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 Disable SSLv2 by default in OpenSSL support.
95 PP/01 Solaris build fix for Oracle's LDAP libraries.
96 Bugzilla 1109, patch from Stephen Usher.
98 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
100 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
103 TF/02 Fix a couple more cases where we did not log the error message
104 when unlink() failed. See also change 4.74-TF/03.
106 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
107 lock up or crash if it happened to be inside a call to libc when it
108 got a SIGUSR1 from exiwhat.
110 The SIGUSR1 handler appends the current process status to the process
111 log which is later printed by exiwhat. It used to use the general
112 purpose logging code to do this, but several functions it calls are
113 not safe for signals.
115 The new output code in the SIGUSR1 handler is specific to the process
116 log, and simple enough that it's easy to inspect for signal safety.
117 Removing some special cases also simplifies the general logging code.
118 Removing the spurious timestamps from the process log simplifies
121 TF/04 Improved ratelimit ACL condition.
123 The /noupdate option has been deprecated in favour of /readonly which
124 has clearer semantics. The /leaky, /strict, and /readonly update modes
125 are mutually exclusive. The update mode is no longer included in the
126 database key; it just determines when the database is updated. (This
127 means that when you upgrde Exim will forget old rate measurements.)
129 Exim now checks that the per_* options are used with an update mode that
130 makes sense for the current ACL. For example, when Exim is processing a
131 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
132 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
133 must specify per_mail/readonly. If you omit the update mode it defaults to
134 /leaky where that makes sense (as before) or /readonly where required.
136 The /noupdate option is now undocumented but still supported for
137 backwards compatibility. It is equivalent to /readonly except that in
138 ACLs where /readonly is required you may specify /leaky/noupdate or
139 /strict/noupdate which are treated the same as /readonly.
141 A useful new feature is the /count= option. This is a generalization
142 of the per_byte option, so that you can measure the throughput of other
143 aggregate values. For example, the per_byte option is now equivalent
144 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
146 The per_rcpt option has been generalized using the /count= mechanism
147 (though it's more complicated than the per_byte equivalence). When it is
148 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
149 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
150 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
151 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
152 Note that using per_rcpt with a non-readonly update mode in more than
153 one ACL will cause the recipients to be double-counted. (The per_mail
154 and per_byte options don't have this problem.)
156 The handling of very low rates has changed slightly. If the computed rate
157 is less than the event's count (usually one) then this event is the first
158 after a long gap. In this case the rate is set to the same as this event's
159 count, so that the first message of a spam run is counted properly.
161 The major new feature is a mechanism for counting the rate of unique
162 events. The new per_addr option counts the number of different
163 recipients that someone has sent messages to in the last time period. It
164 behaves like per_rcpt if all the recipient addresses are different, but
165 duplicate recipient addresses do not increase the measured rate. Like
166 the /count= option this is a general mechanism, so the per_addr option
167 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
168 example, measure the rate that a client uses different sender addresses
169 with the options per_mail/unique=$sender_address. There are further
170 details in the main documentation.
172 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
174 TF/06 Removed a few PCRE remnants.
176 TF/07 Automatically extract Exim's version number from tags in the git
177 repository when doing development or release builds.
179 PP/02 Raise smtp_cmd_buffer_size to 16kB.
180 Bugzilla 879. Patch from Paul Fisher.
182 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
183 Heavily based on revision 40f9a89a from Simon Arlott's tree.
186 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
188 PP/05 Variable $av_failed, true if the AV scanner deferred.
189 Bugzilla 1078. Patch from John Horne.
191 PP/06 Stop make process more reliably on build failure.
192 Bugzilla 1087. Patch from Heiko Schlittermann.
194 PP/07 Make maildir_use_size_file an _expandable_ boolean.
195 Bugzilla 1089. Patch from Heiko Schlittermann.
197 PP/08 Handle ${run} returning more data than OS pipe buffer size.
198 Bugzilla 1131. Patch from Holger Weiß.
200 PP/09 Handle IPv6 addresses with SPF.
201 Bugzilla 860. Patch from Wolfgang Breyha.
203 PP/10 GnuTLS: support TLS 1.2 & 1.1.
205 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
208 PP/11 match_* no longer expand right-hand-side by default.
209 New compile-time build option, EXPAND_LISTMATCH_RHS.
210 New expansion conditions, "inlist", "inlisti".
212 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
214 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
216 PP/14 fix log_write() format string regression from TF/03.
217 Bugzilla 1152. Patch from Dmitry Isaikin.
223 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
225 PP/02 Harmonised TLS library version reporting; only show if debugging.
226 Layout now matches that introduced for other libraries in 4.74 PP/03.
228 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
230 PP/04 New "dns_use_edns0" global option.
232 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
235 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
236 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
238 TK/01 Updated PolarSSL code to 0.14.2.
239 Bugzilla 1097. Patch from Andreas Metzler.
241 PP/07 Catch divide-by-zero in ${eval:...}.
244 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
247 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
248 format-string attack -- SECURITY: remote arbitrary code execution.
250 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
251 time unintentionally subject to list matching rules, letting the header
252 cause arbitrary Exim lookups (of items which can occur in lists, *not*
253 arbitrary string expansion). This allowed for information disclosure.
255 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
256 INT_MIN/-1 -- value coerced to INT_MAX.
262 NM/01 Workround for PCRE version dependency in version reporting
265 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
266 This fixes portability to compilers other than gcc, notably
267 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
269 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
270 makefiles for portability to HP-UX and POSIX correctness.
272 PP/01 Permit LOOKUP_foo enabling on the make command-line.
273 Also via indented variable definition in the Makefile.
274 (Debugging by Oliver Heesakkers).
276 PP/02 Restore caching of spamd results with expanded spamd_address.
277 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
279 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
280 Improves build reliability. Fix from: Frank Elsner
282 NM/02 Fix wide character breakage in the rfc2047 coding
283 Fixes bug 1064. Patch from Andrey N. Oktyabrski
285 NM/03 Allow underscore in dnslist lookups
286 Fixes bug 1026. Patch from Graeme Fowler
288 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
289 Code patches from Adam Ciarcinski of NetBSD.
291 NM/04 Fixed exiqgrep to cope with mailq missing size issue
294 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
295 is logged, to avoid truncation. Patch from John Horne.
297 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
298 Patch from Jakob Hirsch.
300 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
301 SQL string expansion failure details.
302 Patch from Andrey Oktyabrski.
304 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
305 Patch from Simon Arlott.
307 PP/09 New lookups functionality failed to compile on old gcc which rejects
308 extern declarations in function scope.
309 Patch from Oliver Fleischmann
311 PP/10 Use sig_atomic_t for flags set from signal handlers.
312 Check getgroups() return and improve debugging.
313 Fixed developed for diagnosis in bug 927 (which turned out to be
316 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
317 Patch from Mark Zealey.
319 PP/12 Bugzilla 1056: Improved spamd server selection.
320 Patch from Mark Zealey.
322 PP/13 Bugzilla 1086: Deal with maildir quota file races.
323 Based on patch from Heiko Schlittermann.
325 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
326 Patch from Uwe Doering, sign-off by Michael Haardt.
328 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
329 variable declaration deep within a block. Bug and patch from
332 PP/15 lookups-Makefile IRIX compatibilty coercion.
334 PP/16 Make DISABLE_DKIM build knob functional.
336 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
337 Patch by Simon Arlott
339 TF/03 Fix valgrind.h portability to C89 compilers that do not support
340 variable argument macros. Our copy now differs from upstream.
346 TF/01 Failure to get a lock on a hints database can have serious
347 consequences so log it to the panic log.
349 TF/02 Log LMTP confirmation messages in the same way as SMTP,
350 controlled using the smtp_confirmation log selector.
352 TF/03 Include the error message when we fail to unlink a spool file.
354 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
355 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
356 for maintaining out-of-tree patches for some time.
358 PP/01 Bugzilla 139: Documentation and portability issues.
359 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
360 Handle per-OS dynamic-module compilation flags.
362 PP/02 Let /dev/null have normal permissions.
363 The 4.73 fixes were a little too stringent and complained about the
364 permissions on /dev/null. Exempt it from some checks.
365 Reported by Andreas M. Kirchwitz.
367 PP/03 Report version information for many libraries, including
368 Exim version information for dynamically loaded libraries. Created
369 version.h, now support a version extension string for distributors
370 who patch heavily. Dynamic module ABI change.
372 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
373 privilege escalation vulnerability whereby the Exim run-time user
374 can cause root to append content of the attacker's choosing to
377 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
380 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
381 If dropping privileges for untrusted macros, we disabled normal logging
382 on the basis that it would fail; for the Exim run-time user, this is not
383 the case, and it resulted in successful deliveries going unlogged.
384 Fixed. Reported by Andreas Metzler.
390 PP/01 Date: & Message-Id: revert to normally being appended to a message,
391 only prepend for the Resent-* case. Fixes regression introduced in
392 Exim 4.70 by NM/22 for Bugzilla 607.
394 PP/02 Include check_rfc2047_length in configure.default because we're seeing
395 increasing numbers of administrators be bitten by this.
397 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
399 PP/03 Bugzilla 994: added openssl_options main configuration option.
401 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
403 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
405 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
407 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
408 without a peer certificate, leading to a segfault because of an
409 assumption that peers always have certificates. Be a little more
410 paranoid. Problem reported by Martin Tscholak.
412 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
413 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
414 NB: ClamAV planning to remove STREAM in "middle of 2010".
415 CL also introduces -bmalware, various -d+acl logging additions and
416 more caution in buffer sizes.
418 PP/09 Implemented reverse_ip expansion operator.
420 PP/10 Bugzilla 937: provide a "debug" ACL control.
422 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
424 PP/12 Bugzilla 973: Implement --version.
426 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
428 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
430 PP/15 Bugzilla 816: support multiple condition rules on Routers.
432 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
433 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
434 ignore trailing whitespace.
436 JJ/02 prevent non-panic DKIM error from being sent to paniclog
438 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
441 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
442 Notification from Dr Andrew Aitchison.
444 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
445 ExtendedDetectionInfo response format.
446 Notification from John Horne.
448 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
451 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
452 XSL and documented dependency on system catalogs, with examples of how
455 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
458 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
459 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
460 configuration file which is writeable by the Exim user or group.
462 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
463 of configuration files to cover files specified with the -C option if
464 they are going to be used with root privileges, not just the default
467 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
468 option (effectively making it always true).
470 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
471 files to be used while preserving root privileges.
473 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
474 that rogue child processes cannot use them.
476 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
477 run-time user, instead of root.
479 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
480 Exim run-time user without dropping privileges.
482 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
483 result string, instead of calling string_vformat() twice with the same
486 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
487 for other users. Others should always drop root privileges if they use
488 -C on the command line, even for a whitelisted configure file.
490 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
492 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
498 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
499 $data_path, and $header_path variables; fixed documentation bugs and
502 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
503 exipick to access non-standard spools, including the "frozen" queue
506 NM/01 Bugzilla 965: Support mysql stored procedures.
507 Patch from Alain Williams
509 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
511 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
512 Patch from Andreas Metzler
514 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
515 Patch from Kirill Miazine
517 NM/05 Bugzilla 671: Added umask to procmail example.
519 JJ/03 installed exipick 20100323.0, fixing doc bug
521 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
522 directory. Notification and patch from Dan Rosenberg.
524 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
526 TK/02 Improve log output when DKIM signing operation fails.
528 MH/01 Treat the transport option dkim_domain as a colon separated
529 list, not as a single string, and sign the message with each element,
530 omitting multiple occurences of the same signer.
532 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
533 Bugzilla 985, 986. Patch by Simon Arlott
535 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
536 Patch by Simon Arlott
538 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
539 MBX locking. Notification from Dan Rosenberg.
545 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
547 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
549 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
551 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
553 NM/04 Bugzilla 915: Flag broken perl installation during build.
559 TK/01 Added patch by Johannes Berg that expands the main option
560 "spamd_address" if it starts with a dollar sign.
562 TK/02 Write list of recipients to X-Envelope-Sender header when building
563 the mbox-format spool file for content scanning (suggested by Jakob
566 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
567 (http://www.dcc-servers.net/) support via dccifd. Activated by
568 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
570 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
571 by Mark Daniel Reidel <mr@df.eu>.
573 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
574 When building exim an external PCRE library is now needed -
575 PCRE is a system library on the majority of modern systems.
576 See entry on PCRE_LIBS in EDITME file.
578 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
579 conversation. Added nologin parameter to request.
580 Patch contributed by Kirill Miazine.
582 TF/01 Do not log submission mode rewrites if they do not change the address.
584 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
586 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
587 log files in place. Contributed by Roberto Lima.
589 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
591 TF/03 Bugzilla 615: When checking the local_parts router precondition
592 after a local_part_suffix or local_part_prefix option, Exim now
593 does not use the address's named list lookup cache, since this
594 contains cached lookups for the whole local part.
596 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
597 Robert Millan. Documentation is in experimental-spec.txt.
599 TF/04 Bugzilla 668: Fix parallel build (make -j).
601 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
603 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
604 Patch provided by Jan Srzednicki.
606 TF/05 Leading white space used to be stripped from $spam_report which
607 wrecked the formatting. Now it is preserved.
609 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
610 that they are available at delivery time.
612 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
614 TF/08 TLS error reporting now respects the incoming_interface and
615 incoming_port log selectors.
617 TF/09 Produce a more useful error message if an SMTP transport's hosts
618 setting expands to an empty string.
620 NM/06 Bugzilla 744: EXPN did not work under TLS.
621 Patch provided by Phil Pennock.
623 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
624 Patch provided by Richard Godbee.
626 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
627 acl_smtp_notquit, added index entry.
629 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
630 Patch provided by Eugene Bujak.
632 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
633 accept(). Patch provided by Maxim Dounin.
635 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
636 Patch provided by Phil Pennock.
638 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
640 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
641 Patch provided by Brad "anomie" Jorsch.
643 NM/14 Bugzilla 622: Added timeout setting to transport filter.
644 Patch provided by Dean Brooks.
646 TK/05 Add native DKIM support (does not depend on external libraries).
648 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
649 Patch provided by Graeme Fowler.
651 NM/16 Bugzilla 851: Documentation example syntax fix.
653 NM/17 Changed NOTICE file to remove references to embedded PCRE.
655 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
658 NM/19 Bugzilla 745: TLS version reporting.
659 Patch provided by Phil Pennock.
661 NM/20 Bugzilla 167: bool: condition support.
662 Patch provided by Phil Pennock.
664 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
665 clients. Patch provided by Phil Pennock.
667 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
668 Patch provided by Brad "anomie" Jorsch.
670 NM/23 Bugzilla 687: Fix misparses in eximstats.
671 Patch provided by Heiko Schlittermann.
673 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
674 Patch provided by Heiko Schlittermann.
676 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
677 plus update to original patch.
679 NM/26 Bugzilla 799: Documentation correction for ratelimit.
681 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
682 Patch provided by David Brownlee.
684 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
686 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
688 NM/30 Bugzilla 888: TLS documentation bugfixes.
690 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
692 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
693 Unlike the original bugzilla I have changed all shell scripts in src tree.
695 NM/33 Bugzilla 898: Transport filter timeout fix.
696 Patch by Todd Rinaldo.
698 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
699 Patch by Serge Demonchaux.
701 NM/35 Bugzilla 39: Base64 decode bug fixes.
702 Patch by Jakob Hirsch.
704 NM/36 Bugzilla 909: Correct connect() call in dcc code.
706 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
708 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
710 NM/39 Bugzilla 911: Fixed MakeLinks build script.
716 TK/01 Add preliminary DKIM support. Currently requires a forked version of
717 ALT-N's libdkim that I have put here:
718 http://duncanthrax.net/exim-experimental/
720 Note to Michael Haardt: I had to rename some vars in sieve.c. They
721 were called 'true' and it seems that C99 defines that as a reserved
722 keyword to be used with 'bool' variable types. That means you could
723 not include C99-style headers which use bools without triggering
724 build errors in sieve.c.
726 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
727 as mailq or other aliases. Changed the --help handling significantly
728 to do whats expected. exim_usage() emits usage/help information.
730 SC/01 Added the -bylocaldomain option to eximstats.
732 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
734 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
736 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
742 PH/01 Another patch from the Sieve maintainer.
744 PH/02 When an IPv6 address is converted to a string for single-key lookup
745 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
746 dots are used instead of colons so that keys in lsearch files need not
747 contain colons. This was done some time before quoting was made available
748 in lsearch files. However, iplsearch files do require colons in IPv6 keys
749 (notated using the quote facility) so as to distinguish them from IPv4
750 keys. This meant that lookups for IP addresses in host lists did not work
751 for iplsearch lookups.
753 This has been fixed by arranging for IPv6 addresses to be expressed with
754 colons if the lookup type is iplsearch. This is not incompatible, because
755 previously such lookups could never work.
757 The situation is now rather anomolous, since one *can* have colons in
758 ordinary lsearch keys. However, making the change in all cases is
759 incompatible and would probably break a number of configurations.
761 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
764 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
765 conversion specification without a maximum field width, thereby enabling
766 a rogue spamd server to cause a buffer overflow. While nobody in their
767 right mind would setup Exim to query an untrusted spamd server, an
768 attacker that gains access to a server running spamd could potentially
769 exploit this vulnerability to run arbitrary code as the Exim user.
771 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
772 $primary_hostname instead of what libspf2 thinks the hosts name is.
774 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
775 a directory entry by the name of the lookup key. Previously, if a
776 symlink pointed to a non-existing file or a file in a directory that
777 Exim lacked permissions to read, a lookup for a key matching that
778 symlink would fail. Now it is enough that a matching directory entry
779 exists, symlink or not. (Bugzilla 503.)
781 PH/03 The body_linecount and body_zerocount variables are now exported in the
784 PH/04 Added the $dnslist_matched variable.
786 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
787 This means they are set thereafter only if the connection becomes
790 PH/06 Added the client_condition to authenticators so that some can be skipped
791 by clients under certain conditions.
793 PH/07 The error message for a badly-placed control=no_multiline_responses left
794 "_responses" off the end of the name.
796 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
798 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
799 (without spaces) instead of just copying the configuration text.
801 PH/10 Added the /noupdate option to the ratelimit ACL condition.
803 PH/11 Added $max_received_linelength.
805 PH/12 Added +ignore_defer and +include_defer to host lists.
807 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
808 way in which PCRE > 7.0 is built.
810 PH/14 Implemented queue_only_load_latch.
812 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
813 MAIL command. The effect was to mangle the value on 64-bit systems.
815 PH/16 Another patch from the Sieve maintainer.
817 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
819 PH/18 If a system quota error occurred while trying to create the file for
820 a maildir delivery, the message "Mailbox is full" was not appended to the
821 bounce if the delivery eventually timed out. Change 4.67/27 below applied
822 only to a quota excession during the actual writing of the file.
824 PH/19 It seems that peer DN values may contain newlines (and other non-printing
825 characters?) which causes problems in log lines. The DN values are now
826 passed through string_printing() before being added to log lines.
828 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
829 and InterBase are left for another time.)
831 PH/21 Added message_body_newlines option.
833 PH/22 Guard against possible overflow in moan_check_errorcopy().
835 PH/23 POSIX allows open() to be a macro; guard against that.
837 PH/24 If the recipient of an error message contained an @ in the local part
838 (suitably quoted, of course), incorrect values were put in $domain and
839 $local_part during the evaluation of errors_copy.
845 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
846 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
849 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
850 issue a MAIL command.
852 PH/02 In an ACL statement such as
854 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
856 if a client was not listed at all, or was listed with a value other than
857 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
858 the condition was not true (as it should be), so access was not denied.
859 The bug was that the ! inversion was incorrectly passed on to the second
860 item. This has been fixed.
862 PH/03 Added additional dnslists conditions == and =& which are different from
863 = and & when the dns lookup returns more than one IP address.
865 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
866 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
868 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
869 FSYNC, which compiles an option called disable_fsync that allows for
870 bypassing fsync(). The documentation is heavily laced with warnings.
872 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
874 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
875 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
876 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
877 including adding "make clean"; (3) Added -fPIC when compiling the test
878 dynamically loaded module, to get rid of a warning.
880 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
881 message fails, move_frozen_messages = true and ignore_bounce_errors_after
882 = 0s. The bug is otherwise harmless.
884 PH/07 There was a bug in the dovecot authenticator such that the value of
885 $auth1 could be overwritten, and so not correctly preserved, after a
886 successful authentication. This usually meant that the value preserved by
887 the server_setid option was incorrect.
889 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
891 PH/09 Installed PCRE release 7.0.
893 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
894 run for batched SMTP input. It is now run at the start of every message
895 in the batch. While fixing this I discovered that the process information
896 (output by running exiwhat) was not always getting set for -bs and -bS
897 input. This is fixed, and it now also says "batched" for BSMTP.
899 PH/11 Added control=no_pipelining.
901 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
902 patch, slightly modified), and move the expansion of helo_data till after
903 the connection is made in the smtp transport (so it can use these
906 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
908 PH/14 Added log_selector = +pid.
910 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
912 PH/16 Add ${if forany and ${if forall.
914 PH/17 Added dsn_from option to vary the From: line in DSNs.
916 PH/18 Flush SMTP output before performing a callout, unless control =
917 no_callout_flush is set.
919 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
920 was true (the default) a successful delivery failed to delete the retry
921 item, thus causing premature timeout of the address. The bug is now
924 PH/20 Added hosts_avoid_pipelining to the smtp transport.
926 PH/21 Long custom messages for fakedefer and fakereject are now split up
927 into multiline reponses in the same way that messages for "deny" and
928 other ACL rejections are.
930 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
931 with slight modification.
933 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
934 draft, changing the syntax and factoring some duplicate code".
936 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
937 for deliveries of the second and subsequent messages over the same SMTP
940 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
941 ${reduce, with only minor "tidies".
943 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
945 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
946 expansion side effects.
948 PH/27 When a message times out after an over-quota error from an Exim-imposed
949 quota, the bounce message says "mailbox is full". This message was not
950 being given when it was a system quota that was exceeded. It now should
953 MH/03 Made $recipients available in local_scan(). local_scan() already has
954 better access to the recipient list through recipients_list[], but
955 $recipients can be useful in postmaster-provided expansion strings.
957 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
958 in the case of a MAIL command with additional options following the
959 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
960 were accidentally chopped off.
962 PH/29 SMTP synchronization checks are implemented when a command is read -
963 there is a check that no more input is waiting when there shouldn't be
964 any. However, for some commands, a delay in an ACL can mean that it is
965 some time before the response is written. In this time, more input might
966 arrive, invalidly. So now there are extra checks after an ACL has run for
967 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
968 pipelining has not been advertised.
970 PH/30 MH's patch to allow iscntrl() characters to be list separators.
972 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
973 returned in the SMTP response when smtp_return_error_details was false.
976 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
977 instead of the C I/O that was originally supplied, because problems were
980 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
981 Exim which did not show up earlier: it was assuming that a call to
982 SSL_CTX_set_info_callback() might give an error value. In fact, there is
983 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
984 was a macro that became an assignment, so it seemed to work. This has
985 changed to a proper function call with a void return, hence the compile
986 error. Exim's code has been fixed.
988 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
991 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
993 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
995 PH/37 If a message is not accepted after it has had an id assigned (e.g.
996 because it turns out to be too big or there is a timeout) there is no
997 "Completed" line in the log. When some messages of this type were
998 selected by exigrep, they were listed as "not completed". Others were
999 picked up by some special patterns. I have improved the selection
1000 criteria to be more general.
1002 PH/38 The host_find_failed option in the manualroute router can now be set
1003 to "ignore", to completely ignore a host whose IP address cannot be
1004 found. If all hosts are ignored, the behaviour is controlled by the new
1005 host_all_ignored option.
1007 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1008 homing or because of multiple MX records with /mx) generated more than
1009 one IP address, and the following item turned out to be the local host,
1010 all the secondary addresses of the first item were incorrectly removed
1011 from the list, along with the local host and any following hosts (which
1012 is what is supposed to happen).
1014 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1015 whoever called Exim into the -H file. In the case of the daemon it was
1016 behaving confusingly. When first started, it used values for whoever
1017 started the daemon, but after a SIGHUP it used the Exim user (because it
1018 calls itself on a restart). I have changed the code so that it now always
1021 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1022 message are rejected with the same error (e.g. no authentication or bad
1023 sender address), and a DATA command is nevertheless sent (as can happen
1024 with PIPELINING or a stupid MUA), the error message that was given to the
1025 RCPT commands is included in the rejection of the DATA command. This is
1026 intended to be helpful for MUAs that show only the final error to their
1029 PH/42 Another patch from the Sieve maintainer.
1031 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1032 Eximstats - Fixed some broken HTML links and added missing column headers
1034 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1035 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1037 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1039 PH/43 Yet another patch from the Sieve maintainer.
1041 PH/44 I found a way to check for a TCP/IP connection going away before sending
1042 the response to the final '.' that terminates a message, but only in the
1043 case where the client has not sent further data following the '.'
1044 (unfortunately, this is allowed). However, in many cases there won't be
1045 any further data because there won't be any more messages to send. A call
1046 to select() can be used: if it shows that the input is "ready", there is
1047 either input waiting, or the socket has been closed. An attempt to read
1048 the next input character can distinguish the two cases. Previously, Exim
1049 would have sent an OK response which the client would never have see.
1050 This could lead to message repetition. This fix should cure that, at
1051 least in a lot of common cases.
1053 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1054 advertised in response to EHLO.
1060 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1061 fixed by 4.65/MH/01 (is this a record?) are fixed:
1063 (i) An empty string was always treated as zero by the numeric comparison
1064 operators. This behaviour has been restored.
1066 (ii) It is documented that the numeric comparison operators always treat
1067 their arguments as decimal numbers. This was broken in that numbers
1068 starting with 0 were being interpreted as octal.
1070 While fixing these problems I realized that there was another issue that
1071 hadn't been noticed. Values of message_size_limit (both the global option
1072 and the transport option) were treated as octal if they started with 0.
1073 The documentation was vague. These values are now always treated as
1074 decimal, and I will make that clear in the documentation.
1080 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1081 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1084 MH/01 Don't check that the operands of numeric comparison operators are
1085 integers when their expansion is in "skipping" mode (fixes bug
1086 introduced by 4.64-PH/07).
1088 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1089 child addresses, Exim now panics and dies. Previously, because the count
1090 is held in a short int, deliveries were likely to be lost. As such a
1091 large number of recipients for a single message is ridiculous
1092 (performance will be very, very poor), I have chosen to impose a limit
1093 rather than extend the field.
1099 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1100 leftover -K file (the existence of which was triggered by #402).
1101 While we were at it, introduced process PID as part of the -K
1102 filename. This should rule out race conditions when creating
1105 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1106 processing considerably. Previous code took too long for large mails,
1107 triggering a timeout which in turn triggers #401.
1109 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1110 in the DK code in transports.c. sendfile() is not really portable,
1111 hence the _LINUX specificness.
1113 TF/01 In the add_headers option to the mail command in an Exim filter,
1114 there was a bug that Exim would claim a syntax error in any
1115 header after the first one which had an odd number of characters
1118 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1119 callout verification, Exim cached a "reject" for the entire domain. This
1120 is correct for most verifications, but it is not correct for a recipient
1121 verification with use_sender or use_postmaster set, because in that case
1122 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1123 case of MAIL FROM:<> rejection from other early rejections (e.g.
1124 rejection of HELO). When verifying a recipient using a non-null MAIL
1125 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1126 Whatever the result of the callout, the value of the domain cache is
1127 left unchanged (for any other kind of callout, getting as far as trying
1128 RCPT means that the domain itself is ok).
1130 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1133 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1134 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1135 session, the daemon ignores ECONNECT errors and logs others; it now
1136 ignores EPIPE as well.
1138 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1139 (quoted-printable decoding).
1141 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1142 later the small subsequent patch to fix an introduced bug.
1144 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1146 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1148 PH/08 An error is now given if message_size_limit is specified negative.
1150 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1151 to be given (somewhat) arbitrary names.
1153 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1156 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1157 miscellaneous code fixes
1159 PH/10 Added the log_reject_target ACL modifier to specify where to log
1162 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1163 hostname. This is wrong, because it relates to the incoming message (and
1164 probably the interface on which it is arriving) and not to the outgoing
1165 callout (which could be using a different interface). This has been
1166 changed to use the value of the helo_data option from the smtp transport
1167 instead - this is what is used when a message is actually being sent. If
1168 there is no remote transport (possible with a router that sets up host
1169 addresses), $smtp_active_hostname is used.
1171 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1172 tweaks were necessary in order to get it to work (see also 21 below):
1173 (a) The code assumed that strncpy() returns a negative number on buffer
1174 overflow, which isn't the case. Replaced with Exim's string_format()
1176 (b) There were several signed/unsigned issues. I just did the minimum
1177 hacking in of casts. There is scope for a larger refactoring.
1178 (c) The code used strcasecmp() which is not a standard C function.
1179 Replaced with Exim's strcmpic() function.
1180 (d) The code set only $1; it now sets $auth1 as well.
1181 (e) A simple test gave the error "authentication client didn't specify
1182 service in request". It would seem that Dovecot has changed its
1183 interface. Fortunately there's a specification; I followed it and
1184 changed what the client sends and it appears to be working now.
1186 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1189 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1190 address A is aliased to B and C, where B exists and C does not. Without
1191 -v the output is "A verified" because verification stops after a
1192 successful redirection if more than one address is generated. However,
1193 with -v the child addresses are also verified. Exim was outputting "A
1194 failed to verify" and then showing the successful verification for C,
1195 with its parentage. It now outputs "B failed to verify", showing B's
1196 parentage before showing the successful verification of C.
1198 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1199 look up a TXT record in a specific list after matching in a combined
1202 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1203 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1204 they consult the DNS. I had assumed they would set it the way they
1205 wanted; and indeed my experiments on Linux seem to show that in some
1206 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1207 To be on the safe side, however, I have now made the interface to
1208 host_find_byname() similar to host_find_bydns(), with an argument
1209 containing the DNS resolver options. The host_find_byname() function now
1210 sets these options at its start, just as host_find_bydns() does. The smtp
1211 transport options dns_qualify_single and dns_search_parents are passed to
1212 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1213 of host_find_byname() use the default settings of RES_DEFNAMES
1214 (qualify_single) but not RES_DNSRCH (search_parents).
1216 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1217 spool_read_header() do less string testing, by means of a preliminary
1218 switch on the second character of optional "-foo" lines. (This is
1219 overdue, caused by the large number of possibilities that now exist.
1220 Originally there were few.) While I was there, I also converted the
1221 str(n)cmp tests so they don't re-test the leading "-" and the first
1222 character, in the hope this might squeeze out yet more improvement.
1224 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1225 flag allowing group syntax was set by the header_syntax check but not
1226 turned off, possible causing trouble later; (2) The flag was not being
1227 set at all for the header_verify test, causing "group"-style headers to
1228 be rejected. I have now set it in this case, and also caused header_
1229 verify to ignore an empty address taken from a group. While doing this, I
1230 came across some other cases where the code for allowing group syntax
1231 while scanning a header line wasn't quite right (mostly, not resetting
1232 the flag correctly in the right place). These bugs could have caused
1233 trouble for malformed header lines. I hope it is now all correct.
1235 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1236 with the "reply" argument non-NULL. The code, however (which originally
1237 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1238 but it didn't always do it. This confused somebody who was copying the
1239 code for some other use. I have removed all the tests.
1241 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1242 feature that was used to support insecure browsers during the U.S. crypto
1243 embargo. It requires special client support, and Exim is probably the
1244 only MTA that supported it -- and would never use it because real RSA is
1245 always available. This code has been removed, because it had the bad
1246 effect of slowing Exim down by computing (never used) parameters for the
1247 RSA_EXPORT functionality.
1249 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1250 authenticator to fail if there's a tab character in the incoming data
1251 (there should never be unless someone is messing about, as it's supposed
1252 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1253 if the connection is using TLS or if the remote IP is the same as the
1254 local IP, and the "valid-client-cert option" if a client certificate has
1257 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1258 authenticators. This can be used for authorization after authentication
1259 succeeds. (In the case of plaintext, it servers for both authentication
1262 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1263 if any retry times were supplied.
1265 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1266 connection, where there is no client IP address to check. In this
1267 situation, the verify now always succeeds.
1269 PH/25 Applied John Jetmore's -Mset patch.
1271 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1273 PH/27 In a string expansion for a processed (not raw) header when multiple
1274 headers of the same name were present, leading whitespace was being
1275 removed from all of them, but trailing whitespace was being removed only
1276 from the last one. Now trailing whitespace is removed from each header
1277 before concatenation. Completely empty headers in a concatenation (as
1278 before) are ignored.
1280 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1281 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1283 PH/29 [Removed. This was a change that I later backed out, and forgot to
1284 correct the ChangeLog entry (that I had efficiently created) before
1285 committing the later change.]
1287 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1288 address errors (4xx response to RCPT) over the same connection as other
1289 messages routed to the same hosts. Such deliveries are always "forced",
1290 so retry times are not inspected. This resulted in far too many retries
1291 for the affected addresses. The effect occurred only when there were more
1292 hosts than the hosts_max_try setting in the smtp transport when it had
1293 the 4xx errors. Those hosts that it had tried were not added to the list
1294 of hosts for which the message was waiting, so if all were tried, there
1295 was no problem. Two fixes have been applied:
1297 (i) If there are any address or message errors in an SMTP delivery, none
1298 of the hosts (tried or untried) are now added to the list of hosts
1299 for which the message is waiting, so the message should not be a
1300 candidate for sending over the same connection that was used for a
1301 successful delivery of some other message. This seems entirely
1302 reasonable: after all the message is NOT "waiting for some host".
1303 This is so "obvious" that I'm not sure why it wasn't done
1304 previously. Hope I haven't missed anything, but it can't do any
1305 harm, as the worst effect is to miss an optimization.
1307 (ii) If, despite (i), such a delivery is accidentally attempted, the
1308 routing retry time is respected, so at least it doesn't keep
1309 hammering the server.
1311 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1312 in ${readsocket because some servers need this prod.
1314 PH/32 Added some extra debug output when updating a wait-xxx database.
1316 PH/33 The hint "could be header name not terminated by colon", which has been
1317 given for certain expansion errors for a long time, was not being given
1318 for the ${if def:h_colon_omitted{... case.
1320 PH/34 The spec says: "With one important exception, whenever a domain list is
1321 being scanned, $domain contains the subject domain." There was at least
1322 one case where this was not true.
1324 PH/35 The error "getsockname() failed: connection reset by peer" was being
1325 written to the panic log as well as the main log, but it isn't really
1326 panic-worthy as it just means the connection died rather early on. I have
1327 removed the panic log writing for the ECONNRESET error when getsockname()
1330 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1331 runs only) independently of the message's sender address. This meant
1332 that, if the 4xx error was in fact related to the sender, a different
1333 message to the same recipient with a different sender could confuse
1334 things. In particualar, this can happen when sending to a greylisting
1335 server, but other circumstances could also provoke similar problems.
1336 I have changed the default so that the retry time for these errors is now
1337 based a combination of the sender and recipient addresses. This change
1338 can be overridden by setting address_retry_include_sender=false in the
1341 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1342 remote server are returned as part of bounce messages. This was not
1343 happening for LMTP over a pipe (the lmtp transport), but now it is the
1344 same for both kinds of LMTP.
1346 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1347 in header lines that were in fact CNAMEs. This is no longer the case.
1349 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1350 and queue runs started by the daemon processed all messages. This has
1351 been fixed so that -R and -S can now usefully be given with -q<time>.
1353 PH/40 Import PCRE release 6.7 (fixes some bugs).
1355 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1357 PH/42 Give an error if -q is specified more than once.
1359 PH/43 Renamed the variables $interface_address and $interface_port as
1360 $received_ip_address and $received_port, to make it clear that these
1361 values apply to message reception, and not to the outgoing interface when
1362 a message is delivered. (The old names remain recognized, of course.)
1364 PH/44 There was no timeout on the connect() call when using a Unix domain
1365 socket in the ${readsocket expansion. There now is.
1367 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1368 be meaningful with "accept".
1370 SC/01 Eximstats V1.43
1371 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1373 SC/02 Eximstats V1.44
1374 Use a glob alias rather than an array ref in the generated
1375 parser. This improves both readability and performance.
1377 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1378 Collect SpamAssassin and rejection statistics.
1379 Don't display local sender or destination tables unless
1380 there is data to show.
1381 Added average volumes into the top table text output.
1383 SC/04 Eximstats V1.46
1384 Collect data on the number of addresses (recipients)
1385 as well as the number of messages.
1387 SC/05 Eximstats V1.47
1388 Added 'Message too big' to the list of mail rejection
1389 reasons (thanks to Marco Gaiarin).
1391 SC/06 Eximstats V1.48
1392 Mainlog lines which have GMT offsets and are too short to
1393 have a flag are now skipped.
1395 SC/07 Eximstats V1.49 (Alain Williams)
1396 Added the -emptyok flag.
1398 SC/08 Eximstats V1.50
1399 Fixes for obtaining the IP address from reject messages.
1401 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1402 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1403 whitesspace changes from 4.64-PH/27
1405 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1408 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1409 are found, allow negative numbers in numeric criteria)
1411 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1413 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1414 to match changes made in 4.64-PH/43
1416 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1418 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1419 "domain" is actually an IP address, and force a failure. This locks out
1420 those revolvers/nameservers that support "A-for-A" lookups, in
1421 contravention of the specifications.
1423 PH/48 When a host name was looked up from an IP address, and the subsequent
1424 forward lookup of the name timed out, the host name was left in
1425 $sender_host_name, contrary to the specification.
1427 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1428 restricted to single-key lookups, Exim was not diagnosing an error if
1429 * or *@ was used with a query-style lookup.
1431 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1433 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1434 long ago, but noone interested enough thought of it. Let's just say that
1435 the "1.1" means that there are some new functions that weren't there at
1436 some point in the past.
1438 PH/51 Error processing for expansion failure of helo_data from an smtp
1439 transport during callout processing was broken.
1441 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1442 tested/used via the -bh/-bhc/-bs options.
1444 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1445 bug, fixed in subsequent PCRE releases).
1447 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1448 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1450 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1456 SC/01 Use a glob alias rather than an array ref in eximstats generated
1457 parser. This improves both readability and performance.
1459 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1460 Don't display local sender or destination tables in eximstats unless
1461 there is data to show.
1462 Added average volumes into the eximstats top table text output.
1464 SC/03 Collect data on the number of addresses (recipients) as well
1465 as the number of messages in eximstats.
1467 TF/01 Correct an error in the documentation for the redirect router. Exim
1468 does (usually) call initgroups() when daemonizing.
1470 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1471 with consistent privilege compared to when running as a daemon.
1473 TF/03 Note in the spec that $authenticated_id is not set for local
1474 submissions from trusted users.
1476 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1477 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1479 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1480 by adding some example configuration directives to the default
1481 configuration file. A little bit of work is required to uncomment the
1482 directives and define how usernames and passwords are checked, but
1483 there is now a framework to start from.
1485 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1486 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1487 without this. I don't know how relevant this is to other LDAP libraries.
1489 PH/02 Add the verb name to the "unknown ACL verb" error.
1491 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1493 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1495 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1496 directory not expanded when it should be if an expanded home directory
1497 was set for the address (which is overridden by the transport).
1499 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1502 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1503 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1504 because it is too late at that time, and has no effect.
1506 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1507 security issue with \' (bugzilla #107). I could not use the
1508 PQescapeStringConn() function, because it needs a PGconn value as one of
1511 PH/08 When testing addresses using -bt, indicate those final addresses that
1512 are duplicates that would not cause an additional delivery. At least one
1513 person was confused, thinking that -bt output corresponded to deliveries.
1514 (Suppressing duplicates isn't a good idea as you lose the information
1515 about possibly different redirections that led to the duplicates.)
1517 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1518 systems where poll() doesn't work, in particular OS X.
1520 PH/10 Added more information to debugging output for retry time not reached.
1522 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1523 operations in malware.c.
1525 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1528 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1529 syslog_duplication set false, log lines that would normally be written
1530 both the the main log and to the reject log were not written to syslog at
1533 PH/14 In the default configuration, change the use of "message" in ACL warn
1534 statements to "add_header".
1536 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1537 not followed by a command (e.g. "seen endif").
1539 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1540 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1543 PH/17 Added extra conditions to the default value of delay_warning_condition
1547 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1548 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1549 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1552 The Auto-Submitted: and various List- headers are standardised, whereas I
1553 don't think Precedence: ever was.
1555 PH/18 Refactored debugging code in route_finduser() to show more information,
1556 in particular, the error code if getpwnam() issues one.
1558 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1559 This is apparently needed in addition to the PH/07 change above to avoid
1560 any possible encoding problems.
1562 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1563 but not after initializing Perl.
1565 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1566 output them only if debugging. By default they are written stderr,
1567 apparently, which is not desirable.
1569 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1572 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1575 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1577 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1578 authenticated or an ident call has been made. Suppress the default
1579 values for $authenticated_id and $authenticated_sender (but permit -oMai
1580 and -oMas) when testing with -bh.
1582 PH/24 Re-jigged the order of the tests in the default configuration so that the
1583 tests for valid domains and recipients precede the DNS black list and CSA
1584 tests, on the grounds that those ones are more expensive.
1586 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1587 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1588 This bug exists in every version of Exim that I still have, right back to
1591 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1592 However, an attempt to turn on case-sensitivity in a regex key by
1593 including (?-i) didn't work because the subject string was already
1594 lowercased, and the effects were non-intuitive. It turns out that a
1595 one-line patch can be used to allow (?-i) to work as expected.
1601 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1602 other effects) broke the use of negated acl sub-conditions.
1604 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1607 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1608 "Deny" causes Exim to reject the incoming connection with a 554 error.
1609 Unfortunately, if there is a major crisis, such as a disk failure,
1610 tcp-wrappers gives "deny", whereas what one would like would be some
1611 kind of temporary error. A kludge has been added to help with this.
1612 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1613 554 error is used if errno is still zero or contains ENOENT (which occurs
1614 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1617 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1619 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1620 errors. Otherwise a message that provokes a temporary error (when other
1621 messages do not) can cause a whole host to time out.
1623 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1624 addresses were routed directly to files or pipes from a redirect router.
1625 File deliveries just didn't batch; pipe deliveries might have suffered
1628 PH/06 A failure to get a lock for a hints database would erroneously always say
1629 "Failed to get write lock", even when it was really a read lock.
1631 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1632 of 0600. This has been changed to use the value of the lockfile_mode
1633 option (which defaults to 0600).
1635 PH/08 Applied small patch from the Sieve maintainer.
1637 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1638 folder from quota calculations, a direct delivery into this folder messed
1639 up the contents of the maildirsize file. This was because the regex was
1640 used only to exclude .Trash (or whatever) when the size of the mailbox
1641 was calculated. There was no check that a delivery was happening into an
1642 excluded directory. This bug has been fixed by ignoring all quota
1643 processing for deliveries into excluded directories.
1645 PH/10 Added the maildirfolder_create_regex option to appendfile.
1651 PH/01 The code for finding all the local interface addresses on a FreeBSD
1652 system running IPv6 was broken. This may well have applied to all BSD
1653 systems, as well as to others that have similar system calls. The broken
1654 code found IPv4 interfaces correctly, but gave incorrect values for the
1655 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1656 that it would not match correctly against @[] and not recognize the IPv6
1659 PH/02 The ipliteral router was not recognizing addresses of the form user@
1660 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1662 PH/03 Added disable_ipv6.
1664 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1665 decoded form, because it is most often used to construct To: headers
1666 lines in autoreplies, and the decoded form may well be syntactically
1667 invalid. However, $reply_address has leading white space removed, and all
1668 newlines turned into spaces so that the autoreply transport does not
1671 PH/05 If group was specified without a user on a router, and no group or user
1672 was specified on a transport, the group from the router was ignored.
1674 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1675 for visible compile-time settings that can be used to change these
1676 numbers, for those that want even more. Backwards compatibility with old
1677 spool files has been maintained. However, going back to a previous Exim
1678 release will lost any variables that are in spool files.
1680 PH/07 Two small changes when running in the test harness: increase delay when
1681 passing a TCP/IP connection to a new process, in case the original
1682 process has to generate a bounce, and remove special handling of
1683 127.0.0.2 (sic), which is no longer necessary.
1685 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1686 be the same on different OS.
1688 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1691 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1692 whether --show-vars was specified or not
1694 JJ/02 exipick: Added support for new ACL variable spool format introduced
1697 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1698 syntactically invalid From: or Reply-to: line, and a filter used this to
1699 generate an autoreply, and therefore failed to obtain an address for the
1700 autoreply, Exim could try to deliver to a non-existent relative file
1701 name, causing unrelated and misleading errors. What now happens is that
1702 it logs this as a hard delivery error, but does not attempt to create a
1705 PH/11 The exinext utility has a -C option for testing purposes, but although
1706 the given file was scanned by exinext itself; it wasn't being passed on
1707 when Exim was called.
1709 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1710 an end-of-file indication when reading a command response.
1712 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1713 compiled. In many other places in Exim, IPv6 addresses are always
1714 recognized, so I have changed this. It also means that IPv4 domain
1715 literals of the form [IPV4:n.n.n.n] are now always recognized.
1717 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1718 used if the router is not running as root, for example, when verifying at
1719 ACL time, or when using -bh. The debugging output from this situation was
1720 non-existent - all you got was a failure to exec. I have made two
1723 (a) Failures to set uid/gid, the current directory, or a process leader
1724 in a subprocess such as that created by queryprogram now generate
1725 suitable debugging ouput when -d is set.
1727 (b) The queryprogram router detects when it is not running as root,
1728 outputs suitable debugging information if -d is set, and then runs
1729 the subprocess without attempting to change uid/gid.
1731 PH/15 Minor change to Makefile for building test_host (undocumented testing
1734 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1735 additional section of a DNS packet that returns MX or SRV records.
1736 Instead, it always explicitly searches for A/AAAA records. This avoids
1737 major problems that occur when a DNS server includes only records of one
1738 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1739 fixed another bug: if SRV records were looked up and the corresponding
1740 address records were *not* found in the additional section, the port
1741 values from the SRV records were lost.
1743 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1744 using the correct key (the original address) when searching the retry
1745 rules in order to find which one to use for generating the retry hint.
1747 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1748 adding the default one. Similarly, if it contains a Reply-To: header, the
1749 errors_reply_to option, if set, is not used.
1751 PH/19 When calculating a retry time, Exim used to measure the "time since
1752 failure" by looking at the "first failed" field in the retry record. Now
1753 it does not use this if it is later than than the arrival time of the
1754 message. Instead it uses the arrival time. This makes for better
1755 behaviour in cases where some deliveries succeed, thus re-setting the
1756 "first failed" field. An example is a quota failure for a huge message
1757 when small messages continue to be delivered. Without this change, the
1758 "time since failure" will always be short, possible causing more frequent
1759 delivery attempts for the huge message than are intended.
1760 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1762 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1763 $1, $2, $3) because the numerical variables can be reset during some
1764 expansion items (e.g. "match"), thereby losing the authentication data.
1766 PH/21 Make -bV show the size of off_t variables so that the test suite can
1767 decide whether to run tests for quotas > 2G.
1769 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1770 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1771 filecount value is greater than 2G or if a quota value is greater than 2G
1772 on a system where the size of off_t is not greater than 4, a panic error
1775 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1776 never match. The debug and -bh output now contains an explicit error
1777 message indicating a malformed IPv4 address or mask.
1779 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1780 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1781 PH/23 above applies.
1783 PH/25 Do not write to syslog when running in the test harness. The only
1784 occasion when this arises is a failure to open the main or panic logs
1785 (for which there is an explicit test).
1787 PH/26 Added the /no_tell option to "control=freeze".
1789 PH/27 If a host name lookup failed very early in a connection, for example, if
1790 the IP address matched host_lookup and the reverse lookup yielded a name
1791 that did not have a forward lookup, an error message of the form "no IP
1792 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1793 could be logged. Now it outputs the IP address instead of "NULL".
1795 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1796 allows the sender and the authenticated sender to be set when
1797 submitting a message from within Exim. Since child_open_exim() is
1798 documented for local_scan(), the new function should be too.
1800 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1801 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1802 results in an empty string is now treated as unset.
1804 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1806 PH/31 Added sender_verify_fail logging option.
1808 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1809 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1810 tidied the source and removed it altogether.
1812 PH/33 When a queue run was abandoned because the load average was too high, a
1813 log line was always written; now it is written only if the queue_run log
1814 selector is set. In addition, the log line for abandonment now contains
1815 information about the queue run such as the pid. This is always present
1816 in "start" and "stop" lines but was omitted from the "abandon" line.
1818 PH/34 Omit spaces between a header name and the colon in the error message that
1819 is given when verify = headers_syntax fails (if there are lots of them,
1820 the message gets confusing).
1822 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1823 names, as there are now some PTR records that contain slashes. This check
1824 is only to protect against broken name servers that fall over on strange
1825 characters, so the fact that it applies to all lookups doesn't matter.
1827 PH/36 Now that the new test suite is complete, we can remove some of the
1828 special code in Exim that was needed for the old test suite. For example,
1829 sorting DNS records because real resolvers return them in an arbitrary
1830 order. The new test suite's fake resolver always returns records in the
1833 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1834 bounces) except when queue_only is set, to avoid logging races between
1835 the different processes.
1837 PH/38 Panic-die if .include specifies a non-absolute path.
1839 PH/39 A tweak to the "H" retry rule from its user.
1841 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1842 a label. They prevented compilation on older perls.
1844 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1845 a warning to be raised on newish perls.
1847 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1848 on queue. Changes to match documented behaviour of showing count of
1849 messages matching specified criteria.
1851 PH/40 Changed the default ident timeout from 30s to 5s.
1853 PH/41 Added support for the use of login_cap features, on those BSD systems
1854 that have them, for controlling the resources used by pipe deliveries.
1856 PH/42 The content-scanning code uses fopen() to create files in which to put
1857 message data. Previously it was not paying any attention to the mode of
1858 the files. Exim runs with umask(0) because the rest of the code creates
1859 files with open(), and sets the required mode explicitly. Thus, these
1860 files were ending up world-writeable. This was not a big issue, because,
1861 being within the spool directory, they were not world-accessible. I have
1862 created a function called modefopen, which takes an additional mode
1863 argument. It sets umask(777), creates the file, chmods it to the required
1864 mode, then resets the umask. All the relevant calls to fopen() in the
1865 content scanning code have been changed to use this function.
1867 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1868 to 24 hours. This avoids potential overflow problems when processing G
1869 and H retry rules. I suspect nobody ever tinkers with this value.
1871 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1873 PH/45 When the plaintext authenticator is running as a client, the server's
1874 challenges are checked to ensure they are valid base64 strings. By
1875 default, the authentication attempt is cancelled if an invalid string is
1876 received. Setting client_ignore_invalid_base64 true ignores these errors.
1877 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1878 they are received. Thus, the responses can be made to depend on the
1879 challenges. If an invalid string is ignored, an empty string is placed in
1882 PH/46 Messages that are created by the autoreply transport now contains a
1883 References: header, in accordance with RFCs 2822 and 3834.
1885 PH/47 Added authenticated_sender_force to the smtp transport.
1887 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1889 PH/49 Installed latest patch from the Sieve maintainer.
1891 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1892 was also set, the appendfile transport was unnecessarily scanning a
1893 directory of message files (e.g. for maildir delivery) to find the count
1894 of files (along with the size), even though it did not need this
1895 information. It now does the scan only if it needs to find either the
1896 size of the count of files.
1898 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1900 PH/52 Two bugs concerned with error handling when the smtp transport is
1903 (i) Exim was not creating retry information for temporary errors given
1904 for individual recipients after the DATA command when the smtp transport
1905 was used in LMTP mode. This meant that they could be retried too
1906 frequently, and not timed out correctly.
1908 (ii) Exim was setting the flag that allows error details to be returned
1909 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1910 recipients that were returned after the DATA command.
1912 PH/53 This is related to PH/52, but is more general: for any failing address,
1913 when detailed error information was permitted to be returned to the
1914 sender, but the error was temporary, then after the final timeout, only
1915 "retry timeout exceeded" was returned. Now it returns the full error as
1916 well as "retry timeout exceeded".
1918 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1919 do this, and (what is worse) MTAs that accept it.
1921 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1922 will now be deprecated.
1924 PH/56 New os.c-cygwin from the Cygwin maintainer.
1926 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1927 formats (previously only available in exim formats via -bpr, -bpru,
1928 and -bpra. Now also available in native and exiqgrep formats)
1930 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1931 with very large, slow to parse queues
1933 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1935 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1937 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1938 responses to authentication challenges, though it was showing the
1939 challenges; (ii) I've removed the CR characters from the debug output for
1942 PH/58 Allow for the insertion of a newline as well as a space when a string
1943 is turned into more than one encoded-word during RFC 2047 encoding. The
1944 Sieve code now uses this.
1946 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1947 data_4xx, lost_connection, tls_required.
1949 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1950 message was being sent as an SMTP response.
1952 PH/61 Add -l and -k options to exicyclog.
1954 PH/62 When verifying, if an address was redirected to one new address, so that
1955 verification continued, and the new address failed or deferred after
1956 having set something in $address_data, the value of $address_data was not
1957 passed back to the ACL. This was different to the case when no
1958 redirection occurred. The value is now passed back in both cases.
1960 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1961 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1962 use login_cap.h, so on its own it isn't the distinguishing feature. The
1963 new name refers directly to the setclassresources() function.
1965 PH/65 Added configuration files for NetBSD3.
1967 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1969 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1970 is preferred over IPv4.
1972 PH/68 The bounce_return_message and bounce_return_body options were not being
1973 honoured for bounces generated during the reception of non-SMTP messages.
1974 In particular, this applied to messages rejected by the ACL. This bug has
1975 been fixed. However, if bounce_return_message is true and bounce_return_
1976 body is false, the headers that are returned for a non-SMTP message
1977 include only those that have been read before the error was detected.
1978 (In the case of an ACL rejection, they have all been read.)
1980 PH/69 The HTML version of the specification is now built in a directory called
1981 spec_html instead of spec.html, because the latter looks like a path with
1982 a MIME-type, and this confuses some software.
1984 PH/70 Catch two compiler warnings in sieve.c.
1986 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1987 function verify_get_ident() calls ip_connect() to connect a socket, but
1988 if the "connect()" function timed out, ip_connect() used to close the
1989 socket. However, verify_get_ident() also closes the socket later, and in
1990 between Exim writes to the log, which may get opened at this point. When
1991 the socket was closed in ip_connect(), the log could get the same file
1992 descriptor number as the socket. This naturally causes chaos. The fix is
1993 not to close the socket in ip_connect(); the socket should be closed by
1994 the function that creates it. There was only one place in the code where
1995 this was missing, in the iplookup router, which I don't think anybody now
1996 uses, but I've fixed it anyway.
1998 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1999 well as to direct DNS lookups. Otherwise the handling of names in host
2000 lists is inconsistent and therefore confusing.
2006 PH/01 Two changes to the default runtime configuration:
2008 (1) Move the checks for relay_from_hosts and authenticated clients from
2009 after to before the (commented out) DNS black list checks.
2011 (2) Add control=submission to the relay_from_hosts and authenticated
2012 clients checks, on the grounds that messages accepted by these
2013 statements are most likely to be submissions.
2015 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2017 (1) Generate an error if the third argument for the ${prvs expansion is
2020 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2023 (3) Reset the variables that are obtained from the first argument of
2024 ${prvscheck and used in the second argument before leaving the code,
2025 because their memory is reclaimed, so using them afterwards may do
2028 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2029 one (it's much easier than Tom thought :-).
2031 (5) Because of (4), we can now allow for the use of $prvscheck_result
2032 inside the third argument.
2034 PH/03 For some reason, the default setting of PATH when running a command from
2035 a pipe transport was just "/usr/bin". I have changed it to
2038 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2039 anything to be listed in the output from -bV.
2041 PH/05 When a filter generated an autoreply, the entire To: header line was
2042 quoted in the delivery log line, like this:
2044 => >A.N.Other <ano@some.domain> <original@ddress> ...
2046 This has been changed so that it extracts the operative address. There
2047 may be more than one such address. If so, they are comma-separated, like
2050 => >ano@some.domain,ona@other.domain <original@ddress> ...
2052 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2053 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2054 not being looked up in the rDNS to get a host name, Exim was showing the
2055 IP address twice in Received: lines, even though the IP addresses were
2056 identical. For example:
2058 Received: from [1.2.3.4] (helo=[1.2.3.4])
2060 However, if the real host name was known, it was omitting the HELO data
2061 if it matched the actual IP address. This has been tidied up so that it
2062 doesn't show the same IP address twice.
2064 PH/07 When both +timestamp and +memory debugging was on, the value given by
2065 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2066 called by the debug printing, thereby overwriting the timestamp buffer.
2067 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2069 PH/08 When the original message was included in an autoreply transport, it
2070 always said "this is a copy of the message, including all the headers",
2071 even if body_only or headers_only was set. It now gives an appropriate
2074 PH/09 Applied a patch from the Sieve maintainer which:
2076 o fixes some comments
2077 o adds the (disabled) notify extension core
2078 o adds some debug output for the result of if/elsif tests
2079 o points to the current vacation draft in the documentation
2080 and documents the missing references header update
2084 o fixes a bug in processing the envelope test (when testing
2085 multiple envelope elements, the last element determinted the
2088 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2089 Electronic Mail") by including:
2091 Auto-submitted: auto-generated
2093 in the messages that it generates (bounce messages and others, such as
2094 warnings). In the case of bounce messages for non-SMTP mesages, there was
2095 also a typo: it was using "Auto_submitted" (underscore instead of
2096 hyphen). Since every message generated by Exim is necessarily in response
2097 to another message, thes have all been changed to:
2099 Auto-Submitted: auto-replied
2101 in accordance with these statements in the RFC:
2103 The auto-replied keyword:
2105 - SHOULD be used on messages sent in direct response to another
2106 message by an automatic process,
2108 - MUST NOT be used on manually-generated messages,
2110 - MAY be used on Delivery Status Notifications (DSNs) and Message
2111 Disposition Notifications (MDNs),
2113 - MUST NOT be used on messages generated by automatic or periodic
2114 processes, except for messages which are automatic responses to
2117 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2118 to the default Received: header definition.
2120 PH/12 Added log selector acl_warn_skipped (default on).
2122 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2123 variables because (a) they are in the wrong storage pool and (b) even if
2124 they were copied, it wouldn't work properly because of the caching.
2126 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2127 checking when decoding. Apparently there are clients that generate
2128 overlong encoded strings. Why am I not surprised?
2130 PH/15 If the first argument of "${if match_address" was not empty, but did not
2131 contain an "@" character, Exim crashed. Now it writes a panic log message
2132 and treats the condition as false.
2134 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2136 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2137 extension "envelope-auth". The code is finished and in agreement with
2138 other implementations, but there is no documentation so far and in fact,
2139 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2140 not changing the active code.
2142 Print executed "if" and "elsif" statements when debugging is used. This
2143 helps a great deal to understand what a filter does.
2145 Document more things not specified clearly in RFC3028. I had all this
2146 sorted out, when out of a sudden new issues came to my mind. Oops."
2148 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2151 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2152 canonical form (as documented). However, after a host name lookup from
2153 the IP address, check_host() was doing a simple string comparison with
2154 addresses acquired from the DNS when checking that the found name did
2155 have the original IP as one of its addresses. Since any found IPv6
2156 addresses are likely to be in abbreviated form, the comparison could
2157 fail. Luckily, there already exists a function for doing the comparison
2158 by converting both addresses to binary, so now that is used instead of
2159 the text comparison.
2161 PH/20 There was another similar case to PH/19, when a complete host name was
2162 given in a host list; looking up its IP address could give an abbreviated
2163 form, whereas the current host's name might or might not be abbreviated.
2164 The same fix has been applied.
2170 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2171 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2174 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2175 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2177 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2179 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2180 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2181 Also, for "Auto-Submitted", treat anything other than "no" as
2182 non-personal, in accordance with RFC 3834. (Previously it treated
2183 anything starting "auto-" as non-personal.)
2185 TF/01 The control=submission/name=... option had a problem with syntax
2186 errors if the name included a slash character. The /name= option
2187 now slurps the rest of the string, so it can include any characters
2188 but it must come last in the list of options (after /sender_retain
2191 PH/05 Some modifications to the interface to the fake nameserver for the new
2199 TK/01 Added the "success_on_redirect" address verification option. See
2200 NewStuff for rationale and an example.
2202 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2204 PH/02 Patch to exigrep to allow it to work on syslog lines.
2206 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2207 fread() to skip over the body file's header line, because in Cygwin the
2208 header line is locked and is inaccessible.
2210 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2211 co-exist for some time) to make it clear that it is the Exim ID that is
2212 referenced, not the Message-ID: header line.
2214 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2215 string_format() function, because snprintf() does not exist on all
2218 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2219 ${stat: expansion item.
2221 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2222 protocol synchronization error", to keep the pedants happy.
2224 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2225 well as for IRIX systems, when gcc is being used. See the host.c source
2228 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2230 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2233 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2234 transport and to the smtp transport in LMTP mode.
2236 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2238 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2239 run a filter in a subprocess. This could lead to confusion in subsequent
2240 lookups in the parent process. There should also be a search_tidyup() at
2241 the end of the subprocess.
2243 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2244 only if the host matched helo_try_verify_hosts, which caused the
2245 verification to occur when the EHLO/HELO command was issued. The ACL just
2246 tested the remembered result. Now, if a previous verification attempt has
2247 not happened, "verify = helo" does it there and then.
2249 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2251 TK/03 Fix log output including CR from clamd.
2253 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2254 exist provoked a memory error which could cause a segfault.
2256 PH/15 Installed PCRE 6.2
2258 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2260 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2261 of the problem. Specifically, suggested +O2 rather than +O1 for the
2264 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2266 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2267 record, the port was not correctly logged when the outgoing_port log
2268 selector was set (it logged the transort's default port).
2270 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2271 fallback_hosts, and "hosts" in the smtp transport.
2273 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2274 host errors such as "Connection refused".
2276 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2277 authentication with radiusclient 0.4.9:
2279 - Error returned from rc_read_config was caught wrongly
2280 - Username/password not passed on to radius server due to wrong length.
2282 The presumption is that some radiusclient API changes for 4.51/PH/17
2283 were not taken care of correctly. The code is still untested by me (my
2284 Linux distribution still has 0.3.2 of radiusclient), but it was
2285 contributed by a Radius user.
2287 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2288 expanding the "port" option of the smtp transport.
2290 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2291 while reading a MIME header. Thanks to Tom Hughes for a patch.
2293 PH/24 Include config.h inside local_scan.h so that configuration settings are
2296 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2297 that in an ACL for RCPT (for example), you can examine exactly what was
2300 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2301 commands, but it was not correctly comparing the address with the actual
2302 client host address. Thus, it would show the EHLO address in Received:
2303 header lines when this was not necessary.
2305 PH/27 Added the % operator to ${eval:}.
2307 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2308 it should be ignoring failures (because with -C, for example, it has lost
2309 privilege). It wasn't ignoring creation failures other than "already
2312 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2315 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2316 because an input line was too long, either on its own, or by virtue of
2317 too many continuations, the temporary file was not being removed, and the
2318 return code was incorrect.
2320 PH/31 Missing "BOOL" in function definition in filtertest.c.
2322 PH/32 Applied Sieve patches from the maintainer.
2324 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2326 PH/33 Added "verify = not_blind".
2328 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2329 Local/Makefile (with some defaults set). These are used in built scripts
2330 such as exicyclog, but they have never been used in the exim_install
2331 script (though there are many overriding facilities there). I have
2332 arranged that the exim_install script now takes note of these two
2335 PH/35 Installed configuration files for Dragonfly.
2337 PH/36 When a locally submitted message by a trusted user did not contain a
2338 From: header, and the sender address was obtained from -f or from an SMTP
2339 MAIL command, and the trusted user did not use -F to supply a sender
2340 name, $originator_name was incorrectly used when constructing a From:
2341 header. Furthermore, $originator_name was used for submission mode
2342 messages from external hosts without From: headers in a similar way,
2343 which is clearly wrong.
2345 PH/37 Added control=suppress_local_fixups.
2347 PH/38 When log_selector = +received_sender was set, and the addition of the
2348 sender made the log line's construction buffer exactly full, or one byte
2349 less than full, an overflow happened when the terminating "\n" was
2352 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2353 when the result of a list match is failure because a DNS lookup failed.
2355 PH/40 RM_COMMAND is now used in the building process.
2357 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2358 the "build-* directories that it finds.
2360 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2361 address in a domain literal was a prefix of an interface address.
2363 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2364 when verifying a sender address, unless rewrite_headers is false.
2366 PH/44 Wrote a long comment about why errors_to addresses are verified as
2367 recipients, not senders.
2369 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2370 the ratelimit ACL was added.
2372 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2374 PH/46 Added extra information about PostgreSQL errors to the error string.
2376 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2377 avoiding the need to install special zones in a real server. This is
2378 backwards compatible; if it can't find the fake resolver, it drops back.
2379 Thus, both old and new test suites can be run.
2381 TF/02 Added util/ratelimit.pl
2383 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2384 clock is set back in time.
2386 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2387 Candler <B.Candler@pobox.com>.
2389 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2390 OK for addresses that are the result of redirections.
2392 PH/48 A number of further additions for the benefit of the new test suite,
2393 including a fake gethostbyname() that interfaces to the fake DNS resolver
2396 TF/06 The fix for widen_domains has also been applied to qualify_single and
2397 search_parents which are the other dnslookup options that can cause
2400 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2403 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2405 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2406 service" if it encountered an MX record with an empty target hostname.
2407 The message is now "an MX or SRV record indicated no SMTP service".
2409 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2410 if the DNS of the sending site is misconfigured. This is quite a
2411 common situation. This change restores the behaviour of treating a
2412 helo verification defer as a failure.
2414 PH/51 If self=fail was set on a router, the bounce message did not include the
2415 actual error message.
2421 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2423 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2424 command itself ended in error, the underlying message about the transport
2425 filter timeout was being overwritten with the pipe command error. Now the
2426 underlying error message should be appended to the second error message.
2428 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2430 PH/02 Reduce the amount of output that "make" produces by default. Full output
2431 can still be requested.
2433 PH/03 The warning log line about a condition test deferring for a "warn" verb
2434 was being output only once per connection, rather than after each
2435 occurrence (because it was using the same function as for successful
2436 "warn" verbs). This seems wrong, so I have changed it.
2438 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2439 it should not have, which might have caused a crash in the right
2440 circumstances, but probably never did.
2442 PH/04 Installed a modified version of Tony Finch's patch to make submission
2443 mode fix the return path as well as the Sender: header line, and to
2444 add a /name= option so that you can make the user's friendly name appear
2447 TF/03 Added the control = fakedefer ACL modifier.
2449 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2450 Mark Lowes for thorough testing.
2452 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2454 TK/03 Merged latest SRS patch from Miles Wilton.
2456 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2457 with the definition in sysexits.h (which is #included earlier).
2458 Fortunately, Exim does not actually use EX_OK. The code used to try to
2459 preserve the sysexits.h value, by assumimg that macro definitions were
2460 scanned for macro replacements. I have been disabused of this notion,
2461 so now the code just undefines EX_OK before #including unistd.h.
2463 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2464 in the smtp transport. When a block could not be written in a single
2465 write() function, the timeout was being re-applied to each part-write.
2466 This seems wrong - if the receiver was accepting one byte at a time it
2467 would take for ever. The timeout is now adjusted when this happens. It
2468 doesn't have to be particularly precise.
2470 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2471 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2473 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2474 without a domain if the check to <postmaster@domain> fails.
2476 SC/01 Eximstats: added -xls and the ability to specify output files
2477 (patch written by Frank Heydlauf).
2479 SC/02 Eximstats: use FileHandles for outputing results.
2481 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2483 SC/04 Eximstats: fixed display of large numbers with -nvr option
2485 SC/05 Eximstats: fixed merging of reports with empty tables.
2487 SC/06 Eximstats: added the -include_original_destination flag
2489 SC/07 Eximstats: removed tabs and trailing whitespace.
2491 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2493 TK/06 MBOX spool code: Add real "From " MBOX separator line
2494 so the .eml file is really in mbox format (even though
2495 most programs do not really care). Patch from Alex Miller.
2497 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2498 The latter is generated from $received_to and is only set if the
2499 message has one envelope recipient. SA can use these headers,
2500 obviously out-of-the-box. Patch from Alex Miller.
2502 PH/08 The ${def test on a variable was returning false if the variable's
2503 value was "0", contrary to what the specification has always said!
2504 The result should be true unless the variable is empty.
2506 PH/09 The syntax error of a character other than { following "${if
2507 def:variable_name" (after optional whitespace) was not being diagnosed.
2508 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2509 accidental colon was present, for example, could give incorrect results.
2511 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2512 result is used (not including appendfile, where other changes are about
2515 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2516 This involved changing a lot of size variables from int to off_t. It
2517 should work with maildirs and everything.
2519 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2520 spamd dying while we are connected to it.
2522 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2525 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2526 function for BDB 4.3.
2528 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2530 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2531 the sources for PCRE have been re-arranged and more files are now
2534 PH/15 The code I had for printing potentially long long variables in PH/11
2535 above was not the best (it lost precision). The length of off_t variables
2536 is now inspected at build time, and an appropriate printing format (%ld
2537 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2538 to be "long long int" or "long int". This is needed for the internal
2539 formatting function string_vformat().
2541 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2542 the configuration file to be ":syslog", then the script "guesses" where
2543 the logs files are, rather than using the compiled in default. In our
2544 case the guess is not the same as the compiled default, so the script
2545 suddenly stopped working when I started to use syslog. The patch checks
2546 to see if log_file_path is "". If so, it attempts to read it from exim
2547 with no configuration file to get the compiled in version, before it
2548 falls back to the previous guessing code."
2550 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2551 implementing BATV in an Exim configuration. See NewStuff for the gory
2554 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2555 Makefile that are specific to HP-UX.
2557 PH/18 If the "use_postmaster" option was set for a recipient callout together
2558 with the "random" option, the postmaster address was used as the MAIL
2559 FROM address for the random test, but not for the subsequent recipient
2560 test. It is now used for both.
2562 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2563 patch removes a few documentation additions to RFC 3028, because the
2564 latest draft now contains them. It adds the new en;ascii-case comparator
2565 and a new error check for 8bit text in MIME parts. Comparator and
2566 require names are now matched exactly. I enabled the subaddress
2567 extension, but it is not well tested yet (read: it works for me)."
2569 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2570 rework some of the code of TK/09 above to avoid the hardwired use of
2571 "%lld" and "long long". Replaced the call to snprintf() with a call to
2574 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2575 records point to non-existent hosts", "retry timeout exceeded", and
2576 "retry time not reached for any host after a long failure period".
2578 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2579 experimental DomainKeys support:
2581 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2582 (2) On an error such as an illegally used "control", the wrong name for
2583 the control was given.
2585 These problems did NOT occur unless DomainKeys support was compiled.
2587 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2589 PH/24 Added ${if match_ip condition.
2591 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2592 databases so that it will be absolutely obvious if a crash occurs in the
2593 DB library. This is a regular occurrence (often caused by mis-matched
2596 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2597 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2598 picked up on a user's system that detects such things. There doesn't seem
2599 to be a gcc warning option for this - only an attribute that has to be
2600 put on the function's prototype. It seems that in Fedora Core 4 they have
2601 set this on a number of new functions. No doubt there will be more in due
2604 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2605 specify a transport. However, if an address that was verified by such a
2606 router was the subject of a callout, Exim crashed because it tried to
2607 read the rcpt_include_affixes from the non-existent transport. Now it
2608 just assumes that the setting of that option is false. This bug was
2609 introduced by 4.51/PH/31.
2611 PH/28 Changed -d+all to exclude +memory, because that information is very
2612 rarely of interest, but it makes the output a lot bigger. People tend to
2613 do -d+all out of habit.
2615 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2616 code in os-type was giving problems when libc.so lives in lib64, like on
2619 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2620 aren't the modern standard, and it seems that some systems' include files
2621 don't always have them. Exim was already checking for some of the newer
2622 ones like T_AAAA, and defining it itself. I've added checks for all the
2623 record types that Exim uses.
2625 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2626 not automatically generating a new one, as it is supposed to. This
2627 prevented TLS from working. If the file did exist, but contained invalid
2628 data, a new version was generated, as expected. It was only the case of a
2629 non-existent file that was broken.
2631 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2632 with a change in libdomainkeys > 0.64.
2634 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2635 from DNS. If the selector record carries the flag, it now has
2636 precedence over the domain-wide flag.
2638 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2640 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2641 the use of a port name that isn't defined in /etc/services) occurred, the
2642 message was deferred as in a normal delivery, and thus remained on the
2643 spool, instead of being failed because of the mua_wrapper setting. This
2644 is now fixed, and I tidied up some of the mua_wrapper messages at the
2647 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2648 the messages in a hash of arrays rather than using individual hashes.
2649 This is a bit cleaner and results in dramatic memory savings, albeit
2650 at a slight CPU cost.
2652 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2653 as requested by Marc Sherman.
2655 SC/10 Eximstats: added histograms for user specified patterns as requested
2658 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2660 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2661 fopen() in the content-scanning modules that did not already have it.
2667 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2668 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2670 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2672 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2674 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2675 to test Sieve filters that use "vacation".
2677 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2678 that changes the way the GnuTLS parameters are stored in the cache file.
2679 The new format can be generated externally. For backward compatibility,
2680 if the data in the cache doesn't make sense, Exim assumes it has read an
2681 old-format file, and it generates new data and writes a new file. This
2682 means that you can't go back to an older release without removing the
2685 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2686 work if there are any delivery delays because "one_time" forces the
2687 parent to be marked "delivered", so its unseen clone is never tried
2688 again. For this reason, Exim now forbids the simultaneous setting of
2691 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2692 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2693 entry if you want to know the details. The fix, however, neglected to
2694 consider the case when local delivery batching is involved. The test for
2695 "previously delivered" was not happening when checking to see if an
2696 address could be batched with a previous (undelivered) one; under
2697 certain circumstances this could lead to multiple deliveries to the same
2700 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2701 in its include files, and this causes problems building Exim.
2703 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2704 header_syntax) but Exim was just ignoring anything given after a slash.
2705 In particular, this caused confusion with an attempt to use "verify =
2706 reverse_host_lookup/defer_ok". An error is now given when options are
2707 supplied for verify items that do not have them. (Maybe reverse_host_
2708 lookup should have a defer_ok option, but that's a different point.)
2710 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2711 defined by RFC 821) to 2048, because there were problems with some AUTH
2712 commands, and RFC 1869 says the size should be increased for extended
2713 SMTP commands that take arguments.
2715 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2718 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2719 "unknown" error; now it says that the functionality isn't in the binary.
2721 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2722 an address' error message when a string expansion fails (syntax or
2723 whatever). Otherwise the password may appear in the log. Following change
2724 PH/42 below, there is no longer a chance of it appearing in a bounce
2727 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2729 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2730 of its addresses was ever tried. (Bugzilla bug #2.)
2732 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2733 the result incorrectly in the debug output. (It correctly added a newline
2734 to what was transported.)
2736 TF/01 Added $received_time.
2738 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2739 commented out examples of how to interface to a virus scanner and to
2740 SpamAssassin. Also added commented examples of av_scanner and
2741 spamd_address settings.
2743 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2744 and controls are allowed in which ACLs. There were a couple of minor
2745 errors. Some of the entries in the conditions table (which is a table of
2746 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2747 negation of where the condition IS allowed.
2749 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2751 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2752 header file does not have a version number, so I've had to invent a new
2753 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2754 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2755 radiusclient), but it was contributed by a Radius user.
2757 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2758 files or directories, for OpenSSL.
2760 PH/19 When an Exim process that is running as root has to create an Exim log
2761 file, it does so in a subprocess that runs as exim:exim so as to get the
2762 ownership right at creation (otherwise, other Exim processes might see
2763 the file with the wrong ownership). There was no test for failure of this
2764 fork() call, which would lead to the process getting stuck as it waited
2765 for a non-existent subprocess. Forks do occasionally fail when resources
2766 run out. I reviewed all the other calls to fork(); they all seem to check
2769 PH/20 When checking for unexpected SMTP input at connect time (before writing
2770 the banner), Exim was not dealing correctly with a non-positive return
2771 from the read() function. If the client had disconnected by this time,
2772 the result was a log entry for a synchronization error with an empty
2773 string after "input=" when read() returned zero. If read() returned -1
2774 (an event I could not check), uninitialized data bytes were printed.
2775 There were reports of junk text (parts of files, etc) appearing after
2778 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2780 PH/22 Added support for macro redefinition, and (re)definition in between
2781 driver and ACL definitions.
2783 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2784 forgetting to use the resulting value; it was using the unexpanded value.
2786 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2787 hadn't been configured. The fix is from Juergen Kreileder, who
2788 understands it better than I do:
2790 "Here's what I see happening with three configured cyrus_sasl
2791 authenticators configured (plain, login, cram-md5):
2793 On startup auth_cyrus_sasl_init() gets called for each of these.
2794 This means three calls to sasl_listmech() without a specified mech_list.
2795 => SASL tests which mechs of all available mechs actually work
2796 => three warnings about OTP not working
2797 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2799 With the patch, sasl_listmech() also gets called three times. But now
2800 SASL's mech_list option is set to the server_mech specified in the the
2801 authenticator. Or in other words, the answer from sasl_listmech()
2802 gets limited to just the mech you're testing for (which is different
2804 => the return list contains just 'plain' or 'login', 'cram-md5' or
2805 nothing depending on the value of ob->server_mech.
2807 I've just tested the patch: Authentication still works fine,
2808 unavailable mechs specified in the exim configuration are still
2809 caught, and the auth.log warnings about OTP are gone."
2811 PH/25 When debugging is enabled, the contents of the command line are added
2812 to the debugging output, even when log_selector=+arguments is not
2815 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2816 answer is "GNU", and only if the return is "GNU/something" is the answer
2819 PH/27 $acl_verify_message is now set immediately after the failure of a
2820 verification in an ACL, and so is available in subsequent modifiers. In
2821 particular, the message can be preserved by coding like this:
2823 warn !verify = sender
2824 set acl_m0 = $acl_verify_message
2826 Previously, $acl_verify_message was set only while expanding "message"
2827 and "log_message" when a very denied access.
2829 PH/28 Modified OS/os.c-Linux with
2831 -#ifndef OS_LOAD_AVERAGE
2832 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2834 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2835 nomenclature these days.)
2837 PH/29 Installed patch from the Sieve maintainer that adds the options
2838 sieve_useraddress and sieve_subaddress to the redirect router.
2840 PH/30 In these circumstances:
2841 . Two addresses routed to the same list of hosts;
2842 . First host does not offer TLS;
2843 . First host accepts first address;
2844 . First host gives temporary error to second address;
2845 . Second host offers TLS and a TLS session is established;
2846 . Second host accepts second address.
2847 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2848 and peerdn, if requested) that were in fact used only for the second
2851 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2852 attention to any local part prefix or suffix that was matched by the
2853 router that accepted the address. It now behaves in the same way as it
2854 does for delivery: the affixes are removed from the local part unless
2855 rcpt_include_affixes is set on the transport.
2857 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2858 timeout during the DATA phase of an incoming message.
2860 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2861 applied a patch sanctioned by the Sieve maintainer.
2863 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2864 the uid or gid is negative. A case of a negative gid caused this to be
2865 noticed. The fix allows for either to be negative.
2867 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2868 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2871 PH/36 The change PH/12 above was broken. Fixed it.
2873 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2874 the grounds that routing the same address twice would always produce the
2875 same answer. This might have been true once, but it is certainly no
2876 longer true now. Routing a child address may depend on the previous
2877 routing that produced that child. Some complicated redirection strategies
2878 went wrong when messages had multiple recipients, and made Exim's
2879 behaviour dependent on the order in which the addresses were given.
2881 I have moved the duplicate checking until after the routing is complete.
2882 Exim scans the addresses that are assigned to local and remote
2883 transports, and removes any duplicates. This means that more work will be
2884 done, as duplicates will always all be routed, but duplicates are
2885 presumably rare, so I don't expect this is of any significance.
2887 For deliveries to pipes, files, and autoreplies, the duplicate checking
2888 still happens during the routing process, since they are not going to be
2891 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2892 It corrects a timeout issue with spamd. This is Ian's comment: "The
2893 background is that sometimes spamd either never reads data from a
2894 connection it has accepted, or it never writes response data. The exiscan
2895 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2896 blindly assumes that writes won't block so it may never time out."
2898 PH/39 Allow G after quota size as well as K and M.
2900 PH/40 The value set for $authenticated_id in an authenticator may not contain
2901 binary zeroes or newlines because the value is written to log lines and
2902 to spool files. There was no check on this. Now the value is run through
2903 the string_printing() function so that such characters are converted to
2904 printable escape sequences.
2906 PH/41 $message_linecount is a new variable that contains the total number of
2907 lines in the message. Compare $body_linecount, which is the count for the
2910 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2911 bounce and delay warning messages, except in certain special cases, which
2914 (a) An SMTP error message from a remote host;
2915 (b) A message specified in a :fail: redirection;
2916 (c) A message specified in a "fail" command in a system filter;
2917 (d) A message specified in a FAIL return from the queryprogram router;
2918 (e) A message specified by the cannot_route_message router option.
2920 In these cases only, Exim does include the error details in bounce and
2921 warning messages. There are also a few cases where bland messages such
2922 as "unrouteable address" or "local delivery error" are given.
2924 PH/43 $value is now also set for the "else" part of a ${run expansion.
2926 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2927 being worked on, but at least Exim now implements the latest version to
2930 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2931 process to complete was treated as a delivery failure, a timeout while
2932 writing the message to the pipe was logged, but erroneously treated as a
2933 successful delivery. Such timeouts include transport filter timeouts. For
2934 consistency with the overall process timeout, these timeouts are now
2935 treated as errors, giving rise to delivery failures by default. However,
2936 there is now a new Boolean option for the pipe transport called
2937 timeout_defer, which, if set TRUE, converts the failures into defers for
2938 both kinds of timeout. A transport filter timeout is now identified in
2941 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2942 systems where "make" and "gmake" are different, calling "gmake" at top
2943 level broke things. I've arranged for the value of $(MAKE) to be passed
2944 from the Makefile to this script so that it can call the same version of
2948 A note about Exim versions 4.44 and 4.50
2949 ----------------------------------------
2951 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2952 changes of various kinds. As a consequence, a big documentation update was
2953 needed. This delayed the release for rather longer than seemed good, especially
2954 in the light of a couple of (minor) security issues. Therefore, the changes
2955 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2956 release. So 4.44 and 4.50 are in effect two different branches that both start
2959 I have left the 4.50 change log unchanged; it contains all the changes since
2960 4.43. The change log for 4.44 is below; many of its items are identical to
2961 those for 4.50. This seems to be the most sensible way to preserve the
2962 historical information.
2968 1. Minor wording change to the doc/README.SIEVE file.
2970 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2971 computation of the current number of files was incorrect.
2973 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2974 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2977 4. Give more explanation in the error message when the command for a transport
2978 filter fails to execute.
2980 5. There are several places where Exim runs a non-Exim command in a
2981 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2982 was being done only for the command run by the queryprogram router. It is
2983 now done for all such subprocesses. The other cases are: ${run, transport
2984 filters, and the commands run by the lmtp and pipe transports.
2986 6. Added CONFIGURE_GROUP build-time option.
2988 7. Some older OS have a limit of 256 on the maximum number of file
2989 descriptors. Exim was using setrlimit() to set 1000 as a large value
2990 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2991 systems. I've change it so that if it can't get 1000, it tries for 256.
2993 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2994 was an oversight, and furthermore, ever since the addition of extra
2995 controls (e.g. 4.43/32), the checks on when to allow different forms of
2996 "control" were broken. There should now be diagnostics for all cases when a
2997 control that does not make sense is encountered.
2999 9. Added the /retain_sender option to "control=submission".
3001 10. $recipients is now available in the predata ACL (oversight).
3003 11. Tidy the search cache before the fork to do a delivery from a message
3004 received from the command line. Otherwise the child will trigger a lookup
3005 failure and thereby defer the delivery if it tries to use (for example) a
3006 cached ldap connection that the parent has called unbind on.
3008 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3009 of $address_data from the recipient verification was clobbered by the
3010 sender verification.
3012 13. The value of address_data from a sender verification is now available in
3013 $sender_address_data in subsequent conditions in the ACL statement.
3015 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3017 15. Added a new option "connect=<time>" to callout options, to set a different
3020 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3021 was its contents. (It was OK if the option was not defined at all.)
3023 17. A "Completed" log line is now written for messages that are removed from
3024 the spool by the -Mrm option.
3026 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3027 information about exactly what failed.
3029 19. Added -dd to debug only the daemon process.
3031 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3032 handles timeouts, both on the server side and network timeouts. Renamed the
3033 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3035 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3036 It is now set to "smtps".
3038 22. $host_address is now set to the target address during the checking of
3039 ignore_target_hosts.
3041 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3042 being passed; this would have caused $sender_host_name to have been used if
3043 matching the list had actually called for a host name (not very likely,
3044 since this list is usually IP addresses). A host name is now passed as
3047 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3048 code that specifies a non-restarting handler (typically sigaction() in
3049 modern systems) in an attempt to fix a rare and obscure crash bug.
3051 25. Narrowed the window for a race in the daemon that could cause it to ignore
3052 SIGCHLD signals. This is not a major problem, because they are used only to
3053 wake it up if nothing else does.
3055 26. A malformed maildirsize file could cause Exim to calculate negative values
3056 for the mailbox size or file count. Odd effects could occur as a result.
3057 The maildirsize information is now recalculated if the size or filecount
3060 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3061 support for a long time. Removed HAVE_SYS_VFS_H.
3063 28. Installed the latest version of exipick from John Jetmore.
3065 29. In an address list, if the pattern was not a regular expression, an empty
3066 subject address (from a bounce message) matched only if the pattern was an
3067 empty string. Non-empty patterns were not even tested. This was the wrong
3068 because it is perfectly reasonable to use an empty address as part of a
3069 database query. An empty address is now tested by patterns that are
3070 lookups. However, all the other forms of pattern expect the subject to
3071 contain a local part and a domain, and therefore, for them, an empty
3072 address still always fails if the pattern is not itself empty.
3074 30. Exim went into a mad DNS loop when attempting to do a callout where the
3075 host was specified on an smtp transport, and looking it up yielded more
3076 than one IP address.
3078 31. Re-factored the code for checking spool and log partition space into a
3079 function that finds that data and another that does the check. The former
3080 is then used to implement four new variables: $spool_space, $log_space,
3081 $spool_inodes, and $log_inodes.
3083 32. The RFC2047 encoding function was originally intended for short strings
3084 such as real names; it was not keeping to the 75-character limit for
3085 encoded words that the RFC imposes. It now respects the limit, and
3086 generates multiple encoded words if necessary. To be on the safe side, I
3087 have increased the buffer size for the ${rfc2047: expansion operator from
3090 33. It is now permitted to omit both strings after an "if" condition; if the
3091 condition is true, the result is "true". As before, when the second string
3092 is omitted, a false condition yields an empty string. This makes it less
3093 cumbersome to write custom ACL and router conditions.
3095 34. Failure to deliver a bounce message always caused it to be frozen, even if
3096 there was an errors_to setting on the router. The errors_to setting is now
3099 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3100 canonical form (fully expanded) before being placed in
3101 $sender_host_address.
3103 36. The table in the code that translates DNS record types into text (T_A to
3104 "A" for instance) was missing entries for NS and CNAME. It is just possible
3105 that this could have caused confusion if both these types were looked up
3106 for the same domain, because the text type is used as part of Exim's
3107 per-process caching. But the chance of anyone hitting this buglet seems
3110 37. The dnsdb lookup has been extended in a number of ways.
3112 (1) There is a new type, "zns", which walks up the domain tree until it
3113 finds some nameserver records. It should be used with care.
3115 (2) There is a new type, "mxh", which is like "mx" except that it returns
3116 just the host names, not the priorities.
3118 (3) It is now possible to give a list of domains (or IP addresses) to be
3119 looked up. The behaviour when one of the lookups defers can be
3120 controlled by a keyword.
3122 (4) It is now possible to specify the separator character for use when
3123 multiple records are returned.
3125 38. The dnslists ACL condition has been extended: it is now possible to supply
3126 a list of IP addresses and/or domains to be looked up in a particular DNS
3129 39. Added log_selector=+queue_time_overall.
3131 40. When running the queue in the test harness, wait just a tad after forking a
3132 delivery process, to get repeatability of debugging output.
3134 41. Include certificate and key file names in error message when GnuTLS fails
3135 to set them up, because the GnuTLS error message doesn't include the name
3136 of the failing file when there is a problem reading it.
3138 42. Allow both -bf and -bF in the same test run.
3140 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3142 44. The "Exiscan patch" is now merged into the mainline Exim source.
3144 45. Sometimes the final signoff response after QUIT could fail to get
3145 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3146 < 0 before doing a fflush(). This bug looks as though it goes back to the
3147 introduction of TLS in release 3.20, but "sometimes" must have been rare
3148 because the tests only now provoked it.
3150 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3151 (this can affect the format of dates).
3153 47. exim_tidydb, when checking for the continued existence of a message for
3154 which it has found a message-specific retry record, was not finding
3155 messages that were in split spool directories. Consequently, it was
3156 deleting retry records that should have stayed in existence.
3158 48. Steve fixed some bugs in eximstats.
3160 49. The SPA authentication driver was not abandoning authentication and moving
3161 on to the next authenticator when an expansion was forced to fail,
3162 contradicting the general specification for all authenticators. Instead it
3163 was generating a temporary error. It now behaves as specified.
3165 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3166 (the order specifies the preference for clients). The order is now AES256,
3167 AES128, 3DES, ARCFOUR128.
3169 51. Small patch to Sieve code - explicitly set From: when generating an
3172 52. Exim crashed if a remote delivery caused a very long error message to be
3173 recorded - for instance if somebody sent an entire SpamAssassin report back
3174 as a large number of 550 error lines. This bug was coincidentally fixed by
3175 increasing the size of one of Exim's internal buffers (big_buffer) that
3176 happened as part of the Exiscan merge. However, to be on the safe side, I
3177 have made the code more robust (and fixed the comments that describe what
3180 53. Now that there can be additional text after "Completed" in log lines (if
3181 the queue_time_overall log selector is set), a one-byte patch to exigrep
3182 was needed to allow it to recognize "Completed" as not the last thing in
3185 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3186 patch that reportedly fixes this has been added. I am not expert enough to
3187 create a test for it. This is what the patch creator wrote:
3189 "I found a little strange behaviour of ldap code when working with
3190 Windows 2003 AD Domain, where users was placed in more than one
3191 Organization Units. When I tried to give exim partial DN, the exit code
3192 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3193 But simultaneously result of request was absolutely normal ldap result,
3194 so I produce this patch..."
3196 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3197 REFERENCE, so I have modified the code to exclude the patch when that macro
3200 55. Some experimental protocols are using DNS PTR records for new purposes. The
3201 keys for these records are domain names, not reversed IP addresses. The
3202 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3203 leaves it alone. Component reversal etc. now happens only for IP addresses.
3206 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3208 57. Double the size of the debug message buffer (to 2048) so that more of very
3209 long debug lines gets shown.
3211 58. The exicyclog utility now does better if the number of log files to keep
3212 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3214 59. Two changes related to the smtp_active_hostname option:
3216 (1) $smtp_active_hostname is now available as a variable.
3217 (2) The default for smtp_banner uses $smtp_active_hostname instead
3218 of $primary_hostname.
3220 60. The host_aton() function is supposed to be passed a string that is known
3221 to be a valid IP address. However, in the case of IPv6 addresses, it was
3222 not checking this. This is a hostage to fortune. Exim now panics and dies
3223 if the condition is not met. A case was found where this could be provoked
3224 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3225 components; fortuitously, this particular loophole had already been fixed
3226 by change 4.50/55 above.
3228 If there are any other similar loopholes, the new check in host_aton()
3229 itself should stop them being exploited. The report I received stated that
3230 data on the command line could provoke the exploit when Exim was running as
3231 exim, but did not say which command line option was involved. All I could
3232 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3233 running as the user.
3236 61. There was a buffer overflow vulnerability in the SPA authentication code
3237 (which came originally from the Samba project). I have added a test to the
3238 spa_base64_to_bits() function which I hope fixes it.
3241 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3242 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3244 63. The daemon start-up calls getloadavg() while still root for those OS that
3245 need the first call to be done as root, but it missed one case: when
3246 deliver_queue_load_max is set with deliver_drop_privilege. This is
3247 necessary for the benefit of the queue runner, because there is no re-exec
3248 when deliver_drop_privilege is set.
3250 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3251 This has been fixed.
3253 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3254 was in use, was not putting the data itself into the right store pool;
3255 consequently, it could be overwritten for a subsequent message in the same
3256 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3259 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3261 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3262 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3263 the function were treating the return as a boolean value, which happened to
3264 work because 0=false and not-0=true, but is not correct code.
3266 68. The host_aton() function was not handling scoped IPv6 addresses (those
3267 with, for example, "%eth0" on the end) correctly.
3269 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3270 negated items (that is, ~something) in unsigned ints. Some compilers
3271 apparently mutter when there is no cast.
3273 70. If an address verification called from an ACL failed, and did not produce a
3274 user-specific message (i.e. there was only a "system" message), nothing was
3275 put in $acl_verify_message. In this situation, it now puts the system
3278 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3279 session; change 4.31/43 added the unwanted input to the log line - except
3280 that it did not do this in the start of session case. It now does.
3282 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3283 This is wrong and can cause the other end to generate a synchronization
3284 error if it is another Exim or anything else that does the synchronization
3285 check. A QUIT command is no longer sent after a timeout.
3287 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3288 during host lookups.
3290 74. The defer_ok option of callout verification was not working if it was used
3291 when verifying addresses in header lines, that is, for this case:
3293 verify = header_sender/callout=defer_ok
3295 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3296 those file descriptors could be used for SMTP connections. If anything
3297 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3298 could be sent to the SMTP client, causing chaos. The daemon now opens
3299 stdin, stdout, and stderr to /dev/null when it puts itself into the
3302 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3303 log by default. The user can override this with suitable Perl magic.
3305 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3306 the log message when discard triggers, was not working for the DATA ACL or
3307 for the non-SMTP ACL.
3309 78. Error message wording change in sieve.c.
3311 79. If smtp_accept_max_per_host was set, the number of connections could be
3312 restricted to fewer than expected, because the daemon was trying to set up
3313 a new connection before checking whether the processes handling previous
3314 connections had finished. The check for completed processes is now done
3315 earlier. On busy systems, this bug wouldn't be noticed because something
3316 else would have woken the daemon, and it would have reaped the completed
3319 80. If a message was submitted locally by a user whose login name contained one
3320 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3321 It caused a spool format error. I have fixed the spool reading code. A
3322 related problem was that the "from" clause in the Received: line became
3323 illegal because of the space(s). It is now covered by ${quote_local_part.
3325 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3328 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3330 83. Patch from Sieve maintainer to fix unterminated string problem in
3331 "vacation" handling.
3333 84. Some minor changes to the Linux configuration files to help with other
3334 OS variants using glibc.
3336 85. One more patch for Sieve to update vacation handling to latest spec.
3339 ----------------------------------------------------
3340 See the note above about the 4.44 and 4.50 releases.
3341 ----------------------------------------------------
3347 1. Change 4.43/35 introduced a bug that caused file counts to be
3348 incorrectly computed when quota_filecount was set in an appendfile
3351 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3352 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3355 3. Give more explanation in the error message when the command for a transport
3356 filter fails to execute.
3358 4. There are several places where Exim runs a non-Exim command in a
3359 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3360 was being done only for the command run by the queryprogram router. It is
3361 now done for all such subprocesses. The other cases are: ${run, transport
3362 filters, and the commands run by the lmtp and pipe transports.
3364 5. Some older OS have a limit of 256 on the maximum number of file
3365 descriptors. Exim was using setrlimit() to set 1000 as a large value
3366 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3367 systems. I've change it so that if it can't get 1000, it tries for 256.
3369 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3370 was an oversight, and furthermore, ever since the addition of extra
3371 controls (e.g. 4.43/32), the checks on when to allow different forms of
3372 "control" were broken. There should now be diagnostics for all cases when a
3373 control that does not make sense is encountered.
3375 7. $recipients is now available in the predata ACL (oversight).
3377 8. Tidy the search cache before the fork to do a delivery from a message
3378 received from the command line. Otherwise the child will trigger a lookup
3379 failure and thereby defer the delivery if it tries to use (for example) a
3380 cached ldap connection that the parent has called unbind on.
3382 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3383 of $address_data from the recipient verification was clobbered by the
3384 sender verification.
3386 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3387 was its contents. (It was OK if the option was not defined at all.)
3389 11. A "Completed" log line is now written for messages that are removed from
3390 the spool by the -Mrm option.
3392 12. $host_address is now set to the target address during the checking of
3393 ignore_target_hosts.
3395 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3396 being passed; this would have caused $sender_host_name to have been used if
3397 matching the list had actually called for a host name (not very likely,
3398 since this list is usually IP addresses). A host name is now passed as
3401 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3402 code that specifies a non-restarting handler (typically sigaction() in
3403 modern systems) in an attempt to fix a rare and obscure crash bug.
3405 15. Narrowed the window for a race in the daemon that could cause it to ignore
3406 SIGCHLD signals. This is not a major problem, because they are used only to
3407 wake it up if nothing else does.
3409 16. A malformed maildirsize file could cause Exim to calculate negative values
3410 for the mailbox size or file count. Odd effects could occur as a result.
3411 The maildirsize information is now recalculated if the size or filecount
3414 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3415 support for a long time. Removed HAVE_SYS_VFS_H.
3417 18. Updated exipick to current release from John Jetmore.
3419 19. Allow an empty sender to be matched against a lookup in an address list.
3420 Previously the only cases considered were a regular expression, or an
3423 20. Exim went into a mad DNS lookup loop when doing a callout where the
3424 host was specified on the transport, if the DNS lookup yielded more than
3427 21. The RFC2047 encoding function was originally intended for short strings
3428 such as real names; it was not keeping to the 75-character limit for
3429 encoded words that the RFC imposes. It now respects the limit, and
3430 generates multiple encoded words if necessary. To be on the safe side, I
3431 have increased the buffer size for the ${rfc2047: expansion operator from
3434 22. Failure to deliver a bounce message always caused it to be frozen, even if
3435 there was an errors_to setting on the router. The errors_to setting is now
3438 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3439 canonical form (fully expanded) before being placed in
3440 $sender_host_address.
3442 24. Updated eximstats to version 1.33
3444 25. Include certificate and key file names in error message when GnuTLS fails
3445 to set them up, because the GnuTLS error message doesn't include the name
3446 of the failing file when there is a problem reading it.
3448 26. Expand error message when OpenSSL has problems setting up cert/key files.
3451 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3452 (this can affect the format of dates).
3454 28. exim_tidydb, when checking for the continued existence of a message for
3455 which it has found a message-specific retry record, was not finding
3456 messages that were in split spool directories. Consequently, it was
3457 deleting retry records that should have stayed in existence.
3459 29. eximstats updated to version 1.35
3460 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3461 1.35 - bugfix such that pie charts by volume are generated correctly
3463 30. The SPA authentication driver was not abandoning authentication and moving
3464 on to the next authenticator when an expansion was forced to fail,
3465 contradicting the general specification for all authenticators. Instead it
3466 was generating a temporary error. It now behaves as specified.
3468 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3469 (the order specifies the preference for clients). The order is now AES256,
3470 AES128, 3DES, ARCFOUR128.
3472 31. Small patch to Sieve code - explicitly set From: when generating an
3475 32. Exim crashed if a remote delivery caused a very long error message to be
3476 recorded - for instance if somebody sent an entire SpamAssassin report back
3477 as a large number of 550 error lines. This bug was coincidentally fixed by
3478 increasing the size of one of Exim's internal buffers (big_buffer) that
3479 happened as part of the Exiscan merge. However, to be on the safe side, I
3480 have made the code more robust (and fixed the comments that describe what
3483 33. Some experimental protocols are using DNS PTR records for new purposes. The
3484 keys for these records are domain names, not reversed IP addresses. The
3485 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3486 leaves it alone. Component reversal etc. now happens only for IP addresses.
3489 34. The host_aton() function is supposed to be passed a string that is known
3490 to be a valid IP address. However, in the case of IPv6 addresses, it was
3491 not checking this. This is a hostage to fortune. Exim now panics and dies
3492 if the condition is not met. A case was found where this could be provoked
3493 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3494 components; fortuitously, this particular loophole had already been fixed
3495 by change 4.50/55 or 4.44/33 above.
3497 If there are any other similar loopholes, the new check in host_aton()
3498 itself should stop them being exploited. The report I received stated that
3499 data on the command line could provoke the exploit when Exim was running as
3500 exim, but did not say which command line option was involved. All I could
3501 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3502 running as the user.
3505 35. There was a buffer overflow vulnerability in the SPA authentication code
3506 (which came originally from the Samba project). I have added a test to the
3507 spa_base64_to_bits() function which I hope fixes it.
3510 36. The daemon start-up calls getloadavg() while still root for those OS that
3511 need the first call to be done as root, but it missed one case: when
3512 deliver_queue_load_max is set with deliver_drop_privilege. This is
3513 necessary for the benefit of the queue runner, because there is no re-exec
3514 when deliver_drop_privilege is set.
3516 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3517 was in use, was not putting the data itself into the right store pool;
3518 consequently, it could be overwritten for a subsequent message in the same
3519 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3522 38. Sometimes the final signoff response after QUIT could fail to get
3523 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3524 < 0 before doing a fflush(). This bug looks as though it goes back to the
3525 introduction of TLS in release 3.20, but "sometimes" must have been rare
3526 because the tests only now provoked it.
3532 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3533 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3534 Now it returns an integer. A number of calls were still expecting a T/F
3535 return. Fortuitously, in all cases, the tests worked in OK situations,
3536 which is the norm. However, things would have gone wrong on any write
3537 failures on the smtp file descriptor. This function is used when sending
3538 messages over SMTP and also when doing verify callouts.
3540 2. When Exim is called to do synchronous delivery of a locally submitted
3541 message (the -odf or -odi options), it no longer closes stderr before doing
3544 3. Implemented the mua_wrapper option.
3546 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3548 5. Implemented the functions header_remove(), header_testname(),
3549 header_add_at_position(), and receive_remove_recipient(), and exported them
3552 6. If an ACL "warn" statement specified the addition of headers, Exim already
3553 inserted X-ACL-Warn: at the start if there was no header name. However, it
3554 was not making this test for the second and subsequent header lines if
3555 there were newlines in the string. This meant that an invalid header could
3556 be inserted if Exim was badly configured.
3558 7. Allow an ACL "warn" statement to add header lines at the start or after all
3559 the Received: headers, as well as at the end.
3561 8. Added the rcpt_4xx retry error code.
3563 9. Added postmaster_mailfrom=xxx to callout verification option.
3565 10. Added mailfrom=xxxx to the callout verification option, for verify=
3568 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3569 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3571 12. Inserted some casts to stop certain compilers warning when using pointer
3572 differences as field lengths or precisions in printf-type calls (mostly
3573 affecting debugging statements).
3575 13. Added optional readline() support for -be (dynamically loaded).
3577 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3578 same clock tick as a message's arrival, so that its received time was the
3579 same as the "first fail" time on the retry record, and that message
3580 remained on the queue past the ultimate address timeout, every queue runner
3581 would try a delivery (because it was past the ultimate address timeout) but
3582 after another failure, the ultimate address timeout, which should have then
3583 bounced the address, did not kick in. This was a "< instead of <=" error;
3584 in most cases the first failure would have been in the next clock tick
3585 after the received time, and all would be well.
3587 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3588 being recognized when the domain list was tested by the match_domain
3589 condition in an expansion string.
3591 16. Added the ${str2b64: operator.
3593 17. Exim was always calling setrlimit() to set a large limit for the number of
3594 processes, without checking whether the existing limit was already
3595 adequate. (It did check for the limit on file descriptors.) Furthermore,
3596 errors from getrlimit() and setrlimit() were being ignored. Now they are
3597 logged to the main and panic logs, but Exim does carry on, to try to do its
3598 job under whatever limits there are.
3600 18. Imported PCRE 5.0.
3602 19. Trivial typo in log message " temporarily refused connection" (the leading
3605 20. If the log selector return_path_on_delivery was set and an address was
3606 redirected to /dev/null, the delivery process crashed because it assumed
3607 that a return path would always be set for a "successful" delivery. In this
3608 case, the whole delivery is bypassed as an optimization, and therefore no
3611 21. Internal re-arrangement: the function for sending a challenge and reading
3612 a response while authentication was assuming a zero-terminated challenge
3613 string. It's now changed to take a pointer and a length, to allow for
3614 binary data in such strings.
3616 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3618 23. Exim was not respecting finduser_retries when seeking the login of the
3619 uid under which it was called; it was always trying 10 times. (The default
3620 setting of finduser_retries is zero.) Also, it was sleeping after the final
3621 failure, which is pointless.
3623 24. Implemented tls_on_connect_ports.
3625 25. Implemented acl_smtp_predata.
3627 26. If the domain in control=submission is set empty, Exim assumes that the
3628 authenticated id is a complete email address when it generates From: or
3629 Sender: header lines.
3631 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3632 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3633 chown and chgrp in /bin and hostname in /usr/bin.
3635 28. Exim was keeping the "process log" file open after each use, just as it
3636 does for the main log. This opens the possibility of it remaining open for
3637 long periods when the USR1 signal hits a daemon. Occasional processlog
3638 errors were reported, that could have been caused by this. Anyway, it seems
3639 much more sensible not to leave this file open at all, so that is what now
3642 29. The long-running daemon process does not normally write to the log once it
3643 has entered its main loop, and it closes the log before doing so. This is
3644 so that log files can straightforwardly be renamed and moved. However,
3645 there are a couple of unusual error situations where the daemon does write
3646 log entries, and I had neglected to close the log afterwards.
3648 30. The text of an SMTP error response that was received during a remote
3649 delivery was being truncated at 512 bytes. This is too short for some of
3650 the long messages that one sometimes sees. I've increased the limit to
3653 31. It is now possible to make retry rules that apply only when a message has a
3654 specific sender, in particular, an empty sender.
3656 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3657 it possible to be selective about when SMTP synchronization is enforced.
3659 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3661 32. Implemented hosts_connection_nolog.
3663 33. Added an ACL for QUIT.
3665 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3668 35. Added mailbox_size and mailbox_filecount to appendfile.
3670 36. Added control = no_multiline_responses to ACLs.
3672 37. There was a bug in the logic of the code that waits for the clock to tick
3673 in the case where the clock went backwards by a substantial amount such
3674 that the microsecond fraction of "now" was more than the microsecond
3675 fraction of "then" (but the whole seconds number was less).
3677 38. Added support for the libradius Radius client library this is found on
3678 FreeBSD (previously only the radiusclient library was supported).
3684 1. When certain lookups returned multiple values in the form name=value, the
3685 quoting of the values was not always being done properly. Specifically:
3686 (a) If the value started with a double quote, but contained no whitespace,
3688 (b) If the value contained whitespace other than a space character (i.e.
3689 tabs or newlines or carriage returns) it was not quoted.
3690 This fix has been applied to the mysql and pgsql lookups by writing a
3691 separate quoting function and calling it from the lookup code. The fix
3692 should probably also be applied to nisplus, ibase and oracle lookups, but
3693 since I cannot test any of those, I have not disturbed their existing code.
3695 2. A hit in the callout cache for a specific address caused a log line with no
3696 reason for rejecting RCPT. Now it says "Previous (cached) callout
3697 verification failure".
3699 3. There was an off-by-one bug in the queryprogram router. An over-long
3700 return line was truncated at 256 instead of 255 characters, thereby
3701 overflowing its buffer with the terminating zero. As well as fixing this, I
3702 have increased the buffer size to 1024 (and made a note to document this).
3704 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3705 when Exim is waiting for an SMTP response from a remote server, Exim
3706 restarts its select() call on the socket, thereby resetting its timeout.
3707 This is not a problem when such interrupts are rare. Somebody set up a cron
3708 job to run exiwhat every 2 minutes, which is less than the normal select()
3709 timeout (5 or 10 minutes). This meant that the select() timeout never
3710 kicked in because it was always reset. I have fixed this by comparing the
3711 time when an interrupt arrives with the time at the start of the first call
3712 to select(). If more time than the timeout has elapsed, the interrupt is
3713 treated as a timeout.
3715 5. Some internal re-factoring in preparation for the addition of Sieve
3716 extensions (by MH). In particular, the "personal" test is moved to a
3717 separate function, and given an option for scanning Cc: and Bcc: (which is
3718 not set for Exim filters).
3720 6. When Exim created an email address using the login of the caller as the
3721 local part (e.g. when creating a From: or Sender: header line), it was not
3722 quoting the local part when it contained special characters such as @.
3724 7. Installed new OpenBSD configuration files.
3726 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3727 try to make them clearer.
3729 9. Callout options, other than the timeout value, were being ignored when
3730 verifying sender addresses in header lines. For example, when using
3732 verify = header_sender/callout=no_cache
3734 the cache was (incorrectly) being used.
3736 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3737 only the Cygwin environment.
3739 11. When return_path_on_delivery was set as a log selector, if different remote
3740 addresses in the same message used different return paths and parallel
3741 remote delivery occurred, the wrong values would sometimes be logged.
3742 (Whenever a remote delivery process finished, the return path value from
3743 the most recently started remote delivery process was logged.)
3745 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3746 lines when AUTH and/or TLS are in use. This is the "received protocol"
3747 field. Exim used to use "asmtp" for authenticated SMTP, without any
3748 indication (in the protocol name) for TLS use. Now it follows the RFC and
3749 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3750 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3751 names appear in log lines as well as in Received: header lines.
3753 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3754 extensions, and comparison tests, and to fix some bugs.
3756 14. Changes to the "personal" filter test:
3758 (1) The test was buggy in that it was just doing the equivalent of
3759 "contains" tests on header lines. For example, if a user's address was
3760 anne@some.where, the "personal" test would incorrectly be true for
3762 To: susanne@some.where
3764 This test is now done by extracting each address from the header in turn,
3765 and checking the entire address. Other tests that are part of "personal"
3766 are now done using regular expressions (for example, to check local parts
3767 of addresses in From: header lines).
3769 (2) The list of non-personal local parts in From: addresses has been
3770 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3771 taken from the Sieve specification recommendations.
3773 (3) If the message contains any header line starting with "List-" it is
3774 treated as non-personal.
3776 (4) The test for "circular" in the Subject: header line has been removed
3777 because it now seems ill-conceived.
3779 15. Minor typos in src/EDITME comments corrected.
3781 16. Installed latest exipick from John Jetmore.
3783 17. If headers_add on a router specified a text string that was too long for
3784 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3785 of string_sprintf() is now avoided.
3787 18. $message_body_size was not set (it was always zero) when running the DATA
3788 ACL and the local_scan() function.
3790 19. For the "mail" command in an Exim filter, no default was being set for
3791 the once_repeat time, causing a random time value to be used if "once" was
3792 specified. (If the value happened to be <= 0, no repeat happened.) The
3793 default is now 0s, meaning "never repeat". The "vacation" command was OK
3794 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3795 (I found it when inspecting the code).
3797 20. There is now an overall timeout for performing a callout verification. It
3798 defaults to 4 times the callout timeout, which applies to individual SMTP
3799 commands during the callout. The overall timeout applies when there is more
3800 than one host that can be tried. The timeout is checked before trying the
3801 next host. This prevents very long delays if there are a large number of
3802 hosts and all are timing out (e.g. when the network connections are timing
3803 out). The value of the overall timeout can be changed by specifying an
3804 additional sub-option for "callout", called "maxwait". For example:
3806 verify = sender/callout=5s,maxwait=20s
3808 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3809 to the end before writing, but this should make it even safer).
3811 22. Exim was forgetting that it had advertised PIPELINING for the second and
3812 subsequent messages on an SMTP connection. It was also not resetting its
3813 memory on STARTTLS and an internal HELO.
3815 23. When Exim logs an SMTP synchronization error within a session, it now
3816 records whether PIPELINING has been advertised or not.
3818 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3819 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3820 rather than long int.
3822 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3824 26. Added the never_mail option to autoreply.
3830 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3831 crash if the getsockname() call failed; this can happen if a connection is
3832 closed very soon after it is established. The problem was simply in the
3833 order in which certain operations were done, causing Exim to try to write
3834 to the SMTP stream before it had set up the file descriptor. The bug has
3835 been fixed by making things happen in the correct order.
3841 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3842 before the connection was closed, thus losing the rejection response.
3844 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3845 some early Solaris releases, but causes trouble in current releases where
3846 socklen_t is defined.
3848 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3851 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3854 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3855 that releasing the top of it at the end releases what was used for sub-
3856 expansions (unless the block got too big). However, discard this block if
3857 the first thing is a variable or header, so that we can use its block when
3858 it is dynamic (useful for very large $message_headers, for example).
3860 6. Lookups now cache *every* query, not just the most recent. A new, separate
3861 store pool is used for this. It can be recovered when all lookup caches are
3862 flushed. Lookups now release memory at the end of their result strings.
3863 This has involved some general refactoring of the lookup sources.
3865 7. Some code has been added to the store_xxx() functions to reduce the amount
3866 of flapping under certain conditions.
3868 8. log_incoming_interface used to affect only the <= reception log lines. Now
3869 it causes the local interface and port to be added to several more SMTP log
3870 lines, for example "SMTP connection from", and rejection lines.
3872 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3874 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3876 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3877 could be overwritten at the end of the current message (or the start of a
3878 new message if it was set in a HELO ACL). The value is now preserved for
3879 the duration of the SMTP connection.
3881 12. If a transport had a headers_rewrite setting, and a matching header line
3882 contained an unqualified address, that address was qualified, even if it
3883 did not match any rewriting rules. The underlying bug was that the values
3884 of the flags that permit the existence of unqualified sender and recipient
3885 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3886 non-local messages, and by -bnq for local messages) were not being
3887 preserved with the message after it was received.
3889 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3890 "next input=" as part of the text comprising the host identity instead of
3891 the correct text. The code was using the same buffer for two different
3892 strings. However, depending on which order the printing function evaluated
3893 its arguments, the bug did not always show up. Under Linux, for example, my
3894 test suite worked just fine.
3896 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3897 doesn't work with some older versions of Perl. It has been changed to "my",
3898 which in any case is probably the better facility to use.
3900 15. A really picky compiler found some instances of statements for creating
3901 error messages that either had too many or two few arguments for the format
3904 16. The size of the buffer for calls to the DNS resolver has been increased
3905 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3906 for addresses that have a lot of PTR records. This alleviates a problem; it
3907 does not fully solve it.
3909 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3910 buffer now truncates the list and logs the incident, which is the same
3911 action as happens when Exim is looking up a host name and its aliases.
3912 Previously in this situation something unpredictable would happen;
3913 sometimes it was "internal error: store_reset failed".
3915 18. If a server dropped the connection unexpectedly when an Exim client was
3916 using GnuTLS and trying to read a response, the client delivery process
3917 crashed while trying to generate an error log message.
3919 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3920 string, for example:
3922 warn message = H1: something\nH2: something
3924 the text was added as a single header line from Exim's point of view
3925 though it ended up OK in the delivered message. However, searching for the
3926 second and subsequent header lines using $h_h2: did not work. This has been
3927 fixed. Similarly, if a system filter added multiple headers in this way,
3928 the routers could not see them.
3930 20. Expanded the error message when iplsearch is called with an invalid key to
3931 suggest using net-iplsearch in a host list.
3933 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3934 ACLs are no longer actually imposed (and a message to that effect is
3937 22. If a "gecos" field in a passwd entry contained escaped characters, in
3938 particular, if it contained a \" sequence, Exim got it wrong when building
3939 a From: or a Sender: header from that name. A second bug also caused
3940 incorrect handling when an unquoted " was present following a character
3941 that needed quoting.
3943 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3944 was not being matched caselessly.
3946 24. Arranged for all hyphens in the exim.8 source to be escaped with
3949 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3950 properly. Recipient callout cache records were still being keyed to include
3951 the sender, even when use_sender was set false. This led to far more
3952 callouts that were necessary. The sender is no longer included in the key
3953 when use_sender is false.
3955 26. Added "control = submission" modifier to ACLs.
3957 27. Added the ${base62d: operator to decode base 62 numbers.
3959 28. dnsdb lookups can now access SRV records.
3961 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3962 the configuration file.
3964 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3965 (-v) mode. This makes the output for a verbose queue run more intelligible.
3967 31. Added a use_postmaster feature to recipient callouts.
3969 32. Added the $body_zerocount variable, containing the number of binary zero
3970 bytes in the message body.
3972 33. The time of last modification of the "new" subdirectory is now used as the
3973 "mailbox time last read" when there is a quota error for a maildir
3976 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3978 35. Added +ignore_unknown as a special item in host lists.
3980 36. Code for decoding IPv6 addresses in host lists is now included, even if
3981 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3982 address was recognized as an IP address, but was then not correctly decoded
3983 into binary, causing unexpected and incorrect effects when compared with
3990 1. Very minor rewording of debugging text in manualroute to say "list of
3991 hosts" instead of "hostlist".
3993 2. If verify=header_syntax was set, and a header line with an unqualified
3994 address (no domain) and a large number of spaces between the end of the
3995 name and the colon was received, the reception process suffered a buffer
3996 overflow, and (when I tested it) crashed. This was caused by some obsolete
3997 code that should have been removed. The fix is to remove it!
3999 3. When running in the test harness, delay a bit after writing a bounce
4000 message to get a bit more predictability in the log output.
4002 4. Added a call to search_tidyup() just before forking a reception process. In
4003 theory, someone could use a lookup in the expansion of smtp_accept_max_
4004 per_host which, without the tidyup, could leave open a database connection.
4006 5. Added the variables $recipient_data and $sender_data which get set from a
4007 lookup success in an ACL "recipients" or "senders" condition, or a router
4008 "senders" option, similar to $domain_data and $local_part_data.
4010 6. Moved the writing of debug_print from before to after the "senders" test
4013 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4014 problems for message scanning, either using a data ACL, or using
4015 local_scan() because the Received: header was not generated till after they
4016 were called (in order to set the time as the time of reception completion).
4017 I have revised the way this works. The header is now generated after the
4018 body is received, but before the ACL or local_scan() are called. After they
4019 are run, the timestamp in the header is updated.
4025 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4026 before starting a queue runner without re-exec. This happened only when
4027 deliver_drop_privilege was set or when the Exim user was set to root. The
4028 effect of the bug was that timeouts during subsequent deliveries caused
4029 crashes instead of being properly handled. The handler is now left at its
4030 default (and expected) setting.
4032 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4033 message, again when deliver_drop_privilege is set or Exim is run as root.
4034 The bug described in (1) was not present in this case, but the tidying up
4035 of the other signals was missing. I have made the two cases consistent.
4037 3. The ignore_target_hosts setting on a manualroute router was being ignored
4038 for hosts that were looked up using the /MX notation.
4040 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4043 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4044 operated on the sender address. After changing the $sender_address to <>
4045 for the sender address verify, Exim was re-instated it as the original
4046 (before rewriting) address, but remembering that it had rewritten it, so it
4047 wasn't rewriting it again. This bug also had the effect of breaking the
4048 sender address verification caching when the sender address was rewritten.
4050 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4051 This has been changed so that if the ip literal address matches
4052 ignore_target_hosts, the router declines.
4054 7. Added expansion conditions match_domain, match_address, and match_local_
4055 part (NOT match_host).
4057 8. The placeholder for the Received: header didn't have a length field set.
4059 9. Added code to Exim itself and to exim_lock to test for a specific race
4060 condition that could lead to file corruption when using MBX delivery. The
4061 issue is with the lockfile that is created in /tmp. If this file is removed
4062 after a process has opened it but before that process has acquired a lock,
4063 there is the potential for a second process to recreate the file and also
4064 acquire a lock. This could lead to two Exim processes writing to the file
4065 at the same time. The added code performs the same test as UW imapd; it
4066 checks after acquiring the lock that its file descriptor still refers to
4067 the same named file.
4069 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4070 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4073 11. Added the smtp_active_hostname option. If used, this will typically be made
4074 to depend on the incoming interface address. Because $interface_address is
4075 not set up until the daemon has forked a reception process, error responses
4076 that can happen earlier (such as "too many connections") no longer contain
4079 12. If an expansion in a condition on a "warn" statement fails because a lookup
4080 defers, the "warn" statement is abandoned, and the next ACL statement is
4081 processed. Previously this caused the whole ACL to be aborted.
4083 13. Added the iplsearch lookup type.
4085 14. Added ident_timeout as a log selector.
4087 15. Added tls_certificate_verified as a log selector.
4089 16. Added a global option tls_require_ciphers (compare the smtp transport
4090 option of the same name). This controls incoming TLS connections.
4092 17. I finally figured out how to make tls_require_ciphers do a similar thing
4093 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4094 before starting the TLS session.
4096 18. Tabs are now shown as \t in -bP output.
4098 19. If the log selector return_path_on_delivery was set, Exim crashed when
4099 bouncing a message because it had too many Received: header lines.
4101 20. If two routers both had headers_remove settings, and the first one included
4102 a superfluous trailing colon, the final name in the first list and the
4103 first name in the second list were incorrectly joined into one item (with a
4104 colon in the middle).
4110 1. Added -C and -D options to the exinext utility, mainly to make it easier
4111 to include in the automated testing, but these could be helpful when
4112 multiple configurations are in use.
4114 2. The exinext utility was not formatting the output nicely when there was
4115 an alternate port involved in the retry record key, nor when there was a
4116 message id as well (for retries that were specific to a specific message
4117 and a specific host). It was also confused by IPv6 addresses, because of
4118 the additional colons they contain. I have fixed the IPv4 problem, and
4119 patched it up to do a reasonable job for IPv6.
4121 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4122 delivery, the log line now contains "pipelined" if PIPELINING was used.
4124 4. An SMTP transport process used to panic and die if the bind() call to set
4125 an explicit outgoing interface failed. This has been changed; it is now
4126 treated in the same way as a connect() failure.
4128 5. A reference to $sender_host_name in the part of a conditional expansion
4129 that was being skipped was still causing a DNS lookup. This no longer
4132 6. The def: expansion condition was not recognizing references to header lines
4133 that used bh_ and bheader_.
4135 7. Added the _cache feature to named lists.
4137 8. The code for checking quota_filecount in the appendfile transport was
4138 allowing one more file than it should have been.
4140 9. For compatibility with Sendmail, the command line option
4148 and sets the incoming protocol and host name (for trusted callers). The
4149 host name and its colon can be omitted when only the protocol is to be set.
4150 Note the Exim already has two private options, -pd and -ps, that refer to
4151 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4152 "s", but I don't think that's a major issue.
4154 10. A number of refactoring changes to the code, none of which should affect
4157 (a) The number of logging options was getting close to filling up the
4158 32-bit word that was used as a bit map. I have split them into two classes:
4159 those that are passed in the argument to log_write(), and those that are
4160 only ever tested independently outside of that function. These are now in
4161 separate 32-bit words, so there is plenty of room for expansion again.
4162 There is no change in the user interface or the logging behaviour.
4164 (b) When building, for example, log lines, the code previously used a
4165 macro that called string_cat() twice, in order to add two strings. This is
4166 not really sufficiently general. Furthermore, there was one instance where
4167 it was actually wrong because one of the argument was used twice, and in
4168 one call a function was used. (As it happened, calling the function twice
4169 did not affect the overall behaviour.) The macro has been replaced by a
4170 function that can join an arbitrary number of extra strings onto a growing
4173 (c) The code for expansion conditions now uses a table and a binary chop
4174 instead of a serial search (which was left over from when there were very
4175 few conditions). Also, it now recognizes conditions like "pam" even when
4176 the relevant support is not compiled in: a suitably worded error message is
4177 given if an attempt is made to use such a condition.
4179 11. Added ${time_interval:xxxxx}.
4181 12. A bug was causing one of the ddress fields not to be passed back correctly
4182 from remote delivery subprocesses. The field in question was not being
4183 subsequently used, so this caused to problems in practice.
4185 13. Added new log selectors queue_time and deliver_time.
4187 14. Might have fixed a bug in maildirsizefile handling that threw up
4188 "unexpected character" debug warnings, and recalculated the data
4189 unnecessarily. In any case, I expanded the warning message to give more
4192 15. Added the message "Restricted characters in address" to the statements in
4193 the default ACL that block characters like @ and % in local parts.
4195 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4196 Three changes have been made:
4198 (a) There was a serious bug; a negative response to MAIL caused the whole
4199 recipient domain to be cached as invalid, thereby blocking all messages
4200 to all local parts at the same domain, from all senders. This bug has
4201 been fixed. The domain is no longer cached after a negative response to
4202 MAIL if the sender used is not empty.
4204 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4207 (c) A new callout option, "use_sender" has been added for people who want
4208 the modified behaviour.
4214 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4217 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4218 indeed breaks things for older releases.
4220 3. Added additional logging to the case where there is a problem reading data
4221 from a filter that is running in a subprocess using a pipe, in order to
4222 try to track down a specific problem.
4224 4. Testing facility fudge: when running in the test harness and attempting
4225 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4226 getting "No route to host". Convert this to a timeout.
4228 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4231 6. Some OS don't have socklen_t but use size_t instead. This affects the
4232 fifth argument of getsockopt() amongst other things. This is now
4233 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4234 can be set for individual OS. I have set it for SunOS5, OSF1, and
4235 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4236 some earlier ones do not.
4238 7. Change 4.30/15 was not doing the test caselessly.
4240 8. The standard form for an IPv6 address literal was being rejected by address
4241 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4242 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4243 this, as well as the form without the "IPv6" on the front (but only when
4244 address literals are enabled, of course).
4246 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4248 10. Exim crashed if a message with an empty sender address specified by -f
4249 encountered a router with an errors_to setting. This could be provoked only
4250 by a command such as
4254 where an empty string was supplied; "<>" did not hit this bug.
4256 11. Installed PCRE release 4.5.
4258 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4259 remained set. It is now erased.
4261 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4262 times from message ids (which are base 36 rather than the normal 62).
4264 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4265 were being counted as actual protocol errors, and logged if the log
4266 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4267 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4268 and DATA following a set of rejected RCPTs do not count as protocol errors.
4269 In other words, Exim assumes they were pipelined, though this may not
4270 actually be the case. Of course, in all cases the client gets an
4271 appropriate error code.
4273 15. If a lookup fails in an ACL condition, a message about the failure may
4274 be available; it is used if testing the ACL cannot continue, because most
4275 such messages specify what the cause of the deferral is. However, some
4276 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4277 that caused an old message to be retained and used if a later statement
4278 caused a defer, replacing the real cause of the deferral.
4280 16. If an IP address had so many PTR records that the DNS lookup buffer
4281 was not large enough to hold them, Exim could crash while trying to process
4282 the truncated data. It now detects and logs this case.
4284 17. Further to 4.21/58, another change has been made: if (and only if) the
4285 first line of a message (the first header line) ends with CRLF, a bare LF
4286 in a subsequent header line has a space inserted after it, so as not to
4287 terminate the header.
4289 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4290 unnecessarily, used atoi() instead of strtol(), and didn't check the
4291 termination when getting file sizes from file names by regex.
4293 19. Completely re-implemented the support for maildirsize files, in the light
4294 of a number of problems with the previous contributed implementation
4295 (4.30/29). In particular:
4297 . If the quota is zero, the maildirsize file is maintained, but no quota is
4300 . If the maildir directory does not exist, it is created before any attempt
4301 to write a maildirsize file.
4303 . The quota value in the file is just a cache; if the quota is changed in
4304 the transport, the new value overrides.
4306 . A regular expression is available for excluding directories from the
4309 20. The autoreply transport checks the characters in options that define the
4310 message's headers; it allows continued headers, but it was checking with
4311 isspace() after an embedded newline instead of explicitly looking for a
4314 21. If all the "regular" hosts to which an address was routed had passed their
4315 expiry times, and had not reached their retry times, the address was
4316 bounced, even if fallback hosts were defined. Now Exim should go on to try
4319 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4320 equivalent code in the SMTP transport. Some hosts send humungous responses
4321 to HELO/EHLO, more than 1024 it seems.
4323 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4324 gives compiler warnings in some environments. I've now done it "properly",
4327 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4328 (because of problems with the built-in one) was declared to return uschar *
4329 instead of char *, causing compiler failure.
4331 25. Fixed a file descriptor leak when processing alias/forward files.
4333 26. Fixed a minor format string issue in dbfn.c.
4335 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4337 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4338 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4341 29. When a sender address is verified, it is cached, to save repeating the test
4342 when there is more than one recipient in a message. However, when the
4343 verification involves a callout, it is possible for different callout
4344 options to be set for different recipients. It is too complicated to keep
4345 track of this in the cache, so now Exim always runs a verification when a
4346 callout is required, relying on the callout cache for the optimization.
4347 The overhead is duplication of the address routing, but this should not be
4350 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4351 to be verified with callout=postmaster, and the main callout worked but the
4352 postmaster check failed, the verification correctly failed. However, if a
4353 subsequent RCPT command asked for sender verification *without* the
4354 postmaster check, incorrect caching caused this verification also to fail,
4357 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4358 it was not caching the DNS options (qualify_single, search_parents) that
4359 were used when the lookup failed. A subsequent lookup with different
4360 options therefore always gave the same answer, though there were cases
4361 where it should not have. (Example: a "domains = !$mx_any" option on a
4362 dnslookup router: the "domains" option is always processed without any
4363 widening, but the router might have qualify_single set.) Now Exim uses the
4364 cached value only when the same options are set.
4366 32. Added John Jetmore's "exipick" utility to the distribution.
4368 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4369 than a timeout (e.g. a certificate is required, and is not provided), an
4370 Exim server now closes the connection immediately. Previously it waited for
4371 the client to close - but if the client is SSL, it seems that they each
4372 wait for each other, leading to a delay before one of them times out.
4374 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4375 maintained 0.8.x compatibility because I don't think many are using it, and
4376 it is clearly obsolete.
4378 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4381 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4382 client certificate was expired. A simple patch fixes this, though I don't
4383 understand the full logic of why the verify callback is called multiple
4386 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4387 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4388 which causes problems with some clients (such as the Certicom SSL Plus
4389 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4390 disables the coutermeasure allowing Eudora to connect."
4392 38. Exim was not checking that a write() to a log file succeeded. This could
4393 lead to Bad Things if a log got too big, in particular if it hit a file
4394 size limit. Exim now panics and dies if it cannot write to a log file, just
4395 as it does if it cannot open a log file.
4397 39. Modified OS/Makefile-Linux so that it now contains
4399 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4401 The two -D definitions ensure that Exim is compiled with large file
4402 support, which makes it possible to handle log files that are bigger than
4405 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4406 instance) a domain was checked against a named list that involved a lookup,
4407 causing $domain_data to be set, then another domain was checked against the
4408 same list, then the first domain was re-checked, the value of $domain_data
4409 after the final check could be wrong. In particular, if the second check
4410 failed, it could be set empty. This bug probably also applied to
4413 41. The strip_trailing_dot option was not being applied to the address given
4414 with the -f command-line option.
4416 42. The code for reading a message's header from the spool was incrementing
4417 $received_count, but never initializing it. This meant that the value was
4418 incorrect (doubled) while delivering a message in the same process in which
4419 it was received. In the most common configuration of Exim, this never
4420 happens - a fresh exec is done - but it can happen when
4421 deliver_drop_privilege is set.
4423 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4424 it now includes up to 150 characters of the unexpected data in the log
4427 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4428 and building data strings. The size of both of these buffers was 10 000
4429 bytes - far larger than anybody would *ever* want, thought I. Needless to
4430 say, somebody hit the limit. I have increased the maximum line length to
4431 20 000 and the maximum data length of concatenated lines to 100 000. I have
4432 also fixed two bugs, because there was no checking on these buffers. Tsk,
4433 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4434 buffer is too small.
4436 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4437 lsearch lookups. Now it does.
4439 46. When parsing a route_list item in a manualroute router, a fixed-length
4440 buffer was used for the list of hosts. I made this 1024 bytes long,
4441 thinking that nobody would ever have a list of hosts that long. Wrong.
4442 Somebody had a whole pile of complicated expansion conditions, and the
4443 string was silently truncated, leading to an expansion error. It turns out
4444 that it is easier to change to an unlimited length (owing to other changes
4445 that have happened since this code was originally written) than to build
4446 structure for giving a limitation error. The length of the item that
4447 expands into the list of hosts is now unlimited.
4449 47. The lsearch lookup could not handle data where the length of text line was
4450 more than 4095 characters. Such lines were truncated, leading to shortened
4451 data being returned. It should now handle lines of any length.
4453 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4454 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4457 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4458 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4459 no longer happens in comment lines. A list of replacements is now placed
4460 at the head of all of the source files, except those whose only change is
4461 to replace PERL_COMMAND in the very first #! line.
4463 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4464 messages on the queue, with a bottom-up merge sort, using code contributed
4465 by Michael Haardt. This should make operations like -bp somewhat faster on
4466 large queues. It won't affect queue runners, except when queue_run_in_order
4469 51. Installed eximstats 1.31 in the distribution.
4471 52. Added support for SRV lookups to the dnslookup router.
4473 53. If an ACL referred to $message_body or $message_body_end, the value was not
4474 reset for any messages that followed in the same SMTP session.
4476 54. The store-handling optimization for building very long strings was not
4477 differentiating between the different store pools. I don't think this
4478 actually made any difference in practice, but I've tidied it.
4480 55. While running the routers to verify a sender address, $sender_address
4481 was still set to the sender address. This is wrong, because when routing to
4482 send a bounce to the sender, it would be empty. Therefore, I have changed
4483 it so that, while verifying a sender address, $sender_address is set to <>.
4484 (There is no change to what happens when verifying a recipient address.)
4486 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4487 target A or AAAA records (if not already returned) without resetting the
4488 qualify_single or search_parents options of the DNS resolver. These are
4489 inappropriate in this case because the targets of MX and SRV records must
4490 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4491 target that, when qualified, matched something in the local domain. These
4492 two options are now turned off when doing these lookups.
4494 57. It seems that at least some releases of Reiserfs (which does not have the
4495 concept of a fixed number of inodes) returns zero and not -1 for the
4496 number of available inodes. This interacted badly with check_spool_inodes,
4497 which assumed that -1 was the "no such thing" setting. What I have done is
4498 to check that the total number of inodes is greater than zero before doing
4499 the test of how many are available.
4501 58. When a "warn" ACL statement has a log_message modifier, the message is
4502 remembered, and not repeated. This is to avoid a lot of repetition when a
4503 message has many recipients that cause the same warning to be written.
4504 Howewer, Exim was preserving the list of already written lines for an
4505 entire SMTP session, which doesn't seem right. The memory is now reset if a
4506 new message is started.
4508 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4509 debugging output unless log_rewrite was also set.
4511 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4512 of (int)(handle) when we know that handle contains (void *)(-1).
4514 61. The Exim daemon panic-logs an error return when it closes the incoming
4515 connection. However "connection reset by peer" seems to be common, and
4516 isn't really an error worthy of noting specially, so that particular error
4519 62. When Exim is trying to find all the local interfaces, it used to panic and
4520 die if the ioctl to get the interface flags failed. However, it seems that
4521 on at least one OS (Solaris 9) it is possible to have an interface that is
4522 included in the list of interfaces, but for which you get a failure error
4523 for this call. This happens when the interface is not "plumbed" into a
4524 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4525 failure of the "get flags" call assumes that the interface is down.
4527 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4528 makes life easier for people who are doing arithmetic on fields extracted
4529 from dates, where you often get leading zeros that should not be
4530 interpreted as octal.
4532 64. Added qualify_domain to the redirect router, to override the global
4535 65. If a pathologically long header line contained very many addresses (the
4536 report of this problem mentioned 10 000) and each of them was rewritten,
4537 Exim could use up a very large amount of memory. (It kept on making new
4538 copies of the header line as it rewrote, and never released the old ones.)
4539 At the expense of a bit more processing, the header rewriting function has
4540 been changed so that it no longer eats memory in this way.
4542 66. The generation of the Received: header has been moved from the time that a
4543 message starts to be received, to the time that it finishes. The timestamp
4544 in the Received: header should now be very close to that of the <= log
4545 line. There are two side-effects of this change:
4547 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4548 logged header lines no longer include the local Received: line, because
4549 it has not yet been created. The same applies to a copy of the message
4550 that is returned to a non-SMTP sender when a message is rejected.
4552 (b) When a filter file is tested using -bf, no additional Received: header
4553 is added to the test message. After some thought, I decided that this
4556 This change does not affect the value of $received_for. It is still set
4557 after address rewriting, but before local_scan() is called.
4559 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4561 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4562 gave an unhelpful panic error message, and a defer error. I have managed to
4563 change this behaviour so that it now rejects any supplied certificate,
4564 which seems right, as the list of acceptable certificates is empty.
4566 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4567 gave an unhelpful defer error. I have not managed to make this reject any
4568 supplied certificates, but the error message it gives is "no certificate
4569 supplied", which is not helpful.
4571 70. exigrep's output now also includes lines that are not associated with any
4572 message, but which match the given pattern. Implemented by a patch from
4573 Martin Sluka, which also tidied up the Perl a bit.
4575 71. Recipient callout verification, like sender verification, was using <> in
4576 the MAIL FROM command. This isn't really the right thing, since the actual
4577 sender may affect whether the remote host accepts the recipient or not. I
4578 have changed it to use the actual sender in the callout; this means that
4579 the cache record is now keyed on a recipient/sender pair, not just the
4580 recipient address. There doesn't seem to be a real danger of callout loops,
4581 since a callout by the remote host to check the sender would use <>.
4582 [SEE ABOVE: changed after hitting problems.]
4584 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4585 temporary errors. However, in the case of such a code being given after
4586 the end of a data transmission (i.e. after ".") Exim was failing to write
4587 a retry record for the message. (Yes, there was some broken host that was
4588 actually sending 8xx at this point.)
4590 73. An unknown lookup type in a host list could cause Exim to panic-die when
4591 the list was checked. (An example that provoked this was putting <; in the
4592 middle of a list instead of at the start.) If this happened during a DATA
4593 ACL check, a -D file could be left lying around. This kind of configuration
4594 error no longer causes Exim to die; instead it causes a defer errror. The
4595 incident is still logged to the main and panic logs.
4597 74. Buglet left over from Exim 3 conversion. The message "too many messages
4598 in one connection" was written to the rejectlog but not the mainlog, except
4599 when address rewriting (yes!) was being logged.
4601 75. Added write_rejectlog option.
4603 76. When a system filter was run not as root (that is, when system_filter_user
4604 was set), the values of the $n variables were not being returned to the
4605 main process; thus, they were not subsequently available in the $sn
4608 77. Added +return_path_on_delivery log selector.
4610 78. A connection timeout was being treated differently from recipients deferred
4611 when testing hosts_max_try with a message that was older than the host's
4612 retry timeout. (The host should not be counted, thus allowing all hosts to
4613 be tried at least once before bouncing.) This may have been the cause of an
4614 occasionally reported bug whereby a message would remain on the queue
4615 longer than the retry timeout, but would be bounced if a delivery was
4616 forced. I say "may" because I never totally pinned down the problem;
4617 setting up timeout/retry tests is difficult. See also the next item.
4619 79. The ultimate address timeout was not being applied to errors that involved
4620 a combination of host plus message (for example, a timeout on a MAIL
4621 command). When an address resolved to a number of possible hosts, and they
4622 were not all tried for each delivery (e.g. because of hosts_max_try), a
4623 message could remain on the queue longer than the retry timeout.
4625 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4628 81. Fixed an obscure SMTP outgoing bug which required at least the following
4629 conditions: (a) there was another message waiting for the same server;
4630 (b) the server returned 5xx to all RCPT commands in the first message so
4631 that the message was not completed; (c) the server dropped the connection
4632 or gave a negative response to the RSET that Exim sends to abort the
4633 transaction. The observed case was a dropped connection after DATA that had
4634 been sent in pipelining mode. That is, the server had advertised PIPELINING
4635 but was not implementing it correctly. The effect of the bug was incorrect
4636 behaviour, such as trying another host, and this could lead to a crash.
4642 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4643 and daemon.c were passed as pointers to ints; they should have been
4644 pointers to socklen_t variables (which are typically unsigned ints).
4646 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4649 3. Fixed a really odd bug that affected only the testing scheme; patching a
4650 certain fixed string in the binary changed the value of another string that
4651 happened to be identical to the end of the original first string.
4653 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4654 name", it returns that address as the IP address. On some operating
4655 systems (e.g. Solaris), it also passes back the IP address string as the
4656 "host name". However, on others (e.g. Linux), it passes back an empty
4657 string. Exim wasn't checking for this, and was changing the host name to an
4658 empty string, assuming it had been canonicized.
4660 5. Although rare, it is permitted to have more than one PTR record for a given
4661 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4662 all the names associated with an address, because they do in Solaris.
4663 However, it seems that they do not in Linux for data that comes from the
4664 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4665 I found this out when I moved to a new Linux workstation and tried to run
4666 the Exim test suite.
4668 To get round this problem I have changed the code so that it now does its
4669 own call to the DNS to look up PTR records when searching for a host name.
4670 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4671 addresses that are only in /etc/hosts are still found.
4673 This behaviour is, however, controlled by an option called host_lookup_
4674 order, which defaults to "bydns:byaddr". If people want to use the other
4675 order, or indeed, just use one or the other means of lookup, they can
4676 specify it in this variable.
4678 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4679 some operating systems, this comes back from gethostbyaddr() as an empty
4680 string, and this is what Exim used to test for. However, it seems that in
4681 other systems, "." is yielded. Exim now tests for this case too.
4683 7. The values of check_spool_space and check_log_space are now held internally
4684 as a number of kilobytes instead of an absolute number of bytes. If a
4685 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4686 kilobyte. This means that much larger values can be stored.
4688 8. Exim monitor: an attempt to get the action menu when not actually pointing
4689 at a message produces an empty menu entitled "No message selected". This
4690 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4691 no entries in it ("Shell widget menu has zero width and/or height"). So I
4692 have added a single, blank menu entry in this case.
4694 9. Added ${quote_local_part.
4696 10. MIME decoding is now applied to the contents of Subject: header lines when
4699 11. Now that a reference to $sender_host_address automatically causes a reverse
4700 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4701 host lookup before query-style lookups in lists that might use this
4702 variable. This has therefore been abolished, and the "net-" prefix is no
4703 longer necessary for query-style lookups.
4705 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4706 have been a typo for LFLAGS, so it has been changed.
4708 13. The install script calls Exim with "-C /dev/null" in order to find the
4709 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4710 to be output. Howeve, since Exim outputs its version number before the
4711 error, it didn't break the script. It just looked ugly. I fixed this by
4712 always allowing "-C /dev/null" if the caller is root.
4714 14. Ignore overlarge ACL variable number when reading spool file - insurance
4715 against a later release with more variables having written the file.
4717 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4718 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4719 this, as well as the form without the "IPv6" on the front.
4721 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4722 OS/Makefile-Darwin file.
4724 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4725 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4727 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4728 variable. However, some OS use ints for inodes. I've added cast to long int
4729 to get rid of the compiler warning.
4731 19. I had forgotten to lock out "/../" in configuration file names when
4732 ALT_CONFIG_PREFIX was set.
4734 20. Routers used for verification do not need to specify transports. However,
4735 if such a router generated a host list, and callout was configured, Exim
4736 crashed, because it could not find a port number from the (non-existent)
4737 transport. It now assumes port 25 in this circumstance.
4739 21. Added the -t option to exigrep.
4741 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4742 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4743 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4744 them is set, LOOKUP_LSEARCH is forced.
4746 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4748 24. Added sender and host information to the "rejected by local_scan()" log
4749 line; previously there was no indication of these.
4751 25. Added .include_if_exists.
4753 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4754 Linux. It turns out that not all file systems support this. Apparently some
4755 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4756 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4757 not supported on the file descriptor, is now ignored when Exim is trying to
4758 sync a directory. This applies only to Linux.
4760 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4762 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4763 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4766 29. Added support for maildirsize files from supplied patch (modified a bit).
4768 30. The use of :fail: followed by an empty string could lead Exim to respond to
4769 sender verification failures with (e.g.):
4771 550 Verification failed for <xxx>
4772 550 Sender verify failed
4774 where the first response line was missing the '-' that indicates it is not
4775 the final line of the response.
4777 31. The loop for finding the name of the user that called Exim had a hardwired
4778 limit of 10; it now uses the value of finduser_retries, which is used for
4779 all other user lookups.
4781 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4784 33. Exim was neglecting to zero errno before one call of strtol() when
4785 expanding a string and expecting an integer value. On some systems this
4786 resulted in spurious "integer overflow" errors. Also, it was casting the
4787 result into an int without checking.
4789 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4790 did not work. The code looks as if it has *never* worked, though it appears
4791 to have been documented since at least releast 1.62. I have made it work.
4793 35. The "timeout_DNS" error in retry rules, also documented since at least
4794 1.62, also never worked. As it isn't clear exactly what this means, and
4795 clearly it isn't a major issue, I have abolished the feature by treating it
4796 as "timeout", and writing a warning to the main and panic logs.
4798 36. The display of retry rules for -brt wasn't always showing the error code
4801 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4802 timeout_connect_A, timeout_connect_MX.
4804 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4805 to the empty sender.
4807 39. The daemon was not analysing the content of -oX till after it had closed
4808 stderr and disconnected from the controlling terminal. This meant that any
4809 syntax errors were only noted on the panic log, and the return code from
4810 the command was 0. By re-arranging the code a little, I've made the
4811 decoding happen first, so such errors now appear on stderr, and the return
4812 code is 1. However, the actual setting up of the sockets still happens in
4813 the disconnected process, so errors there are still only recorded on the
4816 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4817 connections (as happens on some IP stacks) was logged at start up time as
4818 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4819 it from "IPv6 and IPv4", which means that two separate sockets are being
4822 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4823 says whether AF_INET or AF_INET6 was passed as an argument.
4825 42. Exiwhat output was messed up when time zones were included in log
4828 43. Exiwhat now gives more information about the daemon's listening ports,
4829 and whether -tls-on-connect was used.
4831 44. The "port" option of the smtp transport is now expanded.
4833 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4834 silently ignored. Now an error message is written to the main and panic
4837 46. There's a new ACL modifier called "logwrite" which writes to a log file
4838 as soon as it is encountered.
4840 47. Added $local_user_uid and $local_user_gid at routing time.
4842 48. Exim crashed when trying to verify a sender address that was being
4845 49. Exim was recognizing only a space character after ".include". It now also
4846 recognizes a tab character.
4848 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4849 extracting the relevant information from the specification. The man page no
4850 longer contains scrambled data for the -d option, and I've added a section
4851 at the front about calling Exim under different names.
4853 51. Added "extra_headers" argument to the "mail" command in filter files.
4855 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4858 53. Installed eximstats 1.29.
4860 54. Added transport_filter_timeout as a generic transport option.
4862 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4863 Cc: header lines. This was required by RFC 822, but it not required by RFC
4866 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4867 incoming messages that did not have them. Now it does so only if the
4868 message originates locally, that is, if there is no associated remote host
4869 address. When Resent- header lines are present, this applies to the Resent-
4870 lines rather than the non-Resent- lines.
4872 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4873 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4875 58. Messages for configuration errors now include the name of the main
4876 configuration file - useful now that there may be more than one file in a
4877 list (.included file names were always shown).
4879 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4880 for those rare installations that do not start the daemon as root or run it
4881 setuid root. I've cut out the call to initgroups() if the daemon is not
4884 60. The Exim user and group can now be bound into the binary as text strings
4885 that are looked up at the start of Exim's processing.
4887 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4889 62. Added $mailstore_basename variable.
4891 63. Installed patch to sieve.c from Michael Haardt.
4893 64. When Exim failed to open the panic log after failing to open the main log,
4894 the original message it was trying to log was written to stderr and debug
4895 output, but if they were not available (the usual case in production), it
4896 was lost. Now it is written to syslog before the two lines that record the
4897 failures to open the logs.
4899 65. Users' Exim filters run in subprocesses under the user's uid. It is
4900 possible for a "deliver" command or an alias in a "personal" command to
4901 provoke an address rewrite. If logging of address rewriting is configured,
4902 this fails because the process is not running as root or exim. There may be
4903 a better way of dealing with this, but for the moment (because 4.30 needs
4904 to be released), I have disabled address rewrite logging when running a
4905 filter in a non-root, non-exim process.
4911 1. The buildconfig auxiliary program wasn't quoting the value set for
4912 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4913 not defined. This bug was masked in 4.22 by the effect that was fixed in
4916 2. Some messages that were rejected after a message id was allocated were
4917 shown as "incomplete" by exigrep. It no longer does this for messages that
4918 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4920 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4921 have allow_domain_literals set, the ID did not get logged in the <= line.
4922 Domain literals are now always recognized in Message-ID: header lines.
4924 4. The first argument for a ${extract expansion item is the key name or field
4925 number. Leading and trailing spaces in this item were not being ignored,
4926 causing some misleading effects.
4928 5. When deliver_drop_privilege was set, single queue runner processes started
4929 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4930 same command in the process it spins off) were not dropping privilege.
4932 6. When the daemon running as "exim" started a queue runner, it always
4933 re-executed Exim in the spun-off process. This is a waste of effort when
4934 deliver_drop_privilege is set. The new process now just calls the
4935 queue-runner function directly.
4941 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4944 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4945 ignored. Though the use of -oP was forcing the writing of a pid file, it
4946 was always written to the default place.
4948 3. If the message "no IP address found for host xxxx" is generated during
4949 incoming verification, it is now followed by identification of the incoming
4950 connection (so you can more easily find what provoked it).
4952 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4954 5. Added some features to "harden" Exim a bit more against certain attacks:
4956 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4957 be put in Local/Makefile. This is like the never_users runtime option,
4958 but it cannot be overridden. The default setting is "root".
4960 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4961 prefix string with which any file named in a -C command line option
4964 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4965 is retained for -C and -D only if the caller of Exim is root. Without
4966 it, the exim user may also use -C and -D and retain privilege.
4968 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4969 command line option is disabled.
4971 6. Macro names set by the -D option must start with an upper case letter, just
4972 like macro names defined in the configuration file.
4974 7. Added "dereference=" facility to LDAP.
4976 8. Two instances of the typo "uknown" in the source files are fixed.
4978 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4979 the Configure-Makefile script screwed up while processing it.
4981 10. Incorporated PCRE 4.4.
4983 11. The SMTP synchronization check was not operating right at the start of an
4984 SMTP session. For example, it could not catch a HELO sent before the client
4985 waited for the greeting. There is now a check for outstanding input at the
4986 point when the greeting is written. Because of the duplex, asynchronous
4987 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4988 way, but not yet received, when the check is performed.
4990 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4991 on TCP/IP sockets, because this apparently causes some broken clients to
4994 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4995 unchanged) from the Cygwin maintainer.
4997 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4998 was supported instead of testing for mbx. Effectively a typo.
5000 15. The spa authenticator server code was not checking that the input it
5001 received was valid base64.
5003 16. The debug output line for the "set" modifier in ACLs was not showing the
5004 name of the variable that was being set.
5006 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5008 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5009 lookup on its own. Something else was needed to trigger the lookup. For
5010 example, a match in host_lookup or the need for a host name in a host list.
5011 Now, if $sender_host_name is referenced and the host name has not yet been
5012 looked up, a lookup is performed. If the lookup fails, the variable remains
5013 empty, and $host_lookup_failed is set to "1".
5015 19. Added "eqi" as a case-independent comparison operator.
5017 20. The saslauthd authentication condition could segfault if neither service
5018 nor realm was specified.
5020 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5021 error message that was logged was misleading, and incoming SMTP
5022 connections were dropped. The message is now more accurate, and temporary
5023 errors are given to SMTP connections.
5025 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5026 (except RSET) with a 421 error, until QUIT is received. However, it was
5027 failing to send a response to QUIT.
5029 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5030 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5031 ACL is now run after the helo_try_verify_hosts code.
5033 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5034 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5035 case-independent, so other case variants are also recognized). Apparently
5036 some systems use these upper case variants.
5038 25. If more than two messages were waiting for the same host, and a transport
5039 filter was specified for the transport, Exim sent two messages over the
5040 same TCP/IP connection, and then failed with "socket operation on non-
5041 socket" when it tried to send the third.
5043 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5045 27. The extern definition of crypt16() in expand.c was not being excluded when
5046 the OS had its own crypt16() function.
5048 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5049 as a preferred synonym for return_size_limit, both as an option and as an
5052 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5054 30. Changed the default configuration ACL to relax the local part checking rule
5055 for addresses that are not in any local domains. For these addresses,
5056 slashes and pipe symbols are allowed within local parts, but the sequence
5057 /../ is explicitly forbidden.
5059 31. SPA server authentication was not clearing the challenge buffer before
5062 32. log_message in a "warn" ACL statement was writing to the reject log as
5063 well as to the main log, which contradicts the documentation and doesn't
5064 seem right (because no rejection is happening). So I have stopped it.
5066 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5067 However, I am unable to do any testing of this.
5069 34. Fixed an infelicity in the appendfile transport. When checking directories
5070 for a mailbox, to see if any needed to be created, it was accidentally
5071 using path names with one or more superfluous leading slashes; tracing
5072 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5074 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5075 contents are added to the log line that is written for every discarded
5076 recipient. (Previously a log_message setting was ignored.)
5078 36. The ${quote: operator now quotes the string if it is empty.
5080 37. The install script runs exim in order to find its version number. If for
5081 some reason other than non-existence or emptiness, which it checks, it
5082 could not run './exim', it was installing it with an empty version number,
5083 i.e. as "exim-". This error state is now caught, and the installation is
5086 38. An argument was missing from the function that creates an error message
5087 when Exim fails to connect to the socket for saslauthd authentication.
5088 This could cause Exim to crash, or give a corrupted message.
5090 39. Added isip, isip4, and isip6 to ${if conditions.
5092 40. The ACL variables $acl_xx are now saved with the message, and can be
5093 accessed later in routers, transports, and filters.
5095 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5096 strings in the file are not string-expanded.
5098 42. If a MAIL command specified a SIZE value that was too large to fit into an
5099 int variable, the check against message_size_limit failed. Such values are
5100 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5101 day this will have to be increased, but I don't think I want to be around
5102 when emails are that large.
5109 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5110 iconv() is not standard in FreeBSD.
5112 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5113 IPv6 enabled. The observed symptom was a segmentation fault on return from
5114 the function os_common_find_running_interfaces() in src/os.c.
5116 3. In the check_special_case() function in daemon.c I had used "errno" as an
5117 argument name, which causes warnings on some systems. This was basically a
5118 typo, since it was named "eno" in the comments!
5120 4. The code that waits for the clock to tick (at a resolution of some fraction
5121 of a second) so as to ensure message-id uniqueness was always waiting for
5122 at least one whole tick, when it could have waited for less. [This is
5123 almost certainly not relevant at current processor speeds, where it is
5124 unlikely to ever wait at all. But we try to future-proof.]
5126 5. The function that sleeps for a time interval that includes fractions of a
5127 second contained a race. It did not block SIGALRM between setting the
5128 timer, and suspending (a couple of lines later). If the interval was short
5129 and the sigsuspend() was delayed until after it had expired, the suspension
5130 never ended. On busy systems this could lead to processes getting stuck for
5133 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5134 process, before it forks any delivery processes. The open lookup caching
5135 mechanism meant that the open file or database connection was passed into
5136 the delivery process. The problem was that delivery processes always tidy
5137 up cached lookup data. This could cause a problem for the next delivery
5138 process started by the queue runner, because the external queue runner
5139 process does not know about the closure. So the next delivery process
5140 still has data in the lookup cache. In the case of a file lookup, there was
5141 no problem because closing a file descriptor in a subprocess doesn't affect
5142 the parent. However, if the lookup was caching a connection to a database,
5143 the connection was closed, and the second delivery process was likely to
5144 see errors such as "PGSQL: query failed: server closed the connection
5145 unexpectedly". The problem has been fixed by closing all cached lookups
5146 in a queue runner before running a delivery process.
5148 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5149 seem to have the "const" qualifier which it has on other OS. I've
5152 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5153 *of the same type* (client or server) with the same public name that an
5154 error should be diagnosed.
5156 9. When Exim looked up a host name for an IP address, but failed to find the
5157 original IP address when looking up the host name (a safety check), it
5158 output the message "<ip address> does not match any IP for NULL", which was
5159 confusing, to say the least. The bug was that the host name should have
5160 appeared instead of "NULL".
5162 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5163 user that is built into the binary, and the -C or -D options is used, root
5164 privilege is dropped before the configuration file is read. In addition,
5165 logging is switched to stderr instead of the normal log files. If the
5166 configuration then re-defines the Exim user, the unprivileged environment
5167 is probably not what is expected, so Exim logs a panic warning message (but
5170 However, if deliver_drop_privilege is set, the unprivileged state may well
5171 be exactly what is intended, so the warning has been cut out in that case,
5172 and Exim is allowed to try to write to its normal log files.
5178 1. smtp_return_error_details was not giving details for temporary sender
5179 or receiver verification errors.
5181 2. Diagnose a configuration error if two authenticators have the same public
5184 3. Exim used not to create the message log file for a message until the first
5185 delivery attempt. This could be confusing when incoming messages were held
5186 for policy or load reasons. The message log file is now created at the time
5187 the message is received, and an initial "Received" line is written to it.
5189 4. The automatically generated man page for command line options had a minor
5190 bug that caused no ill effects; however, a more serious problem was that
5191 the procedure for building the man page automatically didn't always
5192 operate. Consequently, release 4.20 contains an out-of-date version. This
5193 shouldn't happen again.
5195 5. When building Exim with embedded Perl support, the script that builds the
5196 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5197 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5199 6. The freeze_tell option was not being used for messages that were frozen on
5200 arrival, either by an ACL or by local_scan().
5202 7. Added the smtp_incomplete_transaction log selector.
5204 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5205 was accepting AUTH without a new EHLO.
5207 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5208 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5211 10. Logging of TCP/IP connections (when configured) now happens in the main
5212 daemon process instead of the child process, so that the TCP/IP connection
5213 count is more accurate (but it can never be perfect).
5215 11. The use of "drop" in a nested ACL was not being handled correctly in the
5216 outer ACL. Now, if condition failure induced by the nested "drop" causes
5217 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5218 or "require"), the connection is dropped.
5220 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5221 that yield "discard" can now be used with an "accept" or a "discard" verb,
5222 but an error is generated for any others (because I can't see a useful way
5223 to define what should happen).
5225 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5226 are now processed in the same way as lines in the Exim configuration file.
5227 In particular, continuation lines are supported.
5229 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5231 15. Added -ti meaning -t -i.
5233 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5234 domains, and warn by logging if others are found.
5236 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5237 returned from GIFCONF when Exim is trying to find the list of interfaces on
5238 a host. The code in os.c has been modified to copy each ifreq to an aligned
5239 structure in all cases.
5241 Also, in some cases, the returned ifreq's were being copied to a 'struct
5242 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5243 means the last couple of bytes of an IPv6 address could be chopped if the
5244 ifreq contained only a normal sockaddr (14 bytes storage).
5246 18. Named domain lists were not supported in the hosts_treat_as_local option.
5247 An entry such as +xxxx was not recognized, and was treated as a literal
5250 19. Ensure that header lines added by a DATA ACL are included in the reject log
5251 if the ACL subsequently rejects the message.
5253 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5254 MD5 (which is deprecated).
5256 21. When testing a filter file using -bf, Exim was writing a message when it
5257 took the sender from a "From " line in the message, but it was not doing so
5258 when it took $return_path from a Return-Path: header line. It now does.
5260 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5261 with a valid header line field name (a series of printing characters
5262 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5264 23. Changed "disc" in the source to "disk" to conform to the documentation and
5265 the book and for uniformity.
5267 24. Ignore Sendmail's -Ooption=value command line item.
5269 25. When execve() failed while trying to run a command in a pipe transport,
5270 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5271 could be confused with a return value of 69 from the command itself. This
5272 has been changed to 127, the value the shell returns if it is asked to run
5273 a non-existent command. The wording for the related log line suggests a
5274 non-existent command as the problem.
5276 26. If received_header_text expands to an empty string, do not add a Received:
5277 header line to the message. (Well, it adds a token one on the spool, but
5278 marks it "old" so that it doesn't get used or transmitted.)
5280 27. Installed eximstats 1.28 (addition of -nt option).
5282 28. There was no check for failure on the call to getsockname() in the daemon
5283 code. This can fail if there is a shortage of resources on the system, with
5284 ENOMEM, for example. A temporary error is now given on failure.
5286 29. Contrary to the C standard, it seems that in some environments, the
5287 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5288 program. Exim now does this explicitly; it affects the formatting of
5289 timestamps using strftime().
5291 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5292 complaints. I've now initialized all the variables, to avoid this.
5294 32. Header lines added by a system filter were not being "seen" during
5295 transport-time rewrites.
5297 33. The info_callback() function passed to OpenSSL is set up with type void
5298 (*)(SSL *, int, int), as described somewhere. However, when calling the
5299 function (actually a macro) that sets it up, the type void(*)() is
5300 expected. I've put in a cast to prevent warnings from picky compilers.
5302 34. If a DNS black list lookup found a CNAME record, but there were no A
5303 records associated with the domain it pointed at, Exim crashed.
5305 35. If a DNS black list lookup returned more than one A record, Exim ignored
5306 all but the first. It now scans all returned addresses if a particular IP
5307 value is being sought. In this situation, the contents of the
5308 $dnslist_value variable are a list of all the addresses, separated by a
5311 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5312 to accept a host name and all its aliases if the forward lookup for any of
5313 them yielded the IP address of the incoming connection. Now it accepts only
5314 those names whose forward lookup yields the correct IP address. Any other
5315 names are discarded. This closes a loophole whereby a rogue DNS
5316 administrator could create reverse DNS records to break through a
5317 wildcarded host restriction in an ACL.
5319 37. If a user filter or a system filter that ran in a subprocess used any of
5320 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5321 the wrong values were passed to the pipe command ($thisaddress had the
5322 value of $0, $0 had the value of $1, etc). This bug was introduced by
5323 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5325 38. Improved the line breaking for long SMTP error messages from ACLs.
5326 Previously, if there was no break point between 40 and 75 characters, Exim
5327 left the rest of the message alone. Two changes have been made: (a) I've
5328 reduced the minimum length to 35 characters; (b) if it can't find a break
5329 point between 35 and 75 characters, it looks ahead and uses the first one
5330 that it finds. This may give the occasional overlong line, but at least the
5331 remaining text gets split now.
5333 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5334 file descriptors might be low, and that setting 1000 would always raise it.
5335 It turns out that in some environments, the limit is already over 1000 and
5336 that lowering it causes trouble. So now Exim takes care not to decrease it.
5338 40. When delivering a message, the value of $return_path is set to $sender_
5339 address at the start of routing (routers may change the value). By an
5340 oversight, this default was not being set up when an address was tested by
5341 -bt or -bv, which affected the outcome if any router or filter referred to
5344 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5345 log only when "message" or "log_message" are set. However, if one of the
5346 conditions was an address verification, or a call to a nested ACL, the
5347 messages generated by the underlying test were being passed through. This
5348 no longer happens. The underlying message is available in $acl_verify_
5349 message for both "message" and "log_message" expansions, so it can be
5350 passed through if needed.
5352 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5353 new expansion $bh_ to give the encoded byte string without charset
5354 translation. Translation happens only if iconv() is available; HAVE_ICONV
5355 indicates this at build time. HEADERS_CHARSET gives the charset to
5356 translate to; headers_charset can change it in the configuration, and
5357 "headers charset" can change it in an individual filter file.
5359 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5360 that creates RFC 2047 encoded "words" labels them as that charset instead
5361 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5362 expansion operator; (ii) when Exim creates a From: line for a local
5363 message; (iii) when a header line is rewritten to include a "phrase" part.
5365 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5366 buggy, causing it to skip the first lines of messages whose message ID
5367 ended in 'D'. This would not have bitten before Exim release 4.14, because
5368 message IDs were unlikely to end in 'D' before then. The effect was to have
5369 incorrect size information for certain domains.
5371 45. #include "config.h" was missing at the start of the crypt16.c module. This
5372 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5375 46. If there was a timeout during a "random" callout check, Exim treated it as
5376 a failure of the random address, and carried on sending RSET and the real
5377 address. If the delay was just some slowness somewhere, the response to the
5378 original RCPT would be taken as a response to RSET and so on, causing
5379 mayhem of various kinds.
5381 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5382 when I implemented it. It didn't allow for the fact that some option values
5383 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5384 the right test for positive values.
5386 48. Domain names in DNS records are case-independent. Exim always looks them up
5387 in lower case. Some resolvers return domain names in exactly the case they
5388 appear in the zone file, that is, they may contain uppercase letters. Not
5389 all resolvers do this - some return always lower case. Exim was treating a
5390 change of case by a resolver as a change of domain, similar to a widening
5391 of a domain abbreviation. This triggered its re-routing code and so it was
5392 trying to route what was effectively the same domain again. This normally
5393 caused routing to fail (because the router wouldn't handle the domain
5394 twice). Now Exim checks for this case specially, and just changes the
5395 casing of the domain that it ultimately uses when it transmits the message
5398 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5401 50. If a filter generated a file delivery with a non-absolute name (possible if
5402 no home directory exists for the router), the forbid_file option was not
5405 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5406 the existing equality matching.
5408 52. Exim was using ints instead of ino_t variables in some places where it was
5409 dealing with inode numbers.
5411 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5412 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5413 called TMPDIR, and if it finds it is different, it changes its value.
5415 54. The smtp_printf() function is now made available to local_scan() so
5416 additional output lines can be written before returning. There is also an
5417 smtp_fflush() function to enable the detection of a dropped connection.
5418 The variables smtp_input and smtp_batched_input are exported to
5421 55. Changed the default runtime configuration: the message "Unknown user"
5422 has been removed from the ACL, and instead placed on the localuser router,
5423 using the cannot_route_message feature. This means that any verification
5424 failures that generate their own messages won't get overridden. Similarly,
5425 the "Unrouteable address" message that was in the ACL for unverifiable
5426 relay addresses has also been removed.
5428 56. Added hosts_avoid_esmtp to the smtp transport.
5430 57. The exicyclog script was not checking for the esoteric option
5431 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5432 will work only if exicyclog is run under the appropriate euid.
5434 58. Following a discussion on the list, the rules by which Exim recognises line
5435 endings on incoming messages have been changed. The -dropcr and drop_cr
5436 options are now no-ops, retained only for backwards compatibility. The
5437 following line terminators are recognized: LF CRLF CR. However, special
5438 processing applies to CR:
5440 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5441 nor a local message in the state where . is a terminator.
5443 (ii) If a bare CR is encountered in a header line, an extra space is added
5444 after the line terminator so as not to end the header. The reasoning
5445 behind this is that bare CRs in header lines are most likely either
5446 to be mistakes, or people trying to play silly games.
5448 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5449 was being incorrectly given as 18 bytes larger than it should have been.
5450 This is a VOB (very old bug).
5452 60. This may never have affected anything current, but just in case it has:
5453 When the local host is found other than at the start of a list of hosts,
5454 the local host, those with the same MX, and any that follow, are discarded.
5455 When the list in question was part of a longer list of hosts, the following
5456 hosts (not currently being processed) were also being discarded. This no
5457 longer happens. I'm not sure if this situation could ever has previously
5460 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5463 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5464 header. This is something that is recommended in a new Internet Draft, and
5465 is something that is documented as being done by Sendmail. There are two
5466 possible values. For messages generated by the autoreply transport, Exim
5469 Auto-Submitted: auto-replied
5471 whereas for all other generated messages (e.g. bounces) it adds
5473 Auto-Submitted: auto-generated
5475 63. The "personal" condition in filters now includes a test for the
5476 Auto-Submitted: header. If it contains the string "auto-" the message it
5477 not considered personal.
5479 64. Added rcpt_include_affixes as a generic transport option.
5481 65. Added queue_only_override (default true).
5483 66. Added the syslog_duplication option.
5485 67. If what should have been the first header line of a message consisted of
5486 a space followed by a colon, Exim was mis-interpreting it as a header line.
5487 It isn't of course - it is syntactically invalid and should therefore be
5488 treated as the start of the message body. The misbehaviour could have
5489 caused a number of strange effects, including loss of data in subsequent
5490 header lines, and spool format errors.
5492 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5493 client host had authenticated. This control can now be exercised by an ACL
5494 for more flexibility.
5496 69. By default, callouts do not happen when testing with -bh. There is now a
5497 variant, -bhc, which does actually run the callout code, including
5498 consulting and updating the callout cache.
5500 70. Added support for saslauthd authentication, courtesy of Alexander
5503 71. If statvfs() failed on the spool or log directories while checking their
5504 size for availability, Exim confusingly gave the error "space shortage".
5505 Furthermore, in debugging mode it crashed with a floating point exception.
5506 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5507 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5508 serious problem, Exim now writes to the main and panic logs when this
5509 happens, with details of the failure. It then refuses to accept the
5510 incoming message, giving the message "spool directory problem" or "log
5511 directory problem" with a 421 code for SMTP messages.
5513 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5514 0, 1, and 2 exist, because some OS complain for execs without them (see
5515 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5516 descriptors. However, the code omitted to check that the open succeeded,
5517 causing mysterious errors if for some reason the permissions on /dev/null
5518 got screwed. Now Exim writes a message to the main and panic logs, and
5519 bombs out if it can't open /dev/null.
5521 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5522 interact so that it is all more flexible. It is supposed to remain
5523 backwards compatible. Also added extra_local_interfaces.
5525 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5526 to bomb out with an assertion failure - to the client this appears as a
5527 connection drop. This problem occurs in the part of the code that was taken
5528 from the Samba project. Fortunately, the assertion is in a very simple
5529 function, so I have fixed this by reproducing the function inline in the
5530 one place where it is called, and arranging for authentication to fail
5531 instead of killing the process with assert().
5533 75. The SPA client code was not working when the server requested OEM rather
5534 than Unicode encoding.
5536 76. Added code to make require_files with a specific uid setting more usable in
5537 the case where statting the file as root fails - usually a non-root-mounted
5538 NFS file system. When this happens and the failure is EACCES, Exim now
5539 forks a subprocess and does the per-uid checking as the relevant uid.
5541 77. Added process_log_path.
5543 78. If log_file_path was not explicitly set, a setting of check_log_space or
5544 check_log_inodes was ignored.
5546 79. If a space check for the spool or log partitions fails, the incident is now
5547 logged. Of course, in the latter case the data may get lost...
5549 80. Added the %p formatting code to string_format() so that it can be used to
5550 print addresses in debug_print(). Adjusted all the address printing in the
5551 debugging in store.c to use %p rather than %d.
5553 81. There was a concern that a line of code in smtp_in.c could overflow a
5554 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5555 initially expressed, the concern was not well-founded, because trailing
5556 spaces are removed early. However, if the trailing spaces were followed by
5557 a NULL, they did not get removed, so the overflow was possible. Two fixes
5560 (a) I re-wrote the offending code in a cleaner fashion.
5561 (b) If an incoming SMTP command contains a NULL character, it is rejected
5564 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5565 runs initgroups(), so that if the Exim user is in any additional groups,
5566 they will be used during message reception.
5572 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob