1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
81 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
82 NULL dereference. Report and patch from Alun Jones.
84 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
85 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
86 Not seeing resolver debug output on NetBSD, but suspect this is a
87 resolver implementation change.
89 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
90 Left warnings. Added "eximon gdb" invocation mode.
92 PP/21 Defaulting "accept_8bitmime" to true, not false.
98 PP/01 Solaris build fix for Oracle's LDAP libraries.
99 Bugzilla 1109, patch from Stephen Usher.
101 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
103 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
106 TF/02 Fix a couple more cases where we did not log the error message
107 when unlink() failed. See also change 4.74-TF/03.
109 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
110 lock up or crash if it happened to be inside a call to libc when it
111 got a SIGUSR1 from exiwhat.
113 The SIGUSR1 handler appends the current process status to the process
114 log which is later printed by exiwhat. It used to use the general
115 purpose logging code to do this, but several functions it calls are
116 not safe for signals.
118 The new output code in the SIGUSR1 handler is specific to the process
119 log, and simple enough that it's easy to inspect for signal safety.
120 Removing some special cases also simplifies the general logging code.
121 Removing the spurious timestamps from the process log simplifies
124 TF/04 Improved ratelimit ACL condition.
126 The /noupdate option has been deprecated in favour of /readonly which
127 has clearer semantics. The /leaky, /strict, and /readonly update modes
128 are mutually exclusive. The update mode is no longer included in the
129 database key; it just determines when the database is updated. (This
130 means that when you upgrde Exim will forget old rate measurements.)
132 Exim now checks that the per_* options are used with an update mode that
133 makes sense for the current ACL. For example, when Exim is processing a
134 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
135 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
136 must specify per_mail/readonly. If you omit the update mode it defaults to
137 /leaky where that makes sense (as before) or /readonly where required.
139 The /noupdate option is now undocumented but still supported for
140 backwards compatibility. It is equivalent to /readonly except that in
141 ACLs where /readonly is required you may specify /leaky/noupdate or
142 /strict/noupdate which are treated the same as /readonly.
144 A useful new feature is the /count= option. This is a generalization
145 of the per_byte option, so that you can measure the throughput of other
146 aggregate values. For example, the per_byte option is now equivalent
147 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
149 The per_rcpt option has been generalized using the /count= mechanism
150 (though it's more complicated than the per_byte equivalence). When it is
151 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
152 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
153 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
154 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
155 Note that using per_rcpt with a non-readonly update mode in more than
156 one ACL will cause the recipients to be double-counted. (The per_mail
157 and per_byte options don't have this problem.)
159 The handling of very low rates has changed slightly. If the computed rate
160 is less than the event's count (usually one) then this event is the first
161 after a long gap. In this case the rate is set to the same as this event's
162 count, so that the first message of a spam run is counted properly.
164 The major new feature is a mechanism for counting the rate of unique
165 events. The new per_addr option counts the number of different
166 recipients that someone has sent messages to in the last time period. It
167 behaves like per_rcpt if all the recipient addresses are different, but
168 duplicate recipient addresses do not increase the measured rate. Like
169 the /count= option this is a general mechanism, so the per_addr option
170 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
171 example, measure the rate that a client uses different sender addresses
172 with the options per_mail/unique=$sender_address. There are further
173 details in the main documentation.
175 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
177 TF/06 Removed a few PCRE remnants.
179 TF/07 Automatically extract Exim's version number from tags in the git
180 repository when doing development or release builds.
182 PP/02 Raise smtp_cmd_buffer_size to 16kB.
183 Bugzilla 879. Patch from Paul Fisher.
185 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
186 Heavily based on revision 40f9a89a from Simon Arlott's tree.
189 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
191 PP/05 Variable $av_failed, true if the AV scanner deferred.
192 Bugzilla 1078. Patch from John Horne.
194 PP/06 Stop make process more reliably on build failure.
195 Bugzilla 1087. Patch from Heiko Schlittermann.
197 PP/07 Make maildir_use_size_file an _expandable_ boolean.
198 Bugzilla 1089. Patch from Heiko Schlittermann.
200 PP/08 Handle ${run} returning more data than OS pipe buffer size.
201 Bugzilla 1131. Patch from Holger Weiß.
203 PP/09 Handle IPv6 addresses with SPF.
204 Bugzilla 860. Patch from Wolfgang Breyha.
206 PP/10 GnuTLS: support TLS 1.2 & 1.1.
208 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
211 PP/11 match_* no longer expand right-hand-side by default.
212 New compile-time build option, EXPAND_LISTMATCH_RHS.
213 New expansion conditions, "inlist", "inlisti".
215 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
217 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
219 PP/14 fix log_write() format string regression from TF/03.
220 Bugzilla 1152. Patch from Dmitry Isaikin.
226 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
228 PP/02 Harmonised TLS library version reporting; only show if debugging.
229 Layout now matches that introduced for other libraries in 4.74 PP/03.
231 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
233 PP/04 New "dns_use_edns0" global option.
235 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
238 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
239 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
241 TK/01 Updated PolarSSL code to 0.14.2.
242 Bugzilla 1097. Patch from Andreas Metzler.
244 PP/07 Catch divide-by-zero in ${eval:...}.
247 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
250 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
251 format-string attack -- SECURITY: remote arbitrary code execution.
253 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
254 time unintentionally subject to list matching rules, letting the header
255 cause arbitrary Exim lookups (of items which can occur in lists, *not*
256 arbitrary string expansion). This allowed for information disclosure.
258 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
259 INT_MIN/-1 -- value coerced to INT_MAX.
265 NM/01 Workround for PCRE version dependency in version reporting
268 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
269 This fixes portability to compilers other than gcc, notably
270 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
272 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
273 makefiles for portability to HP-UX and POSIX correctness.
275 PP/01 Permit LOOKUP_foo enabling on the make command-line.
276 Also via indented variable definition in the Makefile.
277 (Debugging by Oliver Heesakkers).
279 PP/02 Restore caching of spamd results with expanded spamd_address.
280 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
282 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
283 Improves build reliability. Fix from: Frank Elsner
285 NM/02 Fix wide character breakage in the rfc2047 coding
286 Fixes bug 1064. Patch from Andrey N. Oktyabrski
288 NM/03 Allow underscore in dnslist lookups
289 Fixes bug 1026. Patch from Graeme Fowler
291 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
292 Code patches from Adam Ciarcinski of NetBSD.
294 NM/04 Fixed exiqgrep to cope with mailq missing size issue
297 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
298 is logged, to avoid truncation. Patch from John Horne.
300 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
301 Patch from Jakob Hirsch.
303 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
304 SQL string expansion failure details.
305 Patch from Andrey Oktyabrski.
307 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
308 Patch from Simon Arlott.
310 PP/09 New lookups functionality failed to compile on old gcc which rejects
311 extern declarations in function scope.
312 Patch from Oliver Fleischmann
314 PP/10 Use sig_atomic_t for flags set from signal handlers.
315 Check getgroups() return and improve debugging.
316 Fixed developed for diagnosis in bug 927 (which turned out to be
319 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
320 Patch from Mark Zealey.
322 PP/12 Bugzilla 1056: Improved spamd server selection.
323 Patch from Mark Zealey.
325 PP/13 Bugzilla 1086: Deal with maildir quota file races.
326 Based on patch from Heiko Schlittermann.
328 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
329 Patch from Uwe Doering, sign-off by Michael Haardt.
331 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
332 variable declaration deep within a block. Bug and patch from
335 PP/15 lookups-Makefile IRIX compatibilty coercion.
337 PP/16 Make DISABLE_DKIM build knob functional.
339 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
340 Patch by Simon Arlott
342 TF/03 Fix valgrind.h portability to C89 compilers that do not support
343 variable argument macros. Our copy now differs from upstream.
349 TF/01 Failure to get a lock on a hints database can have serious
350 consequences so log it to the panic log.
352 TF/02 Log LMTP confirmation messages in the same way as SMTP,
353 controlled using the smtp_confirmation log selector.
355 TF/03 Include the error message when we fail to unlink a spool file.
357 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
358 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
359 for maintaining out-of-tree patches for some time.
361 PP/01 Bugzilla 139: Documentation and portability issues.
362 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
363 Handle per-OS dynamic-module compilation flags.
365 PP/02 Let /dev/null have normal permissions.
366 The 4.73 fixes were a little too stringent and complained about the
367 permissions on /dev/null. Exempt it from some checks.
368 Reported by Andreas M. Kirchwitz.
370 PP/03 Report version information for many libraries, including
371 Exim version information for dynamically loaded libraries. Created
372 version.h, now support a version extension string for distributors
373 who patch heavily. Dynamic module ABI change.
375 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
376 privilege escalation vulnerability whereby the Exim run-time user
377 can cause root to append content of the attacker's choosing to
380 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
383 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
384 If dropping privileges for untrusted macros, we disabled normal logging
385 on the basis that it would fail; for the Exim run-time user, this is not
386 the case, and it resulted in successful deliveries going unlogged.
387 Fixed. Reported by Andreas Metzler.
393 PP/01 Date: & Message-Id: revert to normally being appended to a message,
394 only prepend for the Resent-* case. Fixes regression introduced in
395 Exim 4.70 by NM/22 for Bugzilla 607.
397 PP/02 Include check_rfc2047_length in configure.default because we're seeing
398 increasing numbers of administrators be bitten by this.
400 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
402 PP/03 Bugzilla 994: added openssl_options main configuration option.
404 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
406 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
408 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
410 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
411 without a peer certificate, leading to a segfault because of an
412 assumption that peers always have certificates. Be a little more
413 paranoid. Problem reported by Martin Tscholak.
415 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
416 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
417 NB: ClamAV planning to remove STREAM in "middle of 2010".
418 CL also introduces -bmalware, various -d+acl logging additions and
419 more caution in buffer sizes.
421 PP/09 Implemented reverse_ip expansion operator.
423 PP/10 Bugzilla 937: provide a "debug" ACL control.
425 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
427 PP/12 Bugzilla 973: Implement --version.
429 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
431 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
433 PP/15 Bugzilla 816: support multiple condition rules on Routers.
435 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
436 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
437 ignore trailing whitespace.
439 JJ/02 prevent non-panic DKIM error from being sent to paniclog
441 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
444 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
445 Notification from Dr Andrew Aitchison.
447 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
448 ExtendedDetectionInfo response format.
449 Notification from John Horne.
451 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
454 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
455 XSL and documented dependency on system catalogs, with examples of how
458 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
461 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
462 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
463 configuration file which is writeable by the Exim user or group.
465 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
466 of configuration files to cover files specified with the -C option if
467 they are going to be used with root privileges, not just the default
470 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
471 option (effectively making it always true).
473 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
474 files to be used while preserving root privileges.
476 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
477 that rogue child processes cannot use them.
479 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
480 run-time user, instead of root.
482 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
483 Exim run-time user without dropping privileges.
485 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
486 result string, instead of calling string_vformat() twice with the same
489 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
490 for other users. Others should always drop root privileges if they use
491 -C on the command line, even for a whitelisted configure file.
493 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
495 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
501 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
502 $data_path, and $header_path variables; fixed documentation bugs and
505 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
506 exipick to access non-standard spools, including the "frozen" queue
509 NM/01 Bugzilla 965: Support mysql stored procedures.
510 Patch from Alain Williams
512 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
514 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
515 Patch from Andreas Metzler
517 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
518 Patch from Kirill Miazine
520 NM/05 Bugzilla 671: Added umask to procmail example.
522 JJ/03 installed exipick 20100323.0, fixing doc bug
524 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
525 directory. Notification and patch from Dan Rosenberg.
527 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
529 TK/02 Improve log output when DKIM signing operation fails.
531 MH/01 Treat the transport option dkim_domain as a colon separated
532 list, not as a single string, and sign the message with each element,
533 omitting multiple occurences of the same signer.
535 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
536 Bugzilla 985, 986. Patch by Simon Arlott
538 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
539 Patch by Simon Arlott
541 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
542 MBX locking. Notification from Dan Rosenberg.
548 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
550 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
552 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
554 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
556 NM/04 Bugzilla 915: Flag broken perl installation during build.
562 TK/01 Added patch by Johannes Berg that expands the main option
563 "spamd_address" if it starts with a dollar sign.
565 TK/02 Write list of recipients to X-Envelope-Sender header when building
566 the mbox-format spool file for content scanning (suggested by Jakob
569 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
570 (http://www.dcc-servers.net/) support via dccifd. Activated by
571 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
573 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
574 by Mark Daniel Reidel <mr@df.eu>.
576 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
577 When building exim an external PCRE library is now needed -
578 PCRE is a system library on the majority of modern systems.
579 See entry on PCRE_LIBS in EDITME file.
581 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
582 conversation. Added nologin parameter to request.
583 Patch contributed by Kirill Miazine.
585 TF/01 Do not log submission mode rewrites if they do not change the address.
587 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
589 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
590 log files in place. Contributed by Roberto Lima.
592 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
594 TF/03 Bugzilla 615: When checking the local_parts router precondition
595 after a local_part_suffix or local_part_prefix option, Exim now
596 does not use the address's named list lookup cache, since this
597 contains cached lookups for the whole local part.
599 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
600 Robert Millan. Documentation is in experimental-spec.txt.
602 TF/04 Bugzilla 668: Fix parallel build (make -j).
604 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
606 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
607 Patch provided by Jan Srzednicki.
609 TF/05 Leading white space used to be stripped from $spam_report which
610 wrecked the formatting. Now it is preserved.
612 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
613 that they are available at delivery time.
615 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
617 TF/08 TLS error reporting now respects the incoming_interface and
618 incoming_port log selectors.
620 TF/09 Produce a more useful error message if an SMTP transport's hosts
621 setting expands to an empty string.
623 NM/06 Bugzilla 744: EXPN did not work under TLS.
624 Patch provided by Phil Pennock.
626 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
627 Patch provided by Richard Godbee.
629 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
630 acl_smtp_notquit, added index entry.
632 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
633 Patch provided by Eugene Bujak.
635 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
636 accept(). Patch provided by Maxim Dounin.
638 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
639 Patch provided by Phil Pennock.
641 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
643 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
644 Patch provided by Brad "anomie" Jorsch.
646 NM/14 Bugzilla 622: Added timeout setting to transport filter.
647 Patch provided by Dean Brooks.
649 TK/05 Add native DKIM support (does not depend on external libraries).
651 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
652 Patch provided by Graeme Fowler.
654 NM/16 Bugzilla 851: Documentation example syntax fix.
656 NM/17 Changed NOTICE file to remove references to embedded PCRE.
658 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
661 NM/19 Bugzilla 745: TLS version reporting.
662 Patch provided by Phil Pennock.
664 NM/20 Bugzilla 167: bool: condition support.
665 Patch provided by Phil Pennock.
667 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
668 clients. Patch provided by Phil Pennock.
670 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
671 Patch provided by Brad "anomie" Jorsch.
673 NM/23 Bugzilla 687: Fix misparses in eximstats.
674 Patch provided by Heiko Schlittermann.
676 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
677 Patch provided by Heiko Schlittermann.
679 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
680 plus update to original patch.
682 NM/26 Bugzilla 799: Documentation correction for ratelimit.
684 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
685 Patch provided by David Brownlee.
687 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
689 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
691 NM/30 Bugzilla 888: TLS documentation bugfixes.
693 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
695 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
696 Unlike the original bugzilla I have changed all shell scripts in src tree.
698 NM/33 Bugzilla 898: Transport filter timeout fix.
699 Patch by Todd Rinaldo.
701 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
702 Patch by Serge Demonchaux.
704 NM/35 Bugzilla 39: Base64 decode bug fixes.
705 Patch by Jakob Hirsch.
707 NM/36 Bugzilla 909: Correct connect() call in dcc code.
709 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
711 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
713 NM/39 Bugzilla 911: Fixed MakeLinks build script.
719 TK/01 Add preliminary DKIM support. Currently requires a forked version of
720 ALT-N's libdkim that I have put here:
721 http://duncanthrax.net/exim-experimental/
723 Note to Michael Haardt: I had to rename some vars in sieve.c. They
724 were called 'true' and it seems that C99 defines that as a reserved
725 keyword to be used with 'bool' variable types. That means you could
726 not include C99-style headers which use bools without triggering
727 build errors in sieve.c.
729 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
730 as mailq or other aliases. Changed the --help handling significantly
731 to do whats expected. exim_usage() emits usage/help information.
733 SC/01 Added the -bylocaldomain option to eximstats.
735 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
737 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
739 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
745 PH/01 Another patch from the Sieve maintainer.
747 PH/02 When an IPv6 address is converted to a string for single-key lookup
748 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
749 dots are used instead of colons so that keys in lsearch files need not
750 contain colons. This was done some time before quoting was made available
751 in lsearch files. However, iplsearch files do require colons in IPv6 keys
752 (notated using the quote facility) so as to distinguish them from IPv4
753 keys. This meant that lookups for IP addresses in host lists did not work
754 for iplsearch lookups.
756 This has been fixed by arranging for IPv6 addresses to be expressed with
757 colons if the lookup type is iplsearch. This is not incompatible, because
758 previously such lookups could never work.
760 The situation is now rather anomolous, since one *can* have colons in
761 ordinary lsearch keys. However, making the change in all cases is
762 incompatible and would probably break a number of configurations.
764 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
767 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
768 conversion specification without a maximum field width, thereby enabling
769 a rogue spamd server to cause a buffer overflow. While nobody in their
770 right mind would setup Exim to query an untrusted spamd server, an
771 attacker that gains access to a server running spamd could potentially
772 exploit this vulnerability to run arbitrary code as the Exim user.
774 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
775 $primary_hostname instead of what libspf2 thinks the hosts name is.
777 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
778 a directory entry by the name of the lookup key. Previously, if a
779 symlink pointed to a non-existing file or a file in a directory that
780 Exim lacked permissions to read, a lookup for a key matching that
781 symlink would fail. Now it is enough that a matching directory entry
782 exists, symlink or not. (Bugzilla 503.)
784 PH/03 The body_linecount and body_zerocount variables are now exported in the
787 PH/04 Added the $dnslist_matched variable.
789 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
790 This means they are set thereafter only if the connection becomes
793 PH/06 Added the client_condition to authenticators so that some can be skipped
794 by clients under certain conditions.
796 PH/07 The error message for a badly-placed control=no_multiline_responses left
797 "_responses" off the end of the name.
799 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
801 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
802 (without spaces) instead of just copying the configuration text.
804 PH/10 Added the /noupdate option to the ratelimit ACL condition.
806 PH/11 Added $max_received_linelength.
808 PH/12 Added +ignore_defer and +include_defer to host lists.
810 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
811 way in which PCRE > 7.0 is built.
813 PH/14 Implemented queue_only_load_latch.
815 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
816 MAIL command. The effect was to mangle the value on 64-bit systems.
818 PH/16 Another patch from the Sieve maintainer.
820 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
822 PH/18 If a system quota error occurred while trying to create the file for
823 a maildir delivery, the message "Mailbox is full" was not appended to the
824 bounce if the delivery eventually timed out. Change 4.67/27 below applied
825 only to a quota excession during the actual writing of the file.
827 PH/19 It seems that peer DN values may contain newlines (and other non-printing
828 characters?) which causes problems in log lines. The DN values are now
829 passed through string_printing() before being added to log lines.
831 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
832 and InterBase are left for another time.)
834 PH/21 Added message_body_newlines option.
836 PH/22 Guard against possible overflow in moan_check_errorcopy().
838 PH/23 POSIX allows open() to be a macro; guard against that.
840 PH/24 If the recipient of an error message contained an @ in the local part
841 (suitably quoted, of course), incorrect values were put in $domain and
842 $local_part during the evaluation of errors_copy.
848 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
849 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
852 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
853 issue a MAIL command.
855 PH/02 In an ACL statement such as
857 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
859 if a client was not listed at all, or was listed with a value other than
860 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
861 the condition was not true (as it should be), so access was not denied.
862 The bug was that the ! inversion was incorrectly passed on to the second
863 item. This has been fixed.
865 PH/03 Added additional dnslists conditions == and =& which are different from
866 = and & when the dns lookup returns more than one IP address.
868 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
869 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
871 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
872 FSYNC, which compiles an option called disable_fsync that allows for
873 bypassing fsync(). The documentation is heavily laced with warnings.
875 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
877 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
878 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
879 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
880 including adding "make clean"; (3) Added -fPIC when compiling the test
881 dynamically loaded module, to get rid of a warning.
883 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
884 message fails, move_frozen_messages = true and ignore_bounce_errors_after
885 = 0s. The bug is otherwise harmless.
887 PH/07 There was a bug in the dovecot authenticator such that the value of
888 $auth1 could be overwritten, and so not correctly preserved, after a
889 successful authentication. This usually meant that the value preserved by
890 the server_setid option was incorrect.
892 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
894 PH/09 Installed PCRE release 7.0.
896 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
897 run for batched SMTP input. It is now run at the start of every message
898 in the batch. While fixing this I discovered that the process information
899 (output by running exiwhat) was not always getting set for -bs and -bS
900 input. This is fixed, and it now also says "batched" for BSMTP.
902 PH/11 Added control=no_pipelining.
904 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
905 patch, slightly modified), and move the expansion of helo_data till after
906 the connection is made in the smtp transport (so it can use these
909 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
911 PH/14 Added log_selector = +pid.
913 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
915 PH/16 Add ${if forany and ${if forall.
917 PH/17 Added dsn_from option to vary the From: line in DSNs.
919 PH/18 Flush SMTP output before performing a callout, unless control =
920 no_callout_flush is set.
922 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
923 was true (the default) a successful delivery failed to delete the retry
924 item, thus causing premature timeout of the address. The bug is now
927 PH/20 Added hosts_avoid_pipelining to the smtp transport.
929 PH/21 Long custom messages for fakedefer and fakereject are now split up
930 into multiline reponses in the same way that messages for "deny" and
931 other ACL rejections are.
933 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
934 with slight modification.
936 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
937 draft, changing the syntax and factoring some duplicate code".
939 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
940 for deliveries of the second and subsequent messages over the same SMTP
943 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
944 ${reduce, with only minor "tidies".
946 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
948 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
949 expansion side effects.
951 PH/27 When a message times out after an over-quota error from an Exim-imposed
952 quota, the bounce message says "mailbox is full". This message was not
953 being given when it was a system quota that was exceeded. It now should
956 MH/03 Made $recipients available in local_scan(). local_scan() already has
957 better access to the recipient list through recipients_list[], but
958 $recipients can be useful in postmaster-provided expansion strings.
960 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
961 in the case of a MAIL command with additional options following the
962 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
963 were accidentally chopped off.
965 PH/29 SMTP synchronization checks are implemented when a command is read -
966 there is a check that no more input is waiting when there shouldn't be
967 any. However, for some commands, a delay in an ACL can mean that it is
968 some time before the response is written. In this time, more input might
969 arrive, invalidly. So now there are extra checks after an ACL has run for
970 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
971 pipelining has not been advertised.
973 PH/30 MH's patch to allow iscntrl() characters to be list separators.
975 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
976 returned in the SMTP response when smtp_return_error_details was false.
979 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
980 instead of the C I/O that was originally supplied, because problems were
983 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
984 Exim which did not show up earlier: it was assuming that a call to
985 SSL_CTX_set_info_callback() might give an error value. In fact, there is
986 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
987 was a macro that became an assignment, so it seemed to work. This has
988 changed to a proper function call with a void return, hence the compile
989 error. Exim's code has been fixed.
991 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
994 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
996 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
998 PH/37 If a message is not accepted after it has had an id assigned (e.g.
999 because it turns out to be too big or there is a timeout) there is no
1000 "Completed" line in the log. When some messages of this type were
1001 selected by exigrep, they were listed as "not completed". Others were
1002 picked up by some special patterns. I have improved the selection
1003 criteria to be more general.
1005 PH/38 The host_find_failed option in the manualroute router can now be set
1006 to "ignore", to completely ignore a host whose IP address cannot be
1007 found. If all hosts are ignored, the behaviour is controlled by the new
1008 host_all_ignored option.
1010 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1011 homing or because of multiple MX records with /mx) generated more than
1012 one IP address, and the following item turned out to be the local host,
1013 all the secondary addresses of the first item were incorrectly removed
1014 from the list, along with the local host and any following hosts (which
1015 is what is supposed to happen).
1017 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1018 whoever called Exim into the -H file. In the case of the daemon it was
1019 behaving confusingly. When first started, it used values for whoever
1020 started the daemon, but after a SIGHUP it used the Exim user (because it
1021 calls itself on a restart). I have changed the code so that it now always
1024 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1025 message are rejected with the same error (e.g. no authentication or bad
1026 sender address), and a DATA command is nevertheless sent (as can happen
1027 with PIPELINING or a stupid MUA), the error message that was given to the
1028 RCPT commands is included in the rejection of the DATA command. This is
1029 intended to be helpful for MUAs that show only the final error to their
1032 PH/42 Another patch from the Sieve maintainer.
1034 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1035 Eximstats - Fixed some broken HTML links and added missing column headers
1037 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1038 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1040 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1042 PH/43 Yet another patch from the Sieve maintainer.
1044 PH/44 I found a way to check for a TCP/IP connection going away before sending
1045 the response to the final '.' that terminates a message, but only in the
1046 case where the client has not sent further data following the '.'
1047 (unfortunately, this is allowed). However, in many cases there won't be
1048 any further data because there won't be any more messages to send. A call
1049 to select() can be used: if it shows that the input is "ready", there is
1050 either input waiting, or the socket has been closed. An attempt to read
1051 the next input character can distinguish the two cases. Previously, Exim
1052 would have sent an OK response which the client would never have see.
1053 This could lead to message repetition. This fix should cure that, at
1054 least in a lot of common cases.
1056 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1057 advertised in response to EHLO.
1063 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1064 fixed by 4.65/MH/01 (is this a record?) are fixed:
1066 (i) An empty string was always treated as zero by the numeric comparison
1067 operators. This behaviour has been restored.
1069 (ii) It is documented that the numeric comparison operators always treat
1070 their arguments as decimal numbers. This was broken in that numbers
1071 starting with 0 were being interpreted as octal.
1073 While fixing these problems I realized that there was another issue that
1074 hadn't been noticed. Values of message_size_limit (both the global option
1075 and the transport option) were treated as octal if they started with 0.
1076 The documentation was vague. These values are now always treated as
1077 decimal, and I will make that clear in the documentation.
1083 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1084 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1087 MH/01 Don't check that the operands of numeric comparison operators are
1088 integers when their expansion is in "skipping" mode (fixes bug
1089 introduced by 4.64-PH/07).
1091 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1092 child addresses, Exim now panics and dies. Previously, because the count
1093 is held in a short int, deliveries were likely to be lost. As such a
1094 large number of recipients for a single message is ridiculous
1095 (performance will be very, very poor), I have chosen to impose a limit
1096 rather than extend the field.
1102 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1103 leftover -K file (the existence of which was triggered by #402).
1104 While we were at it, introduced process PID as part of the -K
1105 filename. This should rule out race conditions when creating
1108 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1109 processing considerably. Previous code took too long for large mails,
1110 triggering a timeout which in turn triggers #401.
1112 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1113 in the DK code in transports.c. sendfile() is not really portable,
1114 hence the _LINUX specificness.
1116 TF/01 In the add_headers option to the mail command in an Exim filter,
1117 there was a bug that Exim would claim a syntax error in any
1118 header after the first one which had an odd number of characters
1121 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1122 callout verification, Exim cached a "reject" for the entire domain. This
1123 is correct for most verifications, but it is not correct for a recipient
1124 verification with use_sender or use_postmaster set, because in that case
1125 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1126 case of MAIL FROM:<> rejection from other early rejections (e.g.
1127 rejection of HELO). When verifying a recipient using a non-null MAIL
1128 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1129 Whatever the result of the callout, the value of the domain cache is
1130 left unchanged (for any other kind of callout, getting as far as trying
1131 RCPT means that the domain itself is ok).
1133 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1136 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1137 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1138 session, the daemon ignores ECONNECT errors and logs others; it now
1139 ignores EPIPE as well.
1141 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1142 (quoted-printable decoding).
1144 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1145 later the small subsequent patch to fix an introduced bug.
1147 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1149 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1151 PH/08 An error is now given if message_size_limit is specified negative.
1153 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1154 to be given (somewhat) arbitrary names.
1156 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1159 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1160 miscellaneous code fixes
1162 PH/10 Added the log_reject_target ACL modifier to specify where to log
1165 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1166 hostname. This is wrong, because it relates to the incoming message (and
1167 probably the interface on which it is arriving) and not to the outgoing
1168 callout (which could be using a different interface). This has been
1169 changed to use the value of the helo_data option from the smtp transport
1170 instead - this is what is used when a message is actually being sent. If
1171 there is no remote transport (possible with a router that sets up host
1172 addresses), $smtp_active_hostname is used.
1174 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1175 tweaks were necessary in order to get it to work (see also 21 below):
1176 (a) The code assumed that strncpy() returns a negative number on buffer
1177 overflow, which isn't the case. Replaced with Exim's string_format()
1179 (b) There were several signed/unsigned issues. I just did the minimum
1180 hacking in of casts. There is scope for a larger refactoring.
1181 (c) The code used strcasecmp() which is not a standard C function.
1182 Replaced with Exim's strcmpic() function.
1183 (d) The code set only $1; it now sets $auth1 as well.
1184 (e) A simple test gave the error "authentication client didn't specify
1185 service in request". It would seem that Dovecot has changed its
1186 interface. Fortunately there's a specification; I followed it and
1187 changed what the client sends and it appears to be working now.
1189 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1192 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1193 address A is aliased to B and C, where B exists and C does not. Without
1194 -v the output is "A verified" because verification stops after a
1195 successful redirection if more than one address is generated. However,
1196 with -v the child addresses are also verified. Exim was outputting "A
1197 failed to verify" and then showing the successful verification for C,
1198 with its parentage. It now outputs "B failed to verify", showing B's
1199 parentage before showing the successful verification of C.
1201 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1202 look up a TXT record in a specific list after matching in a combined
1205 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1206 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1207 they consult the DNS. I had assumed they would set it the way they
1208 wanted; and indeed my experiments on Linux seem to show that in some
1209 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1210 To be on the safe side, however, I have now made the interface to
1211 host_find_byname() similar to host_find_bydns(), with an argument
1212 containing the DNS resolver options. The host_find_byname() function now
1213 sets these options at its start, just as host_find_bydns() does. The smtp
1214 transport options dns_qualify_single and dns_search_parents are passed to
1215 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1216 of host_find_byname() use the default settings of RES_DEFNAMES
1217 (qualify_single) but not RES_DNSRCH (search_parents).
1219 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1220 spool_read_header() do less string testing, by means of a preliminary
1221 switch on the second character of optional "-foo" lines. (This is
1222 overdue, caused by the large number of possibilities that now exist.
1223 Originally there were few.) While I was there, I also converted the
1224 str(n)cmp tests so they don't re-test the leading "-" and the first
1225 character, in the hope this might squeeze out yet more improvement.
1227 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1228 flag allowing group syntax was set by the header_syntax check but not
1229 turned off, possible causing trouble later; (2) The flag was not being
1230 set at all for the header_verify test, causing "group"-style headers to
1231 be rejected. I have now set it in this case, and also caused header_
1232 verify to ignore an empty address taken from a group. While doing this, I
1233 came across some other cases where the code for allowing group syntax
1234 while scanning a header line wasn't quite right (mostly, not resetting
1235 the flag correctly in the right place). These bugs could have caused
1236 trouble for malformed header lines. I hope it is now all correct.
1238 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1239 with the "reply" argument non-NULL. The code, however (which originally
1240 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1241 but it didn't always do it. This confused somebody who was copying the
1242 code for some other use. I have removed all the tests.
1244 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1245 feature that was used to support insecure browsers during the U.S. crypto
1246 embargo. It requires special client support, and Exim is probably the
1247 only MTA that supported it -- and would never use it because real RSA is
1248 always available. This code has been removed, because it had the bad
1249 effect of slowing Exim down by computing (never used) parameters for the
1250 RSA_EXPORT functionality.
1252 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1253 authenticator to fail if there's a tab character in the incoming data
1254 (there should never be unless someone is messing about, as it's supposed
1255 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1256 if the connection is using TLS or if the remote IP is the same as the
1257 local IP, and the "valid-client-cert option" if a client certificate has
1260 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1261 authenticators. This can be used for authorization after authentication
1262 succeeds. (In the case of plaintext, it servers for both authentication
1265 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1266 if any retry times were supplied.
1268 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1269 connection, where there is no client IP address to check. In this
1270 situation, the verify now always succeeds.
1272 PH/25 Applied John Jetmore's -Mset patch.
1274 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1276 PH/27 In a string expansion for a processed (not raw) header when multiple
1277 headers of the same name were present, leading whitespace was being
1278 removed from all of them, but trailing whitespace was being removed only
1279 from the last one. Now trailing whitespace is removed from each header
1280 before concatenation. Completely empty headers in a concatenation (as
1281 before) are ignored.
1283 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1284 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1286 PH/29 [Removed. This was a change that I later backed out, and forgot to
1287 correct the ChangeLog entry (that I had efficiently created) before
1288 committing the later change.]
1290 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1291 address errors (4xx response to RCPT) over the same connection as other
1292 messages routed to the same hosts. Such deliveries are always "forced",
1293 so retry times are not inspected. This resulted in far too many retries
1294 for the affected addresses. The effect occurred only when there were more
1295 hosts than the hosts_max_try setting in the smtp transport when it had
1296 the 4xx errors. Those hosts that it had tried were not added to the list
1297 of hosts for which the message was waiting, so if all were tried, there
1298 was no problem. Two fixes have been applied:
1300 (i) If there are any address or message errors in an SMTP delivery, none
1301 of the hosts (tried or untried) are now added to the list of hosts
1302 for which the message is waiting, so the message should not be a
1303 candidate for sending over the same connection that was used for a
1304 successful delivery of some other message. This seems entirely
1305 reasonable: after all the message is NOT "waiting for some host".
1306 This is so "obvious" that I'm not sure why it wasn't done
1307 previously. Hope I haven't missed anything, but it can't do any
1308 harm, as the worst effect is to miss an optimization.
1310 (ii) If, despite (i), such a delivery is accidentally attempted, the
1311 routing retry time is respected, so at least it doesn't keep
1312 hammering the server.
1314 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1315 in ${readsocket because some servers need this prod.
1317 PH/32 Added some extra debug output when updating a wait-xxx database.
1319 PH/33 The hint "could be header name not terminated by colon", which has been
1320 given for certain expansion errors for a long time, was not being given
1321 for the ${if def:h_colon_omitted{... case.
1323 PH/34 The spec says: "With one important exception, whenever a domain list is
1324 being scanned, $domain contains the subject domain." There was at least
1325 one case where this was not true.
1327 PH/35 The error "getsockname() failed: connection reset by peer" was being
1328 written to the panic log as well as the main log, but it isn't really
1329 panic-worthy as it just means the connection died rather early on. I have
1330 removed the panic log writing for the ECONNRESET error when getsockname()
1333 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1334 runs only) independently of the message's sender address. This meant
1335 that, if the 4xx error was in fact related to the sender, a different
1336 message to the same recipient with a different sender could confuse
1337 things. In particualar, this can happen when sending to a greylisting
1338 server, but other circumstances could also provoke similar problems.
1339 I have changed the default so that the retry time for these errors is now
1340 based a combination of the sender and recipient addresses. This change
1341 can be overridden by setting address_retry_include_sender=false in the
1344 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1345 remote server are returned as part of bounce messages. This was not
1346 happening for LMTP over a pipe (the lmtp transport), but now it is the
1347 same for both kinds of LMTP.
1349 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1350 in header lines that were in fact CNAMEs. This is no longer the case.
1352 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1353 and queue runs started by the daemon processed all messages. This has
1354 been fixed so that -R and -S can now usefully be given with -q<time>.
1356 PH/40 Import PCRE release 6.7 (fixes some bugs).
1358 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1360 PH/42 Give an error if -q is specified more than once.
1362 PH/43 Renamed the variables $interface_address and $interface_port as
1363 $received_ip_address and $received_port, to make it clear that these
1364 values apply to message reception, and not to the outgoing interface when
1365 a message is delivered. (The old names remain recognized, of course.)
1367 PH/44 There was no timeout on the connect() call when using a Unix domain
1368 socket in the ${readsocket expansion. There now is.
1370 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1371 be meaningful with "accept".
1373 SC/01 Eximstats V1.43
1374 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1376 SC/02 Eximstats V1.44
1377 Use a glob alias rather than an array ref in the generated
1378 parser. This improves both readability and performance.
1380 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1381 Collect SpamAssassin and rejection statistics.
1382 Don't display local sender or destination tables unless
1383 there is data to show.
1384 Added average volumes into the top table text output.
1386 SC/04 Eximstats V1.46
1387 Collect data on the number of addresses (recipients)
1388 as well as the number of messages.
1390 SC/05 Eximstats V1.47
1391 Added 'Message too big' to the list of mail rejection
1392 reasons (thanks to Marco Gaiarin).
1394 SC/06 Eximstats V1.48
1395 Mainlog lines which have GMT offsets and are too short to
1396 have a flag are now skipped.
1398 SC/07 Eximstats V1.49 (Alain Williams)
1399 Added the -emptyok flag.
1401 SC/08 Eximstats V1.50
1402 Fixes for obtaining the IP address from reject messages.
1404 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1405 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1406 whitesspace changes from 4.64-PH/27
1408 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1411 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1412 are found, allow negative numbers in numeric criteria)
1414 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1416 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1417 to match changes made in 4.64-PH/43
1419 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1421 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1422 "domain" is actually an IP address, and force a failure. This locks out
1423 those revolvers/nameservers that support "A-for-A" lookups, in
1424 contravention of the specifications.
1426 PH/48 When a host name was looked up from an IP address, and the subsequent
1427 forward lookup of the name timed out, the host name was left in
1428 $sender_host_name, contrary to the specification.
1430 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1431 restricted to single-key lookups, Exim was not diagnosing an error if
1432 * or *@ was used with a query-style lookup.
1434 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1436 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1437 long ago, but noone interested enough thought of it. Let's just say that
1438 the "1.1" means that there are some new functions that weren't there at
1439 some point in the past.
1441 PH/51 Error processing for expansion failure of helo_data from an smtp
1442 transport during callout processing was broken.
1444 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1445 tested/used via the -bh/-bhc/-bs options.
1447 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1448 bug, fixed in subsequent PCRE releases).
1450 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1451 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1453 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1459 SC/01 Use a glob alias rather than an array ref in eximstats generated
1460 parser. This improves both readability and performance.
1462 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1463 Don't display local sender or destination tables in eximstats unless
1464 there is data to show.
1465 Added average volumes into the eximstats top table text output.
1467 SC/03 Collect data on the number of addresses (recipients) as well
1468 as the number of messages in eximstats.
1470 TF/01 Correct an error in the documentation for the redirect router. Exim
1471 does (usually) call initgroups() when daemonizing.
1473 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1474 with consistent privilege compared to when running as a daemon.
1476 TF/03 Note in the spec that $authenticated_id is not set for local
1477 submissions from trusted users.
1479 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1480 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1482 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1483 by adding some example configuration directives to the default
1484 configuration file. A little bit of work is required to uncomment the
1485 directives and define how usernames and passwords are checked, but
1486 there is now a framework to start from.
1488 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1489 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1490 without this. I don't know how relevant this is to other LDAP libraries.
1492 PH/02 Add the verb name to the "unknown ACL verb" error.
1494 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1496 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1498 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1499 directory not expanded when it should be if an expanded home directory
1500 was set for the address (which is overridden by the transport).
1502 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1505 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1506 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1507 because it is too late at that time, and has no effect.
1509 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1510 security issue with \' (bugzilla #107). I could not use the
1511 PQescapeStringConn() function, because it needs a PGconn value as one of
1514 PH/08 When testing addresses using -bt, indicate those final addresses that
1515 are duplicates that would not cause an additional delivery. At least one
1516 person was confused, thinking that -bt output corresponded to deliveries.
1517 (Suppressing duplicates isn't a good idea as you lose the information
1518 about possibly different redirections that led to the duplicates.)
1520 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1521 systems where poll() doesn't work, in particular OS X.
1523 PH/10 Added more information to debugging output for retry time not reached.
1525 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1526 operations in malware.c.
1528 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1531 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1532 syslog_duplication set false, log lines that would normally be written
1533 both the the main log and to the reject log were not written to syslog at
1536 PH/14 In the default configuration, change the use of "message" in ACL warn
1537 statements to "add_header".
1539 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1540 not followed by a command (e.g. "seen endif").
1542 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1543 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1546 PH/17 Added extra conditions to the default value of delay_warning_condition
1550 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1551 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1552 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1555 The Auto-Submitted: and various List- headers are standardised, whereas I
1556 don't think Precedence: ever was.
1558 PH/18 Refactored debugging code in route_finduser() to show more information,
1559 in particular, the error code if getpwnam() issues one.
1561 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1562 This is apparently needed in addition to the PH/07 change above to avoid
1563 any possible encoding problems.
1565 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1566 but not after initializing Perl.
1568 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1569 output them only if debugging. By default they are written stderr,
1570 apparently, which is not desirable.
1572 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1575 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1578 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1580 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1581 authenticated or an ident call has been made. Suppress the default
1582 values for $authenticated_id and $authenticated_sender (but permit -oMai
1583 and -oMas) when testing with -bh.
1585 PH/24 Re-jigged the order of the tests in the default configuration so that the
1586 tests for valid domains and recipients precede the DNS black list and CSA
1587 tests, on the grounds that those ones are more expensive.
1589 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1590 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1591 This bug exists in every version of Exim that I still have, right back to
1594 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1595 However, an attempt to turn on case-sensitivity in a regex key by
1596 including (?-i) didn't work because the subject string was already
1597 lowercased, and the effects were non-intuitive. It turns out that a
1598 one-line patch can be used to allow (?-i) to work as expected.
1604 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1605 other effects) broke the use of negated acl sub-conditions.
1607 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1610 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1611 "Deny" causes Exim to reject the incoming connection with a 554 error.
1612 Unfortunately, if there is a major crisis, such as a disk failure,
1613 tcp-wrappers gives "deny", whereas what one would like would be some
1614 kind of temporary error. A kludge has been added to help with this.
1615 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1616 554 error is used if errno is still zero or contains ENOENT (which occurs
1617 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1620 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1622 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1623 errors. Otherwise a message that provokes a temporary error (when other
1624 messages do not) can cause a whole host to time out.
1626 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1627 addresses were routed directly to files or pipes from a redirect router.
1628 File deliveries just didn't batch; pipe deliveries might have suffered
1631 PH/06 A failure to get a lock for a hints database would erroneously always say
1632 "Failed to get write lock", even when it was really a read lock.
1634 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1635 of 0600. This has been changed to use the value of the lockfile_mode
1636 option (which defaults to 0600).
1638 PH/08 Applied small patch from the Sieve maintainer.
1640 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1641 folder from quota calculations, a direct delivery into this folder messed
1642 up the contents of the maildirsize file. This was because the regex was
1643 used only to exclude .Trash (or whatever) when the size of the mailbox
1644 was calculated. There was no check that a delivery was happening into an
1645 excluded directory. This bug has been fixed by ignoring all quota
1646 processing for deliveries into excluded directories.
1648 PH/10 Added the maildirfolder_create_regex option to appendfile.
1654 PH/01 The code for finding all the local interface addresses on a FreeBSD
1655 system running IPv6 was broken. This may well have applied to all BSD
1656 systems, as well as to others that have similar system calls. The broken
1657 code found IPv4 interfaces correctly, but gave incorrect values for the
1658 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1659 that it would not match correctly against @[] and not recognize the IPv6
1662 PH/02 The ipliteral router was not recognizing addresses of the form user@
1663 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1665 PH/03 Added disable_ipv6.
1667 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1668 decoded form, because it is most often used to construct To: headers
1669 lines in autoreplies, and the decoded form may well be syntactically
1670 invalid. However, $reply_address has leading white space removed, and all
1671 newlines turned into spaces so that the autoreply transport does not
1674 PH/05 If group was specified without a user on a router, and no group or user
1675 was specified on a transport, the group from the router was ignored.
1677 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1678 for visible compile-time settings that can be used to change these
1679 numbers, for those that want even more. Backwards compatibility with old
1680 spool files has been maintained. However, going back to a previous Exim
1681 release will lost any variables that are in spool files.
1683 PH/07 Two small changes when running in the test harness: increase delay when
1684 passing a TCP/IP connection to a new process, in case the original
1685 process has to generate a bounce, and remove special handling of
1686 127.0.0.2 (sic), which is no longer necessary.
1688 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1689 be the same on different OS.
1691 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1694 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1695 whether --show-vars was specified or not
1697 JJ/02 exipick: Added support for new ACL variable spool format introduced
1700 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1701 syntactically invalid From: or Reply-to: line, and a filter used this to
1702 generate an autoreply, and therefore failed to obtain an address for the
1703 autoreply, Exim could try to deliver to a non-existent relative file
1704 name, causing unrelated and misleading errors. What now happens is that
1705 it logs this as a hard delivery error, but does not attempt to create a
1708 PH/11 The exinext utility has a -C option for testing purposes, but although
1709 the given file was scanned by exinext itself; it wasn't being passed on
1710 when Exim was called.
1712 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1713 an end-of-file indication when reading a command response.
1715 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1716 compiled. In many other places in Exim, IPv6 addresses are always
1717 recognized, so I have changed this. It also means that IPv4 domain
1718 literals of the form [IPV4:n.n.n.n] are now always recognized.
1720 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1721 used if the router is not running as root, for example, when verifying at
1722 ACL time, or when using -bh. The debugging output from this situation was
1723 non-existent - all you got was a failure to exec. I have made two
1726 (a) Failures to set uid/gid, the current directory, or a process leader
1727 in a subprocess such as that created by queryprogram now generate
1728 suitable debugging ouput when -d is set.
1730 (b) The queryprogram router detects when it is not running as root,
1731 outputs suitable debugging information if -d is set, and then runs
1732 the subprocess without attempting to change uid/gid.
1734 PH/15 Minor change to Makefile for building test_host (undocumented testing
1737 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1738 additional section of a DNS packet that returns MX or SRV records.
1739 Instead, it always explicitly searches for A/AAAA records. This avoids
1740 major problems that occur when a DNS server includes only records of one
1741 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1742 fixed another bug: if SRV records were looked up and the corresponding
1743 address records were *not* found in the additional section, the port
1744 values from the SRV records were lost.
1746 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1747 using the correct key (the original address) when searching the retry
1748 rules in order to find which one to use for generating the retry hint.
1750 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1751 adding the default one. Similarly, if it contains a Reply-To: header, the
1752 errors_reply_to option, if set, is not used.
1754 PH/19 When calculating a retry time, Exim used to measure the "time since
1755 failure" by looking at the "first failed" field in the retry record. Now
1756 it does not use this if it is later than than the arrival time of the
1757 message. Instead it uses the arrival time. This makes for better
1758 behaviour in cases where some deliveries succeed, thus re-setting the
1759 "first failed" field. An example is a quota failure for a huge message
1760 when small messages continue to be delivered. Without this change, the
1761 "time since failure" will always be short, possible causing more frequent
1762 delivery attempts for the huge message than are intended.
1763 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1765 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1766 $1, $2, $3) because the numerical variables can be reset during some
1767 expansion items (e.g. "match"), thereby losing the authentication data.
1769 PH/21 Make -bV show the size of off_t variables so that the test suite can
1770 decide whether to run tests for quotas > 2G.
1772 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1773 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1774 filecount value is greater than 2G or if a quota value is greater than 2G
1775 on a system where the size of off_t is not greater than 4, a panic error
1778 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1779 never match. The debug and -bh output now contains an explicit error
1780 message indicating a malformed IPv4 address or mask.
1782 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1783 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1784 PH/23 above applies.
1786 PH/25 Do not write to syslog when running in the test harness. The only
1787 occasion when this arises is a failure to open the main or panic logs
1788 (for which there is an explicit test).
1790 PH/26 Added the /no_tell option to "control=freeze".
1792 PH/27 If a host name lookup failed very early in a connection, for example, if
1793 the IP address matched host_lookup and the reverse lookup yielded a name
1794 that did not have a forward lookup, an error message of the form "no IP
1795 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1796 could be logged. Now it outputs the IP address instead of "NULL".
1798 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1799 allows the sender and the authenticated sender to be set when
1800 submitting a message from within Exim. Since child_open_exim() is
1801 documented for local_scan(), the new function should be too.
1803 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1804 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1805 results in an empty string is now treated as unset.
1807 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1809 PH/31 Added sender_verify_fail logging option.
1811 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1812 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1813 tidied the source and removed it altogether.
1815 PH/33 When a queue run was abandoned because the load average was too high, a
1816 log line was always written; now it is written only if the queue_run log
1817 selector is set. In addition, the log line for abandonment now contains
1818 information about the queue run such as the pid. This is always present
1819 in "start" and "stop" lines but was omitted from the "abandon" line.
1821 PH/34 Omit spaces between a header name and the colon in the error message that
1822 is given when verify = headers_syntax fails (if there are lots of them,
1823 the message gets confusing).
1825 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1826 names, as there are now some PTR records that contain slashes. This check
1827 is only to protect against broken name servers that fall over on strange
1828 characters, so the fact that it applies to all lookups doesn't matter.
1830 PH/36 Now that the new test suite is complete, we can remove some of the
1831 special code in Exim that was needed for the old test suite. For example,
1832 sorting DNS records because real resolvers return them in an arbitrary
1833 order. The new test suite's fake resolver always returns records in the
1836 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1837 bounces) except when queue_only is set, to avoid logging races between
1838 the different processes.
1840 PH/38 Panic-die if .include specifies a non-absolute path.
1842 PH/39 A tweak to the "H" retry rule from its user.
1844 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1845 a label. They prevented compilation on older perls.
1847 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1848 a warning to be raised on newish perls.
1850 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1851 on queue. Changes to match documented behaviour of showing count of
1852 messages matching specified criteria.
1854 PH/40 Changed the default ident timeout from 30s to 5s.
1856 PH/41 Added support for the use of login_cap features, on those BSD systems
1857 that have them, for controlling the resources used by pipe deliveries.
1859 PH/42 The content-scanning code uses fopen() to create files in which to put
1860 message data. Previously it was not paying any attention to the mode of
1861 the files. Exim runs with umask(0) because the rest of the code creates
1862 files with open(), and sets the required mode explicitly. Thus, these
1863 files were ending up world-writeable. This was not a big issue, because,
1864 being within the spool directory, they were not world-accessible. I have
1865 created a function called modefopen, which takes an additional mode
1866 argument. It sets umask(777), creates the file, chmods it to the required
1867 mode, then resets the umask. All the relevant calls to fopen() in the
1868 content scanning code have been changed to use this function.
1870 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1871 to 24 hours. This avoids potential overflow problems when processing G
1872 and H retry rules. I suspect nobody ever tinkers with this value.
1874 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1876 PH/45 When the plaintext authenticator is running as a client, the server's
1877 challenges are checked to ensure they are valid base64 strings. By
1878 default, the authentication attempt is cancelled if an invalid string is
1879 received. Setting client_ignore_invalid_base64 true ignores these errors.
1880 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1881 they are received. Thus, the responses can be made to depend on the
1882 challenges. If an invalid string is ignored, an empty string is placed in
1885 PH/46 Messages that are created by the autoreply transport now contains a
1886 References: header, in accordance with RFCs 2822 and 3834.
1888 PH/47 Added authenticated_sender_force to the smtp transport.
1890 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1892 PH/49 Installed latest patch from the Sieve maintainer.
1894 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1895 was also set, the appendfile transport was unnecessarily scanning a
1896 directory of message files (e.g. for maildir delivery) to find the count
1897 of files (along with the size), even though it did not need this
1898 information. It now does the scan only if it needs to find either the
1899 size of the count of files.
1901 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1903 PH/52 Two bugs concerned with error handling when the smtp transport is
1906 (i) Exim was not creating retry information for temporary errors given
1907 for individual recipients after the DATA command when the smtp transport
1908 was used in LMTP mode. This meant that they could be retried too
1909 frequently, and not timed out correctly.
1911 (ii) Exim was setting the flag that allows error details to be returned
1912 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1913 recipients that were returned after the DATA command.
1915 PH/53 This is related to PH/52, but is more general: for any failing address,
1916 when detailed error information was permitted to be returned to the
1917 sender, but the error was temporary, then after the final timeout, only
1918 "retry timeout exceeded" was returned. Now it returns the full error as
1919 well as "retry timeout exceeded".
1921 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1922 do this, and (what is worse) MTAs that accept it.
1924 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1925 will now be deprecated.
1927 PH/56 New os.c-cygwin from the Cygwin maintainer.
1929 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1930 formats (previously only available in exim formats via -bpr, -bpru,
1931 and -bpra. Now also available in native and exiqgrep formats)
1933 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1934 with very large, slow to parse queues
1936 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1938 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1940 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1941 responses to authentication challenges, though it was showing the
1942 challenges; (ii) I've removed the CR characters from the debug output for
1945 PH/58 Allow for the insertion of a newline as well as a space when a string
1946 is turned into more than one encoded-word during RFC 2047 encoding. The
1947 Sieve code now uses this.
1949 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1950 data_4xx, lost_connection, tls_required.
1952 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1953 message was being sent as an SMTP response.
1955 PH/61 Add -l and -k options to exicyclog.
1957 PH/62 When verifying, if an address was redirected to one new address, so that
1958 verification continued, and the new address failed or deferred after
1959 having set something in $address_data, the value of $address_data was not
1960 passed back to the ACL. This was different to the case when no
1961 redirection occurred. The value is now passed back in both cases.
1963 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1964 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1965 use login_cap.h, so on its own it isn't the distinguishing feature. The
1966 new name refers directly to the setclassresources() function.
1968 PH/65 Added configuration files for NetBSD3.
1970 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1972 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1973 is preferred over IPv4.
1975 PH/68 The bounce_return_message and bounce_return_body options were not being
1976 honoured for bounces generated during the reception of non-SMTP messages.
1977 In particular, this applied to messages rejected by the ACL. This bug has
1978 been fixed. However, if bounce_return_message is true and bounce_return_
1979 body is false, the headers that are returned for a non-SMTP message
1980 include only those that have been read before the error was detected.
1981 (In the case of an ACL rejection, they have all been read.)
1983 PH/69 The HTML version of the specification is now built in a directory called
1984 spec_html instead of spec.html, because the latter looks like a path with
1985 a MIME-type, and this confuses some software.
1987 PH/70 Catch two compiler warnings in sieve.c.
1989 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1990 function verify_get_ident() calls ip_connect() to connect a socket, but
1991 if the "connect()" function timed out, ip_connect() used to close the
1992 socket. However, verify_get_ident() also closes the socket later, and in
1993 between Exim writes to the log, which may get opened at this point. When
1994 the socket was closed in ip_connect(), the log could get the same file
1995 descriptor number as the socket. This naturally causes chaos. The fix is
1996 not to close the socket in ip_connect(); the socket should be closed by
1997 the function that creates it. There was only one place in the code where
1998 this was missing, in the iplookup router, which I don't think anybody now
1999 uses, but I've fixed it anyway.
2001 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2002 well as to direct DNS lookups. Otherwise the handling of names in host
2003 lists is inconsistent and therefore confusing.
2009 PH/01 Two changes to the default runtime configuration:
2011 (1) Move the checks for relay_from_hosts and authenticated clients from
2012 after to before the (commented out) DNS black list checks.
2014 (2) Add control=submission to the relay_from_hosts and authenticated
2015 clients checks, on the grounds that messages accepted by these
2016 statements are most likely to be submissions.
2018 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2020 (1) Generate an error if the third argument for the ${prvs expansion is
2023 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2026 (3) Reset the variables that are obtained from the first argument of
2027 ${prvscheck and used in the second argument before leaving the code,
2028 because their memory is reclaimed, so using them afterwards may do
2031 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2032 one (it's much easier than Tom thought :-).
2034 (5) Because of (4), we can now allow for the use of $prvscheck_result
2035 inside the third argument.
2037 PH/03 For some reason, the default setting of PATH when running a command from
2038 a pipe transport was just "/usr/bin". I have changed it to
2041 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2042 anything to be listed in the output from -bV.
2044 PH/05 When a filter generated an autoreply, the entire To: header line was
2045 quoted in the delivery log line, like this:
2047 => >A.N.Other <ano@some.domain> <original@ddress> ...
2049 This has been changed so that it extracts the operative address. There
2050 may be more than one such address. If so, they are comma-separated, like
2053 => >ano@some.domain,ona@other.domain <original@ddress> ...
2055 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2056 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2057 not being looked up in the rDNS to get a host name, Exim was showing the
2058 IP address twice in Received: lines, even though the IP addresses were
2059 identical. For example:
2061 Received: from [1.2.3.4] (helo=[1.2.3.4])
2063 However, if the real host name was known, it was omitting the HELO data
2064 if it matched the actual IP address. This has been tidied up so that it
2065 doesn't show the same IP address twice.
2067 PH/07 When both +timestamp and +memory debugging was on, the value given by
2068 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2069 called by the debug printing, thereby overwriting the timestamp buffer.
2070 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2072 PH/08 When the original message was included in an autoreply transport, it
2073 always said "this is a copy of the message, including all the headers",
2074 even if body_only or headers_only was set. It now gives an appropriate
2077 PH/09 Applied a patch from the Sieve maintainer which:
2079 o fixes some comments
2080 o adds the (disabled) notify extension core
2081 o adds some debug output for the result of if/elsif tests
2082 o points to the current vacation draft in the documentation
2083 and documents the missing references header update
2087 o fixes a bug in processing the envelope test (when testing
2088 multiple envelope elements, the last element determinted the
2091 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2092 Electronic Mail") by including:
2094 Auto-submitted: auto-generated
2096 in the messages that it generates (bounce messages and others, such as
2097 warnings). In the case of bounce messages for non-SMTP mesages, there was
2098 also a typo: it was using "Auto_submitted" (underscore instead of
2099 hyphen). Since every message generated by Exim is necessarily in response
2100 to another message, thes have all been changed to:
2102 Auto-Submitted: auto-replied
2104 in accordance with these statements in the RFC:
2106 The auto-replied keyword:
2108 - SHOULD be used on messages sent in direct response to another
2109 message by an automatic process,
2111 - MUST NOT be used on manually-generated messages,
2113 - MAY be used on Delivery Status Notifications (DSNs) and Message
2114 Disposition Notifications (MDNs),
2116 - MUST NOT be used on messages generated by automatic or periodic
2117 processes, except for messages which are automatic responses to
2120 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2121 to the default Received: header definition.
2123 PH/12 Added log selector acl_warn_skipped (default on).
2125 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2126 variables because (a) they are in the wrong storage pool and (b) even if
2127 they were copied, it wouldn't work properly because of the caching.
2129 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2130 checking when decoding. Apparently there are clients that generate
2131 overlong encoded strings. Why am I not surprised?
2133 PH/15 If the first argument of "${if match_address" was not empty, but did not
2134 contain an "@" character, Exim crashed. Now it writes a panic log message
2135 and treats the condition as false.
2137 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2139 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2140 extension "envelope-auth". The code is finished and in agreement with
2141 other implementations, but there is no documentation so far and in fact,
2142 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2143 not changing the active code.
2145 Print executed "if" and "elsif" statements when debugging is used. This
2146 helps a great deal to understand what a filter does.
2148 Document more things not specified clearly in RFC3028. I had all this
2149 sorted out, when out of a sudden new issues came to my mind. Oops."
2151 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2154 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2155 canonical form (as documented). However, after a host name lookup from
2156 the IP address, check_host() was doing a simple string comparison with
2157 addresses acquired from the DNS when checking that the found name did
2158 have the original IP as one of its addresses. Since any found IPv6
2159 addresses are likely to be in abbreviated form, the comparison could
2160 fail. Luckily, there already exists a function for doing the comparison
2161 by converting both addresses to binary, so now that is used instead of
2162 the text comparison.
2164 PH/20 There was another similar case to PH/19, when a complete host name was
2165 given in a host list; looking up its IP address could give an abbreviated
2166 form, whereas the current host's name might or might not be abbreviated.
2167 The same fix has been applied.
2173 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2174 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2177 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2178 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2180 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2182 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2183 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2184 Also, for "Auto-Submitted", treat anything other than "no" as
2185 non-personal, in accordance with RFC 3834. (Previously it treated
2186 anything starting "auto-" as non-personal.)
2188 TF/01 The control=submission/name=... option had a problem with syntax
2189 errors if the name included a slash character. The /name= option
2190 now slurps the rest of the string, so it can include any characters
2191 but it must come last in the list of options (after /sender_retain
2194 PH/05 Some modifications to the interface to the fake nameserver for the new
2202 TK/01 Added the "success_on_redirect" address verification option. See
2203 NewStuff for rationale and an example.
2205 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2207 PH/02 Patch to exigrep to allow it to work on syslog lines.
2209 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2210 fread() to skip over the body file's header line, because in Cygwin the
2211 header line is locked and is inaccessible.
2213 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2214 co-exist for some time) to make it clear that it is the Exim ID that is
2215 referenced, not the Message-ID: header line.
2217 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2218 string_format() function, because snprintf() does not exist on all
2221 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2222 ${stat: expansion item.
2224 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2225 protocol synchronization error", to keep the pedants happy.
2227 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2228 well as for IRIX systems, when gcc is being used. See the host.c source
2231 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2233 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2236 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2237 transport and to the smtp transport in LMTP mode.
2239 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2241 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2242 run a filter in a subprocess. This could lead to confusion in subsequent
2243 lookups in the parent process. There should also be a search_tidyup() at
2244 the end of the subprocess.
2246 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2247 only if the host matched helo_try_verify_hosts, which caused the
2248 verification to occur when the EHLO/HELO command was issued. The ACL just
2249 tested the remembered result. Now, if a previous verification attempt has
2250 not happened, "verify = helo" does it there and then.
2252 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2254 TK/03 Fix log output including CR from clamd.
2256 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2257 exist provoked a memory error which could cause a segfault.
2259 PH/15 Installed PCRE 6.2
2261 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2263 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2264 of the problem. Specifically, suggested +O2 rather than +O1 for the
2267 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2269 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2270 record, the port was not correctly logged when the outgoing_port log
2271 selector was set (it logged the transort's default port).
2273 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2274 fallback_hosts, and "hosts" in the smtp transport.
2276 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2277 host errors such as "Connection refused".
2279 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2280 authentication with radiusclient 0.4.9:
2282 - Error returned from rc_read_config was caught wrongly
2283 - Username/password not passed on to radius server due to wrong length.
2285 The presumption is that some radiusclient API changes for 4.51/PH/17
2286 were not taken care of correctly. The code is still untested by me (my
2287 Linux distribution still has 0.3.2 of radiusclient), but it was
2288 contributed by a Radius user.
2290 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2291 expanding the "port" option of the smtp transport.
2293 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2294 while reading a MIME header. Thanks to Tom Hughes for a patch.
2296 PH/24 Include config.h inside local_scan.h so that configuration settings are
2299 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2300 that in an ACL for RCPT (for example), you can examine exactly what was
2303 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2304 commands, but it was not correctly comparing the address with the actual
2305 client host address. Thus, it would show the EHLO address in Received:
2306 header lines when this was not necessary.
2308 PH/27 Added the % operator to ${eval:}.
2310 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2311 it should be ignoring failures (because with -C, for example, it has lost
2312 privilege). It wasn't ignoring creation failures other than "already
2315 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2318 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2319 because an input line was too long, either on its own, or by virtue of
2320 too many continuations, the temporary file was not being removed, and the
2321 return code was incorrect.
2323 PH/31 Missing "BOOL" in function definition in filtertest.c.
2325 PH/32 Applied Sieve patches from the maintainer.
2327 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2329 PH/33 Added "verify = not_blind".
2331 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2332 Local/Makefile (with some defaults set). These are used in built scripts
2333 such as exicyclog, but they have never been used in the exim_install
2334 script (though there are many overriding facilities there). I have
2335 arranged that the exim_install script now takes note of these two
2338 PH/35 Installed configuration files for Dragonfly.
2340 PH/36 When a locally submitted message by a trusted user did not contain a
2341 From: header, and the sender address was obtained from -f or from an SMTP
2342 MAIL command, and the trusted user did not use -F to supply a sender
2343 name, $originator_name was incorrectly used when constructing a From:
2344 header. Furthermore, $originator_name was used for submission mode
2345 messages from external hosts without From: headers in a similar way,
2346 which is clearly wrong.
2348 PH/37 Added control=suppress_local_fixups.
2350 PH/38 When log_selector = +received_sender was set, and the addition of the
2351 sender made the log line's construction buffer exactly full, or one byte
2352 less than full, an overflow happened when the terminating "\n" was
2355 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2356 when the result of a list match is failure because a DNS lookup failed.
2358 PH/40 RM_COMMAND is now used in the building process.
2360 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2361 the "build-* directories that it finds.
2363 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2364 address in a domain literal was a prefix of an interface address.
2366 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2367 when verifying a sender address, unless rewrite_headers is false.
2369 PH/44 Wrote a long comment about why errors_to addresses are verified as
2370 recipients, not senders.
2372 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2373 the ratelimit ACL was added.
2375 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2377 PH/46 Added extra information about PostgreSQL errors to the error string.
2379 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2380 avoiding the need to install special zones in a real server. This is
2381 backwards compatible; if it can't find the fake resolver, it drops back.
2382 Thus, both old and new test suites can be run.
2384 TF/02 Added util/ratelimit.pl
2386 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2387 clock is set back in time.
2389 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2390 Candler <B.Candler@pobox.com>.
2392 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2393 OK for addresses that are the result of redirections.
2395 PH/48 A number of further additions for the benefit of the new test suite,
2396 including a fake gethostbyname() that interfaces to the fake DNS resolver
2399 TF/06 The fix for widen_domains has also been applied to qualify_single and
2400 search_parents which are the other dnslookup options that can cause
2403 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2406 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2408 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2409 service" if it encountered an MX record with an empty target hostname.
2410 The message is now "an MX or SRV record indicated no SMTP service".
2412 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2413 if the DNS of the sending site is misconfigured. This is quite a
2414 common situation. This change restores the behaviour of treating a
2415 helo verification defer as a failure.
2417 PH/51 If self=fail was set on a router, the bounce message did not include the
2418 actual error message.
2424 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2426 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2427 command itself ended in error, the underlying message about the transport
2428 filter timeout was being overwritten with the pipe command error. Now the
2429 underlying error message should be appended to the second error message.
2431 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2433 PH/02 Reduce the amount of output that "make" produces by default. Full output
2434 can still be requested.
2436 PH/03 The warning log line about a condition test deferring for a "warn" verb
2437 was being output only once per connection, rather than after each
2438 occurrence (because it was using the same function as for successful
2439 "warn" verbs). This seems wrong, so I have changed it.
2441 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2442 it should not have, which might have caused a crash in the right
2443 circumstances, but probably never did.
2445 PH/04 Installed a modified version of Tony Finch's patch to make submission
2446 mode fix the return path as well as the Sender: header line, and to
2447 add a /name= option so that you can make the user's friendly name appear
2450 TF/03 Added the control = fakedefer ACL modifier.
2452 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2453 Mark Lowes for thorough testing.
2455 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2457 TK/03 Merged latest SRS patch from Miles Wilton.
2459 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2460 with the definition in sysexits.h (which is #included earlier).
2461 Fortunately, Exim does not actually use EX_OK. The code used to try to
2462 preserve the sysexits.h value, by assumimg that macro definitions were
2463 scanned for macro replacements. I have been disabused of this notion,
2464 so now the code just undefines EX_OK before #including unistd.h.
2466 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2467 in the smtp transport. When a block could not be written in a single
2468 write() function, the timeout was being re-applied to each part-write.
2469 This seems wrong - if the receiver was accepting one byte at a time it
2470 would take for ever. The timeout is now adjusted when this happens. It
2471 doesn't have to be particularly precise.
2473 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2474 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2476 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2477 without a domain if the check to <postmaster@domain> fails.
2479 SC/01 Eximstats: added -xls and the ability to specify output files
2480 (patch written by Frank Heydlauf).
2482 SC/02 Eximstats: use FileHandles for outputing results.
2484 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2486 SC/04 Eximstats: fixed display of large numbers with -nvr option
2488 SC/05 Eximstats: fixed merging of reports with empty tables.
2490 SC/06 Eximstats: added the -include_original_destination flag
2492 SC/07 Eximstats: removed tabs and trailing whitespace.
2494 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2496 TK/06 MBOX spool code: Add real "From " MBOX separator line
2497 so the .eml file is really in mbox format (even though
2498 most programs do not really care). Patch from Alex Miller.
2500 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2501 The latter is generated from $received_to and is only set if the
2502 message has one envelope recipient. SA can use these headers,
2503 obviously out-of-the-box. Patch from Alex Miller.
2505 PH/08 The ${def test on a variable was returning false if the variable's
2506 value was "0", contrary to what the specification has always said!
2507 The result should be true unless the variable is empty.
2509 PH/09 The syntax error of a character other than { following "${if
2510 def:variable_name" (after optional whitespace) was not being diagnosed.
2511 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2512 accidental colon was present, for example, could give incorrect results.
2514 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2515 result is used (not including appendfile, where other changes are about
2518 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2519 This involved changing a lot of size variables from int to off_t. It
2520 should work with maildirs and everything.
2522 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2523 spamd dying while we are connected to it.
2525 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2528 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2529 function for BDB 4.3.
2531 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2533 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2534 the sources for PCRE have been re-arranged and more files are now
2537 PH/15 The code I had for printing potentially long long variables in PH/11
2538 above was not the best (it lost precision). The length of off_t variables
2539 is now inspected at build time, and an appropriate printing format (%ld
2540 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2541 to be "long long int" or "long int". This is needed for the internal
2542 formatting function string_vformat().
2544 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2545 the configuration file to be ":syslog", then the script "guesses" where
2546 the logs files are, rather than using the compiled in default. In our
2547 case the guess is not the same as the compiled default, so the script
2548 suddenly stopped working when I started to use syslog. The patch checks
2549 to see if log_file_path is "". If so, it attempts to read it from exim
2550 with no configuration file to get the compiled in version, before it
2551 falls back to the previous guessing code."
2553 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2554 implementing BATV in an Exim configuration. See NewStuff for the gory
2557 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2558 Makefile that are specific to HP-UX.
2560 PH/18 If the "use_postmaster" option was set for a recipient callout together
2561 with the "random" option, the postmaster address was used as the MAIL
2562 FROM address for the random test, but not for the subsequent recipient
2563 test. It is now used for both.
2565 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2566 patch removes a few documentation additions to RFC 3028, because the
2567 latest draft now contains them. It adds the new en;ascii-case comparator
2568 and a new error check for 8bit text in MIME parts. Comparator and
2569 require names are now matched exactly. I enabled the subaddress
2570 extension, but it is not well tested yet (read: it works for me)."
2572 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2573 rework some of the code of TK/09 above to avoid the hardwired use of
2574 "%lld" and "long long". Replaced the call to snprintf() with a call to
2577 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2578 records point to non-existent hosts", "retry timeout exceeded", and
2579 "retry time not reached for any host after a long failure period".
2581 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2582 experimental DomainKeys support:
2584 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2585 (2) On an error such as an illegally used "control", the wrong name for
2586 the control was given.
2588 These problems did NOT occur unless DomainKeys support was compiled.
2590 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2592 PH/24 Added ${if match_ip condition.
2594 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2595 databases so that it will be absolutely obvious if a crash occurs in the
2596 DB library. This is a regular occurrence (often caused by mis-matched
2599 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2600 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2601 picked up on a user's system that detects such things. There doesn't seem
2602 to be a gcc warning option for this - only an attribute that has to be
2603 put on the function's prototype. It seems that in Fedora Core 4 they have
2604 set this on a number of new functions. No doubt there will be more in due
2607 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2608 specify a transport. However, if an address that was verified by such a
2609 router was the subject of a callout, Exim crashed because it tried to
2610 read the rcpt_include_affixes from the non-existent transport. Now it
2611 just assumes that the setting of that option is false. This bug was
2612 introduced by 4.51/PH/31.
2614 PH/28 Changed -d+all to exclude +memory, because that information is very
2615 rarely of interest, but it makes the output a lot bigger. People tend to
2616 do -d+all out of habit.
2618 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2619 code in os-type was giving problems when libc.so lives in lib64, like on
2622 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2623 aren't the modern standard, and it seems that some systems' include files
2624 don't always have them. Exim was already checking for some of the newer
2625 ones like T_AAAA, and defining it itself. I've added checks for all the
2626 record types that Exim uses.
2628 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2629 not automatically generating a new one, as it is supposed to. This
2630 prevented TLS from working. If the file did exist, but contained invalid
2631 data, a new version was generated, as expected. It was only the case of a
2632 non-existent file that was broken.
2634 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2635 with a change in libdomainkeys > 0.64.
2637 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2638 from DNS. If the selector record carries the flag, it now has
2639 precedence over the domain-wide flag.
2641 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2643 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2644 the use of a port name that isn't defined in /etc/services) occurred, the
2645 message was deferred as in a normal delivery, and thus remained on the
2646 spool, instead of being failed because of the mua_wrapper setting. This
2647 is now fixed, and I tidied up some of the mua_wrapper messages at the
2650 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2651 the messages in a hash of arrays rather than using individual hashes.
2652 This is a bit cleaner and results in dramatic memory savings, albeit
2653 at a slight CPU cost.
2655 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2656 as requested by Marc Sherman.
2658 SC/10 Eximstats: added histograms for user specified patterns as requested
2661 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2663 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2664 fopen() in the content-scanning modules that did not already have it.
2670 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2671 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2673 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2675 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2677 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2678 to test Sieve filters that use "vacation".
2680 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2681 that changes the way the GnuTLS parameters are stored in the cache file.
2682 The new format can be generated externally. For backward compatibility,
2683 if the data in the cache doesn't make sense, Exim assumes it has read an
2684 old-format file, and it generates new data and writes a new file. This
2685 means that you can't go back to an older release without removing the
2688 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2689 work if there are any delivery delays because "one_time" forces the
2690 parent to be marked "delivered", so its unseen clone is never tried
2691 again. For this reason, Exim now forbids the simultaneous setting of
2694 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2695 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2696 entry if you want to know the details. The fix, however, neglected to
2697 consider the case when local delivery batching is involved. The test for
2698 "previously delivered" was not happening when checking to see if an
2699 address could be batched with a previous (undelivered) one; under
2700 certain circumstances this could lead to multiple deliveries to the same
2703 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2704 in its include files, and this causes problems building Exim.
2706 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2707 header_syntax) but Exim was just ignoring anything given after a slash.
2708 In particular, this caused confusion with an attempt to use "verify =
2709 reverse_host_lookup/defer_ok". An error is now given when options are
2710 supplied for verify items that do not have them. (Maybe reverse_host_
2711 lookup should have a defer_ok option, but that's a different point.)
2713 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2714 defined by RFC 821) to 2048, because there were problems with some AUTH
2715 commands, and RFC 1869 says the size should be increased for extended
2716 SMTP commands that take arguments.
2718 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2721 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2722 "unknown" error; now it says that the functionality isn't in the binary.
2724 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2725 an address' error message when a string expansion fails (syntax or
2726 whatever). Otherwise the password may appear in the log. Following change
2727 PH/42 below, there is no longer a chance of it appearing in a bounce
2730 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2732 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2733 of its addresses was ever tried. (Bugzilla bug #2.)
2735 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2736 the result incorrectly in the debug output. (It correctly added a newline
2737 to what was transported.)
2739 TF/01 Added $received_time.
2741 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2742 commented out examples of how to interface to a virus scanner and to
2743 SpamAssassin. Also added commented examples of av_scanner and
2744 spamd_address settings.
2746 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2747 and controls are allowed in which ACLs. There were a couple of minor
2748 errors. Some of the entries in the conditions table (which is a table of
2749 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2750 negation of where the condition IS allowed.
2752 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2754 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2755 header file does not have a version number, so I've had to invent a new
2756 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2757 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2758 radiusclient), but it was contributed by a Radius user.
2760 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2761 files or directories, for OpenSSL.
2763 PH/19 When an Exim process that is running as root has to create an Exim log
2764 file, it does so in a subprocess that runs as exim:exim so as to get the
2765 ownership right at creation (otherwise, other Exim processes might see
2766 the file with the wrong ownership). There was no test for failure of this
2767 fork() call, which would lead to the process getting stuck as it waited
2768 for a non-existent subprocess. Forks do occasionally fail when resources
2769 run out. I reviewed all the other calls to fork(); they all seem to check
2772 PH/20 When checking for unexpected SMTP input at connect time (before writing
2773 the banner), Exim was not dealing correctly with a non-positive return
2774 from the read() function. If the client had disconnected by this time,
2775 the result was a log entry for a synchronization error with an empty
2776 string after "input=" when read() returned zero. If read() returned -1
2777 (an event I could not check), uninitialized data bytes were printed.
2778 There were reports of junk text (parts of files, etc) appearing after
2781 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2783 PH/22 Added support for macro redefinition, and (re)definition in between
2784 driver and ACL definitions.
2786 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2787 forgetting to use the resulting value; it was using the unexpanded value.
2789 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2790 hadn't been configured. The fix is from Juergen Kreileder, who
2791 understands it better than I do:
2793 "Here's what I see happening with three configured cyrus_sasl
2794 authenticators configured (plain, login, cram-md5):
2796 On startup auth_cyrus_sasl_init() gets called for each of these.
2797 This means three calls to sasl_listmech() without a specified mech_list.
2798 => SASL tests which mechs of all available mechs actually work
2799 => three warnings about OTP not working
2800 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2802 With the patch, sasl_listmech() also gets called three times. But now
2803 SASL's mech_list option is set to the server_mech specified in the the
2804 authenticator. Or in other words, the answer from sasl_listmech()
2805 gets limited to just the mech you're testing for (which is different
2807 => the return list contains just 'plain' or 'login', 'cram-md5' or
2808 nothing depending on the value of ob->server_mech.
2810 I've just tested the patch: Authentication still works fine,
2811 unavailable mechs specified in the exim configuration are still
2812 caught, and the auth.log warnings about OTP are gone."
2814 PH/25 When debugging is enabled, the contents of the command line are added
2815 to the debugging output, even when log_selector=+arguments is not
2818 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2819 answer is "GNU", and only if the return is "GNU/something" is the answer
2822 PH/27 $acl_verify_message is now set immediately after the failure of a
2823 verification in an ACL, and so is available in subsequent modifiers. In
2824 particular, the message can be preserved by coding like this:
2826 warn !verify = sender
2827 set acl_m0 = $acl_verify_message
2829 Previously, $acl_verify_message was set only while expanding "message"
2830 and "log_message" when a very denied access.
2832 PH/28 Modified OS/os.c-Linux with
2834 -#ifndef OS_LOAD_AVERAGE
2835 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2837 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2838 nomenclature these days.)
2840 PH/29 Installed patch from the Sieve maintainer that adds the options
2841 sieve_useraddress and sieve_subaddress to the redirect router.
2843 PH/30 In these circumstances:
2844 . Two addresses routed to the same list of hosts;
2845 . First host does not offer TLS;
2846 . First host accepts first address;
2847 . First host gives temporary error to second address;
2848 . Second host offers TLS and a TLS session is established;
2849 . Second host accepts second address.
2850 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2851 and peerdn, if requested) that were in fact used only for the second
2854 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2855 attention to any local part prefix or suffix that was matched by the
2856 router that accepted the address. It now behaves in the same way as it
2857 does for delivery: the affixes are removed from the local part unless
2858 rcpt_include_affixes is set on the transport.
2860 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2861 timeout during the DATA phase of an incoming message.
2863 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2864 applied a patch sanctioned by the Sieve maintainer.
2866 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2867 the uid or gid is negative. A case of a negative gid caused this to be
2868 noticed. The fix allows for either to be negative.
2870 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2871 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2874 PH/36 The change PH/12 above was broken. Fixed it.
2876 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2877 the grounds that routing the same address twice would always produce the
2878 same answer. This might have been true once, but it is certainly no
2879 longer true now. Routing a child address may depend on the previous
2880 routing that produced that child. Some complicated redirection strategies
2881 went wrong when messages had multiple recipients, and made Exim's
2882 behaviour dependent on the order in which the addresses were given.
2884 I have moved the duplicate checking until after the routing is complete.
2885 Exim scans the addresses that are assigned to local and remote
2886 transports, and removes any duplicates. This means that more work will be
2887 done, as duplicates will always all be routed, but duplicates are
2888 presumably rare, so I don't expect this is of any significance.
2890 For deliveries to pipes, files, and autoreplies, the duplicate checking
2891 still happens during the routing process, since they are not going to be
2894 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2895 It corrects a timeout issue with spamd. This is Ian's comment: "The
2896 background is that sometimes spamd either never reads data from a
2897 connection it has accepted, or it never writes response data. The exiscan
2898 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2899 blindly assumes that writes won't block so it may never time out."
2901 PH/39 Allow G after quota size as well as K and M.
2903 PH/40 The value set for $authenticated_id in an authenticator may not contain
2904 binary zeroes or newlines because the value is written to log lines and
2905 to spool files. There was no check on this. Now the value is run through
2906 the string_printing() function so that such characters are converted to
2907 printable escape sequences.
2909 PH/41 $message_linecount is a new variable that contains the total number of
2910 lines in the message. Compare $body_linecount, which is the count for the
2913 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2914 bounce and delay warning messages, except in certain special cases, which
2917 (a) An SMTP error message from a remote host;
2918 (b) A message specified in a :fail: redirection;
2919 (c) A message specified in a "fail" command in a system filter;
2920 (d) A message specified in a FAIL return from the queryprogram router;
2921 (e) A message specified by the cannot_route_message router option.
2923 In these cases only, Exim does include the error details in bounce and
2924 warning messages. There are also a few cases where bland messages such
2925 as "unrouteable address" or "local delivery error" are given.
2927 PH/43 $value is now also set for the "else" part of a ${run expansion.
2929 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2930 being worked on, but at least Exim now implements the latest version to
2933 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2934 process to complete was treated as a delivery failure, a timeout while
2935 writing the message to the pipe was logged, but erroneously treated as a
2936 successful delivery. Such timeouts include transport filter timeouts. For
2937 consistency with the overall process timeout, these timeouts are now
2938 treated as errors, giving rise to delivery failures by default. However,
2939 there is now a new Boolean option for the pipe transport called
2940 timeout_defer, which, if set TRUE, converts the failures into defers for
2941 both kinds of timeout. A transport filter timeout is now identified in
2944 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2945 systems where "make" and "gmake" are different, calling "gmake" at top
2946 level broke things. I've arranged for the value of $(MAKE) to be passed
2947 from the Makefile to this script so that it can call the same version of
2951 A note about Exim versions 4.44 and 4.50
2952 ----------------------------------------
2954 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2955 changes of various kinds. As a consequence, a big documentation update was
2956 needed. This delayed the release for rather longer than seemed good, especially
2957 in the light of a couple of (minor) security issues. Therefore, the changes
2958 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2959 release. So 4.44 and 4.50 are in effect two different branches that both start
2962 I have left the 4.50 change log unchanged; it contains all the changes since
2963 4.43. The change log for 4.44 is below; many of its items are identical to
2964 those for 4.50. This seems to be the most sensible way to preserve the
2965 historical information.
2971 1. Minor wording change to the doc/README.SIEVE file.
2973 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2974 computation of the current number of files was incorrect.
2976 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2977 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2980 4. Give more explanation in the error message when the command for a transport
2981 filter fails to execute.
2983 5. There are several places where Exim runs a non-Exim command in a
2984 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2985 was being done only for the command run by the queryprogram router. It is
2986 now done for all such subprocesses. The other cases are: ${run, transport
2987 filters, and the commands run by the lmtp and pipe transports.
2989 6. Added CONFIGURE_GROUP build-time option.
2991 7. Some older OS have a limit of 256 on the maximum number of file
2992 descriptors. Exim was using setrlimit() to set 1000 as a large value
2993 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2994 systems. I've change it so that if it can't get 1000, it tries for 256.
2996 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2997 was an oversight, and furthermore, ever since the addition of extra
2998 controls (e.g. 4.43/32), the checks on when to allow different forms of
2999 "control" were broken. There should now be diagnostics for all cases when a
3000 control that does not make sense is encountered.
3002 9. Added the /retain_sender option to "control=submission".
3004 10. $recipients is now available in the predata ACL (oversight).
3006 11. Tidy the search cache before the fork to do a delivery from a message
3007 received from the command line. Otherwise the child will trigger a lookup
3008 failure and thereby defer the delivery if it tries to use (for example) a
3009 cached ldap connection that the parent has called unbind on.
3011 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3012 of $address_data from the recipient verification was clobbered by the
3013 sender verification.
3015 13. The value of address_data from a sender verification is now available in
3016 $sender_address_data in subsequent conditions in the ACL statement.
3018 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3020 15. Added a new option "connect=<time>" to callout options, to set a different
3023 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3024 was its contents. (It was OK if the option was not defined at all.)
3026 17. A "Completed" log line is now written for messages that are removed from
3027 the spool by the -Mrm option.
3029 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3030 information about exactly what failed.
3032 19. Added -dd to debug only the daemon process.
3034 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3035 handles timeouts, both on the server side and network timeouts. Renamed the
3036 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3038 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3039 It is now set to "smtps".
3041 22. $host_address is now set to the target address during the checking of
3042 ignore_target_hosts.
3044 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3045 being passed; this would have caused $sender_host_name to have been used if
3046 matching the list had actually called for a host name (not very likely,
3047 since this list is usually IP addresses). A host name is now passed as
3050 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3051 code that specifies a non-restarting handler (typically sigaction() in
3052 modern systems) in an attempt to fix a rare and obscure crash bug.
3054 25. Narrowed the window for a race in the daemon that could cause it to ignore
3055 SIGCHLD signals. This is not a major problem, because they are used only to
3056 wake it up if nothing else does.
3058 26. A malformed maildirsize file could cause Exim to calculate negative values
3059 for the mailbox size or file count. Odd effects could occur as a result.
3060 The maildirsize information is now recalculated if the size or filecount
3063 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3064 support for a long time. Removed HAVE_SYS_VFS_H.
3066 28. Installed the latest version of exipick from John Jetmore.
3068 29. In an address list, if the pattern was not a regular expression, an empty
3069 subject address (from a bounce message) matched only if the pattern was an
3070 empty string. Non-empty patterns were not even tested. This was the wrong
3071 because it is perfectly reasonable to use an empty address as part of a
3072 database query. An empty address is now tested by patterns that are
3073 lookups. However, all the other forms of pattern expect the subject to
3074 contain a local part and a domain, and therefore, for them, an empty
3075 address still always fails if the pattern is not itself empty.
3077 30. Exim went into a mad DNS loop when attempting to do a callout where the
3078 host was specified on an smtp transport, and looking it up yielded more
3079 than one IP address.
3081 31. Re-factored the code for checking spool and log partition space into a
3082 function that finds that data and another that does the check. The former
3083 is then used to implement four new variables: $spool_space, $log_space,
3084 $spool_inodes, and $log_inodes.
3086 32. The RFC2047 encoding function was originally intended for short strings
3087 such as real names; it was not keeping to the 75-character limit for
3088 encoded words that the RFC imposes. It now respects the limit, and
3089 generates multiple encoded words if necessary. To be on the safe side, I
3090 have increased the buffer size for the ${rfc2047: expansion operator from
3093 33. It is now permitted to omit both strings after an "if" condition; if the
3094 condition is true, the result is "true". As before, when the second string
3095 is omitted, a false condition yields an empty string. This makes it less
3096 cumbersome to write custom ACL and router conditions.
3098 34. Failure to deliver a bounce message always caused it to be frozen, even if
3099 there was an errors_to setting on the router. The errors_to setting is now
3102 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3103 canonical form (fully expanded) before being placed in
3104 $sender_host_address.
3106 36. The table in the code that translates DNS record types into text (T_A to
3107 "A" for instance) was missing entries for NS and CNAME. It is just possible
3108 that this could have caused confusion if both these types were looked up
3109 for the same domain, because the text type is used as part of Exim's
3110 per-process caching. But the chance of anyone hitting this buglet seems
3113 37. The dnsdb lookup has been extended in a number of ways.
3115 (1) There is a new type, "zns", which walks up the domain tree until it
3116 finds some nameserver records. It should be used with care.
3118 (2) There is a new type, "mxh", which is like "mx" except that it returns
3119 just the host names, not the priorities.
3121 (3) It is now possible to give a list of domains (or IP addresses) to be
3122 looked up. The behaviour when one of the lookups defers can be
3123 controlled by a keyword.
3125 (4) It is now possible to specify the separator character for use when
3126 multiple records are returned.
3128 38. The dnslists ACL condition has been extended: it is now possible to supply
3129 a list of IP addresses and/or domains to be looked up in a particular DNS
3132 39. Added log_selector=+queue_time_overall.
3134 40. When running the queue in the test harness, wait just a tad after forking a
3135 delivery process, to get repeatability of debugging output.
3137 41. Include certificate and key file names in error message when GnuTLS fails
3138 to set them up, because the GnuTLS error message doesn't include the name
3139 of the failing file when there is a problem reading it.
3141 42. Allow both -bf and -bF in the same test run.
3143 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3145 44. The "Exiscan patch" is now merged into the mainline Exim source.
3147 45. Sometimes the final signoff response after QUIT could fail to get
3148 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3149 < 0 before doing a fflush(). This bug looks as though it goes back to the
3150 introduction of TLS in release 3.20, but "sometimes" must have been rare
3151 because the tests only now provoked it.
3153 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3154 (this can affect the format of dates).
3156 47. exim_tidydb, when checking for the continued existence of a message for
3157 which it has found a message-specific retry record, was not finding
3158 messages that were in split spool directories. Consequently, it was
3159 deleting retry records that should have stayed in existence.
3161 48. Steve fixed some bugs in eximstats.
3163 49. The SPA authentication driver was not abandoning authentication and moving
3164 on to the next authenticator when an expansion was forced to fail,
3165 contradicting the general specification for all authenticators. Instead it
3166 was generating a temporary error. It now behaves as specified.
3168 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3169 (the order specifies the preference for clients). The order is now AES256,
3170 AES128, 3DES, ARCFOUR128.
3172 51. Small patch to Sieve code - explicitly set From: when generating an
3175 52. Exim crashed if a remote delivery caused a very long error message to be
3176 recorded - for instance if somebody sent an entire SpamAssassin report back
3177 as a large number of 550 error lines. This bug was coincidentally fixed by
3178 increasing the size of one of Exim's internal buffers (big_buffer) that
3179 happened as part of the Exiscan merge. However, to be on the safe side, I
3180 have made the code more robust (and fixed the comments that describe what
3183 53. Now that there can be additional text after "Completed" in log lines (if
3184 the queue_time_overall log selector is set), a one-byte patch to exigrep
3185 was needed to allow it to recognize "Completed" as not the last thing in
3188 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3189 patch that reportedly fixes this has been added. I am not expert enough to
3190 create a test for it. This is what the patch creator wrote:
3192 "I found a little strange behaviour of ldap code when working with
3193 Windows 2003 AD Domain, where users was placed in more than one
3194 Organization Units. When I tried to give exim partial DN, the exit code
3195 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3196 But simultaneously result of request was absolutely normal ldap result,
3197 so I produce this patch..."
3199 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3200 REFERENCE, so I have modified the code to exclude the patch when that macro
3203 55. Some experimental protocols are using DNS PTR records for new purposes. The
3204 keys for these records are domain names, not reversed IP addresses. The
3205 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3206 leaves it alone. Component reversal etc. now happens only for IP addresses.
3209 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3211 57. Double the size of the debug message buffer (to 2048) so that more of very
3212 long debug lines gets shown.
3214 58. The exicyclog utility now does better if the number of log files to keep
3215 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3217 59. Two changes related to the smtp_active_hostname option:
3219 (1) $smtp_active_hostname is now available as a variable.
3220 (2) The default for smtp_banner uses $smtp_active_hostname instead
3221 of $primary_hostname.
3223 60. The host_aton() function is supposed to be passed a string that is known
3224 to be a valid IP address. However, in the case of IPv6 addresses, it was
3225 not checking this. This is a hostage to fortune. Exim now panics and dies
3226 if the condition is not met. A case was found where this could be provoked
3227 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3228 components; fortuitously, this particular loophole had already been fixed
3229 by change 4.50/55 above.
3231 If there are any other similar loopholes, the new check in host_aton()
3232 itself should stop them being exploited. The report I received stated that
3233 data on the command line could provoke the exploit when Exim was running as
3234 exim, but did not say which command line option was involved. All I could
3235 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3236 running as the user.
3239 61. There was a buffer overflow vulnerability in the SPA authentication code
3240 (which came originally from the Samba project). I have added a test to the
3241 spa_base64_to_bits() function which I hope fixes it.
3244 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3245 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3247 63. The daemon start-up calls getloadavg() while still root for those OS that
3248 need the first call to be done as root, but it missed one case: when
3249 deliver_queue_load_max is set with deliver_drop_privilege. This is
3250 necessary for the benefit of the queue runner, because there is no re-exec
3251 when deliver_drop_privilege is set.
3253 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3254 This has been fixed.
3256 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3257 was in use, was not putting the data itself into the right store pool;
3258 consequently, it could be overwritten for a subsequent message in the same
3259 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3262 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3264 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3265 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3266 the function were treating the return as a boolean value, which happened to
3267 work because 0=false and not-0=true, but is not correct code.
3269 68. The host_aton() function was not handling scoped IPv6 addresses (those
3270 with, for example, "%eth0" on the end) correctly.
3272 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3273 negated items (that is, ~something) in unsigned ints. Some compilers
3274 apparently mutter when there is no cast.
3276 70. If an address verification called from an ACL failed, and did not produce a
3277 user-specific message (i.e. there was only a "system" message), nothing was
3278 put in $acl_verify_message. In this situation, it now puts the system
3281 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3282 session; change 4.31/43 added the unwanted input to the log line - except
3283 that it did not do this in the start of session case. It now does.
3285 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3286 This is wrong and can cause the other end to generate a synchronization
3287 error if it is another Exim or anything else that does the synchronization
3288 check. A QUIT command is no longer sent after a timeout.
3290 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3291 during host lookups.
3293 74. The defer_ok option of callout verification was not working if it was used
3294 when verifying addresses in header lines, that is, for this case:
3296 verify = header_sender/callout=defer_ok
3298 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3299 those file descriptors could be used for SMTP connections. If anything
3300 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3301 could be sent to the SMTP client, causing chaos. The daemon now opens
3302 stdin, stdout, and stderr to /dev/null when it puts itself into the
3305 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3306 log by default. The user can override this with suitable Perl magic.
3308 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3309 the log message when discard triggers, was not working for the DATA ACL or
3310 for the non-SMTP ACL.
3312 78. Error message wording change in sieve.c.
3314 79. If smtp_accept_max_per_host was set, the number of connections could be
3315 restricted to fewer than expected, because the daemon was trying to set up
3316 a new connection before checking whether the processes handling previous
3317 connections had finished. The check for completed processes is now done
3318 earlier. On busy systems, this bug wouldn't be noticed because something
3319 else would have woken the daemon, and it would have reaped the completed
3322 80. If a message was submitted locally by a user whose login name contained one
3323 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3324 It caused a spool format error. I have fixed the spool reading code. A
3325 related problem was that the "from" clause in the Received: line became
3326 illegal because of the space(s). It is now covered by ${quote_local_part.
3328 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3331 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3333 83. Patch from Sieve maintainer to fix unterminated string problem in
3334 "vacation" handling.
3336 84. Some minor changes to the Linux configuration files to help with other
3337 OS variants using glibc.
3339 85. One more patch for Sieve to update vacation handling to latest spec.
3342 ----------------------------------------------------
3343 See the note above about the 4.44 and 4.50 releases.
3344 ----------------------------------------------------
3350 1. Change 4.43/35 introduced a bug that caused file counts to be
3351 incorrectly computed when quota_filecount was set in an appendfile
3354 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3355 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3358 3. Give more explanation in the error message when the command for a transport
3359 filter fails to execute.
3361 4. There are several places where Exim runs a non-Exim command in a
3362 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3363 was being done only for the command run by the queryprogram router. It is
3364 now done for all such subprocesses. The other cases are: ${run, transport
3365 filters, and the commands run by the lmtp and pipe transports.
3367 5. Some older OS have a limit of 256 on the maximum number of file
3368 descriptors. Exim was using setrlimit() to set 1000 as a large value
3369 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3370 systems. I've change it so that if it can't get 1000, it tries for 256.
3372 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3373 was an oversight, and furthermore, ever since the addition of extra
3374 controls (e.g. 4.43/32), the checks on when to allow different forms of
3375 "control" were broken. There should now be diagnostics for all cases when a
3376 control that does not make sense is encountered.
3378 7. $recipients is now available in the predata ACL (oversight).
3380 8. Tidy the search cache before the fork to do a delivery from a message
3381 received from the command line. Otherwise the child will trigger a lookup
3382 failure and thereby defer the delivery if it tries to use (for example) a
3383 cached ldap connection that the parent has called unbind on.
3385 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3386 of $address_data from the recipient verification was clobbered by the
3387 sender verification.
3389 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3390 was its contents. (It was OK if the option was not defined at all.)
3392 11. A "Completed" log line is now written for messages that are removed from
3393 the spool by the -Mrm option.
3395 12. $host_address is now set to the target address during the checking of
3396 ignore_target_hosts.
3398 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3399 being passed; this would have caused $sender_host_name to have been used if
3400 matching the list had actually called for a host name (not very likely,
3401 since this list is usually IP addresses). A host name is now passed as
3404 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3405 code that specifies a non-restarting handler (typically sigaction() in
3406 modern systems) in an attempt to fix a rare and obscure crash bug.
3408 15. Narrowed the window for a race in the daemon that could cause it to ignore
3409 SIGCHLD signals. This is not a major problem, because they are used only to
3410 wake it up if nothing else does.
3412 16. A malformed maildirsize file could cause Exim to calculate negative values
3413 for the mailbox size or file count. Odd effects could occur as a result.
3414 The maildirsize information is now recalculated if the size or filecount
3417 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3418 support for a long time. Removed HAVE_SYS_VFS_H.
3420 18. Updated exipick to current release from John Jetmore.
3422 19. Allow an empty sender to be matched against a lookup in an address list.
3423 Previously the only cases considered were a regular expression, or an
3426 20. Exim went into a mad DNS lookup loop when doing a callout where the
3427 host was specified on the transport, if the DNS lookup yielded more than
3430 21. The RFC2047 encoding function was originally intended for short strings
3431 such as real names; it was not keeping to the 75-character limit for
3432 encoded words that the RFC imposes. It now respects the limit, and
3433 generates multiple encoded words if necessary. To be on the safe side, I
3434 have increased the buffer size for the ${rfc2047: expansion operator from
3437 22. Failure to deliver a bounce message always caused it to be frozen, even if
3438 there was an errors_to setting on the router. The errors_to setting is now
3441 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3442 canonical form (fully expanded) before being placed in
3443 $sender_host_address.
3445 24. Updated eximstats to version 1.33
3447 25. Include certificate and key file names in error message when GnuTLS fails
3448 to set them up, because the GnuTLS error message doesn't include the name
3449 of the failing file when there is a problem reading it.
3451 26. Expand error message when OpenSSL has problems setting up cert/key files.
3454 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3455 (this can affect the format of dates).
3457 28. exim_tidydb, when checking for the continued existence of a message for
3458 which it has found a message-specific retry record, was not finding
3459 messages that were in split spool directories. Consequently, it was
3460 deleting retry records that should have stayed in existence.
3462 29. eximstats updated to version 1.35
3463 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3464 1.35 - bugfix such that pie charts by volume are generated correctly
3466 30. The SPA authentication driver was not abandoning authentication and moving
3467 on to the next authenticator when an expansion was forced to fail,
3468 contradicting the general specification for all authenticators. Instead it
3469 was generating a temporary error. It now behaves as specified.
3471 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3472 (the order specifies the preference for clients). The order is now AES256,
3473 AES128, 3DES, ARCFOUR128.
3475 31. Small patch to Sieve code - explicitly set From: when generating an
3478 32. Exim crashed if a remote delivery caused a very long error message to be
3479 recorded - for instance if somebody sent an entire SpamAssassin report back
3480 as a large number of 550 error lines. This bug was coincidentally fixed by
3481 increasing the size of one of Exim's internal buffers (big_buffer) that
3482 happened as part of the Exiscan merge. However, to be on the safe side, I
3483 have made the code more robust (and fixed the comments that describe what
3486 33. Some experimental protocols are using DNS PTR records for new purposes. The
3487 keys for these records are domain names, not reversed IP addresses. The
3488 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3489 leaves it alone. Component reversal etc. now happens only for IP addresses.
3492 34. The host_aton() function is supposed to be passed a string that is known
3493 to be a valid IP address. However, in the case of IPv6 addresses, it was
3494 not checking this. This is a hostage to fortune. Exim now panics and dies
3495 if the condition is not met. A case was found where this could be provoked
3496 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3497 components; fortuitously, this particular loophole had already been fixed
3498 by change 4.50/55 or 4.44/33 above.
3500 If there are any other similar loopholes, the new check in host_aton()
3501 itself should stop them being exploited. The report I received stated that
3502 data on the command line could provoke the exploit when Exim was running as
3503 exim, but did not say which command line option was involved. All I could
3504 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3505 running as the user.
3508 35. There was a buffer overflow vulnerability in the SPA authentication code
3509 (which came originally from the Samba project). I have added a test to the
3510 spa_base64_to_bits() function which I hope fixes it.
3513 36. The daemon start-up calls getloadavg() while still root for those OS that
3514 need the first call to be done as root, but it missed one case: when
3515 deliver_queue_load_max is set with deliver_drop_privilege. This is
3516 necessary for the benefit of the queue runner, because there is no re-exec
3517 when deliver_drop_privilege is set.
3519 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3520 was in use, was not putting the data itself into the right store pool;
3521 consequently, it could be overwritten for a subsequent message in the same
3522 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3525 38. Sometimes the final signoff response after QUIT could fail to get
3526 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3527 < 0 before doing a fflush(). This bug looks as though it goes back to the
3528 introduction of TLS in release 3.20, but "sometimes" must have been rare
3529 because the tests only now provoked it.
3535 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3536 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3537 Now it returns an integer. A number of calls were still expecting a T/F
3538 return. Fortuitously, in all cases, the tests worked in OK situations,
3539 which is the norm. However, things would have gone wrong on any write
3540 failures on the smtp file descriptor. This function is used when sending
3541 messages over SMTP and also when doing verify callouts.
3543 2. When Exim is called to do synchronous delivery of a locally submitted
3544 message (the -odf or -odi options), it no longer closes stderr before doing
3547 3. Implemented the mua_wrapper option.
3549 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3551 5. Implemented the functions header_remove(), header_testname(),
3552 header_add_at_position(), and receive_remove_recipient(), and exported them
3555 6. If an ACL "warn" statement specified the addition of headers, Exim already
3556 inserted X-ACL-Warn: at the start if there was no header name. However, it
3557 was not making this test for the second and subsequent header lines if
3558 there were newlines in the string. This meant that an invalid header could
3559 be inserted if Exim was badly configured.
3561 7. Allow an ACL "warn" statement to add header lines at the start or after all
3562 the Received: headers, as well as at the end.
3564 8. Added the rcpt_4xx retry error code.
3566 9. Added postmaster_mailfrom=xxx to callout verification option.
3568 10. Added mailfrom=xxxx to the callout verification option, for verify=
3571 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3572 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3574 12. Inserted some casts to stop certain compilers warning when using pointer
3575 differences as field lengths or precisions in printf-type calls (mostly
3576 affecting debugging statements).
3578 13. Added optional readline() support for -be (dynamically loaded).
3580 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3581 same clock tick as a message's arrival, so that its received time was the
3582 same as the "first fail" time on the retry record, and that message
3583 remained on the queue past the ultimate address timeout, every queue runner
3584 would try a delivery (because it was past the ultimate address timeout) but
3585 after another failure, the ultimate address timeout, which should have then
3586 bounced the address, did not kick in. This was a "< instead of <=" error;
3587 in most cases the first failure would have been in the next clock tick
3588 after the received time, and all would be well.
3590 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3591 being recognized when the domain list was tested by the match_domain
3592 condition in an expansion string.
3594 16. Added the ${str2b64: operator.
3596 17. Exim was always calling setrlimit() to set a large limit for the number of
3597 processes, without checking whether the existing limit was already
3598 adequate. (It did check for the limit on file descriptors.) Furthermore,
3599 errors from getrlimit() and setrlimit() were being ignored. Now they are
3600 logged to the main and panic logs, but Exim does carry on, to try to do its
3601 job under whatever limits there are.
3603 18. Imported PCRE 5.0.
3605 19. Trivial typo in log message " temporarily refused connection" (the leading
3608 20. If the log selector return_path_on_delivery was set and an address was
3609 redirected to /dev/null, the delivery process crashed because it assumed
3610 that a return path would always be set for a "successful" delivery. In this
3611 case, the whole delivery is bypassed as an optimization, and therefore no
3614 21. Internal re-arrangement: the function for sending a challenge and reading
3615 a response while authentication was assuming a zero-terminated challenge
3616 string. It's now changed to take a pointer and a length, to allow for
3617 binary data in such strings.
3619 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3621 23. Exim was not respecting finduser_retries when seeking the login of the
3622 uid under which it was called; it was always trying 10 times. (The default
3623 setting of finduser_retries is zero.) Also, it was sleeping after the final
3624 failure, which is pointless.
3626 24. Implemented tls_on_connect_ports.
3628 25. Implemented acl_smtp_predata.
3630 26. If the domain in control=submission is set empty, Exim assumes that the
3631 authenticated id is a complete email address when it generates From: or
3632 Sender: header lines.
3634 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3635 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3636 chown and chgrp in /bin and hostname in /usr/bin.
3638 28. Exim was keeping the "process log" file open after each use, just as it
3639 does for the main log. This opens the possibility of it remaining open for
3640 long periods when the USR1 signal hits a daemon. Occasional processlog
3641 errors were reported, that could have been caused by this. Anyway, it seems
3642 much more sensible not to leave this file open at all, so that is what now
3645 29. The long-running daemon process does not normally write to the log once it
3646 has entered its main loop, and it closes the log before doing so. This is
3647 so that log files can straightforwardly be renamed and moved. However,
3648 there are a couple of unusual error situations where the daemon does write
3649 log entries, and I had neglected to close the log afterwards.
3651 30. The text of an SMTP error response that was received during a remote
3652 delivery was being truncated at 512 bytes. This is too short for some of
3653 the long messages that one sometimes sees. I've increased the limit to
3656 31. It is now possible to make retry rules that apply only when a message has a
3657 specific sender, in particular, an empty sender.
3659 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3660 it possible to be selective about when SMTP synchronization is enforced.
3662 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3664 32. Implemented hosts_connection_nolog.
3666 33. Added an ACL for QUIT.
3668 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3671 35. Added mailbox_size and mailbox_filecount to appendfile.
3673 36. Added control = no_multiline_responses to ACLs.
3675 37. There was a bug in the logic of the code that waits for the clock to tick
3676 in the case where the clock went backwards by a substantial amount such
3677 that the microsecond fraction of "now" was more than the microsecond
3678 fraction of "then" (but the whole seconds number was less).
3680 38. Added support for the libradius Radius client library this is found on
3681 FreeBSD (previously only the radiusclient library was supported).
3687 1. When certain lookups returned multiple values in the form name=value, the
3688 quoting of the values was not always being done properly. Specifically:
3689 (a) If the value started with a double quote, but contained no whitespace,
3691 (b) If the value contained whitespace other than a space character (i.e.
3692 tabs or newlines or carriage returns) it was not quoted.
3693 This fix has been applied to the mysql and pgsql lookups by writing a
3694 separate quoting function and calling it from the lookup code. The fix
3695 should probably also be applied to nisplus, ibase and oracle lookups, but
3696 since I cannot test any of those, I have not disturbed their existing code.
3698 2. A hit in the callout cache for a specific address caused a log line with no
3699 reason for rejecting RCPT. Now it says "Previous (cached) callout
3700 verification failure".
3702 3. There was an off-by-one bug in the queryprogram router. An over-long
3703 return line was truncated at 256 instead of 255 characters, thereby
3704 overflowing its buffer with the terminating zero. As well as fixing this, I
3705 have increased the buffer size to 1024 (and made a note to document this).
3707 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3708 when Exim is waiting for an SMTP response from a remote server, Exim
3709 restarts its select() call on the socket, thereby resetting its timeout.
3710 This is not a problem when such interrupts are rare. Somebody set up a cron
3711 job to run exiwhat every 2 minutes, which is less than the normal select()
3712 timeout (5 or 10 minutes). This meant that the select() timeout never
3713 kicked in because it was always reset. I have fixed this by comparing the
3714 time when an interrupt arrives with the time at the start of the first call
3715 to select(). If more time than the timeout has elapsed, the interrupt is
3716 treated as a timeout.
3718 5. Some internal re-factoring in preparation for the addition of Sieve
3719 extensions (by MH). In particular, the "personal" test is moved to a
3720 separate function, and given an option for scanning Cc: and Bcc: (which is
3721 not set for Exim filters).
3723 6. When Exim created an email address using the login of the caller as the
3724 local part (e.g. when creating a From: or Sender: header line), it was not
3725 quoting the local part when it contained special characters such as @.
3727 7. Installed new OpenBSD configuration files.
3729 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3730 try to make them clearer.
3732 9. Callout options, other than the timeout value, were being ignored when
3733 verifying sender addresses in header lines. For example, when using
3735 verify = header_sender/callout=no_cache
3737 the cache was (incorrectly) being used.
3739 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3740 only the Cygwin environment.
3742 11. When return_path_on_delivery was set as a log selector, if different remote
3743 addresses in the same message used different return paths and parallel
3744 remote delivery occurred, the wrong values would sometimes be logged.
3745 (Whenever a remote delivery process finished, the return path value from
3746 the most recently started remote delivery process was logged.)
3748 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3749 lines when AUTH and/or TLS are in use. This is the "received protocol"
3750 field. Exim used to use "asmtp" for authenticated SMTP, without any
3751 indication (in the protocol name) for TLS use. Now it follows the RFC and
3752 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3753 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3754 names appear in log lines as well as in Received: header lines.
3756 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3757 extensions, and comparison tests, and to fix some bugs.
3759 14. Changes to the "personal" filter test:
3761 (1) The test was buggy in that it was just doing the equivalent of
3762 "contains" tests on header lines. For example, if a user's address was
3763 anne@some.where, the "personal" test would incorrectly be true for
3765 To: susanne@some.where
3767 This test is now done by extracting each address from the header in turn,
3768 and checking the entire address. Other tests that are part of "personal"
3769 are now done using regular expressions (for example, to check local parts
3770 of addresses in From: header lines).
3772 (2) The list of non-personal local parts in From: addresses has been
3773 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3774 taken from the Sieve specification recommendations.
3776 (3) If the message contains any header line starting with "List-" it is
3777 treated as non-personal.
3779 (4) The test for "circular" in the Subject: header line has been removed
3780 because it now seems ill-conceived.
3782 15. Minor typos in src/EDITME comments corrected.
3784 16. Installed latest exipick from John Jetmore.
3786 17. If headers_add on a router specified a text string that was too long for
3787 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3788 of string_sprintf() is now avoided.
3790 18. $message_body_size was not set (it was always zero) when running the DATA
3791 ACL and the local_scan() function.
3793 19. For the "mail" command in an Exim filter, no default was being set for
3794 the once_repeat time, causing a random time value to be used if "once" was
3795 specified. (If the value happened to be <= 0, no repeat happened.) The
3796 default is now 0s, meaning "never repeat". The "vacation" command was OK
3797 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3798 (I found it when inspecting the code).
3800 20. There is now an overall timeout for performing a callout verification. It
3801 defaults to 4 times the callout timeout, which applies to individual SMTP
3802 commands during the callout. The overall timeout applies when there is more
3803 than one host that can be tried. The timeout is checked before trying the
3804 next host. This prevents very long delays if there are a large number of
3805 hosts and all are timing out (e.g. when the network connections are timing
3806 out). The value of the overall timeout can be changed by specifying an
3807 additional sub-option for "callout", called "maxwait". For example:
3809 verify = sender/callout=5s,maxwait=20s
3811 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3812 to the end before writing, but this should make it even safer).
3814 22. Exim was forgetting that it had advertised PIPELINING for the second and
3815 subsequent messages on an SMTP connection. It was also not resetting its
3816 memory on STARTTLS and an internal HELO.
3818 23. When Exim logs an SMTP synchronization error within a session, it now
3819 records whether PIPELINING has been advertised or not.
3821 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3822 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3823 rather than long int.
3825 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3827 26. Added the never_mail option to autoreply.
3833 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3834 crash if the getsockname() call failed; this can happen if a connection is
3835 closed very soon after it is established. The problem was simply in the
3836 order in which certain operations were done, causing Exim to try to write
3837 to the SMTP stream before it had set up the file descriptor. The bug has
3838 been fixed by making things happen in the correct order.
3844 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3845 before the connection was closed, thus losing the rejection response.
3847 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3848 some early Solaris releases, but causes trouble in current releases where
3849 socklen_t is defined.
3851 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3854 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3857 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3858 that releasing the top of it at the end releases what was used for sub-
3859 expansions (unless the block got too big). However, discard this block if
3860 the first thing is a variable or header, so that we can use its block when
3861 it is dynamic (useful for very large $message_headers, for example).
3863 6. Lookups now cache *every* query, not just the most recent. A new, separate
3864 store pool is used for this. It can be recovered when all lookup caches are
3865 flushed. Lookups now release memory at the end of their result strings.
3866 This has involved some general refactoring of the lookup sources.
3868 7. Some code has been added to the store_xxx() functions to reduce the amount
3869 of flapping under certain conditions.
3871 8. log_incoming_interface used to affect only the <= reception log lines. Now
3872 it causes the local interface and port to be added to several more SMTP log
3873 lines, for example "SMTP connection from", and rejection lines.
3875 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3877 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3879 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3880 could be overwritten at the end of the current message (or the start of a
3881 new message if it was set in a HELO ACL). The value is now preserved for
3882 the duration of the SMTP connection.
3884 12. If a transport had a headers_rewrite setting, and a matching header line
3885 contained an unqualified address, that address was qualified, even if it
3886 did not match any rewriting rules. The underlying bug was that the values
3887 of the flags that permit the existence of unqualified sender and recipient
3888 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3889 non-local messages, and by -bnq for local messages) were not being
3890 preserved with the message after it was received.
3892 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3893 "next input=" as part of the text comprising the host identity instead of
3894 the correct text. The code was using the same buffer for two different
3895 strings. However, depending on which order the printing function evaluated
3896 its arguments, the bug did not always show up. Under Linux, for example, my
3897 test suite worked just fine.
3899 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3900 doesn't work with some older versions of Perl. It has been changed to "my",
3901 which in any case is probably the better facility to use.
3903 15. A really picky compiler found some instances of statements for creating
3904 error messages that either had too many or two few arguments for the format
3907 16. The size of the buffer for calls to the DNS resolver has been increased
3908 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3909 for addresses that have a lot of PTR records. This alleviates a problem; it
3910 does not fully solve it.
3912 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3913 buffer now truncates the list and logs the incident, which is the same
3914 action as happens when Exim is looking up a host name and its aliases.
3915 Previously in this situation something unpredictable would happen;
3916 sometimes it was "internal error: store_reset failed".
3918 18. If a server dropped the connection unexpectedly when an Exim client was
3919 using GnuTLS and trying to read a response, the client delivery process
3920 crashed while trying to generate an error log message.
3922 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3923 string, for example:
3925 warn message = H1: something\nH2: something
3927 the text was added as a single header line from Exim's point of view
3928 though it ended up OK in the delivered message. However, searching for the
3929 second and subsequent header lines using $h_h2: did not work. This has been
3930 fixed. Similarly, if a system filter added multiple headers in this way,
3931 the routers could not see them.
3933 20. Expanded the error message when iplsearch is called with an invalid key to
3934 suggest using net-iplsearch in a host list.
3936 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3937 ACLs are no longer actually imposed (and a message to that effect is
3940 22. If a "gecos" field in a passwd entry contained escaped characters, in
3941 particular, if it contained a \" sequence, Exim got it wrong when building
3942 a From: or a Sender: header from that name. A second bug also caused
3943 incorrect handling when an unquoted " was present following a character
3944 that needed quoting.
3946 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3947 was not being matched caselessly.
3949 24. Arranged for all hyphens in the exim.8 source to be escaped with
3952 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3953 properly. Recipient callout cache records were still being keyed to include
3954 the sender, even when use_sender was set false. This led to far more
3955 callouts that were necessary. The sender is no longer included in the key
3956 when use_sender is false.
3958 26. Added "control = submission" modifier to ACLs.
3960 27. Added the ${base62d: operator to decode base 62 numbers.
3962 28. dnsdb lookups can now access SRV records.
3964 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3965 the configuration file.
3967 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3968 (-v) mode. This makes the output for a verbose queue run more intelligible.
3970 31. Added a use_postmaster feature to recipient callouts.
3972 32. Added the $body_zerocount variable, containing the number of binary zero
3973 bytes in the message body.
3975 33. The time of last modification of the "new" subdirectory is now used as the
3976 "mailbox time last read" when there is a quota error for a maildir
3979 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3981 35. Added +ignore_unknown as a special item in host lists.
3983 36. Code for decoding IPv6 addresses in host lists is now included, even if
3984 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3985 address was recognized as an IP address, but was then not correctly decoded
3986 into binary, causing unexpected and incorrect effects when compared with
3993 1. Very minor rewording of debugging text in manualroute to say "list of
3994 hosts" instead of "hostlist".
3996 2. If verify=header_syntax was set, and a header line with an unqualified
3997 address (no domain) and a large number of spaces between the end of the
3998 name and the colon was received, the reception process suffered a buffer
3999 overflow, and (when I tested it) crashed. This was caused by some obsolete
4000 code that should have been removed. The fix is to remove it!
4002 3. When running in the test harness, delay a bit after writing a bounce
4003 message to get a bit more predictability in the log output.
4005 4. Added a call to search_tidyup() just before forking a reception process. In
4006 theory, someone could use a lookup in the expansion of smtp_accept_max_
4007 per_host which, without the tidyup, could leave open a database connection.
4009 5. Added the variables $recipient_data and $sender_data which get set from a
4010 lookup success in an ACL "recipients" or "senders" condition, or a router
4011 "senders" option, similar to $domain_data and $local_part_data.
4013 6. Moved the writing of debug_print from before to after the "senders" test
4016 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4017 problems for message scanning, either using a data ACL, or using
4018 local_scan() because the Received: header was not generated till after they
4019 were called (in order to set the time as the time of reception completion).
4020 I have revised the way this works. The header is now generated after the
4021 body is received, but before the ACL or local_scan() are called. After they
4022 are run, the timestamp in the header is updated.
4028 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4029 before starting a queue runner without re-exec. This happened only when
4030 deliver_drop_privilege was set or when the Exim user was set to root. The
4031 effect of the bug was that timeouts during subsequent deliveries caused
4032 crashes instead of being properly handled. The handler is now left at its
4033 default (and expected) setting.
4035 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4036 message, again when deliver_drop_privilege is set or Exim is run as root.
4037 The bug described in (1) was not present in this case, but the tidying up
4038 of the other signals was missing. I have made the two cases consistent.
4040 3. The ignore_target_hosts setting on a manualroute router was being ignored
4041 for hosts that were looked up using the /MX notation.
4043 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4046 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4047 operated on the sender address. After changing the $sender_address to <>
4048 for the sender address verify, Exim was re-instated it as the original
4049 (before rewriting) address, but remembering that it had rewritten it, so it
4050 wasn't rewriting it again. This bug also had the effect of breaking the
4051 sender address verification caching when the sender address was rewritten.
4053 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4054 This has been changed so that if the ip literal address matches
4055 ignore_target_hosts, the router declines.
4057 7. Added expansion conditions match_domain, match_address, and match_local_
4058 part (NOT match_host).
4060 8. The placeholder for the Received: header didn't have a length field set.
4062 9. Added code to Exim itself and to exim_lock to test for a specific race
4063 condition that could lead to file corruption when using MBX delivery. The
4064 issue is with the lockfile that is created in /tmp. If this file is removed
4065 after a process has opened it but before that process has acquired a lock,
4066 there is the potential for a second process to recreate the file and also
4067 acquire a lock. This could lead to two Exim processes writing to the file
4068 at the same time. The added code performs the same test as UW imapd; it
4069 checks after acquiring the lock that its file descriptor still refers to
4070 the same named file.
4072 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4073 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4076 11. Added the smtp_active_hostname option. If used, this will typically be made
4077 to depend on the incoming interface address. Because $interface_address is
4078 not set up until the daemon has forked a reception process, error responses
4079 that can happen earlier (such as "too many connections") no longer contain
4082 12. If an expansion in a condition on a "warn" statement fails because a lookup
4083 defers, the "warn" statement is abandoned, and the next ACL statement is
4084 processed. Previously this caused the whole ACL to be aborted.
4086 13. Added the iplsearch lookup type.
4088 14. Added ident_timeout as a log selector.
4090 15. Added tls_certificate_verified as a log selector.
4092 16. Added a global option tls_require_ciphers (compare the smtp transport
4093 option of the same name). This controls incoming TLS connections.
4095 17. I finally figured out how to make tls_require_ciphers do a similar thing
4096 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4097 before starting the TLS session.
4099 18. Tabs are now shown as \t in -bP output.
4101 19. If the log selector return_path_on_delivery was set, Exim crashed when
4102 bouncing a message because it had too many Received: header lines.
4104 20. If two routers both had headers_remove settings, and the first one included
4105 a superfluous trailing colon, the final name in the first list and the
4106 first name in the second list were incorrectly joined into one item (with a
4107 colon in the middle).
4113 1. Added -C and -D options to the exinext utility, mainly to make it easier
4114 to include in the automated testing, but these could be helpful when
4115 multiple configurations are in use.
4117 2. The exinext utility was not formatting the output nicely when there was
4118 an alternate port involved in the retry record key, nor when there was a
4119 message id as well (for retries that were specific to a specific message
4120 and a specific host). It was also confused by IPv6 addresses, because of
4121 the additional colons they contain. I have fixed the IPv4 problem, and
4122 patched it up to do a reasonable job for IPv6.
4124 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4125 delivery, the log line now contains "pipelined" if PIPELINING was used.
4127 4. An SMTP transport process used to panic and die if the bind() call to set
4128 an explicit outgoing interface failed. This has been changed; it is now
4129 treated in the same way as a connect() failure.
4131 5. A reference to $sender_host_name in the part of a conditional expansion
4132 that was being skipped was still causing a DNS lookup. This no longer
4135 6. The def: expansion condition was not recognizing references to header lines
4136 that used bh_ and bheader_.
4138 7. Added the _cache feature to named lists.
4140 8. The code for checking quota_filecount in the appendfile transport was
4141 allowing one more file than it should have been.
4143 9. For compatibility with Sendmail, the command line option
4151 and sets the incoming protocol and host name (for trusted callers). The
4152 host name and its colon can be omitted when only the protocol is to be set.
4153 Note the Exim already has two private options, -pd and -ps, that refer to
4154 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4155 "s", but I don't think that's a major issue.
4157 10. A number of refactoring changes to the code, none of which should affect
4160 (a) The number of logging options was getting close to filling up the
4161 32-bit word that was used as a bit map. I have split them into two classes:
4162 those that are passed in the argument to log_write(), and those that are
4163 only ever tested independently outside of that function. These are now in
4164 separate 32-bit words, so there is plenty of room for expansion again.
4165 There is no change in the user interface or the logging behaviour.
4167 (b) When building, for example, log lines, the code previously used a
4168 macro that called string_cat() twice, in order to add two strings. This is
4169 not really sufficiently general. Furthermore, there was one instance where
4170 it was actually wrong because one of the argument was used twice, and in
4171 one call a function was used. (As it happened, calling the function twice
4172 did not affect the overall behaviour.) The macro has been replaced by a
4173 function that can join an arbitrary number of extra strings onto a growing
4176 (c) The code for expansion conditions now uses a table and a binary chop
4177 instead of a serial search (which was left over from when there were very
4178 few conditions). Also, it now recognizes conditions like "pam" even when
4179 the relevant support is not compiled in: a suitably worded error message is
4180 given if an attempt is made to use such a condition.
4182 11. Added ${time_interval:xxxxx}.
4184 12. A bug was causing one of the ddress fields not to be passed back correctly
4185 from remote delivery subprocesses. The field in question was not being
4186 subsequently used, so this caused to problems in practice.
4188 13. Added new log selectors queue_time and deliver_time.
4190 14. Might have fixed a bug in maildirsizefile handling that threw up
4191 "unexpected character" debug warnings, and recalculated the data
4192 unnecessarily. In any case, I expanded the warning message to give more
4195 15. Added the message "Restricted characters in address" to the statements in
4196 the default ACL that block characters like @ and % in local parts.
4198 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4199 Three changes have been made:
4201 (a) There was a serious bug; a negative response to MAIL caused the whole
4202 recipient domain to be cached as invalid, thereby blocking all messages
4203 to all local parts at the same domain, from all senders. This bug has
4204 been fixed. The domain is no longer cached after a negative response to
4205 MAIL if the sender used is not empty.
4207 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4210 (c) A new callout option, "use_sender" has been added for people who want
4211 the modified behaviour.
4217 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4220 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4221 indeed breaks things for older releases.
4223 3. Added additional logging to the case where there is a problem reading data
4224 from a filter that is running in a subprocess using a pipe, in order to
4225 try to track down a specific problem.
4227 4. Testing facility fudge: when running in the test harness and attempting
4228 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4229 getting "No route to host". Convert this to a timeout.
4231 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4234 6. Some OS don't have socklen_t but use size_t instead. This affects the
4235 fifth argument of getsockopt() amongst other things. This is now
4236 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4237 can be set for individual OS. I have set it for SunOS5, OSF1, and
4238 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4239 some earlier ones do not.
4241 7. Change 4.30/15 was not doing the test caselessly.
4243 8. The standard form for an IPv6 address literal was being rejected by address
4244 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4245 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4246 this, as well as the form without the "IPv6" on the front (but only when
4247 address literals are enabled, of course).
4249 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4251 10. Exim crashed if a message with an empty sender address specified by -f
4252 encountered a router with an errors_to setting. This could be provoked only
4253 by a command such as
4257 where an empty string was supplied; "<>" did not hit this bug.
4259 11. Installed PCRE release 4.5.
4261 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4262 remained set. It is now erased.
4264 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4265 times from message ids (which are base 36 rather than the normal 62).
4267 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4268 were being counted as actual protocol errors, and logged if the log
4269 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4270 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4271 and DATA following a set of rejected RCPTs do not count as protocol errors.
4272 In other words, Exim assumes they were pipelined, though this may not
4273 actually be the case. Of course, in all cases the client gets an
4274 appropriate error code.
4276 15. If a lookup fails in an ACL condition, a message about the failure may
4277 be available; it is used if testing the ACL cannot continue, because most
4278 such messages specify what the cause of the deferral is. However, some
4279 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4280 that caused an old message to be retained and used if a later statement
4281 caused a defer, replacing the real cause of the deferral.
4283 16. If an IP address had so many PTR records that the DNS lookup buffer
4284 was not large enough to hold them, Exim could crash while trying to process
4285 the truncated data. It now detects and logs this case.
4287 17. Further to 4.21/58, another change has been made: if (and only if) the
4288 first line of a message (the first header line) ends with CRLF, a bare LF
4289 in a subsequent header line has a space inserted after it, so as not to
4290 terminate the header.
4292 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4293 unnecessarily, used atoi() instead of strtol(), and didn't check the
4294 termination when getting file sizes from file names by regex.
4296 19. Completely re-implemented the support for maildirsize files, in the light
4297 of a number of problems with the previous contributed implementation
4298 (4.30/29). In particular:
4300 . If the quota is zero, the maildirsize file is maintained, but no quota is
4303 . If the maildir directory does not exist, it is created before any attempt
4304 to write a maildirsize file.
4306 . The quota value in the file is just a cache; if the quota is changed in
4307 the transport, the new value overrides.
4309 . A regular expression is available for excluding directories from the
4312 20. The autoreply transport checks the characters in options that define the
4313 message's headers; it allows continued headers, but it was checking with
4314 isspace() after an embedded newline instead of explicitly looking for a
4317 21. If all the "regular" hosts to which an address was routed had passed their
4318 expiry times, and had not reached their retry times, the address was
4319 bounced, even if fallback hosts were defined. Now Exim should go on to try
4322 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4323 equivalent code in the SMTP transport. Some hosts send humungous responses
4324 to HELO/EHLO, more than 1024 it seems.
4326 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4327 gives compiler warnings in some environments. I've now done it "properly",
4330 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4331 (because of problems with the built-in one) was declared to return uschar *
4332 instead of char *, causing compiler failure.
4334 25. Fixed a file descriptor leak when processing alias/forward files.
4336 26. Fixed a minor format string issue in dbfn.c.
4338 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4340 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4341 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4344 29. When a sender address is verified, it is cached, to save repeating the test
4345 when there is more than one recipient in a message. However, when the
4346 verification involves a callout, it is possible for different callout
4347 options to be set for different recipients. It is too complicated to keep
4348 track of this in the cache, so now Exim always runs a verification when a
4349 callout is required, relying on the callout cache for the optimization.
4350 The overhead is duplication of the address routing, but this should not be
4353 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4354 to be verified with callout=postmaster, and the main callout worked but the
4355 postmaster check failed, the verification correctly failed. However, if a
4356 subsequent RCPT command asked for sender verification *without* the
4357 postmaster check, incorrect caching caused this verification also to fail,
4360 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4361 it was not caching the DNS options (qualify_single, search_parents) that
4362 were used when the lookup failed. A subsequent lookup with different
4363 options therefore always gave the same answer, though there were cases
4364 where it should not have. (Example: a "domains = !$mx_any" option on a
4365 dnslookup router: the "domains" option is always processed without any
4366 widening, but the router might have qualify_single set.) Now Exim uses the
4367 cached value only when the same options are set.
4369 32. Added John Jetmore's "exipick" utility to the distribution.
4371 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4372 than a timeout (e.g. a certificate is required, and is not provided), an
4373 Exim server now closes the connection immediately. Previously it waited for
4374 the client to close - but if the client is SSL, it seems that they each
4375 wait for each other, leading to a delay before one of them times out.
4377 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4378 maintained 0.8.x compatibility because I don't think many are using it, and
4379 it is clearly obsolete.
4381 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4384 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4385 client certificate was expired. A simple patch fixes this, though I don't
4386 understand the full logic of why the verify callback is called multiple
4389 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4390 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4391 which causes problems with some clients (such as the Certicom SSL Plus
4392 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4393 disables the coutermeasure allowing Eudora to connect."
4395 38. Exim was not checking that a write() to a log file succeeded. This could
4396 lead to Bad Things if a log got too big, in particular if it hit a file
4397 size limit. Exim now panics and dies if it cannot write to a log file, just
4398 as it does if it cannot open a log file.
4400 39. Modified OS/Makefile-Linux so that it now contains
4402 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4404 The two -D definitions ensure that Exim is compiled with large file
4405 support, which makes it possible to handle log files that are bigger than
4408 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4409 instance) a domain was checked against a named list that involved a lookup,
4410 causing $domain_data to be set, then another domain was checked against the
4411 same list, then the first domain was re-checked, the value of $domain_data
4412 after the final check could be wrong. In particular, if the second check
4413 failed, it could be set empty. This bug probably also applied to
4416 41. The strip_trailing_dot option was not being applied to the address given
4417 with the -f command-line option.
4419 42. The code for reading a message's header from the spool was incrementing
4420 $received_count, but never initializing it. This meant that the value was
4421 incorrect (doubled) while delivering a message in the same process in which
4422 it was received. In the most common configuration of Exim, this never
4423 happens - a fresh exec is done - but it can happen when
4424 deliver_drop_privilege is set.
4426 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4427 it now includes up to 150 characters of the unexpected data in the log
4430 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4431 and building data strings. The size of both of these buffers was 10 000
4432 bytes - far larger than anybody would *ever* want, thought I. Needless to
4433 say, somebody hit the limit. I have increased the maximum line length to
4434 20 000 and the maximum data length of concatenated lines to 100 000. I have
4435 also fixed two bugs, because there was no checking on these buffers. Tsk,
4436 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4437 buffer is too small.
4439 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4440 lsearch lookups. Now it does.
4442 46. When parsing a route_list item in a manualroute router, a fixed-length
4443 buffer was used for the list of hosts. I made this 1024 bytes long,
4444 thinking that nobody would ever have a list of hosts that long. Wrong.
4445 Somebody had a whole pile of complicated expansion conditions, and the
4446 string was silently truncated, leading to an expansion error. It turns out
4447 that it is easier to change to an unlimited length (owing to other changes
4448 that have happened since this code was originally written) than to build
4449 structure for giving a limitation error. The length of the item that
4450 expands into the list of hosts is now unlimited.
4452 47. The lsearch lookup could not handle data where the length of text line was
4453 more than 4095 characters. Such lines were truncated, leading to shortened
4454 data being returned. It should now handle lines of any length.
4456 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4457 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4460 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4461 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4462 no longer happens in comment lines. A list of replacements is now placed
4463 at the head of all of the source files, except those whose only change is
4464 to replace PERL_COMMAND in the very first #! line.
4466 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4467 messages on the queue, with a bottom-up merge sort, using code contributed
4468 by Michael Haardt. This should make operations like -bp somewhat faster on
4469 large queues. It won't affect queue runners, except when queue_run_in_order
4472 51. Installed eximstats 1.31 in the distribution.
4474 52. Added support for SRV lookups to the dnslookup router.
4476 53. If an ACL referred to $message_body or $message_body_end, the value was not
4477 reset for any messages that followed in the same SMTP session.
4479 54. The store-handling optimization for building very long strings was not
4480 differentiating between the different store pools. I don't think this
4481 actually made any difference in practice, but I've tidied it.
4483 55. While running the routers to verify a sender address, $sender_address
4484 was still set to the sender address. This is wrong, because when routing to
4485 send a bounce to the sender, it would be empty. Therefore, I have changed
4486 it so that, while verifying a sender address, $sender_address is set to <>.
4487 (There is no change to what happens when verifying a recipient address.)
4489 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4490 target A or AAAA records (if not already returned) without resetting the
4491 qualify_single or search_parents options of the DNS resolver. These are
4492 inappropriate in this case because the targets of MX and SRV records must
4493 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4494 target that, when qualified, matched something in the local domain. These
4495 two options are now turned off when doing these lookups.
4497 57. It seems that at least some releases of Reiserfs (which does not have the
4498 concept of a fixed number of inodes) returns zero and not -1 for the
4499 number of available inodes. This interacted badly with check_spool_inodes,
4500 which assumed that -1 was the "no such thing" setting. What I have done is
4501 to check that the total number of inodes is greater than zero before doing
4502 the test of how many are available.
4504 58. When a "warn" ACL statement has a log_message modifier, the message is
4505 remembered, and not repeated. This is to avoid a lot of repetition when a
4506 message has many recipients that cause the same warning to be written.
4507 Howewer, Exim was preserving the list of already written lines for an
4508 entire SMTP session, which doesn't seem right. The memory is now reset if a
4509 new message is started.
4511 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4512 debugging output unless log_rewrite was also set.
4514 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4515 of (int)(handle) when we know that handle contains (void *)(-1).
4517 61. The Exim daemon panic-logs an error return when it closes the incoming
4518 connection. However "connection reset by peer" seems to be common, and
4519 isn't really an error worthy of noting specially, so that particular error
4522 62. When Exim is trying to find all the local interfaces, it used to panic and
4523 die if the ioctl to get the interface flags failed. However, it seems that
4524 on at least one OS (Solaris 9) it is possible to have an interface that is
4525 included in the list of interfaces, but for which you get a failure error
4526 for this call. This happens when the interface is not "plumbed" into a
4527 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4528 failure of the "get flags" call assumes that the interface is down.
4530 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4531 makes life easier for people who are doing arithmetic on fields extracted
4532 from dates, where you often get leading zeros that should not be
4533 interpreted as octal.
4535 64. Added qualify_domain to the redirect router, to override the global
4538 65. If a pathologically long header line contained very many addresses (the
4539 report of this problem mentioned 10 000) and each of them was rewritten,
4540 Exim could use up a very large amount of memory. (It kept on making new
4541 copies of the header line as it rewrote, and never released the old ones.)
4542 At the expense of a bit more processing, the header rewriting function has
4543 been changed so that it no longer eats memory in this way.
4545 66. The generation of the Received: header has been moved from the time that a
4546 message starts to be received, to the time that it finishes. The timestamp
4547 in the Received: header should now be very close to that of the <= log
4548 line. There are two side-effects of this change:
4550 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4551 logged header lines no longer include the local Received: line, because
4552 it has not yet been created. The same applies to a copy of the message
4553 that is returned to a non-SMTP sender when a message is rejected.
4555 (b) When a filter file is tested using -bf, no additional Received: header
4556 is added to the test message. After some thought, I decided that this
4559 This change does not affect the value of $received_for. It is still set
4560 after address rewriting, but before local_scan() is called.
4562 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4564 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4565 gave an unhelpful panic error message, and a defer error. I have managed to
4566 change this behaviour so that it now rejects any supplied certificate,
4567 which seems right, as the list of acceptable certificates is empty.
4569 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4570 gave an unhelpful defer error. I have not managed to make this reject any
4571 supplied certificates, but the error message it gives is "no certificate
4572 supplied", which is not helpful.
4574 70. exigrep's output now also includes lines that are not associated with any
4575 message, but which match the given pattern. Implemented by a patch from
4576 Martin Sluka, which also tidied up the Perl a bit.
4578 71. Recipient callout verification, like sender verification, was using <> in
4579 the MAIL FROM command. This isn't really the right thing, since the actual
4580 sender may affect whether the remote host accepts the recipient or not. I
4581 have changed it to use the actual sender in the callout; this means that
4582 the cache record is now keyed on a recipient/sender pair, not just the
4583 recipient address. There doesn't seem to be a real danger of callout loops,
4584 since a callout by the remote host to check the sender would use <>.
4585 [SEE ABOVE: changed after hitting problems.]
4587 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4588 temporary errors. However, in the case of such a code being given after
4589 the end of a data transmission (i.e. after ".") Exim was failing to write
4590 a retry record for the message. (Yes, there was some broken host that was
4591 actually sending 8xx at this point.)
4593 73. An unknown lookup type in a host list could cause Exim to panic-die when
4594 the list was checked. (An example that provoked this was putting <; in the
4595 middle of a list instead of at the start.) If this happened during a DATA
4596 ACL check, a -D file could be left lying around. This kind of configuration
4597 error no longer causes Exim to die; instead it causes a defer errror. The
4598 incident is still logged to the main and panic logs.
4600 74. Buglet left over from Exim 3 conversion. The message "too many messages
4601 in one connection" was written to the rejectlog but not the mainlog, except
4602 when address rewriting (yes!) was being logged.
4604 75. Added write_rejectlog option.
4606 76. When a system filter was run not as root (that is, when system_filter_user
4607 was set), the values of the $n variables were not being returned to the
4608 main process; thus, they were not subsequently available in the $sn
4611 77. Added +return_path_on_delivery log selector.
4613 78. A connection timeout was being treated differently from recipients deferred
4614 when testing hosts_max_try with a message that was older than the host's
4615 retry timeout. (The host should not be counted, thus allowing all hosts to
4616 be tried at least once before bouncing.) This may have been the cause of an
4617 occasionally reported bug whereby a message would remain on the queue
4618 longer than the retry timeout, but would be bounced if a delivery was
4619 forced. I say "may" because I never totally pinned down the problem;
4620 setting up timeout/retry tests is difficult. See also the next item.
4622 79. The ultimate address timeout was not being applied to errors that involved
4623 a combination of host plus message (for example, a timeout on a MAIL
4624 command). When an address resolved to a number of possible hosts, and they
4625 were not all tried for each delivery (e.g. because of hosts_max_try), a
4626 message could remain on the queue longer than the retry timeout.
4628 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4631 81. Fixed an obscure SMTP outgoing bug which required at least the following
4632 conditions: (a) there was another message waiting for the same server;
4633 (b) the server returned 5xx to all RCPT commands in the first message so
4634 that the message was not completed; (c) the server dropped the connection
4635 or gave a negative response to the RSET that Exim sends to abort the
4636 transaction. The observed case was a dropped connection after DATA that had
4637 been sent in pipelining mode. That is, the server had advertised PIPELINING
4638 but was not implementing it correctly. The effect of the bug was incorrect
4639 behaviour, such as trying another host, and this could lead to a crash.
4645 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4646 and daemon.c were passed as pointers to ints; they should have been
4647 pointers to socklen_t variables (which are typically unsigned ints).
4649 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4652 3. Fixed a really odd bug that affected only the testing scheme; patching a
4653 certain fixed string in the binary changed the value of another string that
4654 happened to be identical to the end of the original first string.
4656 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4657 name", it returns that address as the IP address. On some operating
4658 systems (e.g. Solaris), it also passes back the IP address string as the
4659 "host name". However, on others (e.g. Linux), it passes back an empty
4660 string. Exim wasn't checking for this, and was changing the host name to an
4661 empty string, assuming it had been canonicized.
4663 5. Although rare, it is permitted to have more than one PTR record for a given
4664 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4665 all the names associated with an address, because they do in Solaris.
4666 However, it seems that they do not in Linux for data that comes from the
4667 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4668 I found this out when I moved to a new Linux workstation and tried to run
4669 the Exim test suite.
4671 To get round this problem I have changed the code so that it now does its
4672 own call to the DNS to look up PTR records when searching for a host name.
4673 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4674 addresses that are only in /etc/hosts are still found.
4676 This behaviour is, however, controlled by an option called host_lookup_
4677 order, which defaults to "bydns:byaddr". If people want to use the other
4678 order, or indeed, just use one or the other means of lookup, they can
4679 specify it in this variable.
4681 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4682 some operating systems, this comes back from gethostbyaddr() as an empty
4683 string, and this is what Exim used to test for. However, it seems that in
4684 other systems, "." is yielded. Exim now tests for this case too.
4686 7. The values of check_spool_space and check_log_space are now held internally
4687 as a number of kilobytes instead of an absolute number of bytes. If a
4688 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4689 kilobyte. This means that much larger values can be stored.
4691 8. Exim monitor: an attempt to get the action menu when not actually pointing
4692 at a message produces an empty menu entitled "No message selected". This
4693 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4694 no entries in it ("Shell widget menu has zero width and/or height"). So I
4695 have added a single, blank menu entry in this case.
4697 9. Added ${quote_local_part.
4699 10. MIME decoding is now applied to the contents of Subject: header lines when
4702 11. Now that a reference to $sender_host_address automatically causes a reverse
4703 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4704 host lookup before query-style lookups in lists that might use this
4705 variable. This has therefore been abolished, and the "net-" prefix is no
4706 longer necessary for query-style lookups.
4708 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4709 have been a typo for LFLAGS, so it has been changed.
4711 13. The install script calls Exim with "-C /dev/null" in order to find the
4712 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4713 to be output. Howeve, since Exim outputs its version number before the
4714 error, it didn't break the script. It just looked ugly. I fixed this by
4715 always allowing "-C /dev/null" if the caller is root.
4717 14. Ignore overlarge ACL variable number when reading spool file - insurance
4718 against a later release with more variables having written the file.
4720 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4721 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4722 this, as well as the form without the "IPv6" on the front.
4724 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4725 OS/Makefile-Darwin file.
4727 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4728 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4730 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4731 variable. However, some OS use ints for inodes. I've added cast to long int
4732 to get rid of the compiler warning.
4734 19. I had forgotten to lock out "/../" in configuration file names when
4735 ALT_CONFIG_PREFIX was set.
4737 20. Routers used for verification do not need to specify transports. However,
4738 if such a router generated a host list, and callout was configured, Exim
4739 crashed, because it could not find a port number from the (non-existent)
4740 transport. It now assumes port 25 in this circumstance.
4742 21. Added the -t option to exigrep.
4744 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4745 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4746 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4747 them is set, LOOKUP_LSEARCH is forced.
4749 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4751 24. Added sender and host information to the "rejected by local_scan()" log
4752 line; previously there was no indication of these.
4754 25. Added .include_if_exists.
4756 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4757 Linux. It turns out that not all file systems support this. Apparently some
4758 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4759 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4760 not supported on the file descriptor, is now ignored when Exim is trying to
4761 sync a directory. This applies only to Linux.
4763 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4765 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4766 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4769 29. Added support for maildirsize files from supplied patch (modified a bit).
4771 30. The use of :fail: followed by an empty string could lead Exim to respond to
4772 sender verification failures with (e.g.):
4774 550 Verification failed for <xxx>
4775 550 Sender verify failed
4777 where the first response line was missing the '-' that indicates it is not
4778 the final line of the response.
4780 31. The loop for finding the name of the user that called Exim had a hardwired
4781 limit of 10; it now uses the value of finduser_retries, which is used for
4782 all other user lookups.
4784 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4787 33. Exim was neglecting to zero errno before one call of strtol() when
4788 expanding a string and expecting an integer value. On some systems this
4789 resulted in spurious "integer overflow" errors. Also, it was casting the
4790 result into an int without checking.
4792 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4793 did not work. The code looks as if it has *never* worked, though it appears
4794 to have been documented since at least releast 1.62. I have made it work.
4796 35. The "timeout_DNS" error in retry rules, also documented since at least
4797 1.62, also never worked. As it isn't clear exactly what this means, and
4798 clearly it isn't a major issue, I have abolished the feature by treating it
4799 as "timeout", and writing a warning to the main and panic logs.
4801 36. The display of retry rules for -brt wasn't always showing the error code
4804 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4805 timeout_connect_A, timeout_connect_MX.
4807 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4808 to the empty sender.
4810 39. The daemon was not analysing the content of -oX till after it had closed
4811 stderr and disconnected from the controlling terminal. This meant that any
4812 syntax errors were only noted on the panic log, and the return code from
4813 the command was 0. By re-arranging the code a little, I've made the
4814 decoding happen first, so such errors now appear on stderr, and the return
4815 code is 1. However, the actual setting up of the sockets still happens in
4816 the disconnected process, so errors there are still only recorded on the
4819 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4820 connections (as happens on some IP stacks) was logged at start up time as
4821 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4822 it from "IPv6 and IPv4", which means that two separate sockets are being
4825 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4826 says whether AF_INET or AF_INET6 was passed as an argument.
4828 42. Exiwhat output was messed up when time zones were included in log
4831 43. Exiwhat now gives more information about the daemon's listening ports,
4832 and whether -tls-on-connect was used.
4834 44. The "port" option of the smtp transport is now expanded.
4836 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4837 silently ignored. Now an error message is written to the main and panic
4840 46. There's a new ACL modifier called "logwrite" which writes to a log file
4841 as soon as it is encountered.
4843 47. Added $local_user_uid and $local_user_gid at routing time.
4845 48. Exim crashed when trying to verify a sender address that was being
4848 49. Exim was recognizing only a space character after ".include". It now also
4849 recognizes a tab character.
4851 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4852 extracting the relevant information from the specification. The man page no
4853 longer contains scrambled data for the -d option, and I've added a section
4854 at the front about calling Exim under different names.
4856 51. Added "extra_headers" argument to the "mail" command in filter files.
4858 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4861 53. Installed eximstats 1.29.
4863 54. Added transport_filter_timeout as a generic transport option.
4865 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4866 Cc: header lines. This was required by RFC 822, but it not required by RFC
4869 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4870 incoming messages that did not have them. Now it does so only if the
4871 message originates locally, that is, if there is no associated remote host
4872 address. When Resent- header lines are present, this applies to the Resent-
4873 lines rather than the non-Resent- lines.
4875 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4876 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4878 58. Messages for configuration errors now include the name of the main
4879 configuration file - useful now that there may be more than one file in a
4880 list (.included file names were always shown).
4882 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4883 for those rare installations that do not start the daemon as root or run it
4884 setuid root. I've cut out the call to initgroups() if the daemon is not
4887 60. The Exim user and group can now be bound into the binary as text strings
4888 that are looked up at the start of Exim's processing.
4890 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4892 62. Added $mailstore_basename variable.
4894 63. Installed patch to sieve.c from Michael Haardt.
4896 64. When Exim failed to open the panic log after failing to open the main log,
4897 the original message it was trying to log was written to stderr and debug
4898 output, but if they were not available (the usual case in production), it
4899 was lost. Now it is written to syslog before the two lines that record the
4900 failures to open the logs.
4902 65. Users' Exim filters run in subprocesses under the user's uid. It is
4903 possible for a "deliver" command or an alias in a "personal" command to
4904 provoke an address rewrite. If logging of address rewriting is configured,
4905 this fails because the process is not running as root or exim. There may be
4906 a better way of dealing with this, but for the moment (because 4.30 needs
4907 to be released), I have disabled address rewrite logging when running a
4908 filter in a non-root, non-exim process.
4914 1. The buildconfig auxiliary program wasn't quoting the value set for
4915 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4916 not defined. This bug was masked in 4.22 by the effect that was fixed in
4919 2. Some messages that were rejected after a message id was allocated were
4920 shown as "incomplete" by exigrep. It no longer does this for messages that
4921 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4923 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4924 have allow_domain_literals set, the ID did not get logged in the <= line.
4925 Domain literals are now always recognized in Message-ID: header lines.
4927 4. The first argument for a ${extract expansion item is the key name or field
4928 number. Leading and trailing spaces in this item were not being ignored,
4929 causing some misleading effects.
4931 5. When deliver_drop_privilege was set, single queue runner processes started
4932 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4933 same command in the process it spins off) were not dropping privilege.
4935 6. When the daemon running as "exim" started a queue runner, it always
4936 re-executed Exim in the spun-off process. This is a waste of effort when
4937 deliver_drop_privilege is set. The new process now just calls the
4938 queue-runner function directly.
4944 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4947 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4948 ignored. Though the use of -oP was forcing the writing of a pid file, it
4949 was always written to the default place.
4951 3. If the message "no IP address found for host xxxx" is generated during
4952 incoming verification, it is now followed by identification of the incoming
4953 connection (so you can more easily find what provoked it).
4955 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4957 5. Added some features to "harden" Exim a bit more against certain attacks:
4959 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4960 be put in Local/Makefile. This is like the never_users runtime option,
4961 but it cannot be overridden. The default setting is "root".
4963 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4964 prefix string with which any file named in a -C command line option
4967 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4968 is retained for -C and -D only if the caller of Exim is root. Without
4969 it, the exim user may also use -C and -D and retain privilege.
4971 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4972 command line option is disabled.
4974 6. Macro names set by the -D option must start with an upper case letter, just
4975 like macro names defined in the configuration file.
4977 7. Added "dereference=" facility to LDAP.
4979 8. Two instances of the typo "uknown" in the source files are fixed.
4981 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4982 the Configure-Makefile script screwed up while processing it.
4984 10. Incorporated PCRE 4.4.
4986 11. The SMTP synchronization check was not operating right at the start of an
4987 SMTP session. For example, it could not catch a HELO sent before the client
4988 waited for the greeting. There is now a check for outstanding input at the
4989 point when the greeting is written. Because of the duplex, asynchronous
4990 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4991 way, but not yet received, when the check is performed.
4993 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4994 on TCP/IP sockets, because this apparently causes some broken clients to
4997 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4998 unchanged) from the Cygwin maintainer.
5000 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5001 was supported instead of testing for mbx. Effectively a typo.
5003 15. The spa authenticator server code was not checking that the input it
5004 received was valid base64.
5006 16. The debug output line for the "set" modifier in ACLs was not showing the
5007 name of the variable that was being set.
5009 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5011 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5012 lookup on its own. Something else was needed to trigger the lookup. For
5013 example, a match in host_lookup or the need for a host name in a host list.
5014 Now, if $sender_host_name is referenced and the host name has not yet been
5015 looked up, a lookup is performed. If the lookup fails, the variable remains
5016 empty, and $host_lookup_failed is set to "1".
5018 19. Added "eqi" as a case-independent comparison operator.
5020 20. The saslauthd authentication condition could segfault if neither service
5021 nor realm was specified.
5023 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5024 error message that was logged was misleading, and incoming SMTP
5025 connections were dropped. The message is now more accurate, and temporary
5026 errors are given to SMTP connections.
5028 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5029 (except RSET) with a 421 error, until QUIT is received. However, it was
5030 failing to send a response to QUIT.
5032 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5033 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5034 ACL is now run after the helo_try_verify_hosts code.
5036 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5037 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5038 case-independent, so other case variants are also recognized). Apparently
5039 some systems use these upper case variants.
5041 25. If more than two messages were waiting for the same host, and a transport
5042 filter was specified for the transport, Exim sent two messages over the
5043 same TCP/IP connection, and then failed with "socket operation on non-
5044 socket" when it tried to send the third.
5046 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5048 27. The extern definition of crypt16() in expand.c was not being excluded when
5049 the OS had its own crypt16() function.
5051 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5052 as a preferred synonym for return_size_limit, both as an option and as an
5055 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5057 30. Changed the default configuration ACL to relax the local part checking rule
5058 for addresses that are not in any local domains. For these addresses,
5059 slashes and pipe symbols are allowed within local parts, but the sequence
5060 /../ is explicitly forbidden.
5062 31. SPA server authentication was not clearing the challenge buffer before
5065 32. log_message in a "warn" ACL statement was writing to the reject log as
5066 well as to the main log, which contradicts the documentation and doesn't
5067 seem right (because no rejection is happening). So I have stopped it.
5069 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5070 However, I am unable to do any testing of this.
5072 34. Fixed an infelicity in the appendfile transport. When checking directories
5073 for a mailbox, to see if any needed to be created, it was accidentally
5074 using path names with one or more superfluous leading slashes; tracing
5075 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5077 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5078 contents are added to the log line that is written for every discarded
5079 recipient. (Previously a log_message setting was ignored.)
5081 36. The ${quote: operator now quotes the string if it is empty.
5083 37. The install script runs exim in order to find its version number. If for
5084 some reason other than non-existence or emptiness, which it checks, it
5085 could not run './exim', it was installing it with an empty version number,
5086 i.e. as "exim-". This error state is now caught, and the installation is
5089 38. An argument was missing from the function that creates an error message
5090 when Exim fails to connect to the socket for saslauthd authentication.
5091 This could cause Exim to crash, or give a corrupted message.
5093 39. Added isip, isip4, and isip6 to ${if conditions.
5095 40. The ACL variables $acl_xx are now saved with the message, and can be
5096 accessed later in routers, transports, and filters.
5098 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5099 strings in the file are not string-expanded.
5101 42. If a MAIL command specified a SIZE value that was too large to fit into an
5102 int variable, the check against message_size_limit failed. Such values are
5103 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5104 day this will have to be increased, but I don't think I want to be around
5105 when emails are that large.
5112 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5113 iconv() is not standard in FreeBSD.
5115 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5116 IPv6 enabled. The observed symptom was a segmentation fault on return from
5117 the function os_common_find_running_interfaces() in src/os.c.
5119 3. In the check_special_case() function in daemon.c I had used "errno" as an
5120 argument name, which causes warnings on some systems. This was basically a
5121 typo, since it was named "eno" in the comments!
5123 4. The code that waits for the clock to tick (at a resolution of some fraction
5124 of a second) so as to ensure message-id uniqueness was always waiting for
5125 at least one whole tick, when it could have waited for less. [This is
5126 almost certainly not relevant at current processor speeds, where it is
5127 unlikely to ever wait at all. But we try to future-proof.]
5129 5. The function that sleeps for a time interval that includes fractions of a
5130 second contained a race. It did not block SIGALRM between setting the
5131 timer, and suspending (a couple of lines later). If the interval was short
5132 and the sigsuspend() was delayed until after it had expired, the suspension
5133 never ended. On busy systems this could lead to processes getting stuck for
5136 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5137 process, before it forks any delivery processes. The open lookup caching
5138 mechanism meant that the open file or database connection was passed into
5139 the delivery process. The problem was that delivery processes always tidy
5140 up cached lookup data. This could cause a problem for the next delivery
5141 process started by the queue runner, because the external queue runner
5142 process does not know about the closure. So the next delivery process
5143 still has data in the lookup cache. In the case of a file lookup, there was
5144 no problem because closing a file descriptor in a subprocess doesn't affect
5145 the parent. However, if the lookup was caching a connection to a database,
5146 the connection was closed, and the second delivery process was likely to
5147 see errors such as "PGSQL: query failed: server closed the connection
5148 unexpectedly". The problem has been fixed by closing all cached lookups
5149 in a queue runner before running a delivery process.
5151 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5152 seem to have the "const" qualifier which it has on other OS. I've
5155 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5156 *of the same type* (client or server) with the same public name that an
5157 error should be diagnosed.
5159 9. When Exim looked up a host name for an IP address, but failed to find the
5160 original IP address when looking up the host name (a safety check), it
5161 output the message "<ip address> does not match any IP for NULL", which was
5162 confusing, to say the least. The bug was that the host name should have
5163 appeared instead of "NULL".
5165 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5166 user that is built into the binary, and the -C or -D options is used, root
5167 privilege is dropped before the configuration file is read. In addition,
5168 logging is switched to stderr instead of the normal log files. If the
5169 configuration then re-defines the Exim user, the unprivileged environment
5170 is probably not what is expected, so Exim logs a panic warning message (but
5173 However, if deliver_drop_privilege is set, the unprivileged state may well
5174 be exactly what is intended, so the warning has been cut out in that case,
5175 and Exim is allowed to try to write to its normal log files.
5181 1. smtp_return_error_details was not giving details for temporary sender
5182 or receiver verification errors.
5184 2. Diagnose a configuration error if two authenticators have the same public
5187 3. Exim used not to create the message log file for a message until the first
5188 delivery attempt. This could be confusing when incoming messages were held
5189 for policy or load reasons. The message log file is now created at the time
5190 the message is received, and an initial "Received" line is written to it.
5192 4. The automatically generated man page for command line options had a minor
5193 bug that caused no ill effects; however, a more serious problem was that
5194 the procedure for building the man page automatically didn't always
5195 operate. Consequently, release 4.20 contains an out-of-date version. This
5196 shouldn't happen again.
5198 5. When building Exim with embedded Perl support, the script that builds the
5199 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5200 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5202 6. The freeze_tell option was not being used for messages that were frozen on
5203 arrival, either by an ACL or by local_scan().
5205 7. Added the smtp_incomplete_transaction log selector.
5207 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5208 was accepting AUTH without a new EHLO.
5210 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5211 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5214 10. Logging of TCP/IP connections (when configured) now happens in the main
5215 daemon process instead of the child process, so that the TCP/IP connection
5216 count is more accurate (but it can never be perfect).
5218 11. The use of "drop" in a nested ACL was not being handled correctly in the
5219 outer ACL. Now, if condition failure induced by the nested "drop" causes
5220 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5221 or "require"), the connection is dropped.
5223 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5224 that yield "discard" can now be used with an "accept" or a "discard" verb,
5225 but an error is generated for any others (because I can't see a useful way
5226 to define what should happen).
5228 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5229 are now processed in the same way as lines in the Exim configuration file.
5230 In particular, continuation lines are supported.
5232 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5234 15. Added -ti meaning -t -i.
5236 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5237 domains, and warn by logging if others are found.
5239 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5240 returned from GIFCONF when Exim is trying to find the list of interfaces on
5241 a host. The code in os.c has been modified to copy each ifreq to an aligned
5242 structure in all cases.
5244 Also, in some cases, the returned ifreq's were being copied to a 'struct
5245 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5246 means the last couple of bytes of an IPv6 address could be chopped if the
5247 ifreq contained only a normal sockaddr (14 bytes storage).
5249 18. Named domain lists were not supported in the hosts_treat_as_local option.
5250 An entry such as +xxxx was not recognized, and was treated as a literal
5253 19. Ensure that header lines added by a DATA ACL are included in the reject log
5254 if the ACL subsequently rejects the message.
5256 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5257 MD5 (which is deprecated).
5259 21. When testing a filter file using -bf, Exim was writing a message when it
5260 took the sender from a "From " line in the message, but it was not doing so
5261 when it took $return_path from a Return-Path: header line. It now does.
5263 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5264 with a valid header line field name (a series of printing characters
5265 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5267 23. Changed "disc" in the source to "disk" to conform to the documentation and
5268 the book and for uniformity.
5270 24. Ignore Sendmail's -Ooption=value command line item.
5272 25. When execve() failed while trying to run a command in a pipe transport,
5273 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5274 could be confused with a return value of 69 from the command itself. This
5275 has been changed to 127, the value the shell returns if it is asked to run
5276 a non-existent command. The wording for the related log line suggests a
5277 non-existent command as the problem.
5279 26. If received_header_text expands to an empty string, do not add a Received:
5280 header line to the message. (Well, it adds a token one on the spool, but
5281 marks it "old" so that it doesn't get used or transmitted.)
5283 27. Installed eximstats 1.28 (addition of -nt option).
5285 28. There was no check for failure on the call to getsockname() in the daemon
5286 code. This can fail if there is a shortage of resources on the system, with
5287 ENOMEM, for example. A temporary error is now given on failure.
5289 29. Contrary to the C standard, it seems that in some environments, the
5290 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5291 program. Exim now does this explicitly; it affects the formatting of
5292 timestamps using strftime().
5294 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5295 complaints. I've now initialized all the variables, to avoid this.
5297 32. Header lines added by a system filter were not being "seen" during
5298 transport-time rewrites.
5300 33. The info_callback() function passed to OpenSSL is set up with type void
5301 (*)(SSL *, int, int), as described somewhere. However, when calling the
5302 function (actually a macro) that sets it up, the type void(*)() is
5303 expected. I've put in a cast to prevent warnings from picky compilers.
5305 34. If a DNS black list lookup found a CNAME record, but there were no A
5306 records associated with the domain it pointed at, Exim crashed.
5308 35. If a DNS black list lookup returned more than one A record, Exim ignored
5309 all but the first. It now scans all returned addresses if a particular IP
5310 value is being sought. In this situation, the contents of the
5311 $dnslist_value variable are a list of all the addresses, separated by a
5314 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5315 to accept a host name and all its aliases if the forward lookup for any of
5316 them yielded the IP address of the incoming connection. Now it accepts only
5317 those names whose forward lookup yields the correct IP address. Any other
5318 names are discarded. This closes a loophole whereby a rogue DNS
5319 administrator could create reverse DNS records to break through a
5320 wildcarded host restriction in an ACL.
5322 37. If a user filter or a system filter that ran in a subprocess used any of
5323 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5324 the wrong values were passed to the pipe command ($thisaddress had the
5325 value of $0, $0 had the value of $1, etc). This bug was introduced by
5326 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5328 38. Improved the line breaking for long SMTP error messages from ACLs.
5329 Previously, if there was no break point between 40 and 75 characters, Exim
5330 left the rest of the message alone. Two changes have been made: (a) I've
5331 reduced the minimum length to 35 characters; (b) if it can't find a break
5332 point between 35 and 75 characters, it looks ahead and uses the first one
5333 that it finds. This may give the occasional overlong line, but at least the
5334 remaining text gets split now.
5336 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5337 file descriptors might be low, and that setting 1000 would always raise it.
5338 It turns out that in some environments, the limit is already over 1000 and
5339 that lowering it causes trouble. So now Exim takes care not to decrease it.
5341 40. When delivering a message, the value of $return_path is set to $sender_
5342 address at the start of routing (routers may change the value). By an
5343 oversight, this default was not being set up when an address was tested by
5344 -bt or -bv, which affected the outcome if any router or filter referred to
5347 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5348 log only when "message" or "log_message" are set. However, if one of the
5349 conditions was an address verification, or a call to a nested ACL, the
5350 messages generated by the underlying test were being passed through. This
5351 no longer happens. The underlying message is available in $acl_verify_
5352 message for both "message" and "log_message" expansions, so it can be
5353 passed through if needed.
5355 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5356 new expansion $bh_ to give the encoded byte string without charset
5357 translation. Translation happens only if iconv() is available; HAVE_ICONV
5358 indicates this at build time. HEADERS_CHARSET gives the charset to
5359 translate to; headers_charset can change it in the configuration, and
5360 "headers charset" can change it in an individual filter file.
5362 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5363 that creates RFC 2047 encoded "words" labels them as that charset instead
5364 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5365 expansion operator; (ii) when Exim creates a From: line for a local
5366 message; (iii) when a header line is rewritten to include a "phrase" part.
5368 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5369 buggy, causing it to skip the first lines of messages whose message ID
5370 ended in 'D'. This would not have bitten before Exim release 4.14, because
5371 message IDs were unlikely to end in 'D' before then. The effect was to have
5372 incorrect size information for certain domains.
5374 45. #include "config.h" was missing at the start of the crypt16.c module. This
5375 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5378 46. If there was a timeout during a "random" callout check, Exim treated it as
5379 a failure of the random address, and carried on sending RSET and the real
5380 address. If the delay was just some slowness somewhere, the response to the
5381 original RCPT would be taken as a response to RSET and so on, causing
5382 mayhem of various kinds.
5384 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5385 when I implemented it. It didn't allow for the fact that some option values
5386 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5387 the right test for positive values.
5389 48. Domain names in DNS records are case-independent. Exim always looks them up
5390 in lower case. Some resolvers return domain names in exactly the case they
5391 appear in the zone file, that is, they may contain uppercase letters. Not
5392 all resolvers do this - some return always lower case. Exim was treating a
5393 change of case by a resolver as a change of domain, similar to a widening
5394 of a domain abbreviation. This triggered its re-routing code and so it was
5395 trying to route what was effectively the same domain again. This normally
5396 caused routing to fail (because the router wouldn't handle the domain
5397 twice). Now Exim checks for this case specially, and just changes the
5398 casing of the domain that it ultimately uses when it transmits the message
5401 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5404 50. If a filter generated a file delivery with a non-absolute name (possible if
5405 no home directory exists for the router), the forbid_file option was not
5408 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5409 the existing equality matching.
5411 52. Exim was using ints instead of ino_t variables in some places where it was
5412 dealing with inode numbers.
5414 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5415 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5416 called TMPDIR, and if it finds it is different, it changes its value.
5418 54. The smtp_printf() function is now made available to local_scan() so
5419 additional output lines can be written before returning. There is also an
5420 smtp_fflush() function to enable the detection of a dropped connection.
5421 The variables smtp_input and smtp_batched_input are exported to
5424 55. Changed the default runtime configuration: the message "Unknown user"
5425 has been removed from the ACL, and instead placed on the localuser router,
5426 using the cannot_route_message feature. This means that any verification
5427 failures that generate their own messages won't get overridden. Similarly,
5428 the "Unrouteable address" message that was in the ACL for unverifiable
5429 relay addresses has also been removed.
5431 56. Added hosts_avoid_esmtp to the smtp transport.
5433 57. The exicyclog script was not checking for the esoteric option
5434 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5435 will work only if exicyclog is run under the appropriate euid.
5437 58. Following a discussion on the list, the rules by which Exim recognises line
5438 endings on incoming messages have been changed. The -dropcr and drop_cr
5439 options are now no-ops, retained only for backwards compatibility. The
5440 following line terminators are recognized: LF CRLF CR. However, special
5441 processing applies to CR:
5443 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5444 nor a local message in the state where . is a terminator.
5446 (ii) If a bare CR is encountered in a header line, an extra space is added
5447 after the line terminator so as not to end the header. The reasoning
5448 behind this is that bare CRs in header lines are most likely either
5449 to be mistakes, or people trying to play silly games.
5451 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5452 was being incorrectly given as 18 bytes larger than it should have been.
5453 This is a VOB (very old bug).
5455 60. This may never have affected anything current, but just in case it has:
5456 When the local host is found other than at the start of a list of hosts,
5457 the local host, those with the same MX, and any that follow, are discarded.
5458 When the list in question was part of a longer list of hosts, the following
5459 hosts (not currently being processed) were also being discarded. This no
5460 longer happens. I'm not sure if this situation could ever has previously
5463 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5466 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5467 header. This is something that is recommended in a new Internet Draft, and
5468 is something that is documented as being done by Sendmail. There are two
5469 possible values. For messages generated by the autoreply transport, Exim
5472 Auto-Submitted: auto-replied
5474 whereas for all other generated messages (e.g. bounces) it adds
5476 Auto-Submitted: auto-generated
5478 63. The "personal" condition in filters now includes a test for the
5479 Auto-Submitted: header. If it contains the string "auto-" the message it
5480 not considered personal.
5482 64. Added rcpt_include_affixes as a generic transport option.
5484 65. Added queue_only_override (default true).
5486 66. Added the syslog_duplication option.
5488 67. If what should have been the first header line of a message consisted of
5489 a space followed by a colon, Exim was mis-interpreting it as a header line.
5490 It isn't of course - it is syntactically invalid and should therefore be
5491 treated as the start of the message body. The misbehaviour could have
5492 caused a number of strange effects, including loss of data in subsequent
5493 header lines, and spool format errors.
5495 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5496 client host had authenticated. This control can now be exercised by an ACL
5497 for more flexibility.
5499 69. By default, callouts do not happen when testing with -bh. There is now a
5500 variant, -bhc, which does actually run the callout code, including
5501 consulting and updating the callout cache.
5503 70. Added support for saslauthd authentication, courtesy of Alexander
5506 71. If statvfs() failed on the spool or log directories while checking their
5507 size for availability, Exim confusingly gave the error "space shortage".
5508 Furthermore, in debugging mode it crashed with a floating point exception.
5509 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5510 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5511 serious problem, Exim now writes to the main and panic logs when this
5512 happens, with details of the failure. It then refuses to accept the
5513 incoming message, giving the message "spool directory problem" or "log
5514 directory problem" with a 421 code for SMTP messages.
5516 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5517 0, 1, and 2 exist, because some OS complain for execs without them (see
5518 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5519 descriptors. However, the code omitted to check that the open succeeded,
5520 causing mysterious errors if for some reason the permissions on /dev/null
5521 got screwed. Now Exim writes a message to the main and panic logs, and
5522 bombs out if it can't open /dev/null.
5524 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5525 interact so that it is all more flexible. It is supposed to remain
5526 backwards compatible. Also added extra_local_interfaces.
5528 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5529 to bomb out with an assertion failure - to the client this appears as a
5530 connection drop. This problem occurs in the part of the code that was taken
5531 from the Samba project. Fortunately, the assertion is in a very simple
5532 function, so I have fixed this by reproducing the function inline in the
5533 one place where it is called, and arranging for authentication to fail
5534 instead of killing the process with assert().
5536 75. The SPA client code was not working when the server requested OEM rather
5537 than Unicode encoding.
5539 76. Added code to make require_files with a specific uid setting more usable in
5540 the case where statting the file as root fails - usually a non-root-mounted
5541 NFS file system. When this happens and the failure is EACCES, Exim now
5542 forks a subprocess and does the per-uid checking as the relevant uid.
5544 77. Added process_log_path.
5546 78. If log_file_path was not explicitly set, a setting of check_log_space or
5547 check_log_inodes was ignored.
5549 79. If a space check for the spool or log partitions fails, the incident is now
5550 logged. Of course, in the latter case the data may get lost...
5552 80. Added the %p formatting code to string_format() so that it can be used to
5553 print addresses in debug_print(). Adjusted all the address printing in the
5554 debugging in store.c to use %p rather than %d.
5556 81. There was a concern that a line of code in smtp_in.c could overflow a
5557 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5558 initially expressed, the concern was not well-founded, because trailing
5559 spaces are removed early. However, if the trailing spaces were followed by
5560 a NULL, they did not get removed, so the overflow was possible. Two fixes
5563 (a) I re-wrote the offending code in a cleaner fashion.
5564 (b) If an incoming SMTP command contains a NULL character, it is rejected
5567 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5568 runs initgroups(), so that if the Exim user is in any additional groups,
5569 they will be used during message reception.
5575 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob