".
3669 3. There was an off-by-one bug in the queryprogram router. An over-long
3670 return line was truncated at 256 instead of 255 characters, thereby
3671 overflowing its buffer with the terminating zero. As well as fixing this, I
3672 have increased the buffer size to 1024 (and made a note to document this).
3674 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3675 when Exim is waiting for an SMTP response from a remote server, Exim
3676 restarts its select() call on the socket, thereby resetting its timeout.
3677 This is not a problem when such interrupts are rare. Somebody set up a cron
3678 job to run exiwhat every 2 minutes, which is less than the normal select()
3679 timeout (5 or 10 minutes). This meant that the select() timeout never
3680 kicked in because it was always reset. I have fixed this by comparing the
3681 time when an interrupt arrives with the time at the start of the first call
3682 to select(). If more time than the timeout has elapsed, the interrupt is
3683 treated as a timeout.
3685 5. Some internal re-factoring in preparation for the addition of Sieve
3686 extensions (by MH). In particular, the "personal" test is moved to a
3687 separate function, and given an option for scanning Cc: and Bcc: (which is
3688 not set for Exim filters).
3690 6. When Exim created an email address using the login of the caller as the
3691 local part (e.g. when creating a From: or Sender: header line), it was not
3692 quoting the local part when it contained special characters such as @.
3694 7. Installed new OpenBSD configuration files.
3696 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3697 try to make them clearer.
3699 9. Callout options, other than the timeout value, were being ignored when
3700 verifying sender addresses in header lines. For example, when using
3702 verify = header_sender/callout=no_cache
3704 the cache was (incorrectly) being used.
3706 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3707 only the Cygwin environment.
3709 11. When return_path_on_delivery was set as a log selector, if different remote
3710 addresses in the same message used different return paths and parallel
3711 remote delivery occurred, the wrong values would sometimes be logged.
3712 (Whenever a remote delivery process finished, the return path value from
3713 the most recently started remote delivery process was logged.)
3715 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3716 lines when AUTH and/or TLS are in use. This is the "received protocol"
3717 field. Exim used to use "asmtp" for authenticated SMTP, without any
3718 indication (in the protocol name) for TLS use. Now it follows the RFC and
3719 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3720 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3721 names appear in log lines as well as in Received: header lines.
3723 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3724 extensions, and comparison tests, and to fix some bugs.
3726 14. Changes to the "personal" filter test:
3728 (1) The test was buggy in that it was just doing the equivalent of
3729 "contains" tests on header lines. For example, if a user's address was
3730 anne@some.where, the "personal" test would incorrectly be true for
3732 To: susanne@some.where
3734 This test is now done by extracting each address from the header in turn,
3735 and checking the entire address. Other tests that are part of "personal"
3736 are now done using regular expressions (for example, to check local parts
3737 of addresses in From: header lines).
3739 (2) The list of non-personal local parts in From: addresses has been
3740 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3741 taken from the Sieve specification recommendations.
3743 (3) If the message contains any header line starting with "List-" it is
3744 treated as non-personal.
3746 (4) The test for "circular" in the Subject: header line has been removed
3747 because it now seems ill-conceived.
3749 15. Minor typos in src/EDITME comments corrected.
3751 16. Installed latest exipick from John Jetmore.
3753 17. If headers_add on a router specified a text string that was too long for
3754 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3755 of string_sprintf() is now avoided.
3757 18. $message_body_size was not set (it was always zero) when running the DATA
3758 ACL and the local_scan() function.
3760 19. For the "mail" command in an Exim filter, no default was being set for
3761 the once_repeat time, causing a random time value to be used if "once" was
3762 specified. (If the value happened to be <= 0, no repeat happened.) The
3763 default is now 0s, meaning "never repeat". The "vacation" command was OK
3764 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3765 (I found it when inspecting the code).
3767 20. There is now an overall timeout for performing a callout verification. It
3768 defaults to 4 times the callout timeout, which applies to individual SMTP
3769 commands during the callout. The overall timeout applies when there is more
3770 than one host that can be tried. The timeout is checked before trying the
3771 next host. This prevents very long delays if there are a large number of
3772 hosts and all are timing out (e.g. when the network connections are timing
3773 out). The value of the overall timeout can be changed by specifying an
3774 additional sub-option for "callout", called "maxwait". For example:
3776 verify = sender/callout=5s,maxwait=20s
3778 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3779 to the end before writing, but this should make it even safer).
3781 22. Exim was forgetting that it had advertised PIPELINING for the second and
3782 subsequent messages on an SMTP connection. It was also not resetting its
3783 memory on STARTTLS and an internal HELO.
3785 23. When Exim logs an SMTP synchronization error within a session, it now
3786 records whether PIPELINING has been advertised or not.
3788 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3789 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3790 rather than long int.
3792 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3794 26. Added the never_mail option to autoreply.
3800 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3801 crash if the getsockname() call failed; this can happen if a connection is
3802 closed very soon after it is established. The problem was simply in the
3803 order in which certain operations were done, causing Exim to try to write
3804 to the SMTP stream before it had set up the file descriptor. The bug has
3805 been fixed by making things happen in the correct order.
3811 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3812 before the connection was closed, thus losing the rejection response.
3814 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3815 some early Solaris releases, but causes trouble in current releases where
3816 socklen_t is defined.
3818 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3821 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3824 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3825 that releasing the top of it at the end releases what was used for sub-
3826 expansions (unless the block got too big). However, discard this block if
3827 the first thing is a variable or header, so that we can use its block when
3828 it is dynamic (useful for very large $message_headers, for example).
3830 6. Lookups now cache *every* query, not just the most recent. A new, separate
3831 store pool is used for this. It can be recovered when all lookup caches are
3832 flushed. Lookups now release memory at the end of their result strings.
3833 This has involved some general refactoring of the lookup sources.
3835 7. Some code has been added to the store_xxx() functions to reduce the amount
3836 of flapping under certain conditions.
3838 8. log_incoming_interface used to affect only the <= reception log lines. Now
3839 it causes the local interface and port to be added to several more SMTP log
3840 lines, for example "SMTP connection from", and rejection lines.
3842 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3844 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3846 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3847 could be overwritten at the end of the current message (or the start of a
3848 new message if it was set in a HELO ACL). The value is now preserved for
3849 the duration of the SMTP connection.
3851 12. If a transport had a headers_rewrite setting, and a matching header line
3852 contained an unqualified address, that address was qualified, even if it
3853 did not match any rewriting rules. The underlying bug was that the values
3854 of the flags that permit the existence of unqualified sender and recipient
3855 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3856 non-local messages, and by -bnq for local messages) were not being
3857 preserved with the message after it was received.
3859 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3860 "next input=" as part of the text comprising the host identity instead of
3861 the correct text. The code was using the same buffer for two different
3862 strings. However, depending on which order the printing function evaluated
3863 its arguments, the bug did not always show up. Under Linux, for example, my
3864 test suite worked just fine.
3866 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3867 doesn't work with some older versions of Perl. It has been changed to "my",
3868 which in any case is probably the better facility to use.
3870 15. A really picky compiler found some instances of statements for creating
3871 error messages that either had too many or two few arguments for the format
3874 16. The size of the buffer for calls to the DNS resolver has been increased
3875 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3876 for addresses that have a lot of PTR records. This alleviates a problem; it
3877 does not fully solve it.
3879 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3880 buffer now truncates the list and logs the incident, which is the same
3881 action as happens when Exim is looking up a host name and its aliases.
3882 Previously in this situation something unpredictable would happen;
3883 sometimes it was "internal error: store_reset failed".
3885 18. If a server dropped the connection unexpectedly when an Exim client was
3886 using GnuTLS and trying to read a response, the client delivery process
3887 crashed while trying to generate an error log message.
3889 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3890 string, for example:
3892 warn message = H1: something\nH2: something
3894 the text was added as a single header line from Exim's point of view
3895 though it ended up OK in the delivered message. However, searching for the
3896 second and subsequent header lines using $h_h2: did not work. This has been
3897 fixed. Similarly, if a system filter added multiple headers in this way,
3898 the routers could not see them.
3900 20. Expanded the error message when iplsearch is called with an invalid key to
3901 suggest using net-iplsearch in a host list.
3903 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3904 ACLs are no longer actually imposed (and a message to that effect is
3907 22. If a "gecos" field in a passwd entry contained escaped characters, in
3908 particular, if it contained a \" sequence, Exim got it wrong when building
3909 a From: or a Sender: header from that name. A second bug also caused
3910 incorrect handling when an unquoted " was present following a character
3911 that needed quoting.
3913 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3914 was not being matched caselessly.
3916 24. Arranged for all hyphens in the exim.8 source to be escaped with
3919 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3920 properly. Recipient callout cache records were still being keyed to include
3921 the sender, even when use_sender was set false. This led to far more
3922 callouts that were necessary. The sender is no longer included in the key
3923 when use_sender is false.
3925 26. Added "control = submission" modifier to ACLs.
3927 27. Added the ${base62d: operator to decode base 62 numbers.
3929 28. dnsdb lookups can now access SRV records.
3931 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3932 the configuration file.
3934 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3935 (-v) mode. This makes the output for a verbose queue run more intelligible.
3937 31. Added a use_postmaster feature to recipient callouts.
3939 32. Added the $body_zerocount variable, containing the number of binary zero
3940 bytes in the message body.
3942 33. The time of last modification of the "new" subdirectory is now used as the
3943 "mailbox time last read" when there is a quota error for a maildir
3946 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3948 35. Added +ignore_unknown as a special item in host lists.
3950 36. Code for decoding IPv6 addresses in host lists is now included, even if
3951 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3952 address was recognized as an IP address, but was then not correctly decoded
3953 into binary, causing unexpected and incorrect effects when compared with
3960 1. Very minor rewording of debugging text in manualroute to say "list of
3961 hosts" instead of "hostlist".
3963 2. If verify=header_syntax was set, and a header line with an unqualified
3964 address (no domain) and a large number of spaces between the end of the
3965 name and the colon was received, the reception process suffered a buffer
3966 overflow, and (when I tested it) crashed. This was caused by some obsolete
3967 code that should have been removed. The fix is to remove it!
3969 3. When running in the test harness, delay a bit after writing a bounce
3970 message to get a bit more predictability in the log output.
3972 4. Added a call to search_tidyup() just before forking a reception process. In
3973 theory, someone could use a lookup in the expansion of smtp_accept_max_
3974 per_host which, without the tidyup, could leave open a database connection.
3976 5. Added the variables $recipient_data and $sender_data which get set from a
3977 lookup success in an ACL "recipients" or "senders" condition, or a router
3978 "senders" option, similar to $domain_data and $local_part_data.
3980 6. Moved the writing of debug_print from before to after the "senders" test
3983 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3984 problems for message scanning, either using a data ACL, or using
3985 local_scan() because the Received: header was not generated till after they
3986 were called (in order to set the time as the time of reception completion).
3987 I have revised the way this works. The header is now generated after the
3988 body is received, but before the ACL or local_scan() are called. After they
3989 are run, the timestamp in the header is updated.
3995 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3996 before starting a queue runner without re-exec. This happened only when
3997 deliver_drop_privilege was set or when the Exim user was set to root. The
3998 effect of the bug was that timeouts during subsequent deliveries caused
3999 crashes instead of being properly handled. The handler is now left at its
4000 default (and expected) setting.
4002 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4003 message, again when deliver_drop_privilege is set or Exim is run as root.
4004 The bug described in (1) was not present in this case, but the tidying up
4005 of the other signals was missing. I have made the two cases consistent.
4007 3. The ignore_target_hosts setting on a manualroute router was being ignored
4008 for hosts that were looked up using the /MX notation.
4010 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4013 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4014 operated on the sender address. After changing the $sender_address to <>
4015 for the sender address verify, Exim was re-instated it as the original
4016 (before rewriting) address, but remembering that it had rewritten it, so it
4017 wasn't rewriting it again. This bug also had the effect of breaking the
4018 sender address verification caching when the sender address was rewritten.
4020 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4021 This has been changed so that if the ip literal address matches
4022 ignore_target_hosts, the router declines.
4024 7. Added expansion conditions match_domain, match_address, and match_local_
4025 part (NOT match_host).
4027 8. The placeholder for the Received: header didn't have a length field set.
4029 9. Added code to Exim itself and to exim_lock to test for a specific race
4030 condition that could lead to file corruption when using MBX delivery. The
4031 issue is with the lockfile that is created in /tmp. If this file is removed
4032 after a process has opened it but before that process has acquired a lock,
4033 there is the potential for a second process to recreate the file and also
4034 acquire a lock. This could lead to two Exim processes writing to the file
4035 at the same time. The added code performs the same test as UW imapd; it
4036 checks after acquiring the lock that its file descriptor still refers to
4037 the same named file.
4039 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4040 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4043 11. Added the smtp_active_hostname option. If used, this will typically be made
4044 to depend on the incoming interface address. Because $interface_address is
4045 not set up until the daemon has forked a reception process, error responses
4046 that can happen earlier (such as "too many connections") no longer contain
4049 12. If an expansion in a condition on a "warn" statement fails because a lookup
4050 defers, the "warn" statement is abandoned, and the next ACL statement is
4051 processed. Previously this caused the whole ACL to be aborted.
4053 13. Added the iplsearch lookup type.
4055 14. Added ident_timeout as a log selector.
4057 15. Added tls_certificate_verified as a log selector.
4059 16. Added a global option tls_require_ciphers (compare the smtp transport
4060 option of the same name). This controls incoming TLS connections.
4062 17. I finally figured out how to make tls_require_ciphers do a similar thing
4063 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4064 before starting the TLS session.
4066 18. Tabs are now shown as \t in -bP output.
4068 19. If the log selector return_path_on_delivery was set, Exim crashed when
4069 bouncing a message because it had too many Received: header lines.
4071 20. If two routers both had headers_remove settings, and the first one included
4072 a superfluous trailing colon, the final name in the first list and the
4073 first name in the second list were incorrectly joined into one item (with a
4074 colon in the middle).
4080 1. Added -C and -D options to the exinext utility, mainly to make it easier
4081 to include in the automated testing, but these could be helpful when
4082 multiple configurations are in use.
4084 2. The exinext utility was not formatting the output nicely when there was
4085 an alternate port involved in the retry record key, nor when there was a
4086 message id as well (for retries that were specific to a specific message
4087 and a specific host). It was also confused by IPv6 addresses, because of
4088 the additional colons they contain. I have fixed the IPv4 problem, and
4089 patched it up to do a reasonable job for IPv6.
4091 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4092 delivery, the log line now contains "pipelined" if PIPELINING was used.
4094 4. An SMTP transport process used to panic and die if the bind() call to set
4095 an explicit outgoing interface failed. This has been changed; it is now
4096 treated in the same way as a connect() failure.
4098 5. A reference to $sender_host_name in the part of a conditional expansion
4099 that was being skipped was still causing a DNS lookup. This no longer
4102 6. The def: expansion condition was not recognizing references to header lines
4103 that used bh_ and bheader_.
4105 7. Added the _cache feature to named lists.
4107 8. The code for checking quota_filecount in the appendfile transport was
4108 allowing one more file than it should have been.
4110 9. For compatibility with Sendmail, the command line option
4118 and sets the incoming protocol and host name (for trusted callers). The
4119 host name and its colon can be omitted when only the protocol is to be set.
4120 Note the Exim already has two private options, -pd and -ps, that refer to
4121 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4122 "s", but I don't think that's a major issue.
4124 10. A number of refactoring changes to the code, none of which should affect
4127 (a) The number of logging options was getting close to filling up the
4128 32-bit word that was used as a bit map. I have split them into two classes:
4129 those that are passed in the argument to log_write(), and those that are
4130 only ever tested independently outside of that function. These are now in
4131 separate 32-bit words, so there is plenty of room for expansion again.
4132 There is no change in the user interface or the logging behaviour.
4134 (b) When building, for example, log lines, the code previously used a
4135 macro that called string_cat() twice, in order to add two strings. This is
4136 not really sufficiently general. Furthermore, there was one instance where
4137 it was actually wrong because one of the argument was used twice, and in
4138 one call a function was used. (As it happened, calling the function twice
4139 did not affect the overall behaviour.) The macro has been replaced by a
4140 function that can join an arbitrary number of extra strings onto a growing
4143 (c) The code for expansion conditions now uses a table and a binary chop
4144 instead of a serial search (which was left over from when there were very
4145 few conditions). Also, it now recognizes conditions like "pam" even when
4146 the relevant support is not compiled in: a suitably worded error message is
4147 given if an attempt is made to use such a condition.
4149 11. Added ${time_interval:xxxxx}.
4151 12. A bug was causing one of the ddress fields not to be passed back correctly
4152 from remote delivery subprocesses. The field in question was not being
4153 subsequently used, so this caused to problems in practice.
4155 13. Added new log selectors queue_time and deliver_time.
4157 14. Might have fixed a bug in maildirsizefile handling that threw up
4158 "unexpected character" debug warnings, and recalculated the data
4159 unnecessarily. In any case, I expanded the warning message to give more
4162 15. Added the message "Restricted characters in address" to the statements in
4163 the default ACL that block characters like @ and % in local parts.
4165 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4166 Three changes have been made:
4168 (a) There was a serious bug; a negative response to MAIL caused the whole
4169 recipient domain to be cached as invalid, thereby blocking all messages
4170 to all local parts at the same domain, from all senders. This bug has
4171 been fixed. The domain is no longer cached after a negative response to
4172 MAIL if the sender used is not empty.
4174 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4177 (c) A new callout option, "use_sender" has been added for people who want
4178 the modified behaviour.
4184 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4187 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4188 indeed breaks things for older releases.
4190 3. Added additional logging to the case where there is a problem reading data
4191 from a filter that is running in a subprocess using a pipe, in order to
4192 try to track down a specific problem.
4194 4. Testing facility fudge: when running in the test harness and attempting
4195 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4196 getting "No route to host". Convert this to a timeout.
4198 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4201 6. Some OS don't have socklen_t but use size_t instead. This affects the
4202 fifth argument of getsockopt() amongst other things. This is now
4203 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4204 can be set for individual OS. I have set it for SunOS5, OSF1, and
4205 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4206 some earlier ones do not.
4208 7. Change 4.30/15 was not doing the test caselessly.
4210 8. The standard form for an IPv6 address literal was being rejected by address
4211 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4212 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4213 this, as well as the form without the "IPv6" on the front (but only when
4214 address literals are enabled, of course).
4216 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4218 10. Exim crashed if a message with an empty sender address specified by -f
4219 encountered a router with an errors_to setting. This could be provoked only
4220 by a command such as
4224 where an empty string was supplied; "<>" did not hit this bug.
4226 11. Installed PCRE release 4.5.
4228 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4229 remained set. It is now erased.
4231 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4232 times from message ids (which are base 36 rather than the normal 62).
4234 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4235 were being counted as actual protocol errors, and logged if the log
4236 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4237 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4238 and DATA following a set of rejected RCPTs do not count as protocol errors.
4239 In other words, Exim assumes they were pipelined, though this may not
4240 actually be the case. Of course, in all cases the client gets an
4241 appropriate error code.
4243 15. If a lookup fails in an ACL condition, a message about the failure may
4244 be available; it is used if testing the ACL cannot continue, because most
4245 such messages specify what the cause of the deferral is. However, some
4246 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4247 that caused an old message to be retained and used if a later statement
4248 caused a defer, replacing the real cause of the deferral.
4250 16. If an IP address had so many PTR records that the DNS lookup buffer
4251 was not large enough to hold them, Exim could crash while trying to process
4252 the truncated data. It now detects and logs this case.
4254 17. Further to 4.21/58, another change has been made: if (and only if) the
4255 first line of a message (the first header line) ends with CRLF, a bare LF
4256 in a subsequent header line has a space inserted after it, so as not to
4257 terminate the header.
4259 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4260 unnecessarily, used atoi() instead of strtol(), and didn't check the
4261 termination when getting file sizes from file names by regex.
4263 19. Completely re-implemented the support for maildirsize files, in the light
4264 of a number of problems with the previous contributed implementation
4265 (4.30/29). In particular:
4267 . If the quota is zero, the maildirsize file is maintained, but no quota is
4270 . If the maildir directory does not exist, it is created before any attempt
4271 to write a maildirsize file.
4273 . The quota value in the file is just a cache; if the quota is changed in
4274 the transport, the new value overrides.
4276 . A regular expression is available for excluding directories from the
4279 20. The autoreply transport checks the characters in options that define the
4280 message's headers; it allows continued headers, but it was checking with
4281 isspace() after an embedded newline instead of explicitly looking for a
4284 21. If all the "regular" hosts to which an address was routed had passed their
4285 expiry times, and had not reached their retry times, the address was
4286 bounced, even if fallback hosts were defined. Now Exim should go on to try
4289 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4290 equivalent code in the SMTP transport. Some hosts send humungous responses
4291 to HELO/EHLO, more than 1024 it seems.
4293 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4294 gives compiler warnings in some environments. I've now done it "properly",
4297 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4298 (because of problems with the built-in one) was declared to return uschar *
4299 instead of char *, causing compiler failure.
4301 25. Fixed a file descriptor leak when processing alias/forward files.
4303 26. Fixed a minor format string issue in dbfn.c.
4305 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4307 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4308 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4311 29. When a sender address is verified, it is cached, to save repeating the test
4312 when there is more than one recipient in a message. However, when the
4313 verification involves a callout, it is possible for different callout
4314 options to be set for different recipients. It is too complicated to keep
4315 track of this in the cache, so now Exim always runs a verification when a
4316 callout is required, relying on the callout cache for the optimization.
4317 The overhead is duplication of the address routing, but this should not be
4320 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4321 to be verified with callout=postmaster, and the main callout worked but the
4322 postmaster check failed, the verification correctly failed. However, if a
4323 subsequent RCPT command asked for sender verification *without* the
4324 postmaster check, incorrect caching caused this verification also to fail,
4327 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4328 it was not caching the DNS options (qualify_single, search_parents) that
4329 were used when the lookup failed. A subsequent lookup with different
4330 options therefore always gave the same answer, though there were cases
4331 where it should not have. (Example: a "domains = !$mx_any" option on a
4332 dnslookup router: the "domains" option is always processed without any
4333 widening, but the router might have qualify_single set.) Now Exim uses the
4334 cached value only when the same options are set.
4336 32. Added John Jetmore's "exipick" utility to the distribution.
4338 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4339 than a timeout (e.g. a certificate is required, and is not provided), an
4340 Exim server now closes the connection immediately. Previously it waited for
4341 the client to close - but if the client is SSL, it seems that they each
4342 wait for each other, leading to a delay before one of them times out.
4344 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4345 maintained 0.8.x compatibility because I don't think many are using it, and
4346 it is clearly obsolete.
4348 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4351 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4352 client certificate was expired. A simple patch fixes this, though I don't
4353 understand the full logic of why the verify callback is called multiple
4356 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4357 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4358 which causes problems with some clients (such as the Certicom SSL Plus
4359 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4360 disables the coutermeasure allowing Eudora to connect."
4362 38. Exim was not checking that a write() to a log file succeeded. This could
4363 lead to Bad Things if a log got too big, in particular if it hit a file
4364 size limit. Exim now panics and dies if it cannot write to a log file, just
4365 as it does if it cannot open a log file.
4367 39. Modified OS/Makefile-Linux so that it now contains
4369 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4371 The two -D definitions ensure that Exim is compiled with large file
4372 support, which makes it possible to handle log files that are bigger than
4375 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4376 instance) a domain was checked against a named list that involved a lookup,
4377 causing $domain_data to be set, then another domain was checked against the
4378 same list, then the first domain was re-checked, the value of $domain_data
4379 after the final check could be wrong. In particular, if the second check
4380 failed, it could be set empty. This bug probably also applied to
4383 41. The strip_trailing_dot option was not being applied to the address given
4384 with the -f command-line option.
4386 42. The code for reading a message's header from the spool was incrementing
4387 $received_count, but never initializing it. This meant that the value was
4388 incorrect (doubled) while delivering a message in the same process in which
4389 it was received. In the most common configuration of Exim, this never
4390 happens - a fresh exec is done - but it can happen when
4391 deliver_drop_privilege is set.
4393 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4394 it now includes up to 150 characters of the unexpected data in the log
4397 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4398 and building data strings. The size of both of these buffers was 10 000
4399 bytes - far larger than anybody would *ever* want, thought I. Needless to
4400 say, somebody hit the limit. I have increased the maximum line length to
4401 20 000 and the maximum data length of concatenated lines to 100 000. I have
4402 also fixed two bugs, because there was no checking on these buffers. Tsk,
4403 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4404 buffer is too small.
4406 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4407 lsearch lookups. Now it does.
4409 46. When parsing a route_list item in a manualroute router, a fixed-length
4410 buffer was used for the list of hosts. I made this 1024 bytes long,
4411 thinking that nobody would ever have a list of hosts that long. Wrong.
4412 Somebody had a whole pile of complicated expansion conditions, and the
4413 string was silently truncated, leading to an expansion error. It turns out
4414 that it is easier to change to an unlimited length (owing to other changes
4415 that have happened since this code was originally written) than to build
4416 structure for giving a limitation error. The length of the item that
4417 expands into the list of hosts is now unlimited.
4419 47. The lsearch lookup could not handle data where the length of text line was
4420 more than 4095 characters. Such lines were truncated, leading to shortened
4421 data being returned. It should now handle lines of any length.
4423 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4424 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4427 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4428 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4429 no longer happens in comment lines. A list of replacements is now placed
4430 at the head of all of the source files, except those whose only change is
4431 to replace PERL_COMMAND in the very first #! line.
4433 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4434 messages on the queue, with a bottom-up merge sort, using code contributed
4435 by Michael Haardt. This should make operations like -bp somewhat faster on
4436 large queues. It won't affect queue runners, except when queue_run_in_order
4439 51. Installed eximstats 1.31 in the distribution.
4441 52. Added support for SRV lookups to the dnslookup router.
4443 53. If an ACL referred to $message_body or $message_body_end, the value was not
4444 reset for any messages that followed in the same SMTP session.
4446 54. The store-handling optimization for building very long strings was not
4447 differentiating between the different store pools. I don't think this
4448 actually made any difference in practice, but I've tidied it.
4450 55. While running the routers to verify a sender address, $sender_address
4451 was still set to the sender address. This is wrong, because when routing to
4452 send a bounce to the sender, it would be empty. Therefore, I have changed
4453 it so that, while verifying a sender address, $sender_address is set to <>.
4454 (There is no change to what happens when verifying a recipient address.)
4456 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4457 target A or AAAA records (if not already returned) without resetting the
4458 qualify_single or search_parents options of the DNS resolver. These are
4459 inappropriate in this case because the targets of MX and SRV records must
4460 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4461 target that, when qualified, matched something in the local domain. These
4462 two options are now turned off when doing these lookups.
4464 57. It seems that at least some releases of Reiserfs (which does not have the
4465 concept of a fixed number of inodes) returns zero and not -1 for the
4466 number of available inodes. This interacted badly with check_spool_inodes,
4467 which assumed that -1 was the "no such thing" setting. What I have done is
4468 to check that the total number of inodes is greater than zero before doing
4469 the test of how many are available.
4471 58. When a "warn" ACL statement has a log_message modifier, the message is
4472 remembered, and not repeated. This is to avoid a lot of repetition when a
4473 message has many recipients that cause the same warning to be written.
4474 Howewer, Exim was preserving the list of already written lines for an
4475 entire SMTP session, which doesn't seem right. The memory is now reset if a
4476 new message is started.
4478 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4479 debugging output unless log_rewrite was also set.
4481 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4482 of (int)(handle) when we know that handle contains (void *)(-1).
4484 61. The Exim daemon panic-logs an error return when it closes the incoming
4485 connection. However "connection reset by peer" seems to be common, and
4486 isn't really an error worthy of noting specially, so that particular error
4489 62. When Exim is trying to find all the local interfaces, it used to panic and
4490 die if the ioctl to get the interface flags failed. However, it seems that
4491 on at least one OS (Solaris 9) it is possible to have an interface that is
4492 included in the list of interfaces, but for which you get a failure error
4493 for this call. This happens when the interface is not "plumbed" into a
4494 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4495 failure of the "get flags" call assumes that the interface is down.
4497 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4498 makes life easier for people who are doing arithmetic on fields extracted
4499 from dates, where you often get leading zeros that should not be
4500 interpreted as octal.
4502 64. Added qualify_domain to the redirect router, to override the global
4505 65. If a pathologically long header line contained very many addresses (the
4506 report of this problem mentioned 10 000) and each of them was rewritten,
4507 Exim could use up a very large amount of memory. (It kept on making new
4508 copies of the header line as it rewrote, and never released the old ones.)
4509 At the expense of a bit more processing, the header rewriting function has
4510 been changed so that it no longer eats memory in this way.
4512 66. The generation of the Received: header has been moved from the time that a
4513 message starts to be received, to the time that it finishes. The timestamp
4514 in the Received: header should now be very close to that of the <= log
4515 line. There are two side-effects of this change:
4517 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4518 logged header lines no longer include the local Received: line, because
4519 it has not yet been created. The same applies to a copy of the message
4520 that is returned to a non-SMTP sender when a message is rejected.
4522 (b) When a filter file is tested using -bf, no additional Received: header
4523 is added to the test message. After some thought, I decided that this
4526 This change does not affect the value of $received_for. It is still set
4527 after address rewriting, but before local_scan() is called.
4529 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4531 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4532 gave an unhelpful panic error message, and a defer error. I have managed to
4533 change this behaviour so that it now rejects any supplied certificate,
4534 which seems right, as the list of acceptable certificates is empty.
4536 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4537 gave an unhelpful defer error. I have not managed to make this reject any
4538 supplied certificates, but the error message it gives is "no certificate
4539 supplied", which is not helpful.
4541 70. exigrep's output now also includes lines that are not associated with any
4542 message, but which match the given pattern. Implemented by a patch from
4543 Martin Sluka, which also tidied up the Perl a bit.
4545 71. Recipient callout verification, like sender verification, was using <> in
4546 the MAIL FROM command. This isn't really the right thing, since the actual
4547 sender may affect whether the remote host accepts the recipient or not. I
4548 have changed it to use the actual sender in the callout; this means that
4549 the cache record is now keyed on a recipient/sender pair, not just the
4550 recipient address. There doesn't seem to be a real danger of callout loops,
4551 since a callout by the remote host to check the sender would use <>.
4552 [SEE ABOVE: changed after hitting problems.]
4554 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4555 temporary errors. However, in the case of such a code being given after
4556 the end of a data transmission (i.e. after ".") Exim was failing to write
4557 a retry record for the message. (Yes, there was some broken host that was
4558 actually sending 8xx at this point.)
4560 73. An unknown lookup type in a host list could cause Exim to panic-die when
4561 the list was checked. (An example that provoked this was putting <; in the
4562 middle of a list instead of at the start.) If this happened during a DATA
4563 ACL check, a -D file could be left lying around. This kind of configuration
4564 error no longer causes Exim to die; instead it causes a defer errror. The
4565 incident is still logged to the main and panic logs.
4567 74. Buglet left over from Exim 3 conversion. The message "too many messages
4568 in one connection" was written to the rejectlog but not the mainlog, except
4569 when address rewriting (yes!) was being logged.
4571 75. Added write_rejectlog option.
4573 76. When a system filter was run not as root (that is, when system_filter_user
4574 was set), the values of the $n variables were not being returned to the
4575 main process; thus, they were not subsequently available in the $sn
4578 77. Added +return_path_on_delivery log selector.
4580 78. A connection timeout was being treated differently from recipients deferred
4581 when testing hosts_max_try with a message that was older than the host's
4582 retry timeout. (The host should not be counted, thus allowing all hosts to
4583 be tried at least once before bouncing.) This may have been the cause of an
4584 occasionally reported bug whereby a message would remain on the queue
4585 longer than the retry timeout, but would be bounced if a delivery was
4586 forced. I say "may" because I never totally pinned down the problem;
4587 setting up timeout/retry tests is difficult. See also the next item.
4589 79. The ultimate address timeout was not being applied to errors that involved
4590 a combination of host plus message (for example, a timeout on a MAIL
4591 command). When an address resolved to a number of possible hosts, and they
4592 were not all tried for each delivery (e.g. because of hosts_max_try), a
4593 message could remain on the queue longer than the retry timeout.
4595 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4598 81. Fixed an obscure SMTP outgoing bug which required at least the following
4599 conditions: (a) there was another message waiting for the same server;
4600 (b) the server returned 5xx to all RCPT commands in the first message so
4601 that the message was not completed; (c) the server dropped the connection
4602 or gave a negative response to the RSET that Exim sends to abort the
4603 transaction. The observed case was a dropped connection after DATA that had
4604 been sent in pipelining mode. That is, the server had advertised PIPELINING
4605 but was not implementing it correctly. The effect of the bug was incorrect
4606 behaviour, such as trying another host, and this could lead to a crash.
4612 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4613 and daemon.c were passed as pointers to ints; they should have been
4614 pointers to socklen_t variables (which are typically unsigned ints).
4616 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4619 3. Fixed a really odd bug that affected only the testing scheme; patching a
4620 certain fixed string in the binary changed the value of another string that
4621 happened to be identical to the end of the original first string.
4623 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4624 name", it returns that address as the IP address. On some operating
4625 systems (e.g. Solaris), it also passes back the IP address string as the
4626 "host name". However, on others (e.g. Linux), it passes back an empty
4627 string. Exim wasn't checking for this, and was changing the host name to an
4628 empty string, assuming it had been canonicized.
4630 5. Although rare, it is permitted to have more than one PTR record for a given
4631 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4632 all the names associated with an address, because they do in Solaris.
4633 However, it seems that they do not in Linux for data that comes from the
4634 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4635 I found this out when I moved to a new Linux workstation and tried to run
4636 the Exim test suite.
4638 To get round this problem I have changed the code so that it now does its
4639 own call to the DNS to look up PTR records when searching for a host name.
4640 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4641 addresses that are only in /etc/hosts are still found.
4643 This behaviour is, however, controlled by an option called host_lookup_
4644 order, which defaults to "bydns:byaddr". If people want to use the other
4645 order, or indeed, just use one or the other means of lookup, they can
4646 specify it in this variable.
4648 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4649 some operating systems, this comes back from gethostbyaddr() as an empty
4650 string, and this is what Exim used to test for. However, it seems that in
4651 other systems, "." is yielded. Exim now tests for this case too.
4653 7. The values of check_spool_space and check_log_space are now held internally
4654 as a number of kilobytes instead of an absolute number of bytes. If a
4655 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4656 kilobyte. This means that much larger values can be stored.
4658 8. Exim monitor: an attempt to get the action menu when not actually pointing
4659 at a message produces an empty menu entitled "No message selected". This
4660 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4661 no entries in it ("Shell widget menu has zero width and/or height"). So I
4662 have added a single, blank menu entry in this case.
4664 9. Added ${quote_local_part.
4666 10. MIME decoding is now applied to the contents of Subject: header lines when
4669 11. Now that a reference to $sender_host_address automatically causes a reverse
4670 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4671 host lookup before query-style lookups in lists that might use this
4672 variable. This has therefore been abolished, and the "net-" prefix is no
4673 longer necessary for query-style lookups.
4675 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4676 have been a typo for LFLAGS, so it has been changed.
4678 13. The install script calls Exim with "-C /dev/null" in order to find the
4679 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4680 to be output. Howeve, since Exim outputs its version number before the
4681 error, it didn't break the script. It just looked ugly. I fixed this by
4682 always allowing "-C /dev/null" if the caller is root.
4684 14. Ignore overlarge ACL variable number when reading spool file - insurance
4685 against a later release with more variables having written the file.
4687 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4688 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4689 this, as well as the form without the "IPv6" on the front.
4691 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4692 OS/Makefile-Darwin file.
4694 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4695 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4697 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4698 variable. However, some OS use ints for inodes. I've added cast to long int
4699 to get rid of the compiler warning.
4701 19. I had forgotten to lock out "/../" in configuration file names when
4702 ALT_CONFIG_PREFIX was set.
4704 20. Routers used for verification do not need to specify transports. However,
4705 if such a router generated a host list, and callout was configured, Exim
4706 crashed, because it could not find a port number from the (non-existent)
4707 transport. It now assumes port 25 in this circumstance.
4709 21. Added the -t option to exigrep.
4711 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4712 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4713 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4714 them is set, LOOKUP_LSEARCH is forced.
4716 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4718 24. Added sender and host information to the "rejected by local_scan()" log
4719 line; previously there was no indication of these.
4721 25. Added .include_if_exists.
4723 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4724 Linux. It turns out that not all file systems support this. Apparently some
4725 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4726 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4727 not supported on the file descriptor, is now ignored when Exim is trying to
4728 sync a directory. This applies only to Linux.
4730 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4732 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4733 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4736 29. Added support for maildirsize files from supplied patch (modified a bit).
4738 30. The use of :fail: followed by an empty string could lead Exim to respond to
4739 sender verification failures with (e.g.):
4741 550 Verification failed for <xxx>
4742 550 Sender verify failed
4744 where the first response line was missing the '-' that indicates it is not
4745 the final line of the response.
4747 31. The loop for finding the name of the user that called Exim had a hardwired
4748 limit of 10; it now uses the value of finduser_retries, which is used for
4749 all other user lookups.
4751 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4754 33. Exim was neglecting to zero errno before one call of strtol() when
4755 expanding a string and expecting an integer value. On some systems this
4756 resulted in spurious "integer overflow" errors. Also, it was casting the
4757 result into an int without checking.
4759 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4760 did not work. The code looks as if it has *never* worked, though it appears
4761 to have been documented since at least releast 1.62. I have made it work.
4763 35. The "timeout_DNS" error in retry rules, also documented since at least
4764 1.62, also never worked. As it isn't clear exactly what this means, and
4765 clearly it isn't a major issue, I have abolished the feature by treating it
4766 as "timeout", and writing a warning to the main and panic logs.
4768 36. The display of retry rules for -brt wasn't always showing the error code
4771 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4772 timeout_connect_A, timeout_connect_MX.
4774 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4775 to the empty sender.
4777 39. The daemon was not analysing the content of -oX till after it had closed
4778 stderr and disconnected from the controlling terminal. This meant that any
4779 syntax errors were only noted on the panic log, and the return code from
4780 the command was 0. By re-arranging the code a little, I've made the
4781 decoding happen first, so such errors now appear on stderr, and the return
4782 code is 1. However, the actual setting up of the sockets still happens in
4783 the disconnected process, so errors there are still only recorded on the
4786 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4787 connections (as happens on some IP stacks) was logged at start up time as
4788 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4789 it from "IPv6 and IPv4", which means that two separate sockets are being
4792 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4793 says whether AF_INET or AF_INET6 was passed as an argument.
4795 42. Exiwhat output was messed up when time zones were included in log
4798 43. Exiwhat now gives more information about the daemon's listening ports,
4799 and whether -tls-on-connect was used.
4801 44. The "port" option of the smtp transport is now expanded.
4803 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4804 silently ignored. Now an error message is written to the main and panic
4807 46. There's a new ACL modifier called "logwrite" which writes to a log file
4808 as soon as it is encountered.
4810 47. Added $local_user_uid and $local_user_gid at routing time.
4812 48. Exim crashed when trying to verify a sender address that was being
4815 49. Exim was recognizing only a space character after ".include". It now also
4816 recognizes a tab character.
4818 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4819 extracting the relevant information from the specification. The man page no
4820 longer contains scrambled data for the -d option, and I've added a section
4821 at the front about calling Exim under different names.
4823 51. Added "extra_headers" argument to the "mail" command in filter files.
4825 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4828 53. Installed eximstats 1.29.
4830 54. Added transport_filter_timeout as a generic transport option.
4832 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4833 Cc: header lines. This was required by RFC 822, but it not required by RFC
4836 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4837 incoming messages that did not have them. Now it does so only if the
4838 message originates locally, that is, if there is no associated remote host
4839 address. When Resent- header lines are present, this applies to the Resent-
4840 lines rather than the non-Resent- lines.
4842 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4843 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4845 58. Messages for configuration errors now include the name of the main
4846 configuration file - useful now that there may be more than one file in a
4847 list (.included file names were always shown).
4849 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4850 for those rare installations that do not start the daemon as root or run it
4851 setuid root. I've cut out the call to initgroups() if the daemon is not
4854 60. The Exim user and group can now be bound into the binary as text strings
4855 that are looked up at the start of Exim's processing.
4857 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4859 62. Added $mailstore_basename variable.
4861 63. Installed patch to sieve.c from Michael Haardt.
4863 64. When Exim failed to open the panic log after failing to open the main log,
4864 the original message it was trying to log was written to stderr and debug
4865 output, but if they were not available (the usual case in production), it
4866 was lost. Now it is written to syslog before the two lines that record the
4867 failures to open the logs.
4869 65. Users' Exim filters run in subprocesses under the user's uid. It is
4870 possible for a "deliver" command or an alias in a "personal" command to
4871 provoke an address rewrite. If logging of address rewriting is configured,
4872 this fails because the process is not running as root or exim. There may be
4873 a better way of dealing with this, but for the moment (because 4.30 needs
4874 to be released), I have disabled address rewrite logging when running a
4875 filter in a non-root, non-exim process.
4881 1. The buildconfig auxiliary program wasn't quoting the value set for
4882 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4883 not defined. This bug was masked in 4.22 by the effect that was fixed in
4886 2. Some messages that were rejected after a message id was allocated were
4887 shown as "incomplete" by exigrep. It no longer does this for messages that
4888 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4890 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4891 have allow_domain_literals set, the ID did not get logged in the <= line.
4892 Domain literals are now always recognized in Message-ID: header lines.
4894 4. The first argument for a ${extract expansion item is the key name or field
4895 number. Leading and trailing spaces in this item were not being ignored,
4896 causing some misleading effects.
4898 5. When deliver_drop_privilege was set, single queue runner processes started
4899 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4900 same command in the process it spins off) were not dropping privilege.
4902 6. When the daemon running as "exim" started a queue runner, it always
4903 re-executed Exim in the spun-off process. This is a waste of effort when
4904 deliver_drop_privilege is set. The new process now just calls the
4905 queue-runner function directly.
4911 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4914 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4915 ignored. Though the use of -oP was forcing the writing of a pid file, it
4916 was always written to the default place.
4918 3. If the message "no IP address found for host xxxx" is generated during
4919 incoming verification, it is now followed by identification of the incoming
4920 connection (so you can more easily find what provoked it).
4922 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4924 5. Added some features to "harden" Exim a bit more against certain attacks:
4926 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4927 be put in Local/Makefile. This is like the never_users runtime option,
4928 but it cannot be overridden. The default setting is "root".
4930 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4931 prefix string with which any file named in a -C command line option
4934 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4935 is retained for -C and -D only if the caller of Exim is root. Without
4936 it, the exim user may also use -C and -D and retain privilege.
4938 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4939 command line option is disabled.
4941 6. Macro names set by the -D option must start with an upper case letter, just
4942 like macro names defined in the configuration file.
4944 7. Added "dereference=" facility to LDAP.
4946 8. Two instances of the typo "uknown" in the source files are fixed.
4948 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4949 the Configure-Makefile script screwed up while processing it.
4951 10. Incorporated PCRE 4.4.
4953 11. The SMTP synchronization check was not operating right at the start of an
4954 SMTP session. For example, it could not catch a HELO sent before the client
4955 waited for the greeting. There is now a check for outstanding input at the
4956 point when the greeting is written. Because of the duplex, asynchronous
4957 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4958 way, but not yet received, when the check is performed.
4960 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4961 on TCP/IP sockets, because this apparently causes some broken clients to
4964 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4965 unchanged) from the Cygwin maintainer.
4967 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4968 was supported instead of testing for mbx. Effectively a typo.
4970 15. The spa authenticator server code was not checking that the input it
4971 received was valid base64.
4973 16. The debug output line for the "set" modifier in ACLs was not showing the
4974 name of the variable that was being set.
4976 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4978 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4979 lookup on its own. Something else was needed to trigger the lookup. For
4980 example, a match in host_lookup or the need for a host name in a host list.
4981 Now, if $sender_host_name is referenced and the host name has not yet been
4982 looked up, a lookup is performed. If the lookup fails, the variable remains
4983 empty, and $host_lookup_failed is set to "1".
4985 19. Added "eqi" as a case-independent comparison operator.
4987 20. The saslauthd authentication condition could segfault if neither service
4988 nor realm was specified.
4990 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4991 error message that was logged was misleading, and incoming SMTP
4992 connections were dropped. The message is now more accurate, and temporary
4993 errors are given to SMTP connections.
4995 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4996 (except RSET) with a 421 error, until QUIT is received. However, it was
4997 failing to send a response to QUIT.
4999 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5000 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5001 ACL is now run after the helo_try_verify_hosts code.
5003 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5004 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5005 case-independent, so other case variants are also recognized). Apparently
5006 some systems use these upper case variants.
5008 25. If more than two messages were waiting for the same host, and a transport
5009 filter was specified for the transport, Exim sent two messages over the
5010 same TCP/IP connection, and then failed with "socket operation on non-
5011 socket" when it tried to send the third.
5013 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5015 27. The extern definition of crypt16() in expand.c was not being excluded when
5016 the OS had its own crypt16() function.
5018 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5019 as a preferred synonym for return_size_limit, both as an option and as an
5022 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5024 30. Changed the default configuration ACL to relax the local part checking rule
5025 for addresses that are not in any local domains. For these addresses,
5026 slashes and pipe symbols are allowed within local parts, but the sequence
5027 /../ is explicitly forbidden.
5029 31. SPA server authentication was not clearing the challenge buffer before
5032 32. log_message in a "warn" ACL statement was writing to the reject log as
5033 well as to the main log, which contradicts the documentation and doesn't
5034 seem right (because no rejection is happening). So I have stopped it.
5036 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5037 However, I am unable to do any testing of this.
5039 34. Fixed an infelicity in the appendfile transport. When checking directories
5040 for a mailbox, to see if any needed to be created, it was accidentally
5041 using path names with one or more superfluous leading slashes; tracing
5042 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5044 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5045 contents are added to the log line that is written for every discarded
5046 recipient. (Previously a log_message setting was ignored.)
5048 36. The ${quote: operator now quotes the string if it is empty.
5050 37. The install script runs exim in order to find its version number. If for
5051 some reason other than non-existence or emptiness, which it checks, it
5052 could not run './exim', it was installing it with an empty version number,
5053 i.e. as "exim-". This error state is now caught, and the installation is
5056 38. An argument was missing from the function that creates an error message
5057 when Exim fails to connect to the socket for saslauthd authentication.
5058 This could cause Exim to crash, or give a corrupted message.
5060 39. Added isip, isip4, and isip6 to ${if conditions.
5062 40. The ACL variables $acl_xx are now saved with the message, and can be
5063 accessed later in routers, transports, and filters.
5065 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5066 strings in the file are not string-expanded.
5068 42. If a MAIL command specified a SIZE value that was too large to fit into an
5069 int variable, the check against message_size_limit failed. Such values are
5070 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5071 day this will have to be increased, but I don't think I want to be around
5072 when emails are that large.
5079 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5080 iconv() is not standard in FreeBSD.
5082 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5083 IPv6 enabled. The observed symptom was a segmentation fault on return from
5084 the function os_common_find_running_interfaces() in src/os.c.
5086 3. In the check_special_case() function in daemon.c I had used "errno" as an
5087 argument name, which causes warnings on some systems. This was basically a
5088 typo, since it was named "eno" in the comments!
5090 4. The code that waits for the clock to tick (at a resolution of some fraction
5091 of a second) so as to ensure message-id uniqueness was always waiting for
5092 at least one whole tick, when it could have waited for less. [This is
5093 almost certainly not relevant at current processor speeds, where it is
5094 unlikely to ever wait at all. But we try to future-proof.]
5096 5. The function that sleeps for a time interval that includes fractions of a
5097 second contained a race. It did not block SIGALRM between setting the
5098 timer, and suspending (a couple of lines later). If the interval was short
5099 and the sigsuspend() was delayed until after it had expired, the suspension
5100 never ended. On busy systems this could lead to processes getting stuck for
5103 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5104 process, before it forks any delivery processes. The open lookup caching
5105 mechanism meant that the open file or database connection was passed into
5106 the delivery process. The problem was that delivery processes always tidy
5107 up cached lookup data. This could cause a problem for the next delivery
5108 process started by the queue runner, because the external queue runner
5109 process does not know about the closure. So the next delivery process
5110 still has data in the lookup cache. In the case of a file lookup, there was
5111 no problem because closing a file descriptor in a subprocess doesn't affect
5112 the parent. However, if the lookup was caching a connection to a database,
5113 the connection was closed, and the second delivery process was likely to
5114 see errors such as "PGSQL: query failed: server closed the connection
5115 unexpectedly". The problem has been fixed by closing all cached lookups
5116 in a queue runner before running a delivery process.
5118 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5119 seem to have the "const" qualifier which it has on other OS. I've
5122 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5123 *of the same type* (client or server) with the same public name that an
5124 error should be diagnosed.
5126 9. When Exim looked up a host name for an IP address, but failed to find the
5127 original IP address when looking up the host name (a safety check), it
5128 output the message "<ip address> does not match any IP for NULL", which was
5129 confusing, to say the least. The bug was that the host name should have
5130 appeared instead of "NULL".
5132 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5133 user that is built into the binary, and the -C or -D options is used, root
5134 privilege is dropped before the configuration file is read. In addition,
5135 logging is switched to stderr instead of the normal log files. If the
5136 configuration then re-defines the Exim user, the unprivileged environment
5137 is probably not what is expected, so Exim logs a panic warning message (but
5140 However, if deliver_drop_privilege is set, the unprivileged state may well
5141 be exactly what is intended, so the warning has been cut out in that case,
5142 and Exim is allowed to try to write to its normal log files.
5148 1. smtp_return_error_details was not giving details for temporary sender
5149 or receiver verification errors.
5151 2. Diagnose a configuration error if two authenticators have the same public
5154 3. Exim used not to create the message log file for a message until the first
5155 delivery attempt. This could be confusing when incoming messages were held
5156 for policy or load reasons. The message log file is now created at the time
5157 the message is received, and an initial "Received" line is written to it.
5159 4. The automatically generated man page for command line options had a minor
5160 bug that caused no ill effects; however, a more serious problem was that
5161 the procedure for building the man page automatically didn't always
5162 operate. Consequently, release 4.20 contains an out-of-date version. This
5163 shouldn't happen again.
5165 5. When building Exim with embedded Perl support, the script that builds the
5166 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5167 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5169 6. The freeze_tell option was not being used for messages that were frozen on
5170 arrival, either by an ACL or by local_scan().
5172 7. Added the smtp_incomplete_transaction log selector.
5174 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5175 was accepting AUTH without a new EHLO.
5177 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5178 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5181 10. Logging of TCP/IP connections (when configured) now happens in the main
5182 daemon process instead of the child process, so that the TCP/IP connection
5183 count is more accurate (but it can never be perfect).
5185 11. The use of "drop" in a nested ACL was not being handled correctly in the
5186 outer ACL. Now, if condition failure induced by the nested "drop" causes
5187 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5188 or "require"), the connection is dropped.
5190 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5191 that yield "discard" can now be used with an "accept" or a "discard" verb,
5192 but an error is generated for any others (because I can't see a useful way
5193 to define what should happen).
5195 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5196 are now processed in the same way as lines in the Exim configuration file.
5197 In particular, continuation lines are supported.
5199 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5201 15. Added -ti meaning -t -i.
5203 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5204 domains, and warn by logging if others are found.
5206 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5207 returned from GIFCONF when Exim is trying to find the list of interfaces on
5208 a host. The code in os.c has been modified to copy each ifreq to an aligned
5209 structure in all cases.
5211 Also, in some cases, the returned ifreq's were being copied to a 'struct
5212 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5213 means the last couple of bytes of an IPv6 address could be chopped if the
5214 ifreq contained only a normal sockaddr (14 bytes storage).
5216 18. Named domain lists were not supported in the hosts_treat_as_local option.
5217 An entry such as +xxxx was not recognized, and was treated as a literal
5220 19. Ensure that header lines added by a DATA ACL are included in the reject log
5221 if the ACL subsequently rejects the message.
5223 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5224 MD5 (which is deprecated).
5226 21. When testing a filter file using -bf, Exim was writing a message when it
5227 took the sender from a "From " line in the message, but it was not doing so
5228 when it took $return_path from a Return-Path: header line. It now does.
5230 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5231 with a valid header line field name (a series of printing characters
5232 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5234 23. Changed "disc" in the source to "disk" to conform to the documentation and
5235 the book and for uniformity.
5237 24. Ignore Sendmail's -Ooption=value command line item.
5239 25. When execve() failed while trying to run a command in a pipe transport,
5240 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5241 could be confused with a return value of 69 from the command itself. This
5242 has been changed to 127, the value the shell returns if it is asked to run
5243 a non-existent command. The wording for the related log line suggests a
5244 non-existent command as the problem.
5246 26. If received_header_text expands to an empty string, do not add a Received:
5247 header line to the message. (Well, it adds a token one on the spool, but
5248 marks it "old" so that it doesn't get used or transmitted.)
5250 27. Installed eximstats 1.28 (addition of -nt option).
5252 28. There was no check for failure on the call to getsockname() in the daemon
5253 code. This can fail if there is a shortage of resources on the system, with
5254 ENOMEM, for example. A temporary error is now given on failure.
5256 29. Contrary to the C standard, it seems that in some environments, the
5257 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5258 program. Exim now does this explicitly; it affects the formatting of
5259 timestamps using strftime().
5261 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5262 complaints. I've now initialized all the variables, to avoid this.
5264 32. Header lines added by a system filter were not being "seen" during
5265 transport-time rewrites.
5267 33. The info_callback() function passed to OpenSSL is set up with type void
5268 (*)(SSL *, int, int), as described somewhere. However, when calling the
5269 function (actually a macro) that sets it up, the type void(*)() is
5270 expected. I've put in a cast to prevent warnings from picky compilers.
5272 34. If a DNS black list lookup found a CNAME record, but there were no A
5273 records associated with the domain it pointed at, Exim crashed.
5275 35. If a DNS black list lookup returned more than one A record, Exim ignored
5276 all but the first. It now scans all returned addresses if a particular IP
5277 value is being sought. In this situation, the contents of the
5278 $dnslist_value variable are a list of all the addresses, separated by a
5281 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5282 to accept a host name and all its aliases if the forward lookup for any of
5283 them yielded the IP address of the incoming connection. Now it accepts only
5284 those names whose forward lookup yields the correct IP address. Any other
5285 names are discarded. This closes a loophole whereby a rogue DNS
5286 administrator could create reverse DNS records to break through a
5287 wildcarded host restriction in an ACL.
5289 37. If a user filter or a system filter that ran in a subprocess used any of
5290 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5291 the wrong values were passed to the pipe command ($thisaddress had the
5292 value of $0, $0 had the value of $1, etc). This bug was introduced by
5293 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5295 38. Improved the line breaking for long SMTP error messages from ACLs.
5296 Previously, if there was no break point between 40 and 75 characters, Exim
5297 left the rest of the message alone. Two changes have been made: (a) I've
5298 reduced the minimum length to 35 characters; (b) if it can't find a break
5299 point between 35 and 75 characters, it looks ahead and uses the first one
5300 that it finds. This may give the occasional overlong line, but at least the
5301 remaining text gets split now.
5303 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5304 file descriptors might be low, and that setting 1000 would always raise it.
5305 It turns out that in some environments, the limit is already over 1000 and
5306 that lowering it causes trouble. So now Exim takes care not to decrease it.
5308 40. When delivering a message, the value of $return_path is set to $sender_
5309 address at the start of routing (routers may change the value). By an
5310 oversight, this default was not being set up when an address was tested by
5311 -bt or -bv, which affected the outcome if any router or filter referred to
5314 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5315 log only when "message" or "log_message" are set. However, if one of the
5316 conditions was an address verification, or a call to a nested ACL, the
5317 messages generated by the underlying test were being passed through. This
5318 no longer happens. The underlying message is available in $acl_verify_
5319 message for both "message" and "log_message" expansions, so it can be
5320 passed through if needed.
5322 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5323 new expansion $bh_ to give the encoded byte string without charset
5324 translation. Translation happens only if iconv() is available; HAVE_ICONV
5325 indicates this at build time. HEADERS_CHARSET gives the charset to
5326 translate to; headers_charset can change it in the configuration, and
5327 "headers charset" can change it in an individual filter file.
5329 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5330 that creates RFC 2047 encoded "words" labels them as that charset instead
5331 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5332 expansion operator; (ii) when Exim creates a From: line for a local
5333 message; (iii) when a header line is rewritten to include a "phrase" part.
5335 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5336 buggy, causing it to skip the first lines of messages whose message ID
5337 ended in 'D'. This would not have bitten before Exim release 4.14, because
5338 message IDs were unlikely to end in 'D' before then. The effect was to have
5339 incorrect size information for certain domains.
5341 45. #include "config.h" was missing at the start of the crypt16.c module. This
5342 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5345 46. If there was a timeout during a "random" callout check, Exim treated it as
5346 a failure of the random address, and carried on sending RSET and the real
5347 address. If the delay was just some slowness somewhere, the response to the
5348 original RCPT would be taken as a response to RSET and so on, causing
5349 mayhem of various kinds.
5351 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5352 when I implemented it. It didn't allow for the fact that some option values
5353 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5354 the right test for positive values.
5356 48. Domain names in DNS records are case-independent. Exim always looks them up
5357 in lower case. Some resolvers return domain names in exactly the case they
5358 appear in the zone file, that is, they may contain uppercase letters. Not
5359 all resolvers do this - some return always lower case. Exim was treating a
5360 change of case by a resolver as a change of domain, similar to a widening
5361 of a domain abbreviation. This triggered its re-routing code and so it was
5362 trying to route what was effectively the same domain again. This normally
5363 caused routing to fail (because the router wouldn't handle the domain
5364 twice). Now Exim checks for this case specially, and just changes the
5365 casing of the domain that it ultimately uses when it transmits the message
5368 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5371 50. If a filter generated a file delivery with a non-absolute name (possible if
5372 no home directory exists for the router), the forbid_file option was not
5375 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5376 the existing equality matching.
5378 52. Exim was using ints instead of ino_t variables in some places where it was
5379 dealing with inode numbers.
5381 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5382 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5383 called TMPDIR, and if it finds it is different, it changes its value.
5385 54. The smtp_printf() function is now made available to local_scan() so
5386 additional output lines can be written before returning. There is also an
5387 smtp_fflush() function to enable the detection of a dropped connection.
5388 The variables smtp_input and smtp_batched_input are exported to
5391 55. Changed the default runtime configuration: the message "Unknown user"
5392 has been removed from the ACL, and instead placed on the localuser router,
5393 using the cannot_route_message feature. This means that any verification
5394 failures that generate their own messages won't get overridden. Similarly,
5395 the "Unrouteable address" message that was in the ACL for unverifiable
5396 relay addresses has also been removed.
5398 56. Added hosts_avoid_esmtp to the smtp transport.
5400 57. The exicyclog script was not checking for the esoteric option
5401 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5402 will work only if exicyclog is run under the appropriate euid.
5404 58. Following a discussion on the list, the rules by which Exim recognises line
5405 endings on incoming messages have been changed. The -dropcr and drop_cr
5406 options are now no-ops, retained only for backwards compatibility. The
5407 following line terminators are recognized: LF CRLF CR. However, special
5408 processing applies to CR:
5410 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5411 nor a local message in the state where . is a terminator.
5413 (ii) If a bare CR is encountered in a header line, an extra space is added
5414 after the line terminator so as not to end the header. The reasoning
5415 behind this is that bare CRs in header lines are most likely either
5416 to be mistakes, or people trying to play silly games.
5418 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5419 was being incorrectly given as 18 bytes larger than it should have been.
5420 This is a VOB (very old bug).
5422 60. This may never have affected anything current, but just in case it has:
5423 When the local host is found other than at the start of a list of hosts,
5424 the local host, those with the same MX, and any that follow, are discarded.
5425 When the list in question was part of a longer list of hosts, the following
5426 hosts (not currently being processed) were also being discarded. This no
5427 longer happens. I'm not sure if this situation could ever has previously
5430 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5433 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5434 header. This is something that is recommended in a new Internet Draft, and
5435 is something that is documented as being done by Sendmail. There are two
5436 possible values. For messages generated by the autoreply transport, Exim
5439 Auto-Submitted: auto-replied
5441 whereas for all other generated messages (e.g. bounces) it adds
5443 Auto-Submitted: auto-generated
5445 63. The "personal" condition in filters now includes a test for the
5446 Auto-Submitted: header. If it contains the string "auto-" the message it
5447 not considered personal.
5449 64. Added rcpt_include_affixes as a generic transport option.
5451 65. Added queue_only_override (default true).
5453 66. Added the syslog_duplication option.
5455 67. If what should have been the first header line of a message consisted of
5456 a space followed by a colon, Exim was mis-interpreting it as a header line.
5457 It isn't of course - it is syntactically invalid and should therefore be
5458 treated as the start of the message body. The misbehaviour could have
5459 caused a number of strange effects, including loss of data in subsequent
5460 header lines, and spool format errors.
5462 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5463 client host had authenticated. This control can now be exercised by an ACL
5464 for more flexibility.
5466 69. By default, callouts do not happen when testing with -bh. There is now a
5467 variant, -bhc, which does actually run the callout code, including
5468 consulting and updating the callout cache.
5470 70. Added support for saslauthd authentication, courtesy of Alexander
5473 71. If statvfs() failed on the spool or log directories while checking their
5474 size for availability, Exim confusingly gave the error "space shortage".
5475 Furthermore, in debugging mode it crashed with a floating point exception.
5476 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5477 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5478 serious problem, Exim now writes to the main and panic logs when this
5479 happens, with details of the failure. It then refuses to accept the
5480 incoming message, giving the message "spool directory problem" or "log
5481 directory problem" with a 421 code for SMTP messages.
5483 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5484 0, 1, and 2 exist, because some OS complain for execs without them (see
5485 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5486 descriptors. However, the code omitted to check that the open succeeded,
5487 causing mysterious errors if for some reason the permissions on /dev/null
5488 got screwed. Now Exim writes a message to the main and panic logs, and
5489 bombs out if it can't open /dev/null.
5491 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5492 interact so that it is all more flexible. It is supposed to remain
5493 backwards compatible. Also added extra_local_interfaces.
5495 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5496 to bomb out with an assertion failure - to the client this appears as a
5497 connection drop. This problem occurs in the part of the code that was taken
5498 from the Samba project. Fortunately, the assertion is in a very simple
5499 function, so I have fixed this by reproducing the function inline in the
5500 one place where it is called, and arranging for authentication to fail
5501 instead of killing the process with assert().
5503 75. The SPA client code was not working when the server requested OEM rather
5504 than Unicode encoding.
5506 76. Added code to make require_files with a specific uid setting more usable in
5507 the case where statting the file as root fails - usually a non-root-mounted
5508 NFS file system. When this happens and the failure is EACCES, Exim now
5509 forks a subprocess and does the per-uid checking as the relevant uid.
5511 77. Added process_log_path.
5513 78. If log_file_path was not explicitly set, a setting of check_log_space or
5514 check_log_inodes was ignored.
5516 79. If a space check for the spool or log partitions fails, the incident is now
5517 logged. Of course, in the latter case the data may get lost...
5519 80. Added the %p formatting code to string_format() so that it can be used to
5520 print addresses in debug_print(). Adjusted all the address printing in the
5521 debugging in store.c to use %p rather than %d.
5523 81. There was a concern that a line of code in smtp_in.c could overflow a
5524 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5525 initially expressed, the concern was not well-founded, because trailing
5526 spaces are removed early. However, if the trailing spaces were followed by
5527 a NULL, they did not get removed, so the overflow was possible. Two fixes
5530 (a) I re-wrote the offending code in a cleaner fashion.
5531 (b) If an incoming SMTP command contains a NULL character, it is rejected
5534 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5535 runs initgroups(), so that if the Exim user is in any additional groups,
5536 they will be used during message reception.
5542 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob