1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
81 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
82 NULL dereference. Report and patch from Alun Jones.
84 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
85 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
86 Not seeing resolver debug output on NetBSD, but suspect this is a
87 resolver implementation change.
89 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
90 Left warnings. Added "eximon gdb" invocation mode.
92 PP/21 Defaulting "accept_8bitmime" to true, not false.
94 PP/22 Added -bw for inetd wait mode support.
100 PP/01 Solaris build fix for Oracle's LDAP libraries.
101 Bugzilla 1109, patch from Stephen Usher.
103 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
105 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
108 TF/02 Fix a couple more cases where we did not log the error message
109 when unlink() failed. See also change 4.74-TF/03.
111 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
112 lock up or crash if it happened to be inside a call to libc when it
113 got a SIGUSR1 from exiwhat.
115 The SIGUSR1 handler appends the current process status to the process
116 log which is later printed by exiwhat. It used to use the general
117 purpose logging code to do this, but several functions it calls are
118 not safe for signals.
120 The new output code in the SIGUSR1 handler is specific to the process
121 log, and simple enough that it's easy to inspect for signal safety.
122 Removing some special cases also simplifies the general logging code.
123 Removing the spurious timestamps from the process log simplifies
126 TF/04 Improved ratelimit ACL condition.
128 The /noupdate option has been deprecated in favour of /readonly which
129 has clearer semantics. The /leaky, /strict, and /readonly update modes
130 are mutually exclusive. The update mode is no longer included in the
131 database key; it just determines when the database is updated. (This
132 means that when you upgrde Exim will forget old rate measurements.)
134 Exim now checks that the per_* options are used with an update mode that
135 makes sense for the current ACL. For example, when Exim is processing a
136 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
137 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
138 must specify per_mail/readonly. If you omit the update mode it defaults to
139 /leaky where that makes sense (as before) or /readonly where required.
141 The /noupdate option is now undocumented but still supported for
142 backwards compatibility. It is equivalent to /readonly except that in
143 ACLs where /readonly is required you may specify /leaky/noupdate or
144 /strict/noupdate which are treated the same as /readonly.
146 A useful new feature is the /count= option. This is a generalization
147 of the per_byte option, so that you can measure the throughput of other
148 aggregate values. For example, the per_byte option is now equivalent
149 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
151 The per_rcpt option has been generalized using the /count= mechanism
152 (though it's more complicated than the per_byte equivalence). When it is
153 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
154 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
155 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
156 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
157 Note that using per_rcpt with a non-readonly update mode in more than
158 one ACL will cause the recipients to be double-counted. (The per_mail
159 and per_byte options don't have this problem.)
161 The handling of very low rates has changed slightly. If the computed rate
162 is less than the event's count (usually one) then this event is the first
163 after a long gap. In this case the rate is set to the same as this event's
164 count, so that the first message of a spam run is counted properly.
166 The major new feature is a mechanism for counting the rate of unique
167 events. The new per_addr option counts the number of different
168 recipients that someone has sent messages to in the last time period. It
169 behaves like per_rcpt if all the recipient addresses are different, but
170 duplicate recipient addresses do not increase the measured rate. Like
171 the /count= option this is a general mechanism, so the per_addr option
172 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
173 example, measure the rate that a client uses different sender addresses
174 with the options per_mail/unique=$sender_address. There are further
175 details in the main documentation.
177 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
179 TF/06 Removed a few PCRE remnants.
181 TF/07 Automatically extract Exim's version number from tags in the git
182 repository when doing development or release builds.
184 PP/02 Raise smtp_cmd_buffer_size to 16kB.
185 Bugzilla 879. Patch from Paul Fisher.
187 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
188 Heavily based on revision 40f9a89a from Simon Arlott's tree.
191 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
193 PP/05 Variable $av_failed, true if the AV scanner deferred.
194 Bugzilla 1078. Patch from John Horne.
196 PP/06 Stop make process more reliably on build failure.
197 Bugzilla 1087. Patch from Heiko Schlittermann.
199 PP/07 Make maildir_use_size_file an _expandable_ boolean.
200 Bugzilla 1089. Patch from Heiko Schlittermann.
202 PP/08 Handle ${run} returning more data than OS pipe buffer size.
203 Bugzilla 1131. Patch from Holger Weiß.
205 PP/09 Handle IPv6 addresses with SPF.
206 Bugzilla 860. Patch from Wolfgang Breyha.
208 PP/10 GnuTLS: support TLS 1.2 & 1.1.
210 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
213 PP/11 match_* no longer expand right-hand-side by default.
214 New compile-time build option, EXPAND_LISTMATCH_RHS.
215 New expansion conditions, "inlist", "inlisti".
217 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
219 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
221 PP/14 fix log_write() format string regression from TF/03.
222 Bugzilla 1152. Patch from Dmitry Isaikin.
228 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
230 PP/02 Harmonised TLS library version reporting; only show if debugging.
231 Layout now matches that introduced for other libraries in 4.74 PP/03.
233 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
235 PP/04 New "dns_use_edns0" global option.
237 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
240 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
241 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
243 TK/01 Updated PolarSSL code to 0.14.2.
244 Bugzilla 1097. Patch from Andreas Metzler.
246 PP/07 Catch divide-by-zero in ${eval:...}.
249 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
252 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
253 format-string attack -- SECURITY: remote arbitrary code execution.
255 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
256 time unintentionally subject to list matching rules, letting the header
257 cause arbitrary Exim lookups (of items which can occur in lists, *not*
258 arbitrary string expansion). This allowed for information disclosure.
260 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
261 INT_MIN/-1 -- value coerced to INT_MAX.
267 NM/01 Workround for PCRE version dependency in version reporting
270 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
271 This fixes portability to compilers other than gcc, notably
272 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
274 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
275 makefiles for portability to HP-UX and POSIX correctness.
277 PP/01 Permit LOOKUP_foo enabling on the make command-line.
278 Also via indented variable definition in the Makefile.
279 (Debugging by Oliver Heesakkers).
281 PP/02 Restore caching of spamd results with expanded spamd_address.
282 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
284 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
285 Improves build reliability. Fix from: Frank Elsner
287 NM/02 Fix wide character breakage in the rfc2047 coding
288 Fixes bug 1064. Patch from Andrey N. Oktyabrski
290 NM/03 Allow underscore in dnslist lookups
291 Fixes bug 1026. Patch from Graeme Fowler
293 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
294 Code patches from Adam Ciarcinski of NetBSD.
296 NM/04 Fixed exiqgrep to cope with mailq missing size issue
299 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
300 is logged, to avoid truncation. Patch from John Horne.
302 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
303 Patch from Jakob Hirsch.
305 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
306 SQL string expansion failure details.
307 Patch from Andrey Oktyabrski.
309 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
310 Patch from Simon Arlott.
312 PP/09 New lookups functionality failed to compile on old gcc which rejects
313 extern declarations in function scope.
314 Patch from Oliver Fleischmann
316 PP/10 Use sig_atomic_t for flags set from signal handlers.
317 Check getgroups() return and improve debugging.
318 Fixed developed for diagnosis in bug 927 (which turned out to be
321 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
322 Patch from Mark Zealey.
324 PP/12 Bugzilla 1056: Improved spamd server selection.
325 Patch from Mark Zealey.
327 PP/13 Bugzilla 1086: Deal with maildir quota file races.
328 Based on patch from Heiko Schlittermann.
330 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
331 Patch from Uwe Doering, sign-off by Michael Haardt.
333 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
334 variable declaration deep within a block. Bug and patch from
337 PP/15 lookups-Makefile IRIX compatibilty coercion.
339 PP/16 Make DISABLE_DKIM build knob functional.
341 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
342 Patch by Simon Arlott
344 TF/03 Fix valgrind.h portability to C89 compilers that do not support
345 variable argument macros. Our copy now differs from upstream.
351 TF/01 Failure to get a lock on a hints database can have serious
352 consequences so log it to the panic log.
354 TF/02 Log LMTP confirmation messages in the same way as SMTP,
355 controlled using the smtp_confirmation log selector.
357 TF/03 Include the error message when we fail to unlink a spool file.
359 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
360 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
361 for maintaining out-of-tree patches for some time.
363 PP/01 Bugzilla 139: Documentation and portability issues.
364 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
365 Handle per-OS dynamic-module compilation flags.
367 PP/02 Let /dev/null have normal permissions.
368 The 4.73 fixes were a little too stringent and complained about the
369 permissions on /dev/null. Exempt it from some checks.
370 Reported by Andreas M. Kirchwitz.
372 PP/03 Report version information for many libraries, including
373 Exim version information for dynamically loaded libraries. Created
374 version.h, now support a version extension string for distributors
375 who patch heavily. Dynamic module ABI change.
377 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
378 privilege escalation vulnerability whereby the Exim run-time user
379 can cause root to append content of the attacker's choosing to
382 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
385 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
386 If dropping privileges for untrusted macros, we disabled normal logging
387 on the basis that it would fail; for the Exim run-time user, this is not
388 the case, and it resulted in successful deliveries going unlogged.
389 Fixed. Reported by Andreas Metzler.
395 PP/01 Date: & Message-Id: revert to normally being appended to a message,
396 only prepend for the Resent-* case. Fixes regression introduced in
397 Exim 4.70 by NM/22 for Bugzilla 607.
399 PP/02 Include check_rfc2047_length in configure.default because we're seeing
400 increasing numbers of administrators be bitten by this.
402 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
404 PP/03 Bugzilla 994: added openssl_options main configuration option.
406 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
408 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
410 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
412 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
413 without a peer certificate, leading to a segfault because of an
414 assumption that peers always have certificates. Be a little more
415 paranoid. Problem reported by Martin Tscholak.
417 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
418 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
419 NB: ClamAV planning to remove STREAM in "middle of 2010".
420 CL also introduces -bmalware, various -d+acl logging additions and
421 more caution in buffer sizes.
423 PP/09 Implemented reverse_ip expansion operator.
425 PP/10 Bugzilla 937: provide a "debug" ACL control.
427 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
429 PP/12 Bugzilla 973: Implement --version.
431 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
433 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
435 PP/15 Bugzilla 816: support multiple condition rules on Routers.
437 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
438 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
439 ignore trailing whitespace.
441 JJ/02 prevent non-panic DKIM error from being sent to paniclog
443 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
446 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
447 Notification from Dr Andrew Aitchison.
449 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
450 ExtendedDetectionInfo response format.
451 Notification from John Horne.
453 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
456 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
457 XSL and documented dependency on system catalogs, with examples of how
460 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
463 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
464 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
465 configuration file which is writeable by the Exim user or group.
467 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
468 of configuration files to cover files specified with the -C option if
469 they are going to be used with root privileges, not just the default
472 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
473 option (effectively making it always true).
475 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
476 files to be used while preserving root privileges.
478 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
479 that rogue child processes cannot use them.
481 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
482 run-time user, instead of root.
484 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
485 Exim run-time user without dropping privileges.
487 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
488 result string, instead of calling string_vformat() twice with the same
491 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
492 for other users. Others should always drop root privileges if they use
493 -C on the command line, even for a whitelisted configure file.
495 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
497 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
503 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
504 $data_path, and $header_path variables; fixed documentation bugs and
507 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
508 exipick to access non-standard spools, including the "frozen" queue
511 NM/01 Bugzilla 965: Support mysql stored procedures.
512 Patch from Alain Williams
514 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
516 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
517 Patch from Andreas Metzler
519 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
520 Patch from Kirill Miazine
522 NM/05 Bugzilla 671: Added umask to procmail example.
524 JJ/03 installed exipick 20100323.0, fixing doc bug
526 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
527 directory. Notification and patch from Dan Rosenberg.
529 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
531 TK/02 Improve log output when DKIM signing operation fails.
533 MH/01 Treat the transport option dkim_domain as a colon separated
534 list, not as a single string, and sign the message with each element,
535 omitting multiple occurences of the same signer.
537 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
538 Bugzilla 985, 986. Patch by Simon Arlott
540 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
541 Patch by Simon Arlott
543 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
544 MBX locking. Notification from Dan Rosenberg.
550 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
552 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
554 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
556 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
558 NM/04 Bugzilla 915: Flag broken perl installation during build.
564 TK/01 Added patch by Johannes Berg that expands the main option
565 "spamd_address" if it starts with a dollar sign.
567 TK/02 Write list of recipients to X-Envelope-Sender header when building
568 the mbox-format spool file for content scanning (suggested by Jakob
571 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
572 (http://www.dcc-servers.net/) support via dccifd. Activated by
573 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
575 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
576 by Mark Daniel Reidel <mr@df.eu>.
578 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
579 When building exim an external PCRE library is now needed -
580 PCRE is a system library on the majority of modern systems.
581 See entry on PCRE_LIBS in EDITME file.
583 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
584 conversation. Added nologin parameter to request.
585 Patch contributed by Kirill Miazine.
587 TF/01 Do not log submission mode rewrites if they do not change the address.
589 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
591 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
592 log files in place. Contributed by Roberto Lima.
594 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
596 TF/03 Bugzilla 615: When checking the local_parts router precondition
597 after a local_part_suffix or local_part_prefix option, Exim now
598 does not use the address's named list lookup cache, since this
599 contains cached lookups for the whole local part.
601 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
602 Robert Millan. Documentation is in experimental-spec.txt.
604 TF/04 Bugzilla 668: Fix parallel build (make -j).
606 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
608 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
609 Patch provided by Jan Srzednicki.
611 TF/05 Leading white space used to be stripped from $spam_report which
612 wrecked the formatting. Now it is preserved.
614 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
615 that they are available at delivery time.
617 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
619 TF/08 TLS error reporting now respects the incoming_interface and
620 incoming_port log selectors.
622 TF/09 Produce a more useful error message if an SMTP transport's hosts
623 setting expands to an empty string.
625 NM/06 Bugzilla 744: EXPN did not work under TLS.
626 Patch provided by Phil Pennock.
628 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
629 Patch provided by Richard Godbee.
631 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
632 acl_smtp_notquit, added index entry.
634 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
635 Patch provided by Eugene Bujak.
637 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
638 accept(). Patch provided by Maxim Dounin.
640 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
641 Patch provided by Phil Pennock.
643 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
645 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
646 Patch provided by Brad "anomie" Jorsch.
648 NM/14 Bugzilla 622: Added timeout setting to transport filter.
649 Patch provided by Dean Brooks.
651 TK/05 Add native DKIM support (does not depend on external libraries).
653 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
654 Patch provided by Graeme Fowler.
656 NM/16 Bugzilla 851: Documentation example syntax fix.
658 NM/17 Changed NOTICE file to remove references to embedded PCRE.
660 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
663 NM/19 Bugzilla 745: TLS version reporting.
664 Patch provided by Phil Pennock.
666 NM/20 Bugzilla 167: bool: condition support.
667 Patch provided by Phil Pennock.
669 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
670 clients. Patch provided by Phil Pennock.
672 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
673 Patch provided by Brad "anomie" Jorsch.
675 NM/23 Bugzilla 687: Fix misparses in eximstats.
676 Patch provided by Heiko Schlittermann.
678 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
679 Patch provided by Heiko Schlittermann.
681 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
682 plus update to original patch.
684 NM/26 Bugzilla 799: Documentation correction for ratelimit.
686 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
687 Patch provided by David Brownlee.
689 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
691 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
693 NM/30 Bugzilla 888: TLS documentation bugfixes.
695 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
697 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
698 Unlike the original bugzilla I have changed all shell scripts in src tree.
700 NM/33 Bugzilla 898: Transport filter timeout fix.
701 Patch by Todd Rinaldo.
703 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
704 Patch by Serge Demonchaux.
706 NM/35 Bugzilla 39: Base64 decode bug fixes.
707 Patch by Jakob Hirsch.
709 NM/36 Bugzilla 909: Correct connect() call in dcc code.
711 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
713 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
715 NM/39 Bugzilla 911: Fixed MakeLinks build script.
721 TK/01 Add preliminary DKIM support. Currently requires a forked version of
722 ALT-N's libdkim that I have put here:
723 http://duncanthrax.net/exim-experimental/
725 Note to Michael Haardt: I had to rename some vars in sieve.c. They
726 were called 'true' and it seems that C99 defines that as a reserved
727 keyword to be used with 'bool' variable types. That means you could
728 not include C99-style headers which use bools without triggering
729 build errors in sieve.c.
731 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
732 as mailq or other aliases. Changed the --help handling significantly
733 to do whats expected. exim_usage() emits usage/help information.
735 SC/01 Added the -bylocaldomain option to eximstats.
737 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
739 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
741 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
747 PH/01 Another patch from the Sieve maintainer.
749 PH/02 When an IPv6 address is converted to a string for single-key lookup
750 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
751 dots are used instead of colons so that keys in lsearch files need not
752 contain colons. This was done some time before quoting was made available
753 in lsearch files. However, iplsearch files do require colons in IPv6 keys
754 (notated using the quote facility) so as to distinguish them from IPv4
755 keys. This meant that lookups for IP addresses in host lists did not work
756 for iplsearch lookups.
758 This has been fixed by arranging for IPv6 addresses to be expressed with
759 colons if the lookup type is iplsearch. This is not incompatible, because
760 previously such lookups could never work.
762 The situation is now rather anomolous, since one *can* have colons in
763 ordinary lsearch keys. However, making the change in all cases is
764 incompatible and would probably break a number of configurations.
766 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
769 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
770 conversion specification without a maximum field width, thereby enabling
771 a rogue spamd server to cause a buffer overflow. While nobody in their
772 right mind would setup Exim to query an untrusted spamd server, an
773 attacker that gains access to a server running spamd could potentially
774 exploit this vulnerability to run arbitrary code as the Exim user.
776 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
777 $primary_hostname instead of what libspf2 thinks the hosts name is.
779 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
780 a directory entry by the name of the lookup key. Previously, if a
781 symlink pointed to a non-existing file or a file in a directory that
782 Exim lacked permissions to read, a lookup for a key matching that
783 symlink would fail. Now it is enough that a matching directory entry
784 exists, symlink or not. (Bugzilla 503.)
786 PH/03 The body_linecount and body_zerocount variables are now exported in the
789 PH/04 Added the $dnslist_matched variable.
791 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
792 This means they are set thereafter only if the connection becomes
795 PH/06 Added the client_condition to authenticators so that some can be skipped
796 by clients under certain conditions.
798 PH/07 The error message for a badly-placed control=no_multiline_responses left
799 "_responses" off the end of the name.
801 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
803 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
804 (without spaces) instead of just copying the configuration text.
806 PH/10 Added the /noupdate option to the ratelimit ACL condition.
808 PH/11 Added $max_received_linelength.
810 PH/12 Added +ignore_defer and +include_defer to host lists.
812 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
813 way in which PCRE > 7.0 is built.
815 PH/14 Implemented queue_only_load_latch.
817 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
818 MAIL command. The effect was to mangle the value on 64-bit systems.
820 PH/16 Another patch from the Sieve maintainer.
822 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
824 PH/18 If a system quota error occurred while trying to create the file for
825 a maildir delivery, the message "Mailbox is full" was not appended to the
826 bounce if the delivery eventually timed out. Change 4.67/27 below applied
827 only to a quota excession during the actual writing of the file.
829 PH/19 It seems that peer DN values may contain newlines (and other non-printing
830 characters?) which causes problems in log lines. The DN values are now
831 passed through string_printing() before being added to log lines.
833 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
834 and InterBase are left for another time.)
836 PH/21 Added message_body_newlines option.
838 PH/22 Guard against possible overflow in moan_check_errorcopy().
840 PH/23 POSIX allows open() to be a macro; guard against that.
842 PH/24 If the recipient of an error message contained an @ in the local part
843 (suitably quoted, of course), incorrect values were put in $domain and
844 $local_part during the evaluation of errors_copy.
850 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
851 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
854 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
855 issue a MAIL command.
857 PH/02 In an ACL statement such as
859 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
861 if a client was not listed at all, or was listed with a value other than
862 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
863 the condition was not true (as it should be), so access was not denied.
864 The bug was that the ! inversion was incorrectly passed on to the second
865 item. This has been fixed.
867 PH/03 Added additional dnslists conditions == and =& which are different from
868 = and & when the dns lookup returns more than one IP address.
870 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
871 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
873 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
874 FSYNC, which compiles an option called disable_fsync that allows for
875 bypassing fsync(). The documentation is heavily laced with warnings.
877 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
879 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
880 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
881 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
882 including adding "make clean"; (3) Added -fPIC when compiling the test
883 dynamically loaded module, to get rid of a warning.
885 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
886 message fails, move_frozen_messages = true and ignore_bounce_errors_after
887 = 0s. The bug is otherwise harmless.
889 PH/07 There was a bug in the dovecot authenticator such that the value of
890 $auth1 could be overwritten, and so not correctly preserved, after a
891 successful authentication. This usually meant that the value preserved by
892 the server_setid option was incorrect.
894 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
896 PH/09 Installed PCRE release 7.0.
898 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
899 run for batched SMTP input. It is now run at the start of every message
900 in the batch. While fixing this I discovered that the process information
901 (output by running exiwhat) was not always getting set for -bs and -bS
902 input. This is fixed, and it now also says "batched" for BSMTP.
904 PH/11 Added control=no_pipelining.
906 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
907 patch, slightly modified), and move the expansion of helo_data till after
908 the connection is made in the smtp transport (so it can use these
911 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
913 PH/14 Added log_selector = +pid.
915 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
917 PH/16 Add ${if forany and ${if forall.
919 PH/17 Added dsn_from option to vary the From: line in DSNs.
921 PH/18 Flush SMTP output before performing a callout, unless control =
922 no_callout_flush is set.
924 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
925 was true (the default) a successful delivery failed to delete the retry
926 item, thus causing premature timeout of the address. The bug is now
929 PH/20 Added hosts_avoid_pipelining to the smtp transport.
931 PH/21 Long custom messages for fakedefer and fakereject are now split up
932 into multiline reponses in the same way that messages for "deny" and
933 other ACL rejections are.
935 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
936 with slight modification.
938 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
939 draft, changing the syntax and factoring some duplicate code".
941 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
942 for deliveries of the second and subsequent messages over the same SMTP
945 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
946 ${reduce, with only minor "tidies".
948 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
950 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
951 expansion side effects.
953 PH/27 When a message times out after an over-quota error from an Exim-imposed
954 quota, the bounce message says "mailbox is full". This message was not
955 being given when it was a system quota that was exceeded. It now should
958 MH/03 Made $recipients available in local_scan(). local_scan() already has
959 better access to the recipient list through recipients_list[], but
960 $recipients can be useful in postmaster-provided expansion strings.
962 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
963 in the case of a MAIL command with additional options following the
964 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
965 were accidentally chopped off.
967 PH/29 SMTP synchronization checks are implemented when a command is read -
968 there is a check that no more input is waiting when there shouldn't be
969 any. However, for some commands, a delay in an ACL can mean that it is
970 some time before the response is written. In this time, more input might
971 arrive, invalidly. So now there are extra checks after an ACL has run for
972 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
973 pipelining has not been advertised.
975 PH/30 MH's patch to allow iscntrl() characters to be list separators.
977 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
978 returned in the SMTP response when smtp_return_error_details was false.
981 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
982 instead of the C I/O that was originally supplied, because problems were
985 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
986 Exim which did not show up earlier: it was assuming that a call to
987 SSL_CTX_set_info_callback() might give an error value. In fact, there is
988 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
989 was a macro that became an assignment, so it seemed to work. This has
990 changed to a proper function call with a void return, hence the compile
991 error. Exim's code has been fixed.
993 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
996 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
998 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1000 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1001 because it turns out to be too big or there is a timeout) there is no
1002 "Completed" line in the log. When some messages of this type were
1003 selected by exigrep, they were listed as "not completed". Others were
1004 picked up by some special patterns. I have improved the selection
1005 criteria to be more general.
1007 PH/38 The host_find_failed option in the manualroute router can now be set
1008 to "ignore", to completely ignore a host whose IP address cannot be
1009 found. If all hosts are ignored, the behaviour is controlled by the new
1010 host_all_ignored option.
1012 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1013 homing or because of multiple MX records with /mx) generated more than
1014 one IP address, and the following item turned out to be the local host,
1015 all the secondary addresses of the first item were incorrectly removed
1016 from the list, along with the local host and any following hosts (which
1017 is what is supposed to happen).
1019 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1020 whoever called Exim into the -H file. In the case of the daemon it was
1021 behaving confusingly. When first started, it used values for whoever
1022 started the daemon, but after a SIGHUP it used the Exim user (because it
1023 calls itself on a restart). I have changed the code so that it now always
1026 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1027 message are rejected with the same error (e.g. no authentication or bad
1028 sender address), and a DATA command is nevertheless sent (as can happen
1029 with PIPELINING or a stupid MUA), the error message that was given to the
1030 RCPT commands is included in the rejection of the DATA command. This is
1031 intended to be helpful for MUAs that show only the final error to their
1034 PH/42 Another patch from the Sieve maintainer.
1036 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1037 Eximstats - Fixed some broken HTML links and added missing column headers
1039 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1040 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1042 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1044 PH/43 Yet another patch from the Sieve maintainer.
1046 PH/44 I found a way to check for a TCP/IP connection going away before sending
1047 the response to the final '.' that terminates a message, but only in the
1048 case where the client has not sent further data following the '.'
1049 (unfortunately, this is allowed). However, in many cases there won't be
1050 any further data because there won't be any more messages to send. A call
1051 to select() can be used: if it shows that the input is "ready", there is
1052 either input waiting, or the socket has been closed. An attempt to read
1053 the next input character can distinguish the two cases. Previously, Exim
1054 would have sent an OK response which the client would never have see.
1055 This could lead to message repetition. This fix should cure that, at
1056 least in a lot of common cases.
1058 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1059 advertised in response to EHLO.
1065 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1066 fixed by 4.65/MH/01 (is this a record?) are fixed:
1068 (i) An empty string was always treated as zero by the numeric comparison
1069 operators. This behaviour has been restored.
1071 (ii) It is documented that the numeric comparison operators always treat
1072 their arguments as decimal numbers. This was broken in that numbers
1073 starting with 0 were being interpreted as octal.
1075 While fixing these problems I realized that there was another issue that
1076 hadn't been noticed. Values of message_size_limit (both the global option
1077 and the transport option) were treated as octal if they started with 0.
1078 The documentation was vague. These values are now always treated as
1079 decimal, and I will make that clear in the documentation.
1085 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1086 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1089 MH/01 Don't check that the operands of numeric comparison operators are
1090 integers when their expansion is in "skipping" mode (fixes bug
1091 introduced by 4.64-PH/07).
1093 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1094 child addresses, Exim now panics and dies. Previously, because the count
1095 is held in a short int, deliveries were likely to be lost. As such a
1096 large number of recipients for a single message is ridiculous
1097 (performance will be very, very poor), I have chosen to impose a limit
1098 rather than extend the field.
1104 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1105 leftover -K file (the existence of which was triggered by #402).
1106 While we were at it, introduced process PID as part of the -K
1107 filename. This should rule out race conditions when creating
1110 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1111 processing considerably. Previous code took too long for large mails,
1112 triggering a timeout which in turn triggers #401.
1114 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1115 in the DK code in transports.c. sendfile() is not really portable,
1116 hence the _LINUX specificness.
1118 TF/01 In the add_headers option to the mail command in an Exim filter,
1119 there was a bug that Exim would claim a syntax error in any
1120 header after the first one which had an odd number of characters
1123 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1124 callout verification, Exim cached a "reject" for the entire domain. This
1125 is correct for most verifications, but it is not correct for a recipient
1126 verification with use_sender or use_postmaster set, because in that case
1127 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1128 case of MAIL FROM:<> rejection from other early rejections (e.g.
1129 rejection of HELO). When verifying a recipient using a non-null MAIL
1130 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1131 Whatever the result of the callout, the value of the domain cache is
1132 left unchanged (for any other kind of callout, getting as far as trying
1133 RCPT means that the domain itself is ok).
1135 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1138 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1139 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1140 session, the daemon ignores ECONNECT errors and logs others; it now
1141 ignores EPIPE as well.
1143 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1144 (quoted-printable decoding).
1146 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1147 later the small subsequent patch to fix an introduced bug.
1149 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1151 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1153 PH/08 An error is now given if message_size_limit is specified negative.
1155 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1156 to be given (somewhat) arbitrary names.
1158 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1161 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1162 miscellaneous code fixes
1164 PH/10 Added the log_reject_target ACL modifier to specify where to log
1167 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1168 hostname. This is wrong, because it relates to the incoming message (and
1169 probably the interface on which it is arriving) and not to the outgoing
1170 callout (which could be using a different interface). This has been
1171 changed to use the value of the helo_data option from the smtp transport
1172 instead - this is what is used when a message is actually being sent. If
1173 there is no remote transport (possible with a router that sets up host
1174 addresses), $smtp_active_hostname is used.
1176 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1177 tweaks were necessary in order to get it to work (see also 21 below):
1178 (a) The code assumed that strncpy() returns a negative number on buffer
1179 overflow, which isn't the case. Replaced with Exim's string_format()
1181 (b) There were several signed/unsigned issues. I just did the minimum
1182 hacking in of casts. There is scope for a larger refactoring.
1183 (c) The code used strcasecmp() which is not a standard C function.
1184 Replaced with Exim's strcmpic() function.
1185 (d) The code set only $1; it now sets $auth1 as well.
1186 (e) A simple test gave the error "authentication client didn't specify
1187 service in request". It would seem that Dovecot has changed its
1188 interface. Fortunately there's a specification; I followed it and
1189 changed what the client sends and it appears to be working now.
1191 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1194 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1195 address A is aliased to B and C, where B exists and C does not. Without
1196 -v the output is "A verified" because verification stops after a
1197 successful redirection if more than one address is generated. However,
1198 with -v the child addresses are also verified. Exim was outputting "A
1199 failed to verify" and then showing the successful verification for C,
1200 with its parentage. It now outputs "B failed to verify", showing B's
1201 parentage before showing the successful verification of C.
1203 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1204 look up a TXT record in a specific list after matching in a combined
1207 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1208 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1209 they consult the DNS. I had assumed they would set it the way they
1210 wanted; and indeed my experiments on Linux seem to show that in some
1211 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1212 To be on the safe side, however, I have now made the interface to
1213 host_find_byname() similar to host_find_bydns(), with an argument
1214 containing the DNS resolver options. The host_find_byname() function now
1215 sets these options at its start, just as host_find_bydns() does. The smtp
1216 transport options dns_qualify_single and dns_search_parents are passed to
1217 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1218 of host_find_byname() use the default settings of RES_DEFNAMES
1219 (qualify_single) but not RES_DNSRCH (search_parents).
1221 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1222 spool_read_header() do less string testing, by means of a preliminary
1223 switch on the second character of optional "-foo" lines. (This is
1224 overdue, caused by the large number of possibilities that now exist.
1225 Originally there were few.) While I was there, I also converted the
1226 str(n)cmp tests so they don't re-test the leading "-" and the first
1227 character, in the hope this might squeeze out yet more improvement.
1229 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1230 flag allowing group syntax was set by the header_syntax check but not
1231 turned off, possible causing trouble later; (2) The flag was not being
1232 set at all for the header_verify test, causing "group"-style headers to
1233 be rejected. I have now set it in this case, and also caused header_
1234 verify to ignore an empty address taken from a group. While doing this, I
1235 came across some other cases where the code for allowing group syntax
1236 while scanning a header line wasn't quite right (mostly, not resetting
1237 the flag correctly in the right place). These bugs could have caused
1238 trouble for malformed header lines. I hope it is now all correct.
1240 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1241 with the "reply" argument non-NULL. The code, however (which originally
1242 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1243 but it didn't always do it. This confused somebody who was copying the
1244 code for some other use. I have removed all the tests.
1246 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1247 feature that was used to support insecure browsers during the U.S. crypto
1248 embargo. It requires special client support, and Exim is probably the
1249 only MTA that supported it -- and would never use it because real RSA is
1250 always available. This code has been removed, because it had the bad
1251 effect of slowing Exim down by computing (never used) parameters for the
1252 RSA_EXPORT functionality.
1254 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1255 authenticator to fail if there's a tab character in the incoming data
1256 (there should never be unless someone is messing about, as it's supposed
1257 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1258 if the connection is using TLS or if the remote IP is the same as the
1259 local IP, and the "valid-client-cert option" if a client certificate has
1262 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1263 authenticators. This can be used for authorization after authentication
1264 succeeds. (In the case of plaintext, it servers for both authentication
1267 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1268 if any retry times were supplied.
1270 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1271 connection, where there is no client IP address to check. In this
1272 situation, the verify now always succeeds.
1274 PH/25 Applied John Jetmore's -Mset patch.
1276 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1278 PH/27 In a string expansion for a processed (not raw) header when multiple
1279 headers of the same name were present, leading whitespace was being
1280 removed from all of them, but trailing whitespace was being removed only
1281 from the last one. Now trailing whitespace is removed from each header
1282 before concatenation. Completely empty headers in a concatenation (as
1283 before) are ignored.
1285 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1286 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1288 PH/29 [Removed. This was a change that I later backed out, and forgot to
1289 correct the ChangeLog entry (that I had efficiently created) before
1290 committing the later change.]
1292 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1293 address errors (4xx response to RCPT) over the same connection as other
1294 messages routed to the same hosts. Such deliveries are always "forced",
1295 so retry times are not inspected. This resulted in far too many retries
1296 for the affected addresses. The effect occurred only when there were more
1297 hosts than the hosts_max_try setting in the smtp transport when it had
1298 the 4xx errors. Those hosts that it had tried were not added to the list
1299 of hosts for which the message was waiting, so if all were tried, there
1300 was no problem. Two fixes have been applied:
1302 (i) If there are any address or message errors in an SMTP delivery, none
1303 of the hosts (tried or untried) are now added to the list of hosts
1304 for which the message is waiting, so the message should not be a
1305 candidate for sending over the same connection that was used for a
1306 successful delivery of some other message. This seems entirely
1307 reasonable: after all the message is NOT "waiting for some host".
1308 This is so "obvious" that I'm not sure why it wasn't done
1309 previously. Hope I haven't missed anything, but it can't do any
1310 harm, as the worst effect is to miss an optimization.
1312 (ii) If, despite (i), such a delivery is accidentally attempted, the
1313 routing retry time is respected, so at least it doesn't keep
1314 hammering the server.
1316 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1317 in ${readsocket because some servers need this prod.
1319 PH/32 Added some extra debug output when updating a wait-xxx database.
1321 PH/33 The hint "could be header name not terminated by colon", which has been
1322 given for certain expansion errors for a long time, was not being given
1323 for the ${if def:h_colon_omitted{... case.
1325 PH/34 The spec says: "With one important exception, whenever a domain list is
1326 being scanned, $domain contains the subject domain." There was at least
1327 one case where this was not true.
1329 PH/35 The error "getsockname() failed: connection reset by peer" was being
1330 written to the panic log as well as the main log, but it isn't really
1331 panic-worthy as it just means the connection died rather early on. I have
1332 removed the panic log writing for the ECONNRESET error when getsockname()
1335 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1336 runs only) independently of the message's sender address. This meant
1337 that, if the 4xx error was in fact related to the sender, a different
1338 message to the same recipient with a different sender could confuse
1339 things. In particualar, this can happen when sending to a greylisting
1340 server, but other circumstances could also provoke similar problems.
1341 I have changed the default so that the retry time for these errors is now
1342 based a combination of the sender and recipient addresses. This change
1343 can be overridden by setting address_retry_include_sender=false in the
1346 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1347 remote server are returned as part of bounce messages. This was not
1348 happening for LMTP over a pipe (the lmtp transport), but now it is the
1349 same for both kinds of LMTP.
1351 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1352 in header lines that were in fact CNAMEs. This is no longer the case.
1354 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1355 and queue runs started by the daemon processed all messages. This has
1356 been fixed so that -R and -S can now usefully be given with -q<time>.
1358 PH/40 Import PCRE release 6.7 (fixes some bugs).
1360 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1362 PH/42 Give an error if -q is specified more than once.
1364 PH/43 Renamed the variables $interface_address and $interface_port as
1365 $received_ip_address and $received_port, to make it clear that these
1366 values apply to message reception, and not to the outgoing interface when
1367 a message is delivered. (The old names remain recognized, of course.)
1369 PH/44 There was no timeout on the connect() call when using a Unix domain
1370 socket in the ${readsocket expansion. There now is.
1372 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1373 be meaningful with "accept".
1375 SC/01 Eximstats V1.43
1376 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1378 SC/02 Eximstats V1.44
1379 Use a glob alias rather than an array ref in the generated
1380 parser. This improves both readability and performance.
1382 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1383 Collect SpamAssassin and rejection statistics.
1384 Don't display local sender or destination tables unless
1385 there is data to show.
1386 Added average volumes into the top table text output.
1388 SC/04 Eximstats V1.46
1389 Collect data on the number of addresses (recipients)
1390 as well as the number of messages.
1392 SC/05 Eximstats V1.47
1393 Added 'Message too big' to the list of mail rejection
1394 reasons (thanks to Marco Gaiarin).
1396 SC/06 Eximstats V1.48
1397 Mainlog lines which have GMT offsets and are too short to
1398 have a flag are now skipped.
1400 SC/07 Eximstats V1.49 (Alain Williams)
1401 Added the -emptyok flag.
1403 SC/08 Eximstats V1.50
1404 Fixes for obtaining the IP address from reject messages.
1406 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1407 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1408 whitesspace changes from 4.64-PH/27
1410 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1413 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1414 are found, allow negative numbers in numeric criteria)
1416 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1418 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1419 to match changes made in 4.64-PH/43
1421 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1423 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1424 "domain" is actually an IP address, and force a failure. This locks out
1425 those revolvers/nameservers that support "A-for-A" lookups, in
1426 contravention of the specifications.
1428 PH/48 When a host name was looked up from an IP address, and the subsequent
1429 forward lookup of the name timed out, the host name was left in
1430 $sender_host_name, contrary to the specification.
1432 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1433 restricted to single-key lookups, Exim was not diagnosing an error if
1434 * or *@ was used with a query-style lookup.
1436 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1438 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1439 long ago, but noone interested enough thought of it. Let's just say that
1440 the "1.1" means that there are some new functions that weren't there at
1441 some point in the past.
1443 PH/51 Error processing for expansion failure of helo_data from an smtp
1444 transport during callout processing was broken.
1446 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1447 tested/used via the -bh/-bhc/-bs options.
1449 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1450 bug, fixed in subsequent PCRE releases).
1452 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1453 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1455 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1461 SC/01 Use a glob alias rather than an array ref in eximstats generated
1462 parser. This improves both readability and performance.
1464 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1465 Don't display local sender or destination tables in eximstats unless
1466 there is data to show.
1467 Added average volumes into the eximstats top table text output.
1469 SC/03 Collect data on the number of addresses (recipients) as well
1470 as the number of messages in eximstats.
1472 TF/01 Correct an error in the documentation for the redirect router. Exim
1473 does (usually) call initgroups() when daemonizing.
1475 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1476 with consistent privilege compared to when running as a daemon.
1478 TF/03 Note in the spec that $authenticated_id is not set for local
1479 submissions from trusted users.
1481 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1482 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1484 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1485 by adding some example configuration directives to the default
1486 configuration file. A little bit of work is required to uncomment the
1487 directives and define how usernames and passwords are checked, but
1488 there is now a framework to start from.
1490 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1491 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1492 without this. I don't know how relevant this is to other LDAP libraries.
1494 PH/02 Add the verb name to the "unknown ACL verb" error.
1496 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1498 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1500 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1501 directory not expanded when it should be if an expanded home directory
1502 was set for the address (which is overridden by the transport).
1504 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1507 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1508 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1509 because it is too late at that time, and has no effect.
1511 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1512 security issue with \' (bugzilla #107). I could not use the
1513 PQescapeStringConn() function, because it needs a PGconn value as one of
1516 PH/08 When testing addresses using -bt, indicate those final addresses that
1517 are duplicates that would not cause an additional delivery. At least one
1518 person was confused, thinking that -bt output corresponded to deliveries.
1519 (Suppressing duplicates isn't a good idea as you lose the information
1520 about possibly different redirections that led to the duplicates.)
1522 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1523 systems where poll() doesn't work, in particular OS X.
1525 PH/10 Added more information to debugging output for retry time not reached.
1527 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1528 operations in malware.c.
1530 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1533 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1534 syslog_duplication set false, log lines that would normally be written
1535 both the the main log and to the reject log were not written to syslog at
1538 PH/14 In the default configuration, change the use of "message" in ACL warn
1539 statements to "add_header".
1541 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1542 not followed by a command (e.g. "seen endif").
1544 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1545 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1548 PH/17 Added extra conditions to the default value of delay_warning_condition
1552 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1553 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1554 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1557 The Auto-Submitted: and various List- headers are standardised, whereas I
1558 don't think Precedence: ever was.
1560 PH/18 Refactored debugging code in route_finduser() to show more information,
1561 in particular, the error code if getpwnam() issues one.
1563 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1564 This is apparently needed in addition to the PH/07 change above to avoid
1565 any possible encoding problems.
1567 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1568 but not after initializing Perl.
1570 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1571 output them only if debugging. By default they are written stderr,
1572 apparently, which is not desirable.
1574 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1577 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1580 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1582 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1583 authenticated or an ident call has been made. Suppress the default
1584 values for $authenticated_id and $authenticated_sender (but permit -oMai
1585 and -oMas) when testing with -bh.
1587 PH/24 Re-jigged the order of the tests in the default configuration so that the
1588 tests for valid domains and recipients precede the DNS black list and CSA
1589 tests, on the grounds that those ones are more expensive.
1591 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1592 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1593 This bug exists in every version of Exim that I still have, right back to
1596 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1597 However, an attempt to turn on case-sensitivity in a regex key by
1598 including (?-i) didn't work because the subject string was already
1599 lowercased, and the effects were non-intuitive. It turns out that a
1600 one-line patch can be used to allow (?-i) to work as expected.
1606 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1607 other effects) broke the use of negated acl sub-conditions.
1609 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1612 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1613 "Deny" causes Exim to reject the incoming connection with a 554 error.
1614 Unfortunately, if there is a major crisis, such as a disk failure,
1615 tcp-wrappers gives "deny", whereas what one would like would be some
1616 kind of temporary error. A kludge has been added to help with this.
1617 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1618 554 error is used if errno is still zero or contains ENOENT (which occurs
1619 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1622 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1624 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1625 errors. Otherwise a message that provokes a temporary error (when other
1626 messages do not) can cause a whole host to time out.
1628 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1629 addresses were routed directly to files or pipes from a redirect router.
1630 File deliveries just didn't batch; pipe deliveries might have suffered
1633 PH/06 A failure to get a lock for a hints database would erroneously always say
1634 "Failed to get write lock", even when it was really a read lock.
1636 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1637 of 0600. This has been changed to use the value of the lockfile_mode
1638 option (which defaults to 0600).
1640 PH/08 Applied small patch from the Sieve maintainer.
1642 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1643 folder from quota calculations, a direct delivery into this folder messed
1644 up the contents of the maildirsize file. This was because the regex was
1645 used only to exclude .Trash (or whatever) when the size of the mailbox
1646 was calculated. There was no check that a delivery was happening into an
1647 excluded directory. This bug has been fixed by ignoring all quota
1648 processing for deliveries into excluded directories.
1650 PH/10 Added the maildirfolder_create_regex option to appendfile.
1656 PH/01 The code for finding all the local interface addresses on a FreeBSD
1657 system running IPv6 was broken. This may well have applied to all BSD
1658 systems, as well as to others that have similar system calls. The broken
1659 code found IPv4 interfaces correctly, but gave incorrect values for the
1660 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1661 that it would not match correctly against @[] and not recognize the IPv6
1664 PH/02 The ipliteral router was not recognizing addresses of the form user@
1665 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1667 PH/03 Added disable_ipv6.
1669 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1670 decoded form, because it is most often used to construct To: headers
1671 lines in autoreplies, and the decoded form may well be syntactically
1672 invalid. However, $reply_address has leading white space removed, and all
1673 newlines turned into spaces so that the autoreply transport does not
1676 PH/05 If group was specified without a user on a router, and no group or user
1677 was specified on a transport, the group from the router was ignored.
1679 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1680 for visible compile-time settings that can be used to change these
1681 numbers, for those that want even more. Backwards compatibility with old
1682 spool files has been maintained. However, going back to a previous Exim
1683 release will lost any variables that are in spool files.
1685 PH/07 Two small changes when running in the test harness: increase delay when
1686 passing a TCP/IP connection to a new process, in case the original
1687 process has to generate a bounce, and remove special handling of
1688 127.0.0.2 (sic), which is no longer necessary.
1690 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1691 be the same on different OS.
1693 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1696 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1697 whether --show-vars was specified or not
1699 JJ/02 exipick: Added support for new ACL variable spool format introduced
1702 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1703 syntactically invalid From: or Reply-to: line, and a filter used this to
1704 generate an autoreply, and therefore failed to obtain an address for the
1705 autoreply, Exim could try to deliver to a non-existent relative file
1706 name, causing unrelated and misleading errors. What now happens is that
1707 it logs this as a hard delivery error, but does not attempt to create a
1710 PH/11 The exinext utility has a -C option for testing purposes, but although
1711 the given file was scanned by exinext itself; it wasn't being passed on
1712 when Exim was called.
1714 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1715 an end-of-file indication when reading a command response.
1717 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1718 compiled. In many other places in Exim, IPv6 addresses are always
1719 recognized, so I have changed this. It also means that IPv4 domain
1720 literals of the form [IPV4:n.n.n.n] are now always recognized.
1722 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1723 used if the router is not running as root, for example, when verifying at
1724 ACL time, or when using -bh. The debugging output from this situation was
1725 non-existent - all you got was a failure to exec. I have made two
1728 (a) Failures to set uid/gid, the current directory, or a process leader
1729 in a subprocess such as that created by queryprogram now generate
1730 suitable debugging ouput when -d is set.
1732 (b) The queryprogram router detects when it is not running as root,
1733 outputs suitable debugging information if -d is set, and then runs
1734 the subprocess without attempting to change uid/gid.
1736 PH/15 Minor change to Makefile for building test_host (undocumented testing
1739 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1740 additional section of a DNS packet that returns MX or SRV records.
1741 Instead, it always explicitly searches for A/AAAA records. This avoids
1742 major problems that occur when a DNS server includes only records of one
1743 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1744 fixed another bug: if SRV records were looked up and the corresponding
1745 address records were *not* found in the additional section, the port
1746 values from the SRV records were lost.
1748 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1749 using the correct key (the original address) when searching the retry
1750 rules in order to find which one to use for generating the retry hint.
1752 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1753 adding the default one. Similarly, if it contains a Reply-To: header, the
1754 errors_reply_to option, if set, is not used.
1756 PH/19 When calculating a retry time, Exim used to measure the "time since
1757 failure" by looking at the "first failed" field in the retry record. Now
1758 it does not use this if it is later than than the arrival time of the
1759 message. Instead it uses the arrival time. This makes for better
1760 behaviour in cases where some deliveries succeed, thus re-setting the
1761 "first failed" field. An example is a quota failure for a huge message
1762 when small messages continue to be delivered. Without this change, the
1763 "time since failure" will always be short, possible causing more frequent
1764 delivery attempts for the huge message than are intended.
1765 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1767 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1768 $1, $2, $3) because the numerical variables can be reset during some
1769 expansion items (e.g. "match"), thereby losing the authentication data.
1771 PH/21 Make -bV show the size of off_t variables so that the test suite can
1772 decide whether to run tests for quotas > 2G.
1774 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1775 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1776 filecount value is greater than 2G or if a quota value is greater than 2G
1777 on a system where the size of off_t is not greater than 4, a panic error
1780 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1781 never match. The debug and -bh output now contains an explicit error
1782 message indicating a malformed IPv4 address or mask.
1784 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1785 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1786 PH/23 above applies.
1788 PH/25 Do not write to syslog when running in the test harness. The only
1789 occasion when this arises is a failure to open the main or panic logs
1790 (for which there is an explicit test).
1792 PH/26 Added the /no_tell option to "control=freeze".
1794 PH/27 If a host name lookup failed very early in a connection, for example, if
1795 the IP address matched host_lookup and the reverse lookup yielded a name
1796 that did not have a forward lookup, an error message of the form "no IP
1797 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1798 could be logged. Now it outputs the IP address instead of "NULL".
1800 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1801 allows the sender and the authenticated sender to be set when
1802 submitting a message from within Exim. Since child_open_exim() is
1803 documented for local_scan(), the new function should be too.
1805 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1806 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1807 results in an empty string is now treated as unset.
1809 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1811 PH/31 Added sender_verify_fail logging option.
1813 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1814 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1815 tidied the source and removed it altogether.
1817 PH/33 When a queue run was abandoned because the load average was too high, a
1818 log line was always written; now it is written only if the queue_run log
1819 selector is set. In addition, the log line for abandonment now contains
1820 information about the queue run such as the pid. This is always present
1821 in "start" and "stop" lines but was omitted from the "abandon" line.
1823 PH/34 Omit spaces between a header name and the colon in the error message that
1824 is given when verify = headers_syntax fails (if there are lots of them,
1825 the message gets confusing).
1827 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1828 names, as there are now some PTR records that contain slashes. This check
1829 is only to protect against broken name servers that fall over on strange
1830 characters, so the fact that it applies to all lookups doesn't matter.
1832 PH/36 Now that the new test suite is complete, we can remove some of the
1833 special code in Exim that was needed for the old test suite. For example,
1834 sorting DNS records because real resolvers return them in an arbitrary
1835 order. The new test suite's fake resolver always returns records in the
1838 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1839 bounces) except when queue_only is set, to avoid logging races between
1840 the different processes.
1842 PH/38 Panic-die if .include specifies a non-absolute path.
1844 PH/39 A tweak to the "H" retry rule from its user.
1846 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1847 a label. They prevented compilation on older perls.
1849 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1850 a warning to be raised on newish perls.
1852 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1853 on queue. Changes to match documented behaviour of showing count of
1854 messages matching specified criteria.
1856 PH/40 Changed the default ident timeout from 30s to 5s.
1858 PH/41 Added support for the use of login_cap features, on those BSD systems
1859 that have them, for controlling the resources used by pipe deliveries.
1861 PH/42 The content-scanning code uses fopen() to create files in which to put
1862 message data. Previously it was not paying any attention to the mode of
1863 the files. Exim runs with umask(0) because the rest of the code creates
1864 files with open(), and sets the required mode explicitly. Thus, these
1865 files were ending up world-writeable. This was not a big issue, because,
1866 being within the spool directory, they were not world-accessible. I have
1867 created a function called modefopen, which takes an additional mode
1868 argument. It sets umask(777), creates the file, chmods it to the required
1869 mode, then resets the umask. All the relevant calls to fopen() in the
1870 content scanning code have been changed to use this function.
1872 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1873 to 24 hours. This avoids potential overflow problems when processing G
1874 and H retry rules. I suspect nobody ever tinkers with this value.
1876 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1878 PH/45 When the plaintext authenticator is running as a client, the server's
1879 challenges are checked to ensure they are valid base64 strings. By
1880 default, the authentication attempt is cancelled if an invalid string is
1881 received. Setting client_ignore_invalid_base64 true ignores these errors.
1882 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1883 they are received. Thus, the responses can be made to depend on the
1884 challenges. If an invalid string is ignored, an empty string is placed in
1887 PH/46 Messages that are created by the autoreply transport now contains a
1888 References: header, in accordance with RFCs 2822 and 3834.
1890 PH/47 Added authenticated_sender_force to the smtp transport.
1892 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1894 PH/49 Installed latest patch from the Sieve maintainer.
1896 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1897 was also set, the appendfile transport was unnecessarily scanning a
1898 directory of message files (e.g. for maildir delivery) to find the count
1899 of files (along with the size), even though it did not need this
1900 information. It now does the scan only if it needs to find either the
1901 size of the count of files.
1903 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1905 PH/52 Two bugs concerned with error handling when the smtp transport is
1908 (i) Exim was not creating retry information for temporary errors given
1909 for individual recipients after the DATA command when the smtp transport
1910 was used in LMTP mode. This meant that they could be retried too
1911 frequently, and not timed out correctly.
1913 (ii) Exim was setting the flag that allows error details to be returned
1914 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1915 recipients that were returned after the DATA command.
1917 PH/53 This is related to PH/52, but is more general: for any failing address,
1918 when detailed error information was permitted to be returned to the
1919 sender, but the error was temporary, then after the final timeout, only
1920 "retry timeout exceeded" was returned. Now it returns the full error as
1921 well as "retry timeout exceeded".
1923 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1924 do this, and (what is worse) MTAs that accept it.
1926 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1927 will now be deprecated.
1929 PH/56 New os.c-cygwin from the Cygwin maintainer.
1931 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1932 formats (previously only available in exim formats via -bpr, -bpru,
1933 and -bpra. Now also available in native and exiqgrep formats)
1935 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1936 with very large, slow to parse queues
1938 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1940 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1942 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1943 responses to authentication challenges, though it was showing the
1944 challenges; (ii) I've removed the CR characters from the debug output for
1947 PH/58 Allow for the insertion of a newline as well as a space when a string
1948 is turned into more than one encoded-word during RFC 2047 encoding. The
1949 Sieve code now uses this.
1951 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1952 data_4xx, lost_connection, tls_required.
1954 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1955 message was being sent as an SMTP response.
1957 PH/61 Add -l and -k options to exicyclog.
1959 PH/62 When verifying, if an address was redirected to one new address, so that
1960 verification continued, and the new address failed or deferred after
1961 having set something in $address_data, the value of $address_data was not
1962 passed back to the ACL. This was different to the case when no
1963 redirection occurred. The value is now passed back in both cases.
1965 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1966 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1967 use login_cap.h, so on its own it isn't the distinguishing feature. The
1968 new name refers directly to the setclassresources() function.
1970 PH/65 Added configuration files for NetBSD3.
1972 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1974 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1975 is preferred over IPv4.
1977 PH/68 The bounce_return_message and bounce_return_body options were not being
1978 honoured for bounces generated during the reception of non-SMTP messages.
1979 In particular, this applied to messages rejected by the ACL. This bug has
1980 been fixed. However, if bounce_return_message is true and bounce_return_
1981 body is false, the headers that are returned for a non-SMTP message
1982 include only those that have been read before the error was detected.
1983 (In the case of an ACL rejection, they have all been read.)
1985 PH/69 The HTML version of the specification is now built in a directory called
1986 spec_html instead of spec.html, because the latter looks like a path with
1987 a MIME-type, and this confuses some software.
1989 PH/70 Catch two compiler warnings in sieve.c.
1991 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1992 function verify_get_ident() calls ip_connect() to connect a socket, but
1993 if the "connect()" function timed out, ip_connect() used to close the
1994 socket. However, verify_get_ident() also closes the socket later, and in
1995 between Exim writes to the log, which may get opened at this point. When
1996 the socket was closed in ip_connect(), the log could get the same file
1997 descriptor number as the socket. This naturally causes chaos. The fix is
1998 not to close the socket in ip_connect(); the socket should be closed by
1999 the function that creates it. There was only one place in the code where
2000 this was missing, in the iplookup router, which I don't think anybody now
2001 uses, but I've fixed it anyway.
2003 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2004 well as to direct DNS lookups. Otherwise the handling of names in host
2005 lists is inconsistent and therefore confusing.
2011 PH/01 Two changes to the default runtime configuration:
2013 (1) Move the checks for relay_from_hosts and authenticated clients from
2014 after to before the (commented out) DNS black list checks.
2016 (2) Add control=submission to the relay_from_hosts and authenticated
2017 clients checks, on the grounds that messages accepted by these
2018 statements are most likely to be submissions.
2020 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2022 (1) Generate an error if the third argument for the ${prvs expansion is
2025 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2028 (3) Reset the variables that are obtained from the first argument of
2029 ${prvscheck and used in the second argument before leaving the code,
2030 because their memory is reclaimed, so using them afterwards may do
2033 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2034 one (it's much easier than Tom thought :-).
2036 (5) Because of (4), we can now allow for the use of $prvscheck_result
2037 inside the third argument.
2039 PH/03 For some reason, the default setting of PATH when running a command from
2040 a pipe transport was just "/usr/bin". I have changed it to
2043 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2044 anything to be listed in the output from -bV.
2046 PH/05 When a filter generated an autoreply, the entire To: header line was
2047 quoted in the delivery log line, like this:
2049 => >A.N.Other <ano@some.domain> <original@ddress> ...
2051 This has been changed so that it extracts the operative address. There
2052 may be more than one such address. If so, they are comma-separated, like
2055 => >ano@some.domain,ona@other.domain <original@ddress> ...
2057 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2058 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2059 not being looked up in the rDNS to get a host name, Exim was showing the
2060 IP address twice in Received: lines, even though the IP addresses were
2061 identical. For example:
2063 Received: from [1.2.3.4] (helo=[1.2.3.4])
2065 However, if the real host name was known, it was omitting the HELO data
2066 if it matched the actual IP address. This has been tidied up so that it
2067 doesn't show the same IP address twice.
2069 PH/07 When both +timestamp and +memory debugging was on, the value given by
2070 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2071 called by the debug printing, thereby overwriting the timestamp buffer.
2072 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2074 PH/08 When the original message was included in an autoreply transport, it
2075 always said "this is a copy of the message, including all the headers",
2076 even if body_only or headers_only was set. It now gives an appropriate
2079 PH/09 Applied a patch from the Sieve maintainer which:
2081 o fixes some comments
2082 o adds the (disabled) notify extension core
2083 o adds some debug output for the result of if/elsif tests
2084 o points to the current vacation draft in the documentation
2085 and documents the missing references header update
2089 o fixes a bug in processing the envelope test (when testing
2090 multiple envelope elements, the last element determinted the
2093 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2094 Electronic Mail") by including:
2096 Auto-submitted: auto-generated
2098 in the messages that it generates (bounce messages and others, such as
2099 warnings). In the case of bounce messages for non-SMTP mesages, there was
2100 also a typo: it was using "Auto_submitted" (underscore instead of
2101 hyphen). Since every message generated by Exim is necessarily in response
2102 to another message, thes have all been changed to:
2104 Auto-Submitted: auto-replied
2106 in accordance with these statements in the RFC:
2108 The auto-replied keyword:
2110 - SHOULD be used on messages sent in direct response to another
2111 message by an automatic process,
2113 - MUST NOT be used on manually-generated messages,
2115 - MAY be used on Delivery Status Notifications (DSNs) and Message
2116 Disposition Notifications (MDNs),
2118 - MUST NOT be used on messages generated by automatic or periodic
2119 processes, except for messages which are automatic responses to
2122 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2123 to the default Received: header definition.
2125 PH/12 Added log selector acl_warn_skipped (default on).
2127 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2128 variables because (a) they are in the wrong storage pool and (b) even if
2129 they were copied, it wouldn't work properly because of the caching.
2131 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2132 checking when decoding. Apparently there are clients that generate
2133 overlong encoded strings. Why am I not surprised?
2135 PH/15 If the first argument of "${if match_address" was not empty, but did not
2136 contain an "@" character, Exim crashed. Now it writes a panic log message
2137 and treats the condition as false.
2139 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2141 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2142 extension "envelope-auth". The code is finished and in agreement with
2143 other implementations, but there is no documentation so far and in fact,
2144 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2145 not changing the active code.
2147 Print executed "if" and "elsif" statements when debugging is used. This
2148 helps a great deal to understand what a filter does.
2150 Document more things not specified clearly in RFC3028. I had all this
2151 sorted out, when out of a sudden new issues came to my mind. Oops."
2153 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2156 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2157 canonical form (as documented). However, after a host name lookup from
2158 the IP address, check_host() was doing a simple string comparison with
2159 addresses acquired from the DNS when checking that the found name did
2160 have the original IP as one of its addresses. Since any found IPv6
2161 addresses are likely to be in abbreviated form, the comparison could
2162 fail. Luckily, there already exists a function for doing the comparison
2163 by converting both addresses to binary, so now that is used instead of
2164 the text comparison.
2166 PH/20 There was another similar case to PH/19, when a complete host name was
2167 given in a host list; looking up its IP address could give an abbreviated
2168 form, whereas the current host's name might or might not be abbreviated.
2169 The same fix has been applied.
2175 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2176 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2179 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2180 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2182 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2184 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2185 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2186 Also, for "Auto-Submitted", treat anything other than "no" as
2187 non-personal, in accordance with RFC 3834. (Previously it treated
2188 anything starting "auto-" as non-personal.)
2190 TF/01 The control=submission/name=... option had a problem with syntax
2191 errors if the name included a slash character. The /name= option
2192 now slurps the rest of the string, so it can include any characters
2193 but it must come last in the list of options (after /sender_retain
2196 PH/05 Some modifications to the interface to the fake nameserver for the new
2204 TK/01 Added the "success_on_redirect" address verification option. See
2205 NewStuff for rationale and an example.
2207 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2209 PH/02 Patch to exigrep to allow it to work on syslog lines.
2211 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2212 fread() to skip over the body file's header line, because in Cygwin the
2213 header line is locked and is inaccessible.
2215 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2216 co-exist for some time) to make it clear that it is the Exim ID that is
2217 referenced, not the Message-ID: header line.
2219 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2220 string_format() function, because snprintf() does not exist on all
2223 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2224 ${stat: expansion item.
2226 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2227 protocol synchronization error", to keep the pedants happy.
2229 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2230 well as for IRIX systems, when gcc is being used. See the host.c source
2233 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2235 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2238 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2239 transport and to the smtp transport in LMTP mode.
2241 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2243 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2244 run a filter in a subprocess. This could lead to confusion in subsequent
2245 lookups in the parent process. There should also be a search_tidyup() at
2246 the end of the subprocess.
2248 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2249 only if the host matched helo_try_verify_hosts, which caused the
2250 verification to occur when the EHLO/HELO command was issued. The ACL just
2251 tested the remembered result. Now, if a previous verification attempt has
2252 not happened, "verify = helo" does it there and then.
2254 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2256 TK/03 Fix log output including CR from clamd.
2258 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2259 exist provoked a memory error which could cause a segfault.
2261 PH/15 Installed PCRE 6.2
2263 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2265 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2266 of the problem. Specifically, suggested +O2 rather than +O1 for the
2269 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2271 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2272 record, the port was not correctly logged when the outgoing_port log
2273 selector was set (it logged the transort's default port).
2275 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2276 fallback_hosts, and "hosts" in the smtp transport.
2278 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2279 host errors such as "Connection refused".
2281 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2282 authentication with radiusclient 0.4.9:
2284 - Error returned from rc_read_config was caught wrongly
2285 - Username/password not passed on to radius server due to wrong length.
2287 The presumption is that some radiusclient API changes for 4.51/PH/17
2288 were not taken care of correctly. The code is still untested by me (my
2289 Linux distribution still has 0.3.2 of radiusclient), but it was
2290 contributed by a Radius user.
2292 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2293 expanding the "port" option of the smtp transport.
2295 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2296 while reading a MIME header. Thanks to Tom Hughes for a patch.
2298 PH/24 Include config.h inside local_scan.h so that configuration settings are
2301 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2302 that in an ACL for RCPT (for example), you can examine exactly what was
2305 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2306 commands, but it was not correctly comparing the address with the actual
2307 client host address. Thus, it would show the EHLO address in Received:
2308 header lines when this was not necessary.
2310 PH/27 Added the % operator to ${eval:}.
2312 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2313 it should be ignoring failures (because with -C, for example, it has lost
2314 privilege). It wasn't ignoring creation failures other than "already
2317 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2320 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2321 because an input line was too long, either on its own, or by virtue of
2322 too many continuations, the temporary file was not being removed, and the
2323 return code was incorrect.
2325 PH/31 Missing "BOOL" in function definition in filtertest.c.
2327 PH/32 Applied Sieve patches from the maintainer.
2329 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2331 PH/33 Added "verify = not_blind".
2333 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2334 Local/Makefile (with some defaults set). These are used in built scripts
2335 such as exicyclog, but they have never been used in the exim_install
2336 script (though there are many overriding facilities there). I have
2337 arranged that the exim_install script now takes note of these two
2340 PH/35 Installed configuration files for Dragonfly.
2342 PH/36 When a locally submitted message by a trusted user did not contain a
2343 From: header, and the sender address was obtained from -f or from an SMTP
2344 MAIL command, and the trusted user did not use -F to supply a sender
2345 name, $originator_name was incorrectly used when constructing a From:
2346 header. Furthermore, $originator_name was used for submission mode
2347 messages from external hosts without From: headers in a similar way,
2348 which is clearly wrong.
2350 PH/37 Added control=suppress_local_fixups.
2352 PH/38 When log_selector = +received_sender was set, and the addition of the
2353 sender made the log line's construction buffer exactly full, or one byte
2354 less than full, an overflow happened when the terminating "\n" was
2357 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2358 when the result of a list match is failure because a DNS lookup failed.
2360 PH/40 RM_COMMAND is now used in the building process.
2362 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2363 the "build-* directories that it finds.
2365 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2366 address in a domain literal was a prefix of an interface address.
2368 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2369 when verifying a sender address, unless rewrite_headers is false.
2371 PH/44 Wrote a long comment about why errors_to addresses are verified as
2372 recipients, not senders.
2374 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2375 the ratelimit ACL was added.
2377 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2379 PH/46 Added extra information about PostgreSQL errors to the error string.
2381 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2382 avoiding the need to install special zones in a real server. This is
2383 backwards compatible; if it can't find the fake resolver, it drops back.
2384 Thus, both old and new test suites can be run.
2386 TF/02 Added util/ratelimit.pl
2388 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2389 clock is set back in time.
2391 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2392 Candler <B.Candler@pobox.com>.
2394 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2395 OK for addresses that are the result of redirections.
2397 PH/48 A number of further additions for the benefit of the new test suite,
2398 including a fake gethostbyname() that interfaces to the fake DNS resolver
2401 TF/06 The fix for widen_domains has also been applied to qualify_single and
2402 search_parents which are the other dnslookup options that can cause
2405 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2408 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2410 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2411 service" if it encountered an MX record with an empty target hostname.
2412 The message is now "an MX or SRV record indicated no SMTP service".
2414 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2415 if the DNS of the sending site is misconfigured. This is quite a
2416 common situation. This change restores the behaviour of treating a
2417 helo verification defer as a failure.
2419 PH/51 If self=fail was set on a router, the bounce message did not include the
2420 actual error message.
2426 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2428 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2429 command itself ended in error, the underlying message about the transport
2430 filter timeout was being overwritten with the pipe command error. Now the
2431 underlying error message should be appended to the second error message.
2433 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2435 PH/02 Reduce the amount of output that "make" produces by default. Full output
2436 can still be requested.
2438 PH/03 The warning log line about a condition test deferring for a "warn" verb
2439 was being output only once per connection, rather than after each
2440 occurrence (because it was using the same function as for successful
2441 "warn" verbs). This seems wrong, so I have changed it.
2443 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2444 it should not have, which might have caused a crash in the right
2445 circumstances, but probably never did.
2447 PH/04 Installed a modified version of Tony Finch's patch to make submission
2448 mode fix the return path as well as the Sender: header line, and to
2449 add a /name= option so that you can make the user's friendly name appear
2452 TF/03 Added the control = fakedefer ACL modifier.
2454 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2455 Mark Lowes for thorough testing.
2457 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2459 TK/03 Merged latest SRS patch from Miles Wilton.
2461 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2462 with the definition in sysexits.h (which is #included earlier).
2463 Fortunately, Exim does not actually use EX_OK. The code used to try to
2464 preserve the sysexits.h value, by assumimg that macro definitions were
2465 scanned for macro replacements. I have been disabused of this notion,
2466 so now the code just undefines EX_OK before #including unistd.h.
2468 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2469 in the smtp transport. When a block could not be written in a single
2470 write() function, the timeout was being re-applied to each part-write.
2471 This seems wrong - if the receiver was accepting one byte at a time it
2472 would take for ever. The timeout is now adjusted when this happens. It
2473 doesn't have to be particularly precise.
2475 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2476 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2478 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2479 without a domain if the check to <postmaster@domain> fails.
2481 SC/01 Eximstats: added -xls and the ability to specify output files
2482 (patch written by Frank Heydlauf).
2484 SC/02 Eximstats: use FileHandles for outputing results.
2486 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2488 SC/04 Eximstats: fixed display of large numbers with -nvr option
2490 SC/05 Eximstats: fixed merging of reports with empty tables.
2492 SC/06 Eximstats: added the -include_original_destination flag
2494 SC/07 Eximstats: removed tabs and trailing whitespace.
2496 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2498 TK/06 MBOX spool code: Add real "From " MBOX separator line
2499 so the .eml file is really in mbox format (even though
2500 most programs do not really care). Patch from Alex Miller.
2502 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2503 The latter is generated from $received_to and is only set if the
2504 message has one envelope recipient. SA can use these headers,
2505 obviously out-of-the-box. Patch from Alex Miller.
2507 PH/08 The ${def test on a variable was returning false if the variable's
2508 value was "0", contrary to what the specification has always said!
2509 The result should be true unless the variable is empty.
2511 PH/09 The syntax error of a character other than { following "${if
2512 def:variable_name" (after optional whitespace) was not being diagnosed.
2513 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2514 accidental colon was present, for example, could give incorrect results.
2516 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2517 result is used (not including appendfile, where other changes are about
2520 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2521 This involved changing a lot of size variables from int to off_t. It
2522 should work with maildirs and everything.
2524 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2525 spamd dying while we are connected to it.
2527 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2530 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2531 function for BDB 4.3.
2533 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2535 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2536 the sources for PCRE have been re-arranged and more files are now
2539 PH/15 The code I had for printing potentially long long variables in PH/11
2540 above was not the best (it lost precision). The length of off_t variables
2541 is now inspected at build time, and an appropriate printing format (%ld
2542 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2543 to be "long long int" or "long int". This is needed for the internal
2544 formatting function string_vformat().
2546 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2547 the configuration file to be ":syslog", then the script "guesses" where
2548 the logs files are, rather than using the compiled in default. In our
2549 case the guess is not the same as the compiled default, so the script
2550 suddenly stopped working when I started to use syslog. The patch checks
2551 to see if log_file_path is "". If so, it attempts to read it from exim
2552 with no configuration file to get the compiled in version, before it
2553 falls back to the previous guessing code."
2555 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2556 implementing BATV in an Exim configuration. See NewStuff for the gory
2559 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2560 Makefile that are specific to HP-UX.
2562 PH/18 If the "use_postmaster" option was set for a recipient callout together
2563 with the "random" option, the postmaster address was used as the MAIL
2564 FROM address for the random test, but not for the subsequent recipient
2565 test. It is now used for both.
2567 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2568 patch removes a few documentation additions to RFC 3028, because the
2569 latest draft now contains them. It adds the new en;ascii-case comparator
2570 and a new error check for 8bit text in MIME parts. Comparator and
2571 require names are now matched exactly. I enabled the subaddress
2572 extension, but it is not well tested yet (read: it works for me)."
2574 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2575 rework some of the code of TK/09 above to avoid the hardwired use of
2576 "%lld" and "long long". Replaced the call to snprintf() with a call to
2579 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2580 records point to non-existent hosts", "retry timeout exceeded", and
2581 "retry time not reached for any host after a long failure period".
2583 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2584 experimental DomainKeys support:
2586 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2587 (2) On an error such as an illegally used "control", the wrong name for
2588 the control was given.
2590 These problems did NOT occur unless DomainKeys support was compiled.
2592 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2594 PH/24 Added ${if match_ip condition.
2596 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2597 databases so that it will be absolutely obvious if a crash occurs in the
2598 DB library. This is a regular occurrence (often caused by mis-matched
2601 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2602 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2603 picked up on a user's system that detects such things. There doesn't seem
2604 to be a gcc warning option for this - only an attribute that has to be
2605 put on the function's prototype. It seems that in Fedora Core 4 they have
2606 set this on a number of new functions. No doubt there will be more in due
2609 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2610 specify a transport. However, if an address that was verified by such a
2611 router was the subject of a callout, Exim crashed because it tried to
2612 read the rcpt_include_affixes from the non-existent transport. Now it
2613 just assumes that the setting of that option is false. This bug was
2614 introduced by 4.51/PH/31.
2616 PH/28 Changed -d+all to exclude +memory, because that information is very
2617 rarely of interest, but it makes the output a lot bigger. People tend to
2618 do -d+all out of habit.
2620 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2621 code in os-type was giving problems when libc.so lives in lib64, like on
2624 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2625 aren't the modern standard, and it seems that some systems' include files
2626 don't always have them. Exim was already checking for some of the newer
2627 ones like T_AAAA, and defining it itself. I've added checks for all the
2628 record types that Exim uses.
2630 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2631 not automatically generating a new one, as it is supposed to. This
2632 prevented TLS from working. If the file did exist, but contained invalid
2633 data, a new version was generated, as expected. It was only the case of a
2634 non-existent file that was broken.
2636 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2637 with a change in libdomainkeys > 0.64.
2639 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2640 from DNS. If the selector record carries the flag, it now has
2641 precedence over the domain-wide flag.
2643 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2645 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2646 the use of a port name that isn't defined in /etc/services) occurred, the
2647 message was deferred as in a normal delivery, and thus remained on the
2648 spool, instead of being failed because of the mua_wrapper setting. This
2649 is now fixed, and I tidied up some of the mua_wrapper messages at the
2652 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2653 the messages in a hash of arrays rather than using individual hashes.
2654 This is a bit cleaner and results in dramatic memory savings, albeit
2655 at a slight CPU cost.
2657 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2658 as requested by Marc Sherman.
2660 SC/10 Eximstats: added histograms for user specified patterns as requested
2663 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2665 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2666 fopen() in the content-scanning modules that did not already have it.
2672 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2673 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2675 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2677 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2679 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2680 to test Sieve filters that use "vacation".
2682 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2683 that changes the way the GnuTLS parameters are stored in the cache file.
2684 The new format can be generated externally. For backward compatibility,
2685 if the data in the cache doesn't make sense, Exim assumes it has read an
2686 old-format file, and it generates new data and writes a new file. This
2687 means that you can't go back to an older release without removing the
2690 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2691 work if there are any delivery delays because "one_time" forces the
2692 parent to be marked "delivered", so its unseen clone is never tried
2693 again. For this reason, Exim now forbids the simultaneous setting of
2696 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2697 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2698 entry if you want to know the details. The fix, however, neglected to
2699 consider the case when local delivery batching is involved. The test for
2700 "previously delivered" was not happening when checking to see if an
2701 address could be batched with a previous (undelivered) one; under
2702 certain circumstances this could lead to multiple deliveries to the same
2705 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2706 in its include files, and this causes problems building Exim.
2708 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2709 header_syntax) but Exim was just ignoring anything given after a slash.
2710 In particular, this caused confusion with an attempt to use "verify =
2711 reverse_host_lookup/defer_ok". An error is now given when options are
2712 supplied for verify items that do not have them. (Maybe reverse_host_
2713 lookup should have a defer_ok option, but that's a different point.)
2715 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2716 defined by RFC 821) to 2048, because there were problems with some AUTH
2717 commands, and RFC 1869 says the size should be increased for extended
2718 SMTP commands that take arguments.
2720 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2723 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2724 "unknown" error; now it says that the functionality isn't in the binary.
2726 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2727 an address' error message when a string expansion fails (syntax or
2728 whatever). Otherwise the password may appear in the log. Following change
2729 PH/42 below, there is no longer a chance of it appearing in a bounce
2732 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2734 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2735 of its addresses was ever tried. (Bugzilla bug #2.)
2737 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2738 the result incorrectly in the debug output. (It correctly added a newline
2739 to what was transported.)
2741 TF/01 Added $received_time.
2743 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2744 commented out examples of how to interface to a virus scanner and to
2745 SpamAssassin. Also added commented examples of av_scanner and
2746 spamd_address settings.
2748 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2749 and controls are allowed in which ACLs. There were a couple of minor
2750 errors. Some of the entries in the conditions table (which is a table of
2751 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2752 negation of where the condition IS allowed.
2754 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2756 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2757 header file does not have a version number, so I've had to invent a new
2758 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2759 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2760 radiusclient), but it was contributed by a Radius user.
2762 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2763 files or directories, for OpenSSL.
2765 PH/19 When an Exim process that is running as root has to create an Exim log
2766 file, it does so in a subprocess that runs as exim:exim so as to get the
2767 ownership right at creation (otherwise, other Exim processes might see
2768 the file with the wrong ownership). There was no test for failure of this
2769 fork() call, which would lead to the process getting stuck as it waited
2770 for a non-existent subprocess. Forks do occasionally fail when resources
2771 run out. I reviewed all the other calls to fork(); they all seem to check
2774 PH/20 When checking for unexpected SMTP input at connect time (before writing
2775 the banner), Exim was not dealing correctly with a non-positive return
2776 from the read() function. If the client had disconnected by this time,
2777 the result was a log entry for a synchronization error with an empty
2778 string after "input=" when read() returned zero. If read() returned -1
2779 (an event I could not check), uninitialized data bytes were printed.
2780 There were reports of junk text (parts of files, etc) appearing after
2783 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2785 PH/22 Added support for macro redefinition, and (re)definition in between
2786 driver and ACL definitions.
2788 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2789 forgetting to use the resulting value; it was using the unexpanded value.
2791 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2792 hadn't been configured. The fix is from Juergen Kreileder, who
2793 understands it better than I do:
2795 "Here's what I see happening with three configured cyrus_sasl
2796 authenticators configured (plain, login, cram-md5):
2798 On startup auth_cyrus_sasl_init() gets called for each of these.
2799 This means three calls to sasl_listmech() without a specified mech_list.
2800 => SASL tests which mechs of all available mechs actually work
2801 => three warnings about OTP not working
2802 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2804 With the patch, sasl_listmech() also gets called three times. But now
2805 SASL's mech_list option is set to the server_mech specified in the the
2806 authenticator. Or in other words, the answer from sasl_listmech()
2807 gets limited to just the mech you're testing for (which is different
2809 => the return list contains just 'plain' or 'login', 'cram-md5' or
2810 nothing depending on the value of ob->server_mech.
2812 I've just tested the patch: Authentication still works fine,
2813 unavailable mechs specified in the exim configuration are still
2814 caught, and the auth.log warnings about OTP are gone."
2816 PH/25 When debugging is enabled, the contents of the command line are added
2817 to the debugging output, even when log_selector=+arguments is not
2820 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2821 answer is "GNU", and only if the return is "GNU/something" is the answer
2824 PH/27 $acl_verify_message is now set immediately after the failure of a
2825 verification in an ACL, and so is available in subsequent modifiers. In
2826 particular, the message can be preserved by coding like this:
2828 warn !verify = sender
2829 set acl_m0 = $acl_verify_message
2831 Previously, $acl_verify_message was set only while expanding "message"
2832 and "log_message" when a very denied access.
2834 PH/28 Modified OS/os.c-Linux with
2836 -#ifndef OS_LOAD_AVERAGE
2837 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2839 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2840 nomenclature these days.)
2842 PH/29 Installed patch from the Sieve maintainer that adds the options
2843 sieve_useraddress and sieve_subaddress to the redirect router.
2845 PH/30 In these circumstances:
2846 . Two addresses routed to the same list of hosts;
2847 . First host does not offer TLS;
2848 . First host accepts first address;
2849 . First host gives temporary error to second address;
2850 . Second host offers TLS and a TLS session is established;
2851 . Second host accepts second address.
2852 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2853 and peerdn, if requested) that were in fact used only for the second
2856 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2857 attention to any local part prefix or suffix that was matched by the
2858 router that accepted the address. It now behaves in the same way as it
2859 does for delivery: the affixes are removed from the local part unless
2860 rcpt_include_affixes is set on the transport.
2862 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2863 timeout during the DATA phase of an incoming message.
2865 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2866 applied a patch sanctioned by the Sieve maintainer.
2868 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2869 the uid or gid is negative. A case of a negative gid caused this to be
2870 noticed. The fix allows for either to be negative.
2872 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2873 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2876 PH/36 The change PH/12 above was broken. Fixed it.
2878 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2879 the grounds that routing the same address twice would always produce the
2880 same answer. This might have been true once, but it is certainly no
2881 longer true now. Routing a child address may depend on the previous
2882 routing that produced that child. Some complicated redirection strategies
2883 went wrong when messages had multiple recipients, and made Exim's
2884 behaviour dependent on the order in which the addresses were given.
2886 I have moved the duplicate checking until after the routing is complete.
2887 Exim scans the addresses that are assigned to local and remote
2888 transports, and removes any duplicates. This means that more work will be
2889 done, as duplicates will always all be routed, but duplicates are
2890 presumably rare, so I don't expect this is of any significance.
2892 For deliveries to pipes, files, and autoreplies, the duplicate checking
2893 still happens during the routing process, since they are not going to be
2896 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2897 It corrects a timeout issue with spamd. This is Ian's comment: "The
2898 background is that sometimes spamd either never reads data from a
2899 connection it has accepted, or it never writes response data. The exiscan
2900 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2901 blindly assumes that writes won't block so it may never time out."
2903 PH/39 Allow G after quota size as well as K and M.
2905 PH/40 The value set for $authenticated_id in an authenticator may not contain
2906 binary zeroes or newlines because the value is written to log lines and
2907 to spool files. There was no check on this. Now the value is run through
2908 the string_printing() function so that such characters are converted to
2909 printable escape sequences.
2911 PH/41 $message_linecount is a new variable that contains the total number of
2912 lines in the message. Compare $body_linecount, which is the count for the
2915 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2916 bounce and delay warning messages, except in certain special cases, which
2919 (a) An SMTP error message from a remote host;
2920 (b) A message specified in a :fail: redirection;
2921 (c) A message specified in a "fail" command in a system filter;
2922 (d) A message specified in a FAIL return from the queryprogram router;
2923 (e) A message specified by the cannot_route_message router option.
2925 In these cases only, Exim does include the error details in bounce and
2926 warning messages. There are also a few cases where bland messages such
2927 as "unrouteable address" or "local delivery error" are given.
2929 PH/43 $value is now also set for the "else" part of a ${run expansion.
2931 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2932 being worked on, but at least Exim now implements the latest version to
2935 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2936 process to complete was treated as a delivery failure, a timeout while
2937 writing the message to the pipe was logged, but erroneously treated as a
2938 successful delivery. Such timeouts include transport filter timeouts. For
2939 consistency with the overall process timeout, these timeouts are now
2940 treated as errors, giving rise to delivery failures by default. However,
2941 there is now a new Boolean option for the pipe transport called
2942 timeout_defer, which, if set TRUE, converts the failures into defers for
2943 both kinds of timeout. A transport filter timeout is now identified in
2946 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2947 systems where "make" and "gmake" are different, calling "gmake" at top
2948 level broke things. I've arranged for the value of $(MAKE) to be passed
2949 from the Makefile to this script so that it can call the same version of
2953 A note about Exim versions 4.44 and 4.50
2954 ----------------------------------------
2956 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2957 changes of various kinds. As a consequence, a big documentation update was
2958 needed. This delayed the release for rather longer than seemed good, especially
2959 in the light of a couple of (minor) security issues. Therefore, the changes
2960 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2961 release. So 4.44 and 4.50 are in effect two different branches that both start
2964 I have left the 4.50 change log unchanged; it contains all the changes since
2965 4.43. The change log for 4.44 is below; many of its items are identical to
2966 those for 4.50. This seems to be the most sensible way to preserve the
2967 historical information.
2973 1. Minor wording change to the doc/README.SIEVE file.
2975 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2976 computation of the current number of files was incorrect.
2978 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2979 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2982 4. Give more explanation in the error message when the command for a transport
2983 filter fails to execute.
2985 5. There are several places where Exim runs a non-Exim command in a
2986 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2987 was being done only for the command run by the queryprogram router. It is
2988 now done for all such subprocesses. The other cases are: ${run, transport
2989 filters, and the commands run by the lmtp and pipe transports.
2991 6. Added CONFIGURE_GROUP build-time option.
2993 7. Some older OS have a limit of 256 on the maximum number of file
2994 descriptors. Exim was using setrlimit() to set 1000 as a large value
2995 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2996 systems. I've change it so that if it can't get 1000, it tries for 256.
2998 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2999 was an oversight, and furthermore, ever since the addition of extra
3000 controls (e.g. 4.43/32), the checks on when to allow different forms of
3001 "control" were broken. There should now be diagnostics for all cases when a
3002 control that does not make sense is encountered.
3004 9. Added the /retain_sender option to "control=submission".
3006 10. $recipients is now available in the predata ACL (oversight).
3008 11. Tidy the search cache before the fork to do a delivery from a message
3009 received from the command line. Otherwise the child will trigger a lookup
3010 failure and thereby defer the delivery if it tries to use (for example) a
3011 cached ldap connection that the parent has called unbind on.
3013 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3014 of $address_data from the recipient verification was clobbered by the
3015 sender verification.
3017 13. The value of address_data from a sender verification is now available in
3018 $sender_address_data in subsequent conditions in the ACL statement.
3020 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3022 15. Added a new option "connect=<time>" to callout options, to set a different
3025 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3026 was its contents. (It was OK if the option was not defined at all.)
3028 17. A "Completed" log line is now written for messages that are removed from
3029 the spool by the -Mrm option.
3031 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3032 information about exactly what failed.
3034 19. Added -dd to debug only the daemon process.
3036 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3037 handles timeouts, both on the server side and network timeouts. Renamed the
3038 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3040 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3041 It is now set to "smtps".
3043 22. $host_address is now set to the target address during the checking of
3044 ignore_target_hosts.
3046 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3047 being passed; this would have caused $sender_host_name to have been used if
3048 matching the list had actually called for a host name (not very likely,
3049 since this list is usually IP addresses). A host name is now passed as
3052 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3053 code that specifies a non-restarting handler (typically sigaction() in
3054 modern systems) in an attempt to fix a rare and obscure crash bug.
3056 25. Narrowed the window for a race in the daemon that could cause it to ignore
3057 SIGCHLD signals. This is not a major problem, because they are used only to
3058 wake it up if nothing else does.
3060 26. A malformed maildirsize file could cause Exim to calculate negative values
3061 for the mailbox size or file count. Odd effects could occur as a result.
3062 The maildirsize information is now recalculated if the size or filecount
3065 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3066 support for a long time. Removed HAVE_SYS_VFS_H.
3068 28. Installed the latest version of exipick from John Jetmore.
3070 29. In an address list, if the pattern was not a regular expression, an empty
3071 subject address (from a bounce message) matched only if the pattern was an
3072 empty string. Non-empty patterns were not even tested. This was the wrong
3073 because it is perfectly reasonable to use an empty address as part of a
3074 database query. An empty address is now tested by patterns that are
3075 lookups. However, all the other forms of pattern expect the subject to
3076 contain a local part and a domain, and therefore, for them, an empty
3077 address still always fails if the pattern is not itself empty.
3079 30. Exim went into a mad DNS loop when attempting to do a callout where the
3080 host was specified on an smtp transport, and looking it up yielded more
3081 than one IP address.
3083 31. Re-factored the code for checking spool and log partition space into a
3084 function that finds that data and another that does the check. The former
3085 is then used to implement four new variables: $spool_space, $log_space,
3086 $spool_inodes, and $log_inodes.
3088 32. The RFC2047 encoding function was originally intended for short strings
3089 such as real names; it was not keeping to the 75-character limit for
3090 encoded words that the RFC imposes. It now respects the limit, and
3091 generates multiple encoded words if necessary. To be on the safe side, I
3092 have increased the buffer size for the ${rfc2047: expansion operator from
3095 33. It is now permitted to omit both strings after an "if" condition; if the
3096 condition is true, the result is "true". As before, when the second string
3097 is omitted, a false condition yields an empty string. This makes it less
3098 cumbersome to write custom ACL and router conditions.
3100 34. Failure to deliver a bounce message always caused it to be frozen, even if
3101 there was an errors_to setting on the router. The errors_to setting is now
3104 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3105 canonical form (fully expanded) before being placed in
3106 $sender_host_address.
3108 36. The table in the code that translates DNS record types into text (T_A to
3109 "A" for instance) was missing entries for NS and CNAME. It is just possible
3110 that this could have caused confusion if both these types were looked up
3111 for the same domain, because the text type is used as part of Exim's
3112 per-process caching. But the chance of anyone hitting this buglet seems
3115 37. The dnsdb lookup has been extended in a number of ways.
3117 (1) There is a new type, "zns", which walks up the domain tree until it
3118 finds some nameserver records. It should be used with care.
3120 (2) There is a new type, "mxh", which is like "mx" except that it returns
3121 just the host names, not the priorities.
3123 (3) It is now possible to give a list of domains (or IP addresses) to be
3124 looked up. The behaviour when one of the lookups defers can be
3125 controlled by a keyword.
3127 (4) It is now possible to specify the separator character for use when
3128 multiple records are returned.
3130 38. The dnslists ACL condition has been extended: it is now possible to supply
3131 a list of IP addresses and/or domains to be looked up in a particular DNS
3134 39. Added log_selector=+queue_time_overall.
3136 40. When running the queue in the test harness, wait just a tad after forking a
3137 delivery process, to get repeatability of debugging output.
3139 41. Include certificate and key file names in error message when GnuTLS fails
3140 to set them up, because the GnuTLS error message doesn't include the name
3141 of the failing file when there is a problem reading it.
3143 42. Allow both -bf and -bF in the same test run.
3145 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3147 44. The "Exiscan patch" is now merged into the mainline Exim source.
3149 45. Sometimes the final signoff response after QUIT could fail to get
3150 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3151 < 0 before doing a fflush(). This bug looks as though it goes back to the
3152 introduction of TLS in release 3.20, but "sometimes" must have been rare
3153 because the tests only now provoked it.
3155 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3156 (this can affect the format of dates).
3158 47. exim_tidydb, when checking for the continued existence of a message for
3159 which it has found a message-specific retry record, was not finding
3160 messages that were in split spool directories. Consequently, it was
3161 deleting retry records that should have stayed in existence.
3163 48. Steve fixed some bugs in eximstats.
3165 49. The SPA authentication driver was not abandoning authentication and moving
3166 on to the next authenticator when an expansion was forced to fail,
3167 contradicting the general specification for all authenticators. Instead it
3168 was generating a temporary error. It now behaves as specified.
3170 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3171 (the order specifies the preference for clients). The order is now AES256,
3172 AES128, 3DES, ARCFOUR128.
3174 51. Small patch to Sieve code - explicitly set From: when generating an
3177 52. Exim crashed if a remote delivery caused a very long error message to be
3178 recorded - for instance if somebody sent an entire SpamAssassin report back
3179 as a large number of 550 error lines. This bug was coincidentally fixed by
3180 increasing the size of one of Exim's internal buffers (big_buffer) that
3181 happened as part of the Exiscan merge. However, to be on the safe side, I
3182 have made the code more robust (and fixed the comments that describe what
3185 53. Now that there can be additional text after "Completed" in log lines (if
3186 the queue_time_overall log selector is set), a one-byte patch to exigrep
3187 was needed to allow it to recognize "Completed" as not the last thing in
3190 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3191 patch that reportedly fixes this has been added. I am not expert enough to
3192 create a test for it. This is what the patch creator wrote:
3194 "I found a little strange behaviour of ldap code when working with
3195 Windows 2003 AD Domain, where users was placed in more than one
3196 Organization Units. When I tried to give exim partial DN, the exit code
3197 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3198 But simultaneously result of request was absolutely normal ldap result,
3199 so I produce this patch..."
3201 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3202 REFERENCE, so I have modified the code to exclude the patch when that macro
3205 55. Some experimental protocols are using DNS PTR records for new purposes. The
3206 keys for these records are domain names, not reversed IP addresses. The
3207 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3208 leaves it alone. Component reversal etc. now happens only for IP addresses.
3211 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3213 57. Double the size of the debug message buffer (to 2048) so that more of very
3214 long debug lines gets shown.
3216 58. The exicyclog utility now does better if the number of log files to keep
3217 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3219 59. Two changes related to the smtp_active_hostname option:
3221 (1) $smtp_active_hostname is now available as a variable.
3222 (2) The default for smtp_banner uses $smtp_active_hostname instead
3223 of $primary_hostname.
3225 60. The host_aton() function is supposed to be passed a string that is known
3226 to be a valid IP address. However, in the case of IPv6 addresses, it was
3227 not checking this. This is a hostage to fortune. Exim now panics and dies
3228 if the condition is not met. A case was found where this could be provoked
3229 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3230 components; fortuitously, this particular loophole had already been fixed
3231 by change 4.50/55 above.
3233 If there are any other similar loopholes, the new check in host_aton()
3234 itself should stop them being exploited. The report I received stated that
3235 data on the command line could provoke the exploit when Exim was running as
3236 exim, but did not say which command line option was involved. All I could
3237 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3238 running as the user.
3241 61. There was a buffer overflow vulnerability in the SPA authentication code
3242 (which came originally from the Samba project). I have added a test to the
3243 spa_base64_to_bits() function which I hope fixes it.
3246 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3247 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3249 63. The daemon start-up calls getloadavg() while still root for those OS that
3250 need the first call to be done as root, but it missed one case: when
3251 deliver_queue_load_max is set with deliver_drop_privilege. This is
3252 necessary for the benefit of the queue runner, because there is no re-exec
3253 when deliver_drop_privilege is set.
3255 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3256 This has been fixed.
3258 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3259 was in use, was not putting the data itself into the right store pool;
3260 consequently, it could be overwritten for a subsequent message in the same
3261 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3264 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3266 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3267 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3268 the function were treating the return as a boolean value, which happened to
3269 work because 0=false and not-0=true, but is not correct code.
3271 68. The host_aton() function was not handling scoped IPv6 addresses (those
3272 with, for example, "%eth0" on the end) correctly.
3274 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3275 negated items (that is, ~something) in unsigned ints. Some compilers
3276 apparently mutter when there is no cast.
3278 70. If an address verification called from an ACL failed, and did not produce a
3279 user-specific message (i.e. there was only a "system" message), nothing was
3280 put in $acl_verify_message. In this situation, it now puts the system
3283 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3284 session; change 4.31/43 added the unwanted input to the log line - except
3285 that it did not do this in the start of session case. It now does.
3287 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3288 This is wrong and can cause the other end to generate a synchronization
3289 error if it is another Exim or anything else that does the synchronization
3290 check. A QUIT command is no longer sent after a timeout.
3292 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3293 during host lookups.
3295 74. The defer_ok option of callout verification was not working if it was used
3296 when verifying addresses in header lines, that is, for this case:
3298 verify = header_sender/callout=defer_ok
3300 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3301 those file descriptors could be used for SMTP connections. If anything
3302 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3303 could be sent to the SMTP client, causing chaos. The daemon now opens
3304 stdin, stdout, and stderr to /dev/null when it puts itself into the
3307 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3308 log by default. The user can override this with suitable Perl magic.
3310 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3311 the log message when discard triggers, was not working for the DATA ACL or
3312 for the non-SMTP ACL.
3314 78. Error message wording change in sieve.c.
3316 79. If smtp_accept_max_per_host was set, the number of connections could be
3317 restricted to fewer than expected, because the daemon was trying to set up
3318 a new connection before checking whether the processes handling previous
3319 connections had finished. The check for completed processes is now done
3320 earlier. On busy systems, this bug wouldn't be noticed because something
3321 else would have woken the daemon, and it would have reaped the completed
3324 80. If a message was submitted locally by a user whose login name contained one
3325 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3326 It caused a spool format error. I have fixed the spool reading code. A
3327 related problem was that the "from" clause in the Received: line became
3328 illegal because of the space(s). It is now covered by ${quote_local_part.
3330 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3333 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3335 83. Patch from Sieve maintainer to fix unterminated string problem in
3336 "vacation" handling.
3338 84. Some minor changes to the Linux configuration files to help with other
3339 OS variants using glibc.
3341 85. One more patch for Sieve to update vacation handling to latest spec.
3344 ----------------------------------------------------
3345 See the note above about the 4.44 and 4.50 releases.
3346 ----------------------------------------------------
3352 1. Change 4.43/35 introduced a bug that caused file counts to be
3353 incorrectly computed when quota_filecount was set in an appendfile
3356 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3357 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3360 3. Give more explanation in the error message when the command for a transport
3361 filter fails to execute.
3363 4. There are several places where Exim runs a non-Exim command in a
3364 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3365 was being done only for the command run by the queryprogram router. It is
3366 now done for all such subprocesses. The other cases are: ${run, transport
3367 filters, and the commands run by the lmtp and pipe transports.
3369 5. Some older OS have a limit of 256 on the maximum number of file
3370 descriptors. Exim was using setrlimit() to set 1000 as a large value
3371 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3372 systems. I've change it so that if it can't get 1000, it tries for 256.
3374 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3375 was an oversight, and furthermore, ever since the addition of extra
3376 controls (e.g. 4.43/32), the checks on when to allow different forms of
3377 "control" were broken. There should now be diagnostics for all cases when a
3378 control that does not make sense is encountered.
3380 7. $recipients is now available in the predata ACL (oversight).
3382 8. Tidy the search cache before the fork to do a delivery from a message
3383 received from the command line. Otherwise the child will trigger a lookup
3384 failure and thereby defer the delivery if it tries to use (for example) a
3385 cached ldap connection that the parent has called unbind on.
3387 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3388 of $address_data from the recipient verification was clobbered by the
3389 sender verification.
3391 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3392 was its contents. (It was OK if the option was not defined at all.)
3394 11. A "Completed" log line is now written for messages that are removed from
3395 the spool by the -Mrm option.
3397 12. $host_address is now set to the target address during the checking of
3398 ignore_target_hosts.
3400 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3401 being passed; this would have caused $sender_host_name to have been used if
3402 matching the list had actually called for a host name (not very likely,
3403 since this list is usually IP addresses). A host name is now passed as
3406 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3407 code that specifies a non-restarting handler (typically sigaction() in
3408 modern systems) in an attempt to fix a rare and obscure crash bug.
3410 15. Narrowed the window for a race in the daemon that could cause it to ignore
3411 SIGCHLD signals. This is not a major problem, because they are used only to
3412 wake it up if nothing else does.
3414 16. A malformed maildirsize file could cause Exim to calculate negative values
3415 for the mailbox size or file count. Odd effects could occur as a result.
3416 The maildirsize information is now recalculated if the size or filecount
3419 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3420 support for a long time. Removed HAVE_SYS_VFS_H.
3422 18. Updated exipick to current release from John Jetmore.
3424 19. Allow an empty sender to be matched against a lookup in an address list.
3425 Previously the only cases considered were a regular expression, or an
3428 20. Exim went into a mad DNS lookup loop when doing a callout where the
3429 host was specified on the transport, if the DNS lookup yielded more than
3432 21. The RFC2047 encoding function was originally intended for short strings
3433 such as real names; it was not keeping to the 75-character limit for
3434 encoded words that the RFC imposes. It now respects the limit, and
3435 generates multiple encoded words if necessary. To be on the safe side, I
3436 have increased the buffer size for the ${rfc2047: expansion operator from
3439 22. Failure to deliver a bounce message always caused it to be frozen, even if
3440 there was an errors_to setting on the router. The errors_to setting is now
3443 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3444 canonical form (fully expanded) before being placed in
3445 $sender_host_address.
3447 24. Updated eximstats to version 1.33
3449 25. Include certificate and key file names in error message when GnuTLS fails
3450 to set them up, because the GnuTLS error message doesn't include the name
3451 of the failing file when there is a problem reading it.
3453 26. Expand error message when OpenSSL has problems setting up cert/key files.
3456 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3457 (this can affect the format of dates).
3459 28. exim_tidydb, when checking for the continued existence of a message for
3460 which it has found a message-specific retry record, was not finding
3461 messages that were in split spool directories. Consequently, it was
3462 deleting retry records that should have stayed in existence.
3464 29. eximstats updated to version 1.35
3465 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3466 1.35 - bugfix such that pie charts by volume are generated correctly
3468 30. The SPA authentication driver was not abandoning authentication and moving
3469 on to the next authenticator when an expansion was forced to fail,
3470 contradicting the general specification for all authenticators. Instead it
3471 was generating a temporary error. It now behaves as specified.
3473 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3474 (the order specifies the preference for clients). The order is now AES256,
3475 AES128, 3DES, ARCFOUR128.
3477 31. Small patch to Sieve code - explicitly set From: when generating an
3480 32. Exim crashed if a remote delivery caused a very long error message to be
3481 recorded - for instance if somebody sent an entire SpamAssassin report back
3482 as a large number of 550 error lines. This bug was coincidentally fixed by
3483 increasing the size of one of Exim's internal buffers (big_buffer) that
3484 happened as part of the Exiscan merge. However, to be on the safe side, I
3485 have made the code more robust (and fixed the comments that describe what
3488 33. Some experimental protocols are using DNS PTR records for new purposes. The
3489 keys for these records are domain names, not reversed IP addresses. The
3490 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3491 leaves it alone. Component reversal etc. now happens only for IP addresses.
3494 34. The host_aton() function is supposed to be passed a string that is known
3495 to be a valid IP address. However, in the case of IPv6 addresses, it was
3496 not checking this. This is a hostage to fortune. Exim now panics and dies
3497 if the condition is not met. A case was found where this could be provoked
3498 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3499 components; fortuitously, this particular loophole had already been fixed
3500 by change 4.50/55 or 4.44/33 above.
3502 If there are any other similar loopholes, the new check in host_aton()
3503 itself should stop them being exploited. The report I received stated that
3504 data on the command line could provoke the exploit when Exim was running as
3505 exim, but did not say which command line option was involved. All I could
3506 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3507 running as the user.
3510 35. There was a buffer overflow vulnerability in the SPA authentication code
3511 (which came originally from the Samba project). I have added a test to the
3512 spa_base64_to_bits() function which I hope fixes it.
3515 36. The daemon start-up calls getloadavg() while still root for those OS that
3516 need the first call to be done as root, but it missed one case: when
3517 deliver_queue_load_max is set with deliver_drop_privilege. This is
3518 necessary for the benefit of the queue runner, because there is no re-exec
3519 when deliver_drop_privilege is set.
3521 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3522 was in use, was not putting the data itself into the right store pool;
3523 consequently, it could be overwritten for a subsequent message in the same
3524 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3527 38. Sometimes the final signoff response after QUIT could fail to get
3528 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3529 < 0 before doing a fflush(). This bug looks as though it goes back to the
3530 introduction of TLS in release 3.20, but "sometimes" must have been rare
3531 because the tests only now provoked it.
3537 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3538 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3539 Now it returns an integer. A number of calls were still expecting a T/F
3540 return. Fortuitously, in all cases, the tests worked in OK situations,
3541 which is the norm. However, things would have gone wrong on any write
3542 failures on the smtp file descriptor. This function is used when sending
3543 messages over SMTP and also when doing verify callouts.
3545 2. When Exim is called to do synchronous delivery of a locally submitted
3546 message (the -odf or -odi options), it no longer closes stderr before doing
3549 3. Implemented the mua_wrapper option.
3551 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3553 5. Implemented the functions header_remove(), header_testname(),
3554 header_add_at_position(), and receive_remove_recipient(), and exported them
3557 6. If an ACL "warn" statement specified the addition of headers, Exim already
3558 inserted X-ACL-Warn: at the start if there was no header name. However, it
3559 was not making this test for the second and subsequent header lines if
3560 there were newlines in the string. This meant that an invalid header could
3561 be inserted if Exim was badly configured.
3563 7. Allow an ACL "warn" statement to add header lines at the start or after all
3564 the Received: headers, as well as at the end.
3566 8. Added the rcpt_4xx retry error code.
3568 9. Added postmaster_mailfrom=xxx to callout verification option.
3570 10. Added mailfrom=xxxx to the callout verification option, for verify=
3573 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3574 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3576 12. Inserted some casts to stop certain compilers warning when using pointer
3577 differences as field lengths or precisions in printf-type calls (mostly
3578 affecting debugging statements).
3580 13. Added optional readline() support for -be (dynamically loaded).
3582 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3583 same clock tick as a message's arrival, so that its received time was the
3584 same as the "first fail" time on the retry record, and that message
3585 remained on the queue past the ultimate address timeout, every queue runner
3586 would try a delivery (because it was past the ultimate address timeout) but
3587 after another failure, the ultimate address timeout, which should have then
3588 bounced the address, did not kick in. This was a "< instead of <=" error;
3589 in most cases the first failure would have been in the next clock tick
3590 after the received time, and all would be well.
3592 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3593 being recognized when the domain list was tested by the match_domain
3594 condition in an expansion string.
3596 16. Added the ${str2b64: operator.
3598 17. Exim was always calling setrlimit() to set a large limit for the number of
3599 processes, without checking whether the existing limit was already
3600 adequate. (It did check for the limit on file descriptors.) Furthermore,
3601 errors from getrlimit() and setrlimit() were being ignored. Now they are
3602 logged to the main and panic logs, but Exim does carry on, to try to do its
3603 job under whatever limits there are.
3605 18. Imported PCRE 5.0.
3607 19. Trivial typo in log message " temporarily refused connection" (the leading
3610 20. If the log selector return_path_on_delivery was set and an address was
3611 redirected to /dev/null, the delivery process crashed because it assumed
3612 that a return path would always be set for a "successful" delivery. In this
3613 case, the whole delivery is bypassed as an optimization, and therefore no
3616 21. Internal re-arrangement: the function for sending a challenge and reading
3617 a response while authentication was assuming a zero-terminated challenge
3618 string. It's now changed to take a pointer and a length, to allow for
3619 binary data in such strings.
3621 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3623 23. Exim was not respecting finduser_retries when seeking the login of the
3624 uid under which it was called; it was always trying 10 times. (The default
3625 setting of finduser_retries is zero.) Also, it was sleeping after the final
3626 failure, which is pointless.
3628 24. Implemented tls_on_connect_ports.
3630 25. Implemented acl_smtp_predata.
3632 26. If the domain in control=submission is set empty, Exim assumes that the
3633 authenticated id is a complete email address when it generates From: or
3634 Sender: header lines.
3636 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3637 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3638 chown and chgrp in /bin and hostname in /usr/bin.
3640 28. Exim was keeping the "process log" file open after each use, just as it
3641 does for the main log. This opens the possibility of it remaining open for
3642 long periods when the USR1 signal hits a daemon. Occasional processlog
3643 errors were reported, that could have been caused by this. Anyway, it seems
3644 much more sensible not to leave this file open at all, so that is what now
3647 29. The long-running daemon process does not normally write to the log once it
3648 has entered its main loop, and it closes the log before doing so. This is
3649 so that log files can straightforwardly be renamed and moved. However,
3650 there are a couple of unusual error situations where the daemon does write
3651 log entries, and I had neglected to close the log afterwards.
3653 30. The text of an SMTP error response that was received during a remote
3654 delivery was being truncated at 512 bytes. This is too short for some of
3655 the long messages that one sometimes sees. I've increased the limit to
3658 31. It is now possible to make retry rules that apply only when a message has a
3659 specific sender, in particular, an empty sender.
3661 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3662 it possible to be selective about when SMTP synchronization is enforced.
3664 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3666 32. Implemented hosts_connection_nolog.
3668 33. Added an ACL for QUIT.
3670 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3673 35. Added mailbox_size and mailbox_filecount to appendfile.
3675 36. Added control = no_multiline_responses to ACLs.
3677 37. There was a bug in the logic of the code that waits for the clock to tick
3678 in the case where the clock went backwards by a substantial amount such
3679 that the microsecond fraction of "now" was more than the microsecond
3680 fraction of "then" (but the whole seconds number was less).
3682 38. Added support for the libradius Radius client library this is found on
3683 FreeBSD (previously only the radiusclient library was supported).
3689 1. When certain lookups returned multiple values in the form name=value, the
3690 quoting of the values was not always being done properly. Specifically:
3691 (a) If the value started with a double quote, but contained no whitespace,
3693 (b) If the value contained whitespace other than a space character (i.e.
3694 tabs or newlines or carriage returns) it was not quoted.
3695 This fix has been applied to the mysql and pgsql lookups by writing a
3696 separate quoting function and calling it from the lookup code. The fix
3697 should probably also be applied to nisplus, ibase and oracle lookups, but
3698 since I cannot test any of those, I have not disturbed their existing code.
3700 2. A hit in the callout cache for a specific address caused a log line with no
3701 reason for rejecting RCPT. Now it says "Previous (cached) callout
3702 verification failure".
3704 3. There was an off-by-one bug in the queryprogram router. An over-long
3705 return line was truncated at 256 instead of 255 characters, thereby
3706 overflowing its buffer with the terminating zero. As well as fixing this, I
3707 have increased the buffer size to 1024 (and made a note to document this).
3709 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3710 when Exim is waiting for an SMTP response from a remote server, Exim
3711 restarts its select() call on the socket, thereby resetting its timeout.
3712 This is not a problem when such interrupts are rare. Somebody set up a cron
3713 job to run exiwhat every 2 minutes, which is less than the normal select()
3714 timeout (5 or 10 minutes). This meant that the select() timeout never
3715 kicked in because it was always reset. I have fixed this by comparing the
3716 time when an interrupt arrives with the time at the start of the first call
3717 to select(). If more time than the timeout has elapsed, the interrupt is
3718 treated as a timeout.
3720 5. Some internal re-factoring in preparation for the addition of Sieve
3721 extensions (by MH). In particular, the "personal" test is moved to a
3722 separate function, and given an option for scanning Cc: and Bcc: (which is
3723 not set for Exim filters).
3725 6. When Exim created an email address using the login of the caller as the
3726 local part (e.g. when creating a From: or Sender: header line), it was not
3727 quoting the local part when it contained special characters such as @.
3729 7. Installed new OpenBSD configuration files.
3731 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3732 try to make them clearer.
3734 9. Callout options, other than the timeout value, were being ignored when
3735 verifying sender addresses in header lines. For example, when using
3737 verify = header_sender/callout=no_cache
3739 the cache was (incorrectly) being used.
3741 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3742 only the Cygwin environment.
3744 11. When return_path_on_delivery was set as a log selector, if different remote
3745 addresses in the same message used different return paths and parallel
3746 remote delivery occurred, the wrong values would sometimes be logged.
3747 (Whenever a remote delivery process finished, the return path value from
3748 the most recently started remote delivery process was logged.)
3750 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3751 lines when AUTH and/or TLS are in use. This is the "received protocol"
3752 field. Exim used to use "asmtp" for authenticated SMTP, without any
3753 indication (in the protocol name) for TLS use. Now it follows the RFC and
3754 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3755 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3756 names appear in log lines as well as in Received: header lines.
3758 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3759 extensions, and comparison tests, and to fix some bugs.
3761 14. Changes to the "personal" filter test:
3763 (1) The test was buggy in that it was just doing the equivalent of
3764 "contains" tests on header lines. For example, if a user's address was
3765 anne@some.where, the "personal" test would incorrectly be true for
3767 To: susanne@some.where
3769 This test is now done by extracting each address from the header in turn,
3770 and checking the entire address. Other tests that are part of "personal"
3771 are now done using regular expressions (for example, to check local parts
3772 of addresses in From: header lines).
3774 (2) The list of non-personal local parts in From: addresses has been
3775 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3776 taken from the Sieve specification recommendations.
3778 (3) If the message contains any header line starting with "List-" it is
3779 treated as non-personal.
3781 (4) The test for "circular" in the Subject: header line has been removed
3782 because it now seems ill-conceived.
3784 15. Minor typos in src/EDITME comments corrected.
3786 16. Installed latest exipick from John Jetmore.
3788 17. If headers_add on a router specified a text string that was too long for
3789 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3790 of string_sprintf() is now avoided.
3792 18. $message_body_size was not set (it was always zero) when running the DATA
3793 ACL and the local_scan() function.
3795 19. For the "mail" command in an Exim filter, no default was being set for
3796 the once_repeat time, causing a random time value to be used if "once" was
3797 specified. (If the value happened to be <= 0, no repeat happened.) The
3798 default is now 0s, meaning "never repeat". The "vacation" command was OK
3799 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3800 (I found it when inspecting the code).
3802 20. There is now an overall timeout for performing a callout verification. It
3803 defaults to 4 times the callout timeout, which applies to individual SMTP
3804 commands during the callout. The overall timeout applies when there is more
3805 than one host that can be tried. The timeout is checked before trying the
3806 next host. This prevents very long delays if there are a large number of
3807 hosts and all are timing out (e.g. when the network connections are timing
3808 out). The value of the overall timeout can be changed by specifying an
3809 additional sub-option for "callout", called "maxwait". For example:
3811 verify = sender/callout=5s,maxwait=20s
3813 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3814 to the end before writing, but this should make it even safer).
3816 22. Exim was forgetting that it had advertised PIPELINING for the second and
3817 subsequent messages on an SMTP connection. It was also not resetting its
3818 memory on STARTTLS and an internal HELO.
3820 23. When Exim logs an SMTP synchronization error within a session, it now
3821 records whether PIPELINING has been advertised or not.
3823 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3824 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3825 rather than long int.
3827 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3829 26. Added the never_mail option to autoreply.
3835 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3836 crash if the getsockname() call failed; this can happen if a connection is
3837 closed very soon after it is established. The problem was simply in the
3838 order in which certain operations were done, causing Exim to try to write
3839 to the SMTP stream before it had set up the file descriptor. The bug has
3840 been fixed by making things happen in the correct order.
3846 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3847 before the connection was closed, thus losing the rejection response.
3849 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3850 some early Solaris releases, but causes trouble in current releases where
3851 socklen_t is defined.
3853 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3856 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3859 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3860 that releasing the top of it at the end releases what was used for sub-
3861 expansions (unless the block got too big). However, discard this block if
3862 the first thing is a variable or header, so that we can use its block when
3863 it is dynamic (useful for very large $message_headers, for example).
3865 6. Lookups now cache *every* query, not just the most recent. A new, separate
3866 store pool is used for this. It can be recovered when all lookup caches are
3867 flushed. Lookups now release memory at the end of their result strings.
3868 This has involved some general refactoring of the lookup sources.
3870 7. Some code has been added to the store_xxx() functions to reduce the amount
3871 of flapping under certain conditions.
3873 8. log_incoming_interface used to affect only the <= reception log lines. Now
3874 it causes the local interface and port to be added to several more SMTP log
3875 lines, for example "SMTP connection from", and rejection lines.
3877 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3879 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3881 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3882 could be overwritten at the end of the current message (or the start of a
3883 new message if it was set in a HELO ACL). The value is now preserved for
3884 the duration of the SMTP connection.
3886 12. If a transport had a headers_rewrite setting, and a matching header line
3887 contained an unqualified address, that address was qualified, even if it
3888 did not match any rewriting rules. The underlying bug was that the values
3889 of the flags that permit the existence of unqualified sender and recipient
3890 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3891 non-local messages, and by -bnq for local messages) were not being
3892 preserved with the message after it was received.
3894 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3895 "next input=" as part of the text comprising the host identity instead of
3896 the correct text. The code was using the same buffer for two different
3897 strings. However, depending on which order the printing function evaluated
3898 its arguments, the bug did not always show up. Under Linux, for example, my
3899 test suite worked just fine.
3901 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3902 doesn't work with some older versions of Perl. It has been changed to "my",
3903 which in any case is probably the better facility to use.
3905 15. A really picky compiler found some instances of statements for creating
3906 error messages that either had too many or two few arguments for the format
3909 16. The size of the buffer for calls to the DNS resolver has been increased
3910 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3911 for addresses that have a lot of PTR records. This alleviates a problem; it
3912 does not fully solve it.
3914 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3915 buffer now truncates the list and logs the incident, which is the same
3916 action as happens when Exim is looking up a host name and its aliases.
3917 Previously in this situation something unpredictable would happen;
3918 sometimes it was "internal error: store_reset failed".
3920 18. If a server dropped the connection unexpectedly when an Exim client was
3921 using GnuTLS and trying to read a response, the client delivery process
3922 crashed while trying to generate an error log message.
3924 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3925 string, for example:
3927 warn message = H1: something\nH2: something
3929 the text was added as a single header line from Exim's point of view
3930 though it ended up OK in the delivered message. However, searching for the
3931 second and subsequent header lines using $h_h2: did not work. This has been
3932 fixed. Similarly, if a system filter added multiple headers in this way,
3933 the routers could not see them.
3935 20. Expanded the error message when iplsearch is called with an invalid key to
3936 suggest using net-iplsearch in a host list.
3938 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3939 ACLs are no longer actually imposed (and a message to that effect is
3942 22. If a "gecos" field in a passwd entry contained escaped characters, in
3943 particular, if it contained a \" sequence, Exim got it wrong when building
3944 a From: or a Sender: header from that name. A second bug also caused
3945 incorrect handling when an unquoted " was present following a character
3946 that needed quoting.
3948 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3949 was not being matched caselessly.
3951 24. Arranged for all hyphens in the exim.8 source to be escaped with
3954 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3955 properly. Recipient callout cache records were still being keyed to include
3956 the sender, even when use_sender was set false. This led to far more
3957 callouts that were necessary. The sender is no longer included in the key
3958 when use_sender is false.
3960 26. Added "control = submission" modifier to ACLs.
3962 27. Added the ${base62d: operator to decode base 62 numbers.
3964 28. dnsdb lookups can now access SRV records.
3966 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3967 the configuration file.
3969 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3970 (-v) mode. This makes the output for a verbose queue run more intelligible.
3972 31. Added a use_postmaster feature to recipient callouts.
3974 32. Added the $body_zerocount variable, containing the number of binary zero
3975 bytes in the message body.
3977 33. The time of last modification of the "new" subdirectory is now used as the
3978 "mailbox time last read" when there is a quota error for a maildir
3981 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3983 35. Added +ignore_unknown as a special item in host lists.
3985 36. Code for decoding IPv6 addresses in host lists is now included, even if
3986 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3987 address was recognized as an IP address, but was then not correctly decoded
3988 into binary, causing unexpected and incorrect effects when compared with
3995 1. Very minor rewording of debugging text in manualroute to say "list of
3996 hosts" instead of "hostlist".
3998 2. If verify=header_syntax was set, and a header line with an unqualified
3999 address (no domain) and a large number of spaces between the end of the
4000 name and the colon was received, the reception process suffered a buffer
4001 overflow, and (when I tested it) crashed. This was caused by some obsolete
4002 code that should have been removed. The fix is to remove it!
4004 3. When running in the test harness, delay a bit after writing a bounce
4005 message to get a bit more predictability in the log output.
4007 4. Added a call to search_tidyup() just before forking a reception process. In
4008 theory, someone could use a lookup in the expansion of smtp_accept_max_
4009 per_host which, without the tidyup, could leave open a database connection.
4011 5. Added the variables $recipient_data and $sender_data which get set from a
4012 lookup success in an ACL "recipients" or "senders" condition, or a router
4013 "senders" option, similar to $domain_data and $local_part_data.
4015 6. Moved the writing of debug_print from before to after the "senders" test
4018 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4019 problems for message scanning, either using a data ACL, or using
4020 local_scan() because the Received: header was not generated till after they
4021 were called (in order to set the time as the time of reception completion).
4022 I have revised the way this works. The header is now generated after the
4023 body is received, but before the ACL or local_scan() are called. After they
4024 are run, the timestamp in the header is updated.
4030 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4031 before starting a queue runner without re-exec. This happened only when
4032 deliver_drop_privilege was set or when the Exim user was set to root. The
4033 effect of the bug was that timeouts during subsequent deliveries caused
4034 crashes instead of being properly handled. The handler is now left at its
4035 default (and expected) setting.
4037 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4038 message, again when deliver_drop_privilege is set or Exim is run as root.
4039 The bug described in (1) was not present in this case, but the tidying up
4040 of the other signals was missing. I have made the two cases consistent.
4042 3. The ignore_target_hosts setting on a manualroute router was being ignored
4043 for hosts that were looked up using the /MX notation.
4045 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4048 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4049 operated on the sender address. After changing the $sender_address to <>
4050 for the sender address verify, Exim was re-instated it as the original
4051 (before rewriting) address, but remembering that it had rewritten it, so it
4052 wasn't rewriting it again. This bug also had the effect of breaking the
4053 sender address verification caching when the sender address was rewritten.
4055 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4056 This has been changed so that if the ip literal address matches
4057 ignore_target_hosts, the router declines.
4059 7. Added expansion conditions match_domain, match_address, and match_local_
4060 part (NOT match_host).
4062 8. The placeholder for the Received: header didn't have a length field set.
4064 9. Added code to Exim itself and to exim_lock to test for a specific race
4065 condition that could lead to file corruption when using MBX delivery. The
4066 issue is with the lockfile that is created in /tmp. If this file is removed
4067 after a process has opened it but before that process has acquired a lock,
4068 there is the potential for a second process to recreate the file and also
4069 acquire a lock. This could lead to two Exim processes writing to the file
4070 at the same time. The added code performs the same test as UW imapd; it
4071 checks after acquiring the lock that its file descriptor still refers to
4072 the same named file.
4074 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4075 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4078 11. Added the smtp_active_hostname option. If used, this will typically be made
4079 to depend on the incoming interface address. Because $interface_address is
4080 not set up until the daemon has forked a reception process, error responses
4081 that can happen earlier (such as "too many connections") no longer contain
4084 12. If an expansion in a condition on a "warn" statement fails because a lookup
4085 defers, the "warn" statement is abandoned, and the next ACL statement is
4086 processed. Previously this caused the whole ACL to be aborted.
4088 13. Added the iplsearch lookup type.
4090 14. Added ident_timeout as a log selector.
4092 15. Added tls_certificate_verified as a log selector.
4094 16. Added a global option tls_require_ciphers (compare the smtp transport
4095 option of the same name). This controls incoming TLS connections.
4097 17. I finally figured out how to make tls_require_ciphers do a similar thing
4098 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4099 before starting the TLS session.
4101 18. Tabs are now shown as \t in -bP output.
4103 19. If the log selector return_path_on_delivery was set, Exim crashed when
4104 bouncing a message because it had too many Received: header lines.
4106 20. If two routers both had headers_remove settings, and the first one included
4107 a superfluous trailing colon, the final name in the first list and the
4108 first name in the second list were incorrectly joined into one item (with a
4109 colon in the middle).
4115 1. Added -C and -D options to the exinext utility, mainly to make it easier
4116 to include in the automated testing, but these could be helpful when
4117 multiple configurations are in use.
4119 2. The exinext utility was not formatting the output nicely when there was
4120 an alternate port involved in the retry record key, nor when there was a
4121 message id as well (for retries that were specific to a specific message
4122 and a specific host). It was also confused by IPv6 addresses, because of
4123 the additional colons they contain. I have fixed the IPv4 problem, and
4124 patched it up to do a reasonable job for IPv6.
4126 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4127 delivery, the log line now contains "pipelined" if PIPELINING was used.
4129 4. An SMTP transport process used to panic and die if the bind() call to set
4130 an explicit outgoing interface failed. This has been changed; it is now
4131 treated in the same way as a connect() failure.
4133 5. A reference to $sender_host_name in the part of a conditional expansion
4134 that was being skipped was still causing a DNS lookup. This no longer
4137 6. The def: expansion condition was not recognizing references to header lines
4138 that used bh_ and bheader_.
4140 7. Added the _cache feature to named lists.
4142 8. The code for checking quota_filecount in the appendfile transport was
4143 allowing one more file than it should have been.
4145 9. For compatibility with Sendmail, the command line option
4153 and sets the incoming protocol and host name (for trusted callers). The
4154 host name and its colon can be omitted when only the protocol is to be set.
4155 Note the Exim already has two private options, -pd and -ps, that refer to
4156 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4157 "s", but I don't think that's a major issue.
4159 10. A number of refactoring changes to the code, none of which should affect
4162 (a) The number of logging options was getting close to filling up the
4163 32-bit word that was used as a bit map. I have split them into two classes:
4164 those that are passed in the argument to log_write(), and those that are
4165 only ever tested independently outside of that function. These are now in
4166 separate 32-bit words, so there is plenty of room for expansion again.
4167 There is no change in the user interface or the logging behaviour.
4169 (b) When building, for example, log lines, the code previously used a
4170 macro that called string_cat() twice, in order to add two strings. This is
4171 not really sufficiently general. Furthermore, there was one instance where
4172 it was actually wrong because one of the argument was used twice, and in
4173 one call a function was used. (As it happened, calling the function twice
4174 did not affect the overall behaviour.) The macro has been replaced by a
4175 function that can join an arbitrary number of extra strings onto a growing
4178 (c) The code for expansion conditions now uses a table and a binary chop
4179 instead of a serial search (which was left over from when there were very
4180 few conditions). Also, it now recognizes conditions like "pam" even when
4181 the relevant support is not compiled in: a suitably worded error message is
4182 given if an attempt is made to use such a condition.
4184 11. Added ${time_interval:xxxxx}.
4186 12. A bug was causing one of the ddress fields not to be passed back correctly
4187 from remote delivery subprocesses. The field in question was not being
4188 subsequently used, so this caused to problems in practice.
4190 13. Added new log selectors queue_time and deliver_time.
4192 14. Might have fixed a bug in maildirsizefile handling that threw up
4193 "unexpected character" debug warnings, and recalculated the data
4194 unnecessarily. In any case, I expanded the warning message to give more
4197 15. Added the message "Restricted characters in address" to the statements in
4198 the default ACL that block characters like @ and % in local parts.
4200 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4201 Three changes have been made:
4203 (a) There was a serious bug; a negative response to MAIL caused the whole
4204 recipient domain to be cached as invalid, thereby blocking all messages
4205 to all local parts at the same domain, from all senders. This bug has
4206 been fixed. The domain is no longer cached after a negative response to
4207 MAIL if the sender used is not empty.
4209 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4212 (c) A new callout option, "use_sender" has been added for people who want
4213 the modified behaviour.
4219 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4222 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4223 indeed breaks things for older releases.
4225 3. Added additional logging to the case where there is a problem reading data
4226 from a filter that is running in a subprocess using a pipe, in order to
4227 try to track down a specific problem.
4229 4. Testing facility fudge: when running in the test harness and attempting
4230 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4231 getting "No route to host". Convert this to a timeout.
4233 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4236 6. Some OS don't have socklen_t but use size_t instead. This affects the
4237 fifth argument of getsockopt() amongst other things. This is now
4238 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4239 can be set for individual OS. I have set it for SunOS5, OSF1, and
4240 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4241 some earlier ones do not.
4243 7. Change 4.30/15 was not doing the test caselessly.
4245 8. The standard form for an IPv6 address literal was being rejected by address
4246 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4247 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4248 this, as well as the form without the "IPv6" on the front (but only when
4249 address literals are enabled, of course).
4251 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4253 10. Exim crashed if a message with an empty sender address specified by -f
4254 encountered a router with an errors_to setting. This could be provoked only
4255 by a command such as
4259 where an empty string was supplied; "<>" did not hit this bug.
4261 11. Installed PCRE release 4.5.
4263 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4264 remained set. It is now erased.
4266 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4267 times from message ids (which are base 36 rather than the normal 62).
4269 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4270 were being counted as actual protocol errors, and logged if the log
4271 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4272 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4273 and DATA following a set of rejected RCPTs do not count as protocol errors.
4274 In other words, Exim assumes they were pipelined, though this may not
4275 actually be the case. Of course, in all cases the client gets an
4276 appropriate error code.
4278 15. If a lookup fails in an ACL condition, a message about the failure may
4279 be available; it is used if testing the ACL cannot continue, because most
4280 such messages specify what the cause of the deferral is. However, some
4281 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4282 that caused an old message to be retained and used if a later statement
4283 caused a defer, replacing the real cause of the deferral.
4285 16. If an IP address had so many PTR records that the DNS lookup buffer
4286 was not large enough to hold them, Exim could crash while trying to process
4287 the truncated data. It now detects and logs this case.
4289 17. Further to 4.21/58, another change has been made: if (and only if) the
4290 first line of a message (the first header line) ends with CRLF, a bare LF
4291 in a subsequent header line has a space inserted after it, so as not to
4292 terminate the header.
4294 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4295 unnecessarily, used atoi() instead of strtol(), and didn't check the
4296 termination when getting file sizes from file names by regex.
4298 19. Completely re-implemented the support for maildirsize files, in the light
4299 of a number of problems with the previous contributed implementation
4300 (4.30/29). In particular:
4302 . If the quota is zero, the maildirsize file is maintained, but no quota is
4305 . If the maildir directory does not exist, it is created before any attempt
4306 to write a maildirsize file.
4308 . The quota value in the file is just a cache; if the quota is changed in
4309 the transport, the new value overrides.
4311 . A regular expression is available for excluding directories from the
4314 20. The autoreply transport checks the characters in options that define the
4315 message's headers; it allows continued headers, but it was checking with
4316 isspace() after an embedded newline instead of explicitly looking for a
4319 21. If all the "regular" hosts to which an address was routed had passed their
4320 expiry times, and had not reached their retry times, the address was
4321 bounced, even if fallback hosts were defined. Now Exim should go on to try
4324 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4325 equivalent code in the SMTP transport. Some hosts send humungous responses
4326 to HELO/EHLO, more than 1024 it seems.
4328 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4329 gives compiler warnings in some environments. I've now done it "properly",
4332 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4333 (because of problems with the built-in one) was declared to return uschar *
4334 instead of char *, causing compiler failure.
4336 25. Fixed a file descriptor leak when processing alias/forward files.
4338 26. Fixed a minor format string issue in dbfn.c.
4340 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4342 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4343 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4346 29. When a sender address is verified, it is cached, to save repeating the test
4347 when there is more than one recipient in a message. However, when the
4348 verification involves a callout, it is possible for different callout
4349 options to be set for different recipients. It is too complicated to keep
4350 track of this in the cache, so now Exim always runs a verification when a
4351 callout is required, relying on the callout cache for the optimization.
4352 The overhead is duplication of the address routing, but this should not be
4355 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4356 to be verified with callout=postmaster, and the main callout worked but the
4357 postmaster check failed, the verification correctly failed. However, if a
4358 subsequent RCPT command asked for sender verification *without* the
4359 postmaster check, incorrect caching caused this verification also to fail,
4362 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4363 it was not caching the DNS options (qualify_single, search_parents) that
4364 were used when the lookup failed. A subsequent lookup with different
4365 options therefore always gave the same answer, though there were cases
4366 where it should not have. (Example: a "domains = !$mx_any" option on a
4367 dnslookup router: the "domains" option is always processed without any
4368 widening, but the router might have qualify_single set.) Now Exim uses the
4369 cached value only when the same options are set.
4371 32. Added John Jetmore's "exipick" utility to the distribution.
4373 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4374 than a timeout (e.g. a certificate is required, and is not provided), an
4375 Exim server now closes the connection immediately. Previously it waited for
4376 the client to close - but if the client is SSL, it seems that they each
4377 wait for each other, leading to a delay before one of them times out.
4379 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4380 maintained 0.8.x compatibility because I don't think many are using it, and
4381 it is clearly obsolete.
4383 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4386 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4387 client certificate was expired. A simple patch fixes this, though I don't
4388 understand the full logic of why the verify callback is called multiple
4391 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4392 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4393 which causes problems with some clients (such as the Certicom SSL Plus
4394 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4395 disables the coutermeasure allowing Eudora to connect."
4397 38. Exim was not checking that a write() to a log file succeeded. This could
4398 lead to Bad Things if a log got too big, in particular if it hit a file
4399 size limit. Exim now panics and dies if it cannot write to a log file, just
4400 as it does if it cannot open a log file.
4402 39. Modified OS/Makefile-Linux so that it now contains
4404 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4406 The two -D definitions ensure that Exim is compiled with large file
4407 support, which makes it possible to handle log files that are bigger than
4410 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4411 instance) a domain was checked against a named list that involved a lookup,
4412 causing $domain_data to be set, then another domain was checked against the
4413 same list, then the first domain was re-checked, the value of $domain_data
4414 after the final check could be wrong. In particular, if the second check
4415 failed, it could be set empty. This bug probably also applied to
4418 41. The strip_trailing_dot option was not being applied to the address given
4419 with the -f command-line option.
4421 42. The code for reading a message's header from the spool was incrementing
4422 $received_count, but never initializing it. This meant that the value was
4423 incorrect (doubled) while delivering a message in the same process in which
4424 it was received. In the most common configuration of Exim, this never
4425 happens - a fresh exec is done - but it can happen when
4426 deliver_drop_privilege is set.
4428 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4429 it now includes up to 150 characters of the unexpected data in the log
4432 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4433 and building data strings. The size of both of these buffers was 10 000
4434 bytes - far larger than anybody would *ever* want, thought I. Needless to
4435 say, somebody hit the limit. I have increased the maximum line length to
4436 20 000 and the maximum data length of concatenated lines to 100 000. I have
4437 also fixed two bugs, because there was no checking on these buffers. Tsk,
4438 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4439 buffer is too small.
4441 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4442 lsearch lookups. Now it does.
4444 46. When parsing a route_list item in a manualroute router, a fixed-length
4445 buffer was used for the list of hosts. I made this 1024 bytes long,
4446 thinking that nobody would ever have a list of hosts that long. Wrong.
4447 Somebody had a whole pile of complicated expansion conditions, and the
4448 string was silently truncated, leading to an expansion error. It turns out
4449 that it is easier to change to an unlimited length (owing to other changes
4450 that have happened since this code was originally written) than to build
4451 structure for giving a limitation error. The length of the item that
4452 expands into the list of hosts is now unlimited.
4454 47. The lsearch lookup could not handle data where the length of text line was
4455 more than 4095 characters. Such lines were truncated, leading to shortened
4456 data being returned. It should now handle lines of any length.
4458 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4459 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4462 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4463 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4464 no longer happens in comment lines. A list of replacements is now placed
4465 at the head of all of the source files, except those whose only change is
4466 to replace PERL_COMMAND in the very first #! line.
4468 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4469 messages on the queue, with a bottom-up merge sort, using code contributed
4470 by Michael Haardt. This should make operations like -bp somewhat faster on
4471 large queues. It won't affect queue runners, except when queue_run_in_order
4474 51. Installed eximstats 1.31 in the distribution.
4476 52. Added support for SRV lookups to the dnslookup router.
4478 53. If an ACL referred to $message_body or $message_body_end, the value was not
4479 reset for any messages that followed in the same SMTP session.
4481 54. The store-handling optimization for building very long strings was not
4482 differentiating between the different store pools. I don't think this
4483 actually made any difference in practice, but I've tidied it.
4485 55. While running the routers to verify a sender address, $sender_address
4486 was still set to the sender address. This is wrong, because when routing to
4487 send a bounce to the sender, it would be empty. Therefore, I have changed
4488 it so that, while verifying a sender address, $sender_address is set to <>.
4489 (There is no change to what happens when verifying a recipient address.)
4491 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4492 target A or AAAA records (if not already returned) without resetting the
4493 qualify_single or search_parents options of the DNS resolver. These are
4494 inappropriate in this case because the targets of MX and SRV records must
4495 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4496 target that, when qualified, matched something in the local domain. These
4497 two options are now turned off when doing these lookups.
4499 57. It seems that at least some releases of Reiserfs (which does not have the
4500 concept of a fixed number of inodes) returns zero and not -1 for the
4501 number of available inodes. This interacted badly with check_spool_inodes,
4502 which assumed that -1 was the "no such thing" setting. What I have done is
4503 to check that the total number of inodes is greater than zero before doing
4504 the test of how many are available.
4506 58. When a "warn" ACL statement has a log_message modifier, the message is
4507 remembered, and not repeated. This is to avoid a lot of repetition when a
4508 message has many recipients that cause the same warning to be written.
4509 Howewer, Exim was preserving the list of already written lines for an
4510 entire SMTP session, which doesn't seem right. The memory is now reset if a
4511 new message is started.
4513 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4514 debugging output unless log_rewrite was also set.
4516 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4517 of (int)(handle) when we know that handle contains (void *)(-1).
4519 61. The Exim daemon panic-logs an error return when it closes the incoming
4520 connection. However "connection reset by peer" seems to be common, and
4521 isn't really an error worthy of noting specially, so that particular error
4524 62. When Exim is trying to find all the local interfaces, it used to panic and
4525 die if the ioctl to get the interface flags failed. However, it seems that
4526 on at least one OS (Solaris 9) it is possible to have an interface that is
4527 included in the list of interfaces, but for which you get a failure error
4528 for this call. This happens when the interface is not "plumbed" into a
4529 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4530 failure of the "get flags" call assumes that the interface is down.
4532 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4533 makes life easier for people who are doing arithmetic on fields extracted
4534 from dates, where you often get leading zeros that should not be
4535 interpreted as octal.
4537 64. Added qualify_domain to the redirect router, to override the global
4540 65. If a pathologically long header line contained very many addresses (the
4541 report of this problem mentioned 10 000) and each of them was rewritten,
4542 Exim could use up a very large amount of memory. (It kept on making new
4543 copies of the header line as it rewrote, and never released the old ones.)
4544 At the expense of a bit more processing, the header rewriting function has
4545 been changed so that it no longer eats memory in this way.
4547 66. The generation of the Received: header has been moved from the time that a
4548 message starts to be received, to the time that it finishes. The timestamp
4549 in the Received: header should now be very close to that of the <= log
4550 line. There are two side-effects of this change:
4552 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4553 logged header lines no longer include the local Received: line, because
4554 it has not yet been created. The same applies to a copy of the message
4555 that is returned to a non-SMTP sender when a message is rejected.
4557 (b) When a filter file is tested using -bf, no additional Received: header
4558 is added to the test message. After some thought, I decided that this
4561 This change does not affect the value of $received_for. It is still set
4562 after address rewriting, but before local_scan() is called.
4564 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4566 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4567 gave an unhelpful panic error message, and a defer error. I have managed to
4568 change this behaviour so that it now rejects any supplied certificate,
4569 which seems right, as the list of acceptable certificates is empty.
4571 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4572 gave an unhelpful defer error. I have not managed to make this reject any
4573 supplied certificates, but the error message it gives is "no certificate
4574 supplied", which is not helpful.
4576 70. exigrep's output now also includes lines that are not associated with any
4577 message, but which match the given pattern. Implemented by a patch from
4578 Martin Sluka, which also tidied up the Perl a bit.
4580 71. Recipient callout verification, like sender verification, was using <> in
4581 the MAIL FROM command. This isn't really the right thing, since the actual
4582 sender may affect whether the remote host accepts the recipient or not. I
4583 have changed it to use the actual sender in the callout; this means that
4584 the cache record is now keyed on a recipient/sender pair, not just the
4585 recipient address. There doesn't seem to be a real danger of callout loops,
4586 since a callout by the remote host to check the sender would use <>.
4587 [SEE ABOVE: changed after hitting problems.]
4589 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4590 temporary errors. However, in the case of such a code being given after
4591 the end of a data transmission (i.e. after ".") Exim was failing to write
4592 a retry record for the message. (Yes, there was some broken host that was
4593 actually sending 8xx at this point.)
4595 73. An unknown lookup type in a host list could cause Exim to panic-die when
4596 the list was checked. (An example that provoked this was putting <; in the
4597 middle of a list instead of at the start.) If this happened during a DATA
4598 ACL check, a -D file could be left lying around. This kind of configuration
4599 error no longer causes Exim to die; instead it causes a defer errror. The
4600 incident is still logged to the main and panic logs.
4602 74. Buglet left over from Exim 3 conversion. The message "too many messages
4603 in one connection" was written to the rejectlog but not the mainlog, except
4604 when address rewriting (yes!) was being logged.
4606 75. Added write_rejectlog option.
4608 76. When a system filter was run not as root (that is, when system_filter_user
4609 was set), the values of the $n variables were not being returned to the
4610 main process; thus, they were not subsequently available in the $sn
4613 77. Added +return_path_on_delivery log selector.
4615 78. A connection timeout was being treated differently from recipients deferred
4616 when testing hosts_max_try with a message that was older than the host's
4617 retry timeout. (The host should not be counted, thus allowing all hosts to
4618 be tried at least once before bouncing.) This may have been the cause of an
4619 occasionally reported bug whereby a message would remain on the queue
4620 longer than the retry timeout, but would be bounced if a delivery was
4621 forced. I say "may" because I never totally pinned down the problem;
4622 setting up timeout/retry tests is difficult. See also the next item.
4624 79. The ultimate address timeout was not being applied to errors that involved
4625 a combination of host plus message (for example, a timeout on a MAIL
4626 command). When an address resolved to a number of possible hosts, and they
4627 were not all tried for each delivery (e.g. because of hosts_max_try), a
4628 message could remain on the queue longer than the retry timeout.
4630 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4633 81. Fixed an obscure SMTP outgoing bug which required at least the following
4634 conditions: (a) there was another message waiting for the same server;
4635 (b) the server returned 5xx to all RCPT commands in the first message so
4636 that the message was not completed; (c) the server dropped the connection
4637 or gave a negative response to the RSET that Exim sends to abort the
4638 transaction. The observed case was a dropped connection after DATA that had
4639 been sent in pipelining mode. That is, the server had advertised PIPELINING
4640 but was not implementing it correctly. The effect of the bug was incorrect
4641 behaviour, such as trying another host, and this could lead to a crash.
4647 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4648 and daemon.c were passed as pointers to ints; they should have been
4649 pointers to socklen_t variables (which are typically unsigned ints).
4651 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4654 3. Fixed a really odd bug that affected only the testing scheme; patching a
4655 certain fixed string in the binary changed the value of another string that
4656 happened to be identical to the end of the original first string.
4658 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4659 name", it returns that address as the IP address. On some operating
4660 systems (e.g. Solaris), it also passes back the IP address string as the
4661 "host name". However, on others (e.g. Linux), it passes back an empty
4662 string. Exim wasn't checking for this, and was changing the host name to an
4663 empty string, assuming it had been canonicized.
4665 5. Although rare, it is permitted to have more than one PTR record for a given
4666 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4667 all the names associated with an address, because they do in Solaris.
4668 However, it seems that they do not in Linux for data that comes from the
4669 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4670 I found this out when I moved to a new Linux workstation and tried to run
4671 the Exim test suite.
4673 To get round this problem I have changed the code so that it now does its
4674 own call to the DNS to look up PTR records when searching for a host name.
4675 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4676 addresses that are only in /etc/hosts are still found.
4678 This behaviour is, however, controlled by an option called host_lookup_
4679 order, which defaults to "bydns:byaddr". If people want to use the other
4680 order, or indeed, just use one or the other means of lookup, they can
4681 specify it in this variable.
4683 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4684 some operating systems, this comes back from gethostbyaddr() as an empty
4685 string, and this is what Exim used to test for. However, it seems that in
4686 other systems, "." is yielded. Exim now tests for this case too.
4688 7. The values of check_spool_space and check_log_space are now held internally
4689 as a number of kilobytes instead of an absolute number of bytes. If a
4690 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4691 kilobyte. This means that much larger values can be stored.
4693 8. Exim monitor: an attempt to get the action menu when not actually pointing
4694 at a message produces an empty menu entitled "No message selected". This
4695 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4696 no entries in it ("Shell widget menu has zero width and/or height"). So I
4697 have added a single, blank menu entry in this case.
4699 9. Added ${quote_local_part.
4701 10. MIME decoding is now applied to the contents of Subject: header lines when
4704 11. Now that a reference to $sender_host_address automatically causes a reverse
4705 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4706 host lookup before query-style lookups in lists that might use this
4707 variable. This has therefore been abolished, and the "net-" prefix is no
4708 longer necessary for query-style lookups.
4710 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4711 have been a typo for LFLAGS, so it has been changed.
4713 13. The install script calls Exim with "-C /dev/null" in order to find the
4714 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4715 to be output. Howeve, since Exim outputs its version number before the
4716 error, it didn't break the script. It just looked ugly. I fixed this by
4717 always allowing "-C /dev/null" if the caller is root.
4719 14. Ignore overlarge ACL variable number when reading spool file - insurance
4720 against a later release with more variables having written the file.
4722 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4723 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4724 this, as well as the form without the "IPv6" on the front.
4726 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4727 OS/Makefile-Darwin file.
4729 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4730 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4732 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4733 variable. However, some OS use ints for inodes. I've added cast to long int
4734 to get rid of the compiler warning.
4736 19. I had forgotten to lock out "/../" in configuration file names when
4737 ALT_CONFIG_PREFIX was set.
4739 20. Routers used for verification do not need to specify transports. However,
4740 if such a router generated a host list, and callout was configured, Exim
4741 crashed, because it could not find a port number from the (non-existent)
4742 transport. It now assumes port 25 in this circumstance.
4744 21. Added the -t option to exigrep.
4746 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4747 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4748 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4749 them is set, LOOKUP_LSEARCH is forced.
4751 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4753 24. Added sender and host information to the "rejected by local_scan()" log
4754 line; previously there was no indication of these.
4756 25. Added .include_if_exists.
4758 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4759 Linux. It turns out that not all file systems support this. Apparently some
4760 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4761 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4762 not supported on the file descriptor, is now ignored when Exim is trying to
4763 sync a directory. This applies only to Linux.
4765 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4767 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4768 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4771 29. Added support for maildirsize files from supplied patch (modified a bit).
4773 30. The use of :fail: followed by an empty string could lead Exim to respond to
4774 sender verification failures with (e.g.):
4776 550 Verification failed for <xxx>
4777 550 Sender verify failed
4779 where the first response line was missing the '-' that indicates it is not
4780 the final line of the response.
4782 31. The loop for finding the name of the user that called Exim had a hardwired
4783 limit of 10; it now uses the value of finduser_retries, which is used for
4784 all other user lookups.
4786 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4789 33. Exim was neglecting to zero errno before one call of strtol() when
4790 expanding a string and expecting an integer value. On some systems this
4791 resulted in spurious "integer overflow" errors. Also, it was casting the
4792 result into an int without checking.
4794 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4795 did not work. The code looks as if it has *never* worked, though it appears
4796 to have been documented since at least releast 1.62. I have made it work.
4798 35. The "timeout_DNS" error in retry rules, also documented since at least
4799 1.62, also never worked. As it isn't clear exactly what this means, and
4800 clearly it isn't a major issue, I have abolished the feature by treating it
4801 as "timeout", and writing a warning to the main and panic logs.
4803 36. The display of retry rules for -brt wasn't always showing the error code
4806 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4807 timeout_connect_A, timeout_connect_MX.
4809 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4810 to the empty sender.
4812 39. The daemon was not analysing the content of -oX till after it had closed
4813 stderr and disconnected from the controlling terminal. This meant that any
4814 syntax errors were only noted on the panic log, and the return code from
4815 the command was 0. By re-arranging the code a little, I've made the
4816 decoding happen first, so such errors now appear on stderr, and the return
4817 code is 1. However, the actual setting up of the sockets still happens in
4818 the disconnected process, so errors there are still only recorded on the
4821 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4822 connections (as happens on some IP stacks) was logged at start up time as
4823 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4824 it from "IPv6 and IPv4", which means that two separate sockets are being
4827 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4828 says whether AF_INET or AF_INET6 was passed as an argument.
4830 42. Exiwhat output was messed up when time zones were included in log
4833 43. Exiwhat now gives more information about the daemon's listening ports,
4834 and whether -tls-on-connect was used.
4836 44. The "port" option of the smtp transport is now expanded.
4838 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4839 silently ignored. Now an error message is written to the main and panic
4842 46. There's a new ACL modifier called "logwrite" which writes to a log file
4843 as soon as it is encountered.
4845 47. Added $local_user_uid and $local_user_gid at routing time.
4847 48. Exim crashed when trying to verify a sender address that was being
4850 49. Exim was recognizing only a space character after ".include". It now also
4851 recognizes a tab character.
4853 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4854 extracting the relevant information from the specification. The man page no
4855 longer contains scrambled data for the -d option, and I've added a section
4856 at the front about calling Exim under different names.
4858 51. Added "extra_headers" argument to the "mail" command in filter files.
4860 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4863 53. Installed eximstats 1.29.
4865 54. Added transport_filter_timeout as a generic transport option.
4867 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4868 Cc: header lines. This was required by RFC 822, but it not required by RFC
4871 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4872 incoming messages that did not have them. Now it does so only if the
4873 message originates locally, that is, if there is no associated remote host
4874 address. When Resent- header lines are present, this applies to the Resent-
4875 lines rather than the non-Resent- lines.
4877 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4878 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4880 58. Messages for configuration errors now include the name of the main
4881 configuration file - useful now that there may be more than one file in a
4882 list (.included file names were always shown).
4884 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4885 for those rare installations that do not start the daemon as root or run it
4886 setuid root. I've cut out the call to initgroups() if the daemon is not
4889 60. The Exim user and group can now be bound into the binary as text strings
4890 that are looked up at the start of Exim's processing.
4892 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4894 62. Added $mailstore_basename variable.
4896 63. Installed patch to sieve.c from Michael Haardt.
4898 64. When Exim failed to open the panic log after failing to open the main log,
4899 the original message it was trying to log was written to stderr and debug
4900 output, but if they were not available (the usual case in production), it
4901 was lost. Now it is written to syslog before the two lines that record the
4902 failures to open the logs.
4904 65. Users' Exim filters run in subprocesses under the user's uid. It is
4905 possible for a "deliver" command or an alias in a "personal" command to
4906 provoke an address rewrite. If logging of address rewriting is configured,
4907 this fails because the process is not running as root or exim. There may be
4908 a better way of dealing with this, but for the moment (because 4.30 needs
4909 to be released), I have disabled address rewrite logging when running a
4910 filter in a non-root, non-exim process.
4916 1. The buildconfig auxiliary program wasn't quoting the value set for
4917 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4918 not defined. This bug was masked in 4.22 by the effect that was fixed in
4921 2. Some messages that were rejected after a message id was allocated were
4922 shown as "incomplete" by exigrep. It no longer does this for messages that
4923 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4925 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4926 have allow_domain_literals set, the ID did not get logged in the <= line.
4927 Domain literals are now always recognized in Message-ID: header lines.
4929 4. The first argument for a ${extract expansion item is the key name or field
4930 number. Leading and trailing spaces in this item were not being ignored,
4931 causing some misleading effects.
4933 5. When deliver_drop_privilege was set, single queue runner processes started
4934 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4935 same command in the process it spins off) were not dropping privilege.
4937 6. When the daemon running as "exim" started a queue runner, it always
4938 re-executed Exim in the spun-off process. This is a waste of effort when
4939 deliver_drop_privilege is set. The new process now just calls the
4940 queue-runner function directly.
4946 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4949 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4950 ignored. Though the use of -oP was forcing the writing of a pid file, it
4951 was always written to the default place.
4953 3. If the message "no IP address found for host xxxx" is generated during
4954 incoming verification, it is now followed by identification of the incoming
4955 connection (so you can more easily find what provoked it).
4957 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4959 5. Added some features to "harden" Exim a bit more against certain attacks:
4961 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4962 be put in Local/Makefile. This is like the never_users runtime option,
4963 but it cannot be overridden. The default setting is "root".
4965 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4966 prefix string with which any file named in a -C command line option
4969 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4970 is retained for -C and -D only if the caller of Exim is root. Without
4971 it, the exim user may also use -C and -D and retain privilege.
4973 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4974 command line option is disabled.
4976 6. Macro names set by the -D option must start with an upper case letter, just
4977 like macro names defined in the configuration file.
4979 7. Added "dereference=" facility to LDAP.
4981 8. Two instances of the typo "uknown" in the source files are fixed.
4983 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4984 the Configure-Makefile script screwed up while processing it.
4986 10. Incorporated PCRE 4.4.
4988 11. The SMTP synchronization check was not operating right at the start of an
4989 SMTP session. For example, it could not catch a HELO sent before the client
4990 waited for the greeting. There is now a check for outstanding input at the
4991 point when the greeting is written. Because of the duplex, asynchronous
4992 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4993 way, but not yet received, when the check is performed.
4995 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4996 on TCP/IP sockets, because this apparently causes some broken clients to
4999 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5000 unchanged) from the Cygwin maintainer.
5002 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5003 was supported instead of testing for mbx. Effectively a typo.
5005 15. The spa authenticator server code was not checking that the input it
5006 received was valid base64.
5008 16. The debug output line for the "set" modifier in ACLs was not showing the
5009 name of the variable that was being set.
5011 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5013 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5014 lookup on its own. Something else was needed to trigger the lookup. For
5015 example, a match in host_lookup or the need for a host name in a host list.
5016 Now, if $sender_host_name is referenced and the host name has not yet been
5017 looked up, a lookup is performed. If the lookup fails, the variable remains
5018 empty, and $host_lookup_failed is set to "1".
5020 19. Added "eqi" as a case-independent comparison operator.
5022 20. The saslauthd authentication condition could segfault if neither service
5023 nor realm was specified.
5025 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5026 error message that was logged was misleading, and incoming SMTP
5027 connections were dropped. The message is now more accurate, and temporary
5028 errors are given to SMTP connections.
5030 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5031 (except RSET) with a 421 error, until QUIT is received. However, it was
5032 failing to send a response to QUIT.
5034 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5035 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5036 ACL is now run after the helo_try_verify_hosts code.
5038 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5039 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5040 case-independent, so other case variants are also recognized). Apparently
5041 some systems use these upper case variants.
5043 25. If more than two messages were waiting for the same host, and a transport
5044 filter was specified for the transport, Exim sent two messages over the
5045 same TCP/IP connection, and then failed with "socket operation on non-
5046 socket" when it tried to send the third.
5048 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5050 27. The extern definition of crypt16() in expand.c was not being excluded when
5051 the OS had its own crypt16() function.
5053 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5054 as a preferred synonym for return_size_limit, both as an option and as an
5057 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5059 30. Changed the default configuration ACL to relax the local part checking rule
5060 for addresses that are not in any local domains. For these addresses,
5061 slashes and pipe symbols are allowed within local parts, but the sequence
5062 /../ is explicitly forbidden.
5064 31. SPA server authentication was not clearing the challenge buffer before
5067 32. log_message in a "warn" ACL statement was writing to the reject log as
5068 well as to the main log, which contradicts the documentation and doesn't
5069 seem right (because no rejection is happening). So I have stopped it.
5071 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5072 However, I am unable to do any testing of this.
5074 34. Fixed an infelicity in the appendfile transport. When checking directories
5075 for a mailbox, to see if any needed to be created, it was accidentally
5076 using path names with one or more superfluous leading slashes; tracing
5077 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5079 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5080 contents are added to the log line that is written for every discarded
5081 recipient. (Previously a log_message setting was ignored.)
5083 36. The ${quote: operator now quotes the string if it is empty.
5085 37. The install script runs exim in order to find its version number. If for
5086 some reason other than non-existence or emptiness, which it checks, it
5087 could not run './exim', it was installing it with an empty version number,
5088 i.e. as "exim-". This error state is now caught, and the installation is
5091 38. An argument was missing from the function that creates an error message
5092 when Exim fails to connect to the socket for saslauthd authentication.
5093 This could cause Exim to crash, or give a corrupted message.
5095 39. Added isip, isip4, and isip6 to ${if conditions.
5097 40. The ACL variables $acl_xx are now saved with the message, and can be
5098 accessed later in routers, transports, and filters.
5100 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5101 strings in the file are not string-expanded.
5103 42. If a MAIL command specified a SIZE value that was too large to fit into an
5104 int variable, the check against message_size_limit failed. Such values are
5105 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5106 day this will have to be increased, but I don't think I want to be around
5107 when emails are that large.
5114 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5115 iconv() is not standard in FreeBSD.
5117 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5118 IPv6 enabled. The observed symptom was a segmentation fault on return from
5119 the function os_common_find_running_interfaces() in src/os.c.
5121 3. In the check_special_case() function in daemon.c I had used "errno" as an
5122 argument name, which causes warnings on some systems. This was basically a
5123 typo, since it was named "eno" in the comments!
5125 4. The code that waits for the clock to tick (at a resolution of some fraction
5126 of a second) so as to ensure message-id uniqueness was always waiting for
5127 at least one whole tick, when it could have waited for less. [This is
5128 almost certainly not relevant at current processor speeds, where it is
5129 unlikely to ever wait at all. But we try to future-proof.]
5131 5. The function that sleeps for a time interval that includes fractions of a
5132 second contained a race. It did not block SIGALRM between setting the
5133 timer, and suspending (a couple of lines later). If the interval was short
5134 and the sigsuspend() was delayed until after it had expired, the suspension
5135 never ended. On busy systems this could lead to processes getting stuck for
5138 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5139 process, before it forks any delivery processes. The open lookup caching
5140 mechanism meant that the open file or database connection was passed into
5141 the delivery process. The problem was that delivery processes always tidy
5142 up cached lookup data. This could cause a problem for the next delivery
5143 process started by the queue runner, because the external queue runner
5144 process does not know about the closure. So the next delivery process
5145 still has data in the lookup cache. In the case of a file lookup, there was
5146 no problem because closing a file descriptor in a subprocess doesn't affect
5147 the parent. However, if the lookup was caching a connection to a database,
5148 the connection was closed, and the second delivery process was likely to
5149 see errors such as "PGSQL: query failed: server closed the connection
5150 unexpectedly". The problem has been fixed by closing all cached lookups
5151 in a queue runner before running a delivery process.
5153 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5154 seem to have the "const" qualifier which it has on other OS. I've
5157 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5158 *of the same type* (client or server) with the same public name that an
5159 error should be diagnosed.
5161 9. When Exim looked up a host name for an IP address, but failed to find the
5162 original IP address when looking up the host name (a safety check), it
5163 output the message "<ip address> does not match any IP for NULL", which was
5164 confusing, to say the least. The bug was that the host name should have
5165 appeared instead of "NULL".
5167 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5168 user that is built into the binary, and the -C or -D options is used, root
5169 privilege is dropped before the configuration file is read. In addition,
5170 logging is switched to stderr instead of the normal log files. If the
5171 configuration then re-defines the Exim user, the unprivileged environment
5172 is probably not what is expected, so Exim logs a panic warning message (but
5175 However, if deliver_drop_privilege is set, the unprivileged state may well
5176 be exactly what is intended, so the warning has been cut out in that case,
5177 and Exim is allowed to try to write to its normal log files.
5183 1. smtp_return_error_details was not giving details for temporary sender
5184 or receiver verification errors.
5186 2. Diagnose a configuration error if two authenticators have the same public
5189 3. Exim used not to create the message log file for a message until the first
5190 delivery attempt. This could be confusing when incoming messages were held
5191 for policy or load reasons. The message log file is now created at the time
5192 the message is received, and an initial "Received" line is written to it.
5194 4. The automatically generated man page for command line options had a minor
5195 bug that caused no ill effects; however, a more serious problem was that
5196 the procedure for building the man page automatically didn't always
5197 operate. Consequently, release 4.20 contains an out-of-date version. This
5198 shouldn't happen again.
5200 5. When building Exim with embedded Perl support, the script that builds the
5201 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5202 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5204 6. The freeze_tell option was not being used for messages that were frozen on
5205 arrival, either by an ACL or by local_scan().
5207 7. Added the smtp_incomplete_transaction log selector.
5209 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5210 was accepting AUTH without a new EHLO.
5212 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5213 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5216 10. Logging of TCP/IP connections (when configured) now happens in the main
5217 daemon process instead of the child process, so that the TCP/IP connection
5218 count is more accurate (but it can never be perfect).
5220 11. The use of "drop" in a nested ACL was not being handled correctly in the
5221 outer ACL. Now, if condition failure induced by the nested "drop" causes
5222 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5223 or "require"), the connection is dropped.
5225 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5226 that yield "discard" can now be used with an "accept" or a "discard" verb,
5227 but an error is generated for any others (because I can't see a useful way
5228 to define what should happen).
5230 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5231 are now processed in the same way as lines in the Exim configuration file.
5232 In particular, continuation lines are supported.
5234 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5236 15. Added -ti meaning -t -i.
5238 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5239 domains, and warn by logging if others are found.
5241 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5242 returned from GIFCONF when Exim is trying to find the list of interfaces on
5243 a host. The code in os.c has been modified to copy each ifreq to an aligned
5244 structure in all cases.
5246 Also, in some cases, the returned ifreq's were being copied to a 'struct
5247 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5248 means the last couple of bytes of an IPv6 address could be chopped if the
5249 ifreq contained only a normal sockaddr (14 bytes storage).
5251 18. Named domain lists were not supported in the hosts_treat_as_local option.
5252 An entry such as +xxxx was not recognized, and was treated as a literal
5255 19. Ensure that header lines added by a DATA ACL are included in the reject log
5256 if the ACL subsequently rejects the message.
5258 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5259 MD5 (which is deprecated).
5261 21. When testing a filter file using -bf, Exim was writing a message when it
5262 took the sender from a "From " line in the message, but it was not doing so
5263 when it took $return_path from a Return-Path: header line. It now does.
5265 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5266 with a valid header line field name (a series of printing characters
5267 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5269 23. Changed "disc" in the source to "disk" to conform to the documentation and
5270 the book and for uniformity.
5272 24. Ignore Sendmail's -Ooption=value command line item.
5274 25. When execve() failed while trying to run a command in a pipe transport,
5275 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5276 could be confused with a return value of 69 from the command itself. This
5277 has been changed to 127, the value the shell returns if it is asked to run
5278 a non-existent command. The wording for the related log line suggests a
5279 non-existent command as the problem.
5281 26. If received_header_text expands to an empty string, do not add a Received:
5282 header line to the message. (Well, it adds a token one on the spool, but
5283 marks it "old" so that it doesn't get used or transmitted.)
5285 27. Installed eximstats 1.28 (addition of -nt option).
5287 28. There was no check for failure on the call to getsockname() in the daemon
5288 code. This can fail if there is a shortage of resources on the system, with
5289 ENOMEM, for example. A temporary error is now given on failure.
5291 29. Contrary to the C standard, it seems that in some environments, the
5292 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5293 program. Exim now does this explicitly; it affects the formatting of
5294 timestamps using strftime().
5296 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5297 complaints. I've now initialized all the variables, to avoid this.
5299 32. Header lines added by a system filter were not being "seen" during
5300 transport-time rewrites.
5302 33. The info_callback() function passed to OpenSSL is set up with type void
5303 (*)(SSL *, int, int), as described somewhere. However, when calling the
5304 function (actually a macro) that sets it up, the type void(*)() is
5305 expected. I've put in a cast to prevent warnings from picky compilers.
5307 34. If a DNS black list lookup found a CNAME record, but there were no A
5308 records associated with the domain it pointed at, Exim crashed.
5310 35. If a DNS black list lookup returned more than one A record, Exim ignored
5311 all but the first. It now scans all returned addresses if a particular IP
5312 value is being sought. In this situation, the contents of the
5313 $dnslist_value variable are a list of all the addresses, separated by a
5316 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5317 to accept a host name and all its aliases if the forward lookup for any of
5318 them yielded the IP address of the incoming connection. Now it accepts only
5319 those names whose forward lookup yields the correct IP address. Any other
5320 names are discarded. This closes a loophole whereby a rogue DNS
5321 administrator could create reverse DNS records to break through a
5322 wildcarded host restriction in an ACL.
5324 37. If a user filter or a system filter that ran in a subprocess used any of
5325 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5326 the wrong values were passed to the pipe command ($thisaddress had the
5327 value of $0, $0 had the value of $1, etc). This bug was introduced by
5328 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5330 38. Improved the line breaking for long SMTP error messages from ACLs.
5331 Previously, if there was no break point between 40 and 75 characters, Exim
5332 left the rest of the message alone. Two changes have been made: (a) I've
5333 reduced the minimum length to 35 characters; (b) if it can't find a break
5334 point between 35 and 75 characters, it looks ahead and uses the first one
5335 that it finds. This may give the occasional overlong line, but at least the
5336 remaining text gets split now.
5338 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5339 file descriptors might be low, and that setting 1000 would always raise it.
5340 It turns out that in some environments, the limit is already over 1000 and
5341 that lowering it causes trouble. So now Exim takes care not to decrease it.
5343 40. When delivering a message, the value of $return_path is set to $sender_
5344 address at the start of routing (routers may change the value). By an
5345 oversight, this default was not being set up when an address was tested by
5346 -bt or -bv, which affected the outcome if any router or filter referred to
5349 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5350 log only when "message" or "log_message" are set. However, if one of the
5351 conditions was an address verification, or a call to a nested ACL, the
5352 messages generated by the underlying test were being passed through. This
5353 no longer happens. The underlying message is available in $acl_verify_
5354 message for both "message" and "log_message" expansions, so it can be
5355 passed through if needed.
5357 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5358 new expansion $bh_ to give the encoded byte string without charset
5359 translation. Translation happens only if iconv() is available; HAVE_ICONV
5360 indicates this at build time. HEADERS_CHARSET gives the charset to
5361 translate to; headers_charset can change it in the configuration, and
5362 "headers charset" can change it in an individual filter file.
5364 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5365 that creates RFC 2047 encoded "words" labels them as that charset instead
5366 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5367 expansion operator; (ii) when Exim creates a From: line for a local
5368 message; (iii) when a header line is rewritten to include a "phrase" part.
5370 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5371 buggy, causing it to skip the first lines of messages whose message ID
5372 ended in 'D'. This would not have bitten before Exim release 4.14, because
5373 message IDs were unlikely to end in 'D' before then. The effect was to have
5374 incorrect size information for certain domains.
5376 45. #include "config.h" was missing at the start of the crypt16.c module. This
5377 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5380 46. If there was a timeout during a "random" callout check, Exim treated it as
5381 a failure of the random address, and carried on sending RSET and the real
5382 address. If the delay was just some slowness somewhere, the response to the
5383 original RCPT would be taken as a response to RSET and so on, causing
5384 mayhem of various kinds.
5386 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5387 when I implemented it. It didn't allow for the fact that some option values
5388 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5389 the right test for positive values.
5391 48. Domain names in DNS records are case-independent. Exim always looks them up
5392 in lower case. Some resolvers return domain names in exactly the case they
5393 appear in the zone file, that is, they may contain uppercase letters. Not
5394 all resolvers do this - some return always lower case. Exim was treating a
5395 change of case by a resolver as a change of domain, similar to a widening
5396 of a domain abbreviation. This triggered its re-routing code and so it was
5397 trying to route what was effectively the same domain again. This normally
5398 caused routing to fail (because the router wouldn't handle the domain
5399 twice). Now Exim checks for this case specially, and just changes the
5400 casing of the domain that it ultimately uses when it transmits the message
5403 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5406 50. If a filter generated a file delivery with a non-absolute name (possible if
5407 no home directory exists for the router), the forbid_file option was not
5410 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5411 the existing equality matching.
5413 52. Exim was using ints instead of ino_t variables in some places where it was
5414 dealing with inode numbers.
5416 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5417 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5418 called TMPDIR, and if it finds it is different, it changes its value.
5420 54. The smtp_printf() function is now made available to local_scan() so
5421 additional output lines can be written before returning. There is also an
5422 smtp_fflush() function to enable the detection of a dropped connection.
5423 The variables smtp_input and smtp_batched_input are exported to
5426 55. Changed the default runtime configuration: the message "Unknown user"
5427 has been removed from the ACL, and instead placed on the localuser router,
5428 using the cannot_route_message feature. This means that any verification
5429 failures that generate their own messages won't get overridden. Similarly,
5430 the "Unrouteable address" message that was in the ACL for unverifiable
5431 relay addresses has also been removed.
5433 56. Added hosts_avoid_esmtp to the smtp transport.
5435 57. The exicyclog script was not checking for the esoteric option
5436 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5437 will work only if exicyclog is run under the appropriate euid.
5439 58. Following a discussion on the list, the rules by which Exim recognises line
5440 endings on incoming messages have been changed. The -dropcr and drop_cr
5441 options are now no-ops, retained only for backwards compatibility. The
5442 following line terminators are recognized: LF CRLF CR. However, special
5443 processing applies to CR:
5445 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5446 nor a local message in the state where . is a terminator.
5448 (ii) If a bare CR is encountered in a header line, an extra space is added
5449 after the line terminator so as not to end the header. The reasoning
5450 behind this is that bare CRs in header lines are most likely either
5451 to be mistakes, or people trying to play silly games.
5453 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5454 was being incorrectly given as 18 bytes larger than it should have been.
5455 This is a VOB (very old bug).
5457 60. This may never have affected anything current, but just in case it has:
5458 When the local host is found other than at the start of a list of hosts,
5459 the local host, those with the same MX, and any that follow, are discarded.
5460 When the list in question was part of a longer list of hosts, the following
5461 hosts (not currently being processed) were also being discarded. This no
5462 longer happens. I'm not sure if this situation could ever has previously
5465 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5468 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5469 header. This is something that is recommended in a new Internet Draft, and
5470 is something that is documented as being done by Sendmail. There are two
5471 possible values. For messages generated by the autoreply transport, Exim
5474 Auto-Submitted: auto-replied
5476 whereas for all other generated messages (e.g. bounces) it adds
5478 Auto-Submitted: auto-generated
5480 63. The "personal" condition in filters now includes a test for the
5481 Auto-Submitted: header. If it contains the string "auto-" the message it
5482 not considered personal.
5484 64. Added rcpt_include_affixes as a generic transport option.
5486 65. Added queue_only_override (default true).
5488 66. Added the syslog_duplication option.
5490 67. If what should have been the first header line of a message consisted of
5491 a space followed by a colon, Exim was mis-interpreting it as a header line.
5492 It isn't of course - it is syntactically invalid and should therefore be
5493 treated as the start of the message body. The misbehaviour could have
5494 caused a number of strange effects, including loss of data in subsequent
5495 header lines, and spool format errors.
5497 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5498 client host had authenticated. This control can now be exercised by an ACL
5499 for more flexibility.
5501 69. By default, callouts do not happen when testing with -bh. There is now a
5502 variant, -bhc, which does actually run the callout code, including
5503 consulting and updating the callout cache.
5505 70. Added support for saslauthd authentication, courtesy of Alexander
5508 71. If statvfs() failed on the spool or log directories while checking their
5509 size for availability, Exim confusingly gave the error "space shortage".
5510 Furthermore, in debugging mode it crashed with a floating point exception.
5511 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5512 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5513 serious problem, Exim now writes to the main and panic logs when this
5514 happens, with details of the failure. It then refuses to accept the
5515 incoming message, giving the message "spool directory problem" or "log
5516 directory problem" with a 421 code for SMTP messages.
5518 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5519 0, 1, and 2 exist, because some OS complain for execs without them (see
5520 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5521 descriptors. However, the code omitted to check that the open succeeded,
5522 causing mysterious errors if for some reason the permissions on /dev/null
5523 got screwed. Now Exim writes a message to the main and panic logs, and
5524 bombs out if it can't open /dev/null.
5526 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5527 interact so that it is all more flexible. It is supposed to remain
5528 backwards compatible. Also added extra_local_interfaces.
5530 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5531 to bomb out with an assertion failure - to the client this appears as a
5532 connection drop. This problem occurs in the part of the code that was taken
5533 from the Samba project. Fortunately, the assertion is in a very simple
5534 function, so I have fixed this by reproducing the function inline in the
5535 one place where it is called, and arranging for authentication to fail
5536 instead of killing the process with assert().
5538 75. The SPA client code was not working when the server requested OEM rather
5539 than Unicode encoding.
5541 76. Added code to make require_files with a specific uid setting more usable in
5542 the case where statting the file as root fails - usually a non-root-mounted
5543 NFS file system. When this happens and the failure is EACCES, Exim now
5544 forks a subprocess and does the per-uid checking as the relevant uid.
5546 77. Added process_log_path.
5548 78. If log_file_path was not explicitly set, a setting of check_log_space or
5549 check_log_inodes was ignored.
5551 79. If a space check for the spool or log partitions fails, the incident is now
5552 logged. Of course, in the latter case the data may get lost...
5554 80. Added the %p formatting code to string_format() so that it can be used to
5555 print addresses in debug_print(). Adjusted all the address printing in the
5556 debugging in store.c to use %p rather than %d.
5558 81. There was a concern that a line of code in smtp_in.c could overflow a
5559 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5560 initially expressed, the concern was not well-founded, because trailing
5561 spaces are removed early. However, if the trailing spaces were followed by
5562 a NULL, they did not get removed, so the overflow was possible. Two fixes
5565 (a) I re-wrote the offending code in a cleaner fashion.
5566 (b) If an incoming SMTP command contains a NULL character, it is rejected
5569 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5570 runs initgroups(), so that if the Exim user is in any additional groups,
5571 they will be used during message reception.
5577 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob