1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
81 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
82 NULL dereference. Report and patch from Alun Jones.
84 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
85 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
86 Not seeing resolver debug output on NetBSD, but suspect this is a
87 resolver implementation change.
89 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
90 Left warnings. Added "eximon gdb" invocation mode.
92 PP/21 Defaulting "accept_8bitmime" to true, not false.
94 PP/22 Added -bw for inetd wait mode support.
96 PP/23 Safety checks against connection reuse where transport options present
97 a different identity to a remote host for two different messages (in
98 expansion of interface, helo_data, tls_certificate).
105 PP/01 Solaris build fix for Oracle's LDAP libraries.
106 Bugzilla 1109, patch from Stephen Usher.
108 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
110 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
113 TF/02 Fix a couple more cases where we did not log the error message
114 when unlink() failed. See also change 4.74-TF/03.
116 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
117 lock up or crash if it happened to be inside a call to libc when it
118 got a SIGUSR1 from exiwhat.
120 The SIGUSR1 handler appends the current process status to the process
121 log which is later printed by exiwhat. It used to use the general
122 purpose logging code to do this, but several functions it calls are
123 not safe for signals.
125 The new output code in the SIGUSR1 handler is specific to the process
126 log, and simple enough that it's easy to inspect for signal safety.
127 Removing some special cases also simplifies the general logging code.
128 Removing the spurious timestamps from the process log simplifies
131 TF/04 Improved ratelimit ACL condition.
133 The /noupdate option has been deprecated in favour of /readonly which
134 has clearer semantics. The /leaky, /strict, and /readonly update modes
135 are mutually exclusive. The update mode is no longer included in the
136 database key; it just determines when the database is updated. (This
137 means that when you upgrde Exim will forget old rate measurements.)
139 Exim now checks that the per_* options are used with an update mode that
140 makes sense for the current ACL. For example, when Exim is processing a
141 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
142 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
143 must specify per_mail/readonly. If you omit the update mode it defaults to
144 /leaky where that makes sense (as before) or /readonly where required.
146 The /noupdate option is now undocumented but still supported for
147 backwards compatibility. It is equivalent to /readonly except that in
148 ACLs where /readonly is required you may specify /leaky/noupdate or
149 /strict/noupdate which are treated the same as /readonly.
151 A useful new feature is the /count= option. This is a generalization
152 of the per_byte option, so that you can measure the throughput of other
153 aggregate values. For example, the per_byte option is now equivalent
154 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
156 The per_rcpt option has been generalized using the /count= mechanism
157 (though it's more complicated than the per_byte equivalence). When it is
158 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
159 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
160 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
161 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
162 Note that using per_rcpt with a non-readonly update mode in more than
163 one ACL will cause the recipients to be double-counted. (The per_mail
164 and per_byte options don't have this problem.)
166 The handling of very low rates has changed slightly. If the computed rate
167 is less than the event's count (usually one) then this event is the first
168 after a long gap. In this case the rate is set to the same as this event's
169 count, so that the first message of a spam run is counted properly.
171 The major new feature is a mechanism for counting the rate of unique
172 events. The new per_addr option counts the number of different
173 recipients that someone has sent messages to in the last time period. It
174 behaves like per_rcpt if all the recipient addresses are different, but
175 duplicate recipient addresses do not increase the measured rate. Like
176 the /count= option this is a general mechanism, so the per_addr option
177 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
178 example, measure the rate that a client uses different sender addresses
179 with the options per_mail/unique=$sender_address. There are further
180 details in the main documentation.
182 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
184 TF/06 Removed a few PCRE remnants.
186 TF/07 Automatically extract Exim's version number from tags in the git
187 repository when doing development or release builds.
189 PP/02 Raise smtp_cmd_buffer_size to 16kB.
190 Bugzilla 879. Patch from Paul Fisher.
192 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
193 Heavily based on revision 40f9a89a from Simon Arlott's tree.
196 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
198 PP/05 Variable $av_failed, true if the AV scanner deferred.
199 Bugzilla 1078. Patch from John Horne.
201 PP/06 Stop make process more reliably on build failure.
202 Bugzilla 1087. Patch from Heiko Schlittermann.
204 PP/07 Make maildir_use_size_file an _expandable_ boolean.
205 Bugzilla 1089. Patch from Heiko Schlittermann.
207 PP/08 Handle ${run} returning more data than OS pipe buffer size.
208 Bugzilla 1131. Patch from Holger Weiß.
210 PP/09 Handle IPv6 addresses with SPF.
211 Bugzilla 860. Patch from Wolfgang Breyha.
213 PP/10 GnuTLS: support TLS 1.2 & 1.1.
215 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
218 PP/11 match_* no longer expand right-hand-side by default.
219 New compile-time build option, EXPAND_LISTMATCH_RHS.
220 New expansion conditions, "inlist", "inlisti".
222 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
224 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
226 PP/14 fix log_write() format string regression from TF/03.
227 Bugzilla 1152. Patch from Dmitry Isaikin.
233 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
235 PP/02 Harmonised TLS library version reporting; only show if debugging.
236 Layout now matches that introduced for other libraries in 4.74 PP/03.
238 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
240 PP/04 New "dns_use_edns0" global option.
242 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
245 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
246 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
248 TK/01 Updated PolarSSL code to 0.14.2.
249 Bugzilla 1097. Patch from Andreas Metzler.
251 PP/07 Catch divide-by-zero in ${eval:...}.
254 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
257 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
258 format-string attack -- SECURITY: remote arbitrary code execution.
260 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
261 time unintentionally subject to list matching rules, letting the header
262 cause arbitrary Exim lookups (of items which can occur in lists, *not*
263 arbitrary string expansion). This allowed for information disclosure.
265 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
266 INT_MIN/-1 -- value coerced to INT_MAX.
272 NM/01 Workround for PCRE version dependency in version reporting
275 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
276 This fixes portability to compilers other than gcc, notably
277 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
279 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
280 makefiles for portability to HP-UX and POSIX correctness.
282 PP/01 Permit LOOKUP_foo enabling on the make command-line.
283 Also via indented variable definition in the Makefile.
284 (Debugging by Oliver Heesakkers).
286 PP/02 Restore caching of spamd results with expanded spamd_address.
287 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
289 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
290 Improves build reliability. Fix from: Frank Elsner
292 NM/02 Fix wide character breakage in the rfc2047 coding
293 Fixes bug 1064. Patch from Andrey N. Oktyabrski
295 NM/03 Allow underscore in dnslist lookups
296 Fixes bug 1026. Patch from Graeme Fowler
298 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
299 Code patches from Adam Ciarcinski of NetBSD.
301 NM/04 Fixed exiqgrep to cope with mailq missing size issue
304 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
305 is logged, to avoid truncation. Patch from John Horne.
307 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
308 Patch from Jakob Hirsch.
310 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
311 SQL string expansion failure details.
312 Patch from Andrey Oktyabrski.
314 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
315 Patch from Simon Arlott.
317 PP/09 New lookups functionality failed to compile on old gcc which rejects
318 extern declarations in function scope.
319 Patch from Oliver Fleischmann
321 PP/10 Use sig_atomic_t for flags set from signal handlers.
322 Check getgroups() return and improve debugging.
323 Fixed developed for diagnosis in bug 927 (which turned out to be
326 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
327 Patch from Mark Zealey.
329 PP/12 Bugzilla 1056: Improved spamd server selection.
330 Patch from Mark Zealey.
332 PP/13 Bugzilla 1086: Deal with maildir quota file races.
333 Based on patch from Heiko Schlittermann.
335 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
336 Patch from Uwe Doering, sign-off by Michael Haardt.
338 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
339 variable declaration deep within a block. Bug and patch from
342 PP/15 lookups-Makefile IRIX compatibilty coercion.
344 PP/16 Make DISABLE_DKIM build knob functional.
346 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
347 Patch by Simon Arlott
349 TF/03 Fix valgrind.h portability to C89 compilers that do not support
350 variable argument macros. Our copy now differs from upstream.
356 TF/01 Failure to get a lock on a hints database can have serious
357 consequences so log it to the panic log.
359 TF/02 Log LMTP confirmation messages in the same way as SMTP,
360 controlled using the smtp_confirmation log selector.
362 TF/03 Include the error message when we fail to unlink a spool file.
364 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
365 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
366 for maintaining out-of-tree patches for some time.
368 PP/01 Bugzilla 139: Documentation and portability issues.
369 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
370 Handle per-OS dynamic-module compilation flags.
372 PP/02 Let /dev/null have normal permissions.
373 The 4.73 fixes were a little too stringent and complained about the
374 permissions on /dev/null. Exempt it from some checks.
375 Reported by Andreas M. Kirchwitz.
377 PP/03 Report version information for many libraries, including
378 Exim version information for dynamically loaded libraries. Created
379 version.h, now support a version extension string for distributors
380 who patch heavily. Dynamic module ABI change.
382 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
383 privilege escalation vulnerability whereby the Exim run-time user
384 can cause root to append content of the attacker's choosing to
387 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
390 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
391 If dropping privileges for untrusted macros, we disabled normal logging
392 on the basis that it would fail; for the Exim run-time user, this is not
393 the case, and it resulted in successful deliveries going unlogged.
394 Fixed. Reported by Andreas Metzler.
400 PP/01 Date: & Message-Id: revert to normally being appended to a message,
401 only prepend for the Resent-* case. Fixes regression introduced in
402 Exim 4.70 by NM/22 for Bugzilla 607.
404 PP/02 Include check_rfc2047_length in configure.default because we're seeing
405 increasing numbers of administrators be bitten by this.
407 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
409 PP/03 Bugzilla 994: added openssl_options main configuration option.
411 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
413 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
415 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
417 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
418 without a peer certificate, leading to a segfault because of an
419 assumption that peers always have certificates. Be a little more
420 paranoid. Problem reported by Martin Tscholak.
422 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
423 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
424 NB: ClamAV planning to remove STREAM in "middle of 2010".
425 CL also introduces -bmalware, various -d+acl logging additions and
426 more caution in buffer sizes.
428 PP/09 Implemented reverse_ip expansion operator.
430 PP/10 Bugzilla 937: provide a "debug" ACL control.
432 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
434 PP/12 Bugzilla 973: Implement --version.
436 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
438 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
440 PP/15 Bugzilla 816: support multiple condition rules on Routers.
442 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
443 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
444 ignore trailing whitespace.
446 JJ/02 prevent non-panic DKIM error from being sent to paniclog
448 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
451 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
452 Notification from Dr Andrew Aitchison.
454 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
455 ExtendedDetectionInfo response format.
456 Notification from John Horne.
458 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
461 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
462 XSL and documented dependency on system catalogs, with examples of how
465 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
468 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
469 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
470 configuration file which is writeable by the Exim user or group.
472 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
473 of configuration files to cover files specified with the -C option if
474 they are going to be used with root privileges, not just the default
477 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
478 option (effectively making it always true).
480 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
481 files to be used while preserving root privileges.
483 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
484 that rogue child processes cannot use them.
486 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
487 run-time user, instead of root.
489 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
490 Exim run-time user without dropping privileges.
492 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
493 result string, instead of calling string_vformat() twice with the same
496 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
497 for other users. Others should always drop root privileges if they use
498 -C on the command line, even for a whitelisted configure file.
500 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
502 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
508 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
509 $data_path, and $header_path variables; fixed documentation bugs and
512 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
513 exipick to access non-standard spools, including the "frozen" queue
516 NM/01 Bugzilla 965: Support mysql stored procedures.
517 Patch from Alain Williams
519 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
521 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
522 Patch from Andreas Metzler
524 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
525 Patch from Kirill Miazine
527 NM/05 Bugzilla 671: Added umask to procmail example.
529 JJ/03 installed exipick 20100323.0, fixing doc bug
531 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
532 directory. Notification and patch from Dan Rosenberg.
534 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
536 TK/02 Improve log output when DKIM signing operation fails.
538 MH/01 Treat the transport option dkim_domain as a colon separated
539 list, not as a single string, and sign the message with each element,
540 omitting multiple occurences of the same signer.
542 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
543 Bugzilla 985, 986. Patch by Simon Arlott
545 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
546 Patch by Simon Arlott
548 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
549 MBX locking. Notification from Dan Rosenberg.
555 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
557 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
559 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
561 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
563 NM/04 Bugzilla 915: Flag broken perl installation during build.
569 TK/01 Added patch by Johannes Berg that expands the main option
570 "spamd_address" if it starts with a dollar sign.
572 TK/02 Write list of recipients to X-Envelope-Sender header when building
573 the mbox-format spool file for content scanning (suggested by Jakob
576 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
577 (http://www.dcc-servers.net/) support via dccifd. Activated by
578 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
580 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
581 by Mark Daniel Reidel <mr@df.eu>.
583 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
584 When building exim an external PCRE library is now needed -
585 PCRE is a system library on the majority of modern systems.
586 See entry on PCRE_LIBS in EDITME file.
588 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
589 conversation. Added nologin parameter to request.
590 Patch contributed by Kirill Miazine.
592 TF/01 Do not log submission mode rewrites if they do not change the address.
594 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
596 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
597 log files in place. Contributed by Roberto Lima.
599 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
601 TF/03 Bugzilla 615: When checking the local_parts router precondition
602 after a local_part_suffix or local_part_prefix option, Exim now
603 does not use the address's named list lookup cache, since this
604 contains cached lookups for the whole local part.
606 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
607 Robert Millan. Documentation is in experimental-spec.txt.
609 TF/04 Bugzilla 668: Fix parallel build (make -j).
611 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
613 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
614 Patch provided by Jan Srzednicki.
616 TF/05 Leading white space used to be stripped from $spam_report which
617 wrecked the formatting. Now it is preserved.
619 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
620 that they are available at delivery time.
622 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
624 TF/08 TLS error reporting now respects the incoming_interface and
625 incoming_port log selectors.
627 TF/09 Produce a more useful error message if an SMTP transport's hosts
628 setting expands to an empty string.
630 NM/06 Bugzilla 744: EXPN did not work under TLS.
631 Patch provided by Phil Pennock.
633 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
634 Patch provided by Richard Godbee.
636 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
637 acl_smtp_notquit, added index entry.
639 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
640 Patch provided by Eugene Bujak.
642 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
643 accept(). Patch provided by Maxim Dounin.
645 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
646 Patch provided by Phil Pennock.
648 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
650 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
651 Patch provided by Brad "anomie" Jorsch.
653 NM/14 Bugzilla 622: Added timeout setting to transport filter.
654 Patch provided by Dean Brooks.
656 TK/05 Add native DKIM support (does not depend on external libraries).
658 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
659 Patch provided by Graeme Fowler.
661 NM/16 Bugzilla 851: Documentation example syntax fix.
663 NM/17 Changed NOTICE file to remove references to embedded PCRE.
665 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
668 NM/19 Bugzilla 745: TLS version reporting.
669 Patch provided by Phil Pennock.
671 NM/20 Bugzilla 167: bool: condition support.
672 Patch provided by Phil Pennock.
674 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
675 clients. Patch provided by Phil Pennock.
677 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
678 Patch provided by Brad "anomie" Jorsch.
680 NM/23 Bugzilla 687: Fix misparses in eximstats.
681 Patch provided by Heiko Schlittermann.
683 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
684 Patch provided by Heiko Schlittermann.
686 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
687 plus update to original patch.
689 NM/26 Bugzilla 799: Documentation correction for ratelimit.
691 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
692 Patch provided by David Brownlee.
694 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
696 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
698 NM/30 Bugzilla 888: TLS documentation bugfixes.
700 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
702 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
703 Unlike the original bugzilla I have changed all shell scripts in src tree.
705 NM/33 Bugzilla 898: Transport filter timeout fix.
706 Patch by Todd Rinaldo.
708 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
709 Patch by Serge Demonchaux.
711 NM/35 Bugzilla 39: Base64 decode bug fixes.
712 Patch by Jakob Hirsch.
714 NM/36 Bugzilla 909: Correct connect() call in dcc code.
716 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
718 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
720 NM/39 Bugzilla 911: Fixed MakeLinks build script.
726 TK/01 Add preliminary DKIM support. Currently requires a forked version of
727 ALT-N's libdkim that I have put here:
728 http://duncanthrax.net/exim-experimental/
730 Note to Michael Haardt: I had to rename some vars in sieve.c. They
731 were called 'true' and it seems that C99 defines that as a reserved
732 keyword to be used with 'bool' variable types. That means you could
733 not include C99-style headers which use bools without triggering
734 build errors in sieve.c.
736 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
737 as mailq or other aliases. Changed the --help handling significantly
738 to do whats expected. exim_usage() emits usage/help information.
740 SC/01 Added the -bylocaldomain option to eximstats.
742 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
744 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
746 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
752 PH/01 Another patch from the Sieve maintainer.
754 PH/02 When an IPv6 address is converted to a string for single-key lookup
755 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
756 dots are used instead of colons so that keys in lsearch files need not
757 contain colons. This was done some time before quoting was made available
758 in lsearch files. However, iplsearch files do require colons in IPv6 keys
759 (notated using the quote facility) so as to distinguish them from IPv4
760 keys. This meant that lookups for IP addresses in host lists did not work
761 for iplsearch lookups.
763 This has been fixed by arranging for IPv6 addresses to be expressed with
764 colons if the lookup type is iplsearch. This is not incompatible, because
765 previously such lookups could never work.
767 The situation is now rather anomolous, since one *can* have colons in
768 ordinary lsearch keys. However, making the change in all cases is
769 incompatible and would probably break a number of configurations.
771 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
774 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
775 conversion specification without a maximum field width, thereby enabling
776 a rogue spamd server to cause a buffer overflow. While nobody in their
777 right mind would setup Exim to query an untrusted spamd server, an
778 attacker that gains access to a server running spamd could potentially
779 exploit this vulnerability to run arbitrary code as the Exim user.
781 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
782 $primary_hostname instead of what libspf2 thinks the hosts name is.
784 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
785 a directory entry by the name of the lookup key. Previously, if a
786 symlink pointed to a non-existing file or a file in a directory that
787 Exim lacked permissions to read, a lookup for a key matching that
788 symlink would fail. Now it is enough that a matching directory entry
789 exists, symlink or not. (Bugzilla 503.)
791 PH/03 The body_linecount and body_zerocount variables are now exported in the
794 PH/04 Added the $dnslist_matched variable.
796 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
797 This means they are set thereafter only if the connection becomes
800 PH/06 Added the client_condition to authenticators so that some can be skipped
801 by clients under certain conditions.
803 PH/07 The error message for a badly-placed control=no_multiline_responses left
804 "_responses" off the end of the name.
806 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
808 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
809 (without spaces) instead of just copying the configuration text.
811 PH/10 Added the /noupdate option to the ratelimit ACL condition.
813 PH/11 Added $max_received_linelength.
815 PH/12 Added +ignore_defer and +include_defer to host lists.
817 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
818 way in which PCRE > 7.0 is built.
820 PH/14 Implemented queue_only_load_latch.
822 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
823 MAIL command. The effect was to mangle the value on 64-bit systems.
825 PH/16 Another patch from the Sieve maintainer.
827 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
829 PH/18 If a system quota error occurred while trying to create the file for
830 a maildir delivery, the message "Mailbox is full" was not appended to the
831 bounce if the delivery eventually timed out. Change 4.67/27 below applied
832 only to a quota excession during the actual writing of the file.
834 PH/19 It seems that peer DN values may contain newlines (and other non-printing
835 characters?) which causes problems in log lines. The DN values are now
836 passed through string_printing() before being added to log lines.
838 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
839 and InterBase are left for another time.)
841 PH/21 Added message_body_newlines option.
843 PH/22 Guard against possible overflow in moan_check_errorcopy().
845 PH/23 POSIX allows open() to be a macro; guard against that.
847 PH/24 If the recipient of an error message contained an @ in the local part
848 (suitably quoted, of course), incorrect values were put in $domain and
849 $local_part during the evaluation of errors_copy.
855 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
856 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
859 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
860 issue a MAIL command.
862 PH/02 In an ACL statement such as
864 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
866 if a client was not listed at all, or was listed with a value other than
867 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
868 the condition was not true (as it should be), so access was not denied.
869 The bug was that the ! inversion was incorrectly passed on to the second
870 item. This has been fixed.
872 PH/03 Added additional dnslists conditions == and =& which are different from
873 = and & when the dns lookup returns more than one IP address.
875 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
876 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
878 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
879 FSYNC, which compiles an option called disable_fsync that allows for
880 bypassing fsync(). The documentation is heavily laced with warnings.
882 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
884 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
885 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
886 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
887 including adding "make clean"; (3) Added -fPIC when compiling the test
888 dynamically loaded module, to get rid of a warning.
890 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
891 message fails, move_frozen_messages = true and ignore_bounce_errors_after
892 = 0s. The bug is otherwise harmless.
894 PH/07 There was a bug in the dovecot authenticator such that the value of
895 $auth1 could be overwritten, and so not correctly preserved, after a
896 successful authentication. This usually meant that the value preserved by
897 the server_setid option was incorrect.
899 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
901 PH/09 Installed PCRE release 7.0.
903 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
904 run for batched SMTP input. It is now run at the start of every message
905 in the batch. While fixing this I discovered that the process information
906 (output by running exiwhat) was not always getting set for -bs and -bS
907 input. This is fixed, and it now also says "batched" for BSMTP.
909 PH/11 Added control=no_pipelining.
911 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
912 patch, slightly modified), and move the expansion of helo_data till after
913 the connection is made in the smtp transport (so it can use these
916 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
918 PH/14 Added log_selector = +pid.
920 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
922 PH/16 Add ${if forany and ${if forall.
924 PH/17 Added dsn_from option to vary the From: line in DSNs.
926 PH/18 Flush SMTP output before performing a callout, unless control =
927 no_callout_flush is set.
929 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
930 was true (the default) a successful delivery failed to delete the retry
931 item, thus causing premature timeout of the address. The bug is now
934 PH/20 Added hosts_avoid_pipelining to the smtp transport.
936 PH/21 Long custom messages for fakedefer and fakereject are now split up
937 into multiline reponses in the same way that messages for "deny" and
938 other ACL rejections are.
940 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
941 with slight modification.
943 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
944 draft, changing the syntax and factoring some duplicate code".
946 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
947 for deliveries of the second and subsequent messages over the same SMTP
950 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
951 ${reduce, with only minor "tidies".
953 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
955 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
956 expansion side effects.
958 PH/27 When a message times out after an over-quota error from an Exim-imposed
959 quota, the bounce message says "mailbox is full". This message was not
960 being given when it was a system quota that was exceeded. It now should
963 MH/03 Made $recipients available in local_scan(). local_scan() already has
964 better access to the recipient list through recipients_list[], but
965 $recipients can be useful in postmaster-provided expansion strings.
967 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
968 in the case of a MAIL command with additional options following the
969 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
970 were accidentally chopped off.
972 PH/29 SMTP synchronization checks are implemented when a command is read -
973 there is a check that no more input is waiting when there shouldn't be
974 any. However, for some commands, a delay in an ACL can mean that it is
975 some time before the response is written. In this time, more input might
976 arrive, invalidly. So now there are extra checks after an ACL has run for
977 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
978 pipelining has not been advertised.
980 PH/30 MH's patch to allow iscntrl() characters to be list separators.
982 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
983 returned in the SMTP response when smtp_return_error_details was false.
986 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
987 instead of the C I/O that was originally supplied, because problems were
990 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
991 Exim which did not show up earlier: it was assuming that a call to
992 SSL_CTX_set_info_callback() might give an error value. In fact, there is
993 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
994 was a macro that became an assignment, so it seemed to work. This has
995 changed to a proper function call with a void return, hence the compile
996 error. Exim's code has been fixed.
998 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1001 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1003 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1005 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1006 because it turns out to be too big or there is a timeout) there is no
1007 "Completed" line in the log. When some messages of this type were
1008 selected by exigrep, they were listed as "not completed". Others were
1009 picked up by some special patterns. I have improved the selection
1010 criteria to be more general.
1012 PH/38 The host_find_failed option in the manualroute router can now be set
1013 to "ignore", to completely ignore a host whose IP address cannot be
1014 found. If all hosts are ignored, the behaviour is controlled by the new
1015 host_all_ignored option.
1017 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1018 homing or because of multiple MX records with /mx) generated more than
1019 one IP address, and the following item turned out to be the local host,
1020 all the secondary addresses of the first item were incorrectly removed
1021 from the list, along with the local host and any following hosts (which
1022 is what is supposed to happen).
1024 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1025 whoever called Exim into the -H file. In the case of the daemon it was
1026 behaving confusingly. When first started, it used values for whoever
1027 started the daemon, but after a SIGHUP it used the Exim user (because it
1028 calls itself on a restart). I have changed the code so that it now always
1031 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1032 message are rejected with the same error (e.g. no authentication or bad
1033 sender address), and a DATA command is nevertheless sent (as can happen
1034 with PIPELINING or a stupid MUA), the error message that was given to the
1035 RCPT commands is included in the rejection of the DATA command. This is
1036 intended to be helpful for MUAs that show only the final error to their
1039 PH/42 Another patch from the Sieve maintainer.
1041 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1042 Eximstats - Fixed some broken HTML links and added missing column headers
1044 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1045 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1047 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1049 PH/43 Yet another patch from the Sieve maintainer.
1051 PH/44 I found a way to check for a TCP/IP connection going away before sending
1052 the response to the final '.' that terminates a message, but only in the
1053 case where the client has not sent further data following the '.'
1054 (unfortunately, this is allowed). However, in many cases there won't be
1055 any further data because there won't be any more messages to send. A call
1056 to select() can be used: if it shows that the input is "ready", there is
1057 either input waiting, or the socket has been closed. An attempt to read
1058 the next input character can distinguish the two cases. Previously, Exim
1059 would have sent an OK response which the client would never have see.
1060 This could lead to message repetition. This fix should cure that, at
1061 least in a lot of common cases.
1063 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1064 advertised in response to EHLO.
1070 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1071 fixed by 4.65/MH/01 (is this a record?) are fixed:
1073 (i) An empty string was always treated as zero by the numeric comparison
1074 operators. This behaviour has been restored.
1076 (ii) It is documented that the numeric comparison operators always treat
1077 their arguments as decimal numbers. This was broken in that numbers
1078 starting with 0 were being interpreted as octal.
1080 While fixing these problems I realized that there was another issue that
1081 hadn't been noticed. Values of message_size_limit (both the global option
1082 and the transport option) were treated as octal if they started with 0.
1083 The documentation was vague. These values are now always treated as
1084 decimal, and I will make that clear in the documentation.
1090 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1091 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1094 MH/01 Don't check that the operands of numeric comparison operators are
1095 integers when their expansion is in "skipping" mode (fixes bug
1096 introduced by 4.64-PH/07).
1098 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1099 child addresses, Exim now panics and dies. Previously, because the count
1100 is held in a short int, deliveries were likely to be lost. As such a
1101 large number of recipients for a single message is ridiculous
1102 (performance will be very, very poor), I have chosen to impose a limit
1103 rather than extend the field.
1109 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1110 leftover -K file (the existence of which was triggered by #402).
1111 While we were at it, introduced process PID as part of the -K
1112 filename. This should rule out race conditions when creating
1115 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1116 processing considerably. Previous code took too long for large mails,
1117 triggering a timeout which in turn triggers #401.
1119 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1120 in the DK code in transports.c. sendfile() is not really portable,
1121 hence the _LINUX specificness.
1123 TF/01 In the add_headers option to the mail command in an Exim filter,
1124 there was a bug that Exim would claim a syntax error in any
1125 header after the first one which had an odd number of characters
1128 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1129 callout verification, Exim cached a "reject" for the entire domain. This
1130 is correct for most verifications, but it is not correct for a recipient
1131 verification with use_sender or use_postmaster set, because in that case
1132 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1133 case of MAIL FROM:<> rejection from other early rejections (e.g.
1134 rejection of HELO). When verifying a recipient using a non-null MAIL
1135 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1136 Whatever the result of the callout, the value of the domain cache is
1137 left unchanged (for any other kind of callout, getting as far as trying
1138 RCPT means that the domain itself is ok).
1140 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1143 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1144 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1145 session, the daemon ignores ECONNECT errors and logs others; it now
1146 ignores EPIPE as well.
1148 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1149 (quoted-printable decoding).
1151 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1152 later the small subsequent patch to fix an introduced bug.
1154 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1156 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1158 PH/08 An error is now given if message_size_limit is specified negative.
1160 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1161 to be given (somewhat) arbitrary names.
1163 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1166 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1167 miscellaneous code fixes
1169 PH/10 Added the log_reject_target ACL modifier to specify where to log
1172 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1173 hostname. This is wrong, because it relates to the incoming message (and
1174 probably the interface on which it is arriving) and not to the outgoing
1175 callout (which could be using a different interface). This has been
1176 changed to use the value of the helo_data option from the smtp transport
1177 instead - this is what is used when a message is actually being sent. If
1178 there is no remote transport (possible with a router that sets up host
1179 addresses), $smtp_active_hostname is used.
1181 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1182 tweaks were necessary in order to get it to work (see also 21 below):
1183 (a) The code assumed that strncpy() returns a negative number on buffer
1184 overflow, which isn't the case. Replaced with Exim's string_format()
1186 (b) There were several signed/unsigned issues. I just did the minimum
1187 hacking in of casts. There is scope for a larger refactoring.
1188 (c) The code used strcasecmp() which is not a standard C function.
1189 Replaced with Exim's strcmpic() function.
1190 (d) The code set only $1; it now sets $auth1 as well.
1191 (e) A simple test gave the error "authentication client didn't specify
1192 service in request". It would seem that Dovecot has changed its
1193 interface. Fortunately there's a specification; I followed it and
1194 changed what the client sends and it appears to be working now.
1196 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1199 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1200 address A is aliased to B and C, where B exists and C does not. Without
1201 -v the output is "A verified" because verification stops after a
1202 successful redirection if more than one address is generated. However,
1203 with -v the child addresses are also verified. Exim was outputting "A
1204 failed to verify" and then showing the successful verification for C,
1205 with its parentage. It now outputs "B failed to verify", showing B's
1206 parentage before showing the successful verification of C.
1208 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1209 look up a TXT record in a specific list after matching in a combined
1212 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1213 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1214 they consult the DNS. I had assumed they would set it the way they
1215 wanted; and indeed my experiments on Linux seem to show that in some
1216 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1217 To be on the safe side, however, I have now made the interface to
1218 host_find_byname() similar to host_find_bydns(), with an argument
1219 containing the DNS resolver options. The host_find_byname() function now
1220 sets these options at its start, just as host_find_bydns() does. The smtp
1221 transport options dns_qualify_single and dns_search_parents are passed to
1222 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1223 of host_find_byname() use the default settings of RES_DEFNAMES
1224 (qualify_single) but not RES_DNSRCH (search_parents).
1226 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1227 spool_read_header() do less string testing, by means of a preliminary
1228 switch on the second character of optional "-foo" lines. (This is
1229 overdue, caused by the large number of possibilities that now exist.
1230 Originally there were few.) While I was there, I also converted the
1231 str(n)cmp tests so they don't re-test the leading "-" and the first
1232 character, in the hope this might squeeze out yet more improvement.
1234 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1235 flag allowing group syntax was set by the header_syntax check but not
1236 turned off, possible causing trouble later; (2) The flag was not being
1237 set at all for the header_verify test, causing "group"-style headers to
1238 be rejected. I have now set it in this case, and also caused header_
1239 verify to ignore an empty address taken from a group. While doing this, I
1240 came across some other cases where the code for allowing group syntax
1241 while scanning a header line wasn't quite right (mostly, not resetting
1242 the flag correctly in the right place). These bugs could have caused
1243 trouble for malformed header lines. I hope it is now all correct.
1245 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1246 with the "reply" argument non-NULL. The code, however (which originally
1247 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1248 but it didn't always do it. This confused somebody who was copying the
1249 code for some other use. I have removed all the tests.
1251 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1252 feature that was used to support insecure browsers during the U.S. crypto
1253 embargo. It requires special client support, and Exim is probably the
1254 only MTA that supported it -- and would never use it because real RSA is
1255 always available. This code has been removed, because it had the bad
1256 effect of slowing Exim down by computing (never used) parameters for the
1257 RSA_EXPORT functionality.
1259 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1260 authenticator to fail if there's a tab character in the incoming data
1261 (there should never be unless someone is messing about, as it's supposed
1262 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1263 if the connection is using TLS or if the remote IP is the same as the
1264 local IP, and the "valid-client-cert option" if a client certificate has
1267 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1268 authenticators. This can be used for authorization after authentication
1269 succeeds. (In the case of plaintext, it servers for both authentication
1272 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1273 if any retry times were supplied.
1275 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1276 connection, where there is no client IP address to check. In this
1277 situation, the verify now always succeeds.
1279 PH/25 Applied John Jetmore's -Mset patch.
1281 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1283 PH/27 In a string expansion for a processed (not raw) header when multiple
1284 headers of the same name were present, leading whitespace was being
1285 removed from all of them, but trailing whitespace was being removed only
1286 from the last one. Now trailing whitespace is removed from each header
1287 before concatenation. Completely empty headers in a concatenation (as
1288 before) are ignored.
1290 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1291 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1293 PH/29 [Removed. This was a change that I later backed out, and forgot to
1294 correct the ChangeLog entry (that I had efficiently created) before
1295 committing the later change.]
1297 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1298 address errors (4xx response to RCPT) over the same connection as other
1299 messages routed to the same hosts. Such deliveries are always "forced",
1300 so retry times are not inspected. This resulted in far too many retries
1301 for the affected addresses. The effect occurred only when there were more
1302 hosts than the hosts_max_try setting in the smtp transport when it had
1303 the 4xx errors. Those hosts that it had tried were not added to the list
1304 of hosts for which the message was waiting, so if all were tried, there
1305 was no problem. Two fixes have been applied:
1307 (i) If there are any address or message errors in an SMTP delivery, none
1308 of the hosts (tried or untried) are now added to the list of hosts
1309 for which the message is waiting, so the message should not be a
1310 candidate for sending over the same connection that was used for a
1311 successful delivery of some other message. This seems entirely
1312 reasonable: after all the message is NOT "waiting for some host".
1313 This is so "obvious" that I'm not sure why it wasn't done
1314 previously. Hope I haven't missed anything, but it can't do any
1315 harm, as the worst effect is to miss an optimization.
1317 (ii) If, despite (i), such a delivery is accidentally attempted, the
1318 routing retry time is respected, so at least it doesn't keep
1319 hammering the server.
1321 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1322 in ${readsocket because some servers need this prod.
1324 PH/32 Added some extra debug output when updating a wait-xxx database.
1326 PH/33 The hint "could be header name not terminated by colon", which has been
1327 given for certain expansion errors for a long time, was not being given
1328 for the ${if def:h_colon_omitted{... case.
1330 PH/34 The spec says: "With one important exception, whenever a domain list is
1331 being scanned, $domain contains the subject domain." There was at least
1332 one case where this was not true.
1334 PH/35 The error "getsockname() failed: connection reset by peer" was being
1335 written to the panic log as well as the main log, but it isn't really
1336 panic-worthy as it just means the connection died rather early on. I have
1337 removed the panic log writing for the ECONNRESET error when getsockname()
1340 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1341 runs only) independently of the message's sender address. This meant
1342 that, if the 4xx error was in fact related to the sender, a different
1343 message to the same recipient with a different sender could confuse
1344 things. In particualar, this can happen when sending to a greylisting
1345 server, but other circumstances could also provoke similar problems.
1346 I have changed the default so that the retry time for these errors is now
1347 based a combination of the sender and recipient addresses. This change
1348 can be overridden by setting address_retry_include_sender=false in the
1351 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1352 remote server are returned as part of bounce messages. This was not
1353 happening for LMTP over a pipe (the lmtp transport), but now it is the
1354 same for both kinds of LMTP.
1356 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1357 in header lines that were in fact CNAMEs. This is no longer the case.
1359 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1360 and queue runs started by the daemon processed all messages. This has
1361 been fixed so that -R and -S can now usefully be given with -q<time>.
1363 PH/40 Import PCRE release 6.7 (fixes some bugs).
1365 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1367 PH/42 Give an error if -q is specified more than once.
1369 PH/43 Renamed the variables $interface_address and $interface_port as
1370 $received_ip_address and $received_port, to make it clear that these
1371 values apply to message reception, and not to the outgoing interface when
1372 a message is delivered. (The old names remain recognized, of course.)
1374 PH/44 There was no timeout on the connect() call when using a Unix domain
1375 socket in the ${readsocket expansion. There now is.
1377 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1378 be meaningful with "accept".
1380 SC/01 Eximstats V1.43
1381 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1383 SC/02 Eximstats V1.44
1384 Use a glob alias rather than an array ref in the generated
1385 parser. This improves both readability and performance.
1387 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1388 Collect SpamAssassin and rejection statistics.
1389 Don't display local sender or destination tables unless
1390 there is data to show.
1391 Added average volumes into the top table text output.
1393 SC/04 Eximstats V1.46
1394 Collect data on the number of addresses (recipients)
1395 as well as the number of messages.
1397 SC/05 Eximstats V1.47
1398 Added 'Message too big' to the list of mail rejection
1399 reasons (thanks to Marco Gaiarin).
1401 SC/06 Eximstats V1.48
1402 Mainlog lines which have GMT offsets and are too short to
1403 have a flag are now skipped.
1405 SC/07 Eximstats V1.49 (Alain Williams)
1406 Added the -emptyok flag.
1408 SC/08 Eximstats V1.50
1409 Fixes for obtaining the IP address from reject messages.
1411 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1412 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1413 whitesspace changes from 4.64-PH/27
1415 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1418 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1419 are found, allow negative numbers in numeric criteria)
1421 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1423 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1424 to match changes made in 4.64-PH/43
1426 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1428 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1429 "domain" is actually an IP address, and force a failure. This locks out
1430 those revolvers/nameservers that support "A-for-A" lookups, in
1431 contravention of the specifications.
1433 PH/48 When a host name was looked up from an IP address, and the subsequent
1434 forward lookup of the name timed out, the host name was left in
1435 $sender_host_name, contrary to the specification.
1437 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1438 restricted to single-key lookups, Exim was not diagnosing an error if
1439 * or *@ was used with a query-style lookup.
1441 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1443 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1444 long ago, but noone interested enough thought of it. Let's just say that
1445 the "1.1" means that there are some new functions that weren't there at
1446 some point in the past.
1448 PH/51 Error processing for expansion failure of helo_data from an smtp
1449 transport during callout processing was broken.
1451 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1452 tested/used via the -bh/-bhc/-bs options.
1454 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1455 bug, fixed in subsequent PCRE releases).
1457 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1458 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1460 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1466 SC/01 Use a glob alias rather than an array ref in eximstats generated
1467 parser. This improves both readability and performance.
1469 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1470 Don't display local sender or destination tables in eximstats unless
1471 there is data to show.
1472 Added average volumes into the eximstats top table text output.
1474 SC/03 Collect data on the number of addresses (recipients) as well
1475 as the number of messages in eximstats.
1477 TF/01 Correct an error in the documentation for the redirect router. Exim
1478 does (usually) call initgroups() when daemonizing.
1480 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1481 with consistent privilege compared to when running as a daemon.
1483 TF/03 Note in the spec that $authenticated_id is not set for local
1484 submissions from trusted users.
1486 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1487 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1489 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1490 by adding some example configuration directives to the default
1491 configuration file. A little bit of work is required to uncomment the
1492 directives and define how usernames and passwords are checked, but
1493 there is now a framework to start from.
1495 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1496 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1497 without this. I don't know how relevant this is to other LDAP libraries.
1499 PH/02 Add the verb name to the "unknown ACL verb" error.
1501 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1503 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1505 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1506 directory not expanded when it should be if an expanded home directory
1507 was set for the address (which is overridden by the transport).
1509 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1512 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1513 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1514 because it is too late at that time, and has no effect.
1516 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1517 security issue with \' (bugzilla #107). I could not use the
1518 PQescapeStringConn() function, because it needs a PGconn value as one of
1521 PH/08 When testing addresses using -bt, indicate those final addresses that
1522 are duplicates that would not cause an additional delivery. At least one
1523 person was confused, thinking that -bt output corresponded to deliveries.
1524 (Suppressing duplicates isn't a good idea as you lose the information
1525 about possibly different redirections that led to the duplicates.)
1527 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1528 systems where poll() doesn't work, in particular OS X.
1530 PH/10 Added more information to debugging output for retry time not reached.
1532 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1533 operations in malware.c.
1535 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1538 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1539 syslog_duplication set false, log lines that would normally be written
1540 both the the main log and to the reject log were not written to syslog at
1543 PH/14 In the default configuration, change the use of "message" in ACL warn
1544 statements to "add_header".
1546 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1547 not followed by a command (e.g. "seen endif").
1549 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1550 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1553 PH/17 Added extra conditions to the default value of delay_warning_condition
1557 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1558 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1559 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1562 The Auto-Submitted: and various List- headers are standardised, whereas I
1563 don't think Precedence: ever was.
1565 PH/18 Refactored debugging code in route_finduser() to show more information,
1566 in particular, the error code if getpwnam() issues one.
1568 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1569 This is apparently needed in addition to the PH/07 change above to avoid
1570 any possible encoding problems.
1572 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1573 but not after initializing Perl.
1575 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1576 output them only if debugging. By default they are written stderr,
1577 apparently, which is not desirable.
1579 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1582 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1585 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1587 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1588 authenticated or an ident call has been made. Suppress the default
1589 values for $authenticated_id and $authenticated_sender (but permit -oMai
1590 and -oMas) when testing with -bh.
1592 PH/24 Re-jigged the order of the tests in the default configuration so that the
1593 tests for valid domains and recipients precede the DNS black list and CSA
1594 tests, on the grounds that those ones are more expensive.
1596 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1597 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1598 This bug exists in every version of Exim that I still have, right back to
1601 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1602 However, an attempt to turn on case-sensitivity in a regex key by
1603 including (?-i) didn't work because the subject string was already
1604 lowercased, and the effects were non-intuitive. It turns out that a
1605 one-line patch can be used to allow (?-i) to work as expected.
1611 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1612 other effects) broke the use of negated acl sub-conditions.
1614 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1617 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1618 "Deny" causes Exim to reject the incoming connection with a 554 error.
1619 Unfortunately, if there is a major crisis, such as a disk failure,
1620 tcp-wrappers gives "deny", whereas what one would like would be some
1621 kind of temporary error. A kludge has been added to help with this.
1622 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1623 554 error is used if errno is still zero or contains ENOENT (which occurs
1624 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1627 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1629 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1630 errors. Otherwise a message that provokes a temporary error (when other
1631 messages do not) can cause a whole host to time out.
1633 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1634 addresses were routed directly to files or pipes from a redirect router.
1635 File deliveries just didn't batch; pipe deliveries might have suffered
1638 PH/06 A failure to get a lock for a hints database would erroneously always say
1639 "Failed to get write lock", even when it was really a read lock.
1641 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1642 of 0600. This has been changed to use the value of the lockfile_mode
1643 option (which defaults to 0600).
1645 PH/08 Applied small patch from the Sieve maintainer.
1647 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1648 folder from quota calculations, a direct delivery into this folder messed
1649 up the contents of the maildirsize file. This was because the regex was
1650 used only to exclude .Trash (or whatever) when the size of the mailbox
1651 was calculated. There was no check that a delivery was happening into an
1652 excluded directory. This bug has been fixed by ignoring all quota
1653 processing for deliveries into excluded directories.
1655 PH/10 Added the maildirfolder_create_regex option to appendfile.
1661 PH/01 The code for finding all the local interface addresses on a FreeBSD
1662 system running IPv6 was broken. This may well have applied to all BSD
1663 systems, as well as to others that have similar system calls. The broken
1664 code found IPv4 interfaces correctly, but gave incorrect values for the
1665 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1666 that it would not match correctly against @[] and not recognize the IPv6
1669 PH/02 The ipliteral router was not recognizing addresses of the form user@
1670 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1672 PH/03 Added disable_ipv6.
1674 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1675 decoded form, because it is most often used to construct To: headers
1676 lines in autoreplies, and the decoded form may well be syntactically
1677 invalid. However, $reply_address has leading white space removed, and all
1678 newlines turned into spaces so that the autoreply transport does not
1681 PH/05 If group was specified without a user on a router, and no group or user
1682 was specified on a transport, the group from the router was ignored.
1684 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1685 for visible compile-time settings that can be used to change these
1686 numbers, for those that want even more. Backwards compatibility with old
1687 spool files has been maintained. However, going back to a previous Exim
1688 release will lost any variables that are in spool files.
1690 PH/07 Two small changes when running in the test harness: increase delay when
1691 passing a TCP/IP connection to a new process, in case the original
1692 process has to generate a bounce, and remove special handling of
1693 127.0.0.2 (sic), which is no longer necessary.
1695 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1696 be the same on different OS.
1698 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1701 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1702 whether --show-vars was specified or not
1704 JJ/02 exipick: Added support for new ACL variable spool format introduced
1707 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1708 syntactically invalid From: or Reply-to: line, and a filter used this to
1709 generate an autoreply, and therefore failed to obtain an address for the
1710 autoreply, Exim could try to deliver to a non-existent relative file
1711 name, causing unrelated and misleading errors. What now happens is that
1712 it logs this as a hard delivery error, but does not attempt to create a
1715 PH/11 The exinext utility has a -C option for testing purposes, but although
1716 the given file was scanned by exinext itself; it wasn't being passed on
1717 when Exim was called.
1719 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1720 an end-of-file indication when reading a command response.
1722 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1723 compiled. In many other places in Exim, IPv6 addresses are always
1724 recognized, so I have changed this. It also means that IPv4 domain
1725 literals of the form [IPV4:n.n.n.n] are now always recognized.
1727 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1728 used if the router is not running as root, for example, when verifying at
1729 ACL time, or when using -bh. The debugging output from this situation was
1730 non-existent - all you got was a failure to exec. I have made two
1733 (a) Failures to set uid/gid, the current directory, or a process leader
1734 in a subprocess such as that created by queryprogram now generate
1735 suitable debugging ouput when -d is set.
1737 (b) The queryprogram router detects when it is not running as root,
1738 outputs suitable debugging information if -d is set, and then runs
1739 the subprocess without attempting to change uid/gid.
1741 PH/15 Minor change to Makefile for building test_host (undocumented testing
1744 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1745 additional section of a DNS packet that returns MX or SRV records.
1746 Instead, it always explicitly searches for A/AAAA records. This avoids
1747 major problems that occur when a DNS server includes only records of one
1748 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1749 fixed another bug: if SRV records were looked up and the corresponding
1750 address records were *not* found in the additional section, the port
1751 values from the SRV records were lost.
1753 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1754 using the correct key (the original address) when searching the retry
1755 rules in order to find which one to use for generating the retry hint.
1757 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1758 adding the default one. Similarly, if it contains a Reply-To: header, the
1759 errors_reply_to option, if set, is not used.
1761 PH/19 When calculating a retry time, Exim used to measure the "time since
1762 failure" by looking at the "first failed" field in the retry record. Now
1763 it does not use this if it is later than than the arrival time of the
1764 message. Instead it uses the arrival time. This makes for better
1765 behaviour in cases where some deliveries succeed, thus re-setting the
1766 "first failed" field. An example is a quota failure for a huge message
1767 when small messages continue to be delivered. Without this change, the
1768 "time since failure" will always be short, possible causing more frequent
1769 delivery attempts for the huge message than are intended.
1770 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1772 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1773 $1, $2, $3) because the numerical variables can be reset during some
1774 expansion items (e.g. "match"), thereby losing the authentication data.
1776 PH/21 Make -bV show the size of off_t variables so that the test suite can
1777 decide whether to run tests for quotas > 2G.
1779 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1780 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1781 filecount value is greater than 2G or if a quota value is greater than 2G
1782 on a system where the size of off_t is not greater than 4, a panic error
1785 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1786 never match. The debug and -bh output now contains an explicit error
1787 message indicating a malformed IPv4 address or mask.
1789 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1790 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1791 PH/23 above applies.
1793 PH/25 Do not write to syslog when running in the test harness. The only
1794 occasion when this arises is a failure to open the main or panic logs
1795 (for which there is an explicit test).
1797 PH/26 Added the /no_tell option to "control=freeze".
1799 PH/27 If a host name lookup failed very early in a connection, for example, if
1800 the IP address matched host_lookup and the reverse lookup yielded a name
1801 that did not have a forward lookup, an error message of the form "no IP
1802 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1803 could be logged. Now it outputs the IP address instead of "NULL".
1805 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1806 allows the sender and the authenticated sender to be set when
1807 submitting a message from within Exim. Since child_open_exim() is
1808 documented for local_scan(), the new function should be too.
1810 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1811 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1812 results in an empty string is now treated as unset.
1814 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1816 PH/31 Added sender_verify_fail logging option.
1818 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1819 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1820 tidied the source and removed it altogether.
1822 PH/33 When a queue run was abandoned because the load average was too high, a
1823 log line was always written; now it is written only if the queue_run log
1824 selector is set. In addition, the log line for abandonment now contains
1825 information about the queue run such as the pid. This is always present
1826 in "start" and "stop" lines but was omitted from the "abandon" line.
1828 PH/34 Omit spaces between a header name and the colon in the error message that
1829 is given when verify = headers_syntax fails (if there are lots of them,
1830 the message gets confusing).
1832 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1833 names, as there are now some PTR records that contain slashes. This check
1834 is only to protect against broken name servers that fall over on strange
1835 characters, so the fact that it applies to all lookups doesn't matter.
1837 PH/36 Now that the new test suite is complete, we can remove some of the
1838 special code in Exim that was needed for the old test suite. For example,
1839 sorting DNS records because real resolvers return them in an arbitrary
1840 order. The new test suite's fake resolver always returns records in the
1843 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1844 bounces) except when queue_only is set, to avoid logging races between
1845 the different processes.
1847 PH/38 Panic-die if .include specifies a non-absolute path.
1849 PH/39 A tweak to the "H" retry rule from its user.
1851 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1852 a label. They prevented compilation on older perls.
1854 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1855 a warning to be raised on newish perls.
1857 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1858 on queue. Changes to match documented behaviour of showing count of
1859 messages matching specified criteria.
1861 PH/40 Changed the default ident timeout from 30s to 5s.
1863 PH/41 Added support for the use of login_cap features, on those BSD systems
1864 that have them, for controlling the resources used by pipe deliveries.
1866 PH/42 The content-scanning code uses fopen() to create files in which to put
1867 message data. Previously it was not paying any attention to the mode of
1868 the files. Exim runs with umask(0) because the rest of the code creates
1869 files with open(), and sets the required mode explicitly. Thus, these
1870 files were ending up world-writeable. This was not a big issue, because,
1871 being within the spool directory, they were not world-accessible. I have
1872 created a function called modefopen, which takes an additional mode
1873 argument. It sets umask(777), creates the file, chmods it to the required
1874 mode, then resets the umask. All the relevant calls to fopen() in the
1875 content scanning code have been changed to use this function.
1877 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1878 to 24 hours. This avoids potential overflow problems when processing G
1879 and H retry rules. I suspect nobody ever tinkers with this value.
1881 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1883 PH/45 When the plaintext authenticator is running as a client, the server's
1884 challenges are checked to ensure they are valid base64 strings. By
1885 default, the authentication attempt is cancelled if an invalid string is
1886 received. Setting client_ignore_invalid_base64 true ignores these errors.
1887 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1888 they are received. Thus, the responses can be made to depend on the
1889 challenges. If an invalid string is ignored, an empty string is placed in
1892 PH/46 Messages that are created by the autoreply transport now contains a
1893 References: header, in accordance with RFCs 2822 and 3834.
1895 PH/47 Added authenticated_sender_force to the smtp transport.
1897 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1899 PH/49 Installed latest patch from the Sieve maintainer.
1901 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1902 was also set, the appendfile transport was unnecessarily scanning a
1903 directory of message files (e.g. for maildir delivery) to find the count
1904 of files (along with the size), even though it did not need this
1905 information. It now does the scan only if it needs to find either the
1906 size of the count of files.
1908 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1910 PH/52 Two bugs concerned with error handling when the smtp transport is
1913 (i) Exim was not creating retry information for temporary errors given
1914 for individual recipients after the DATA command when the smtp transport
1915 was used in LMTP mode. This meant that they could be retried too
1916 frequently, and not timed out correctly.
1918 (ii) Exim was setting the flag that allows error details to be returned
1919 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1920 recipients that were returned after the DATA command.
1922 PH/53 This is related to PH/52, but is more general: for any failing address,
1923 when detailed error information was permitted to be returned to the
1924 sender, but the error was temporary, then after the final timeout, only
1925 "retry timeout exceeded" was returned. Now it returns the full error as
1926 well as "retry timeout exceeded".
1928 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1929 do this, and (what is worse) MTAs that accept it.
1931 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1932 will now be deprecated.
1934 PH/56 New os.c-cygwin from the Cygwin maintainer.
1936 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1937 formats (previously only available in exim formats via -bpr, -bpru,
1938 and -bpra. Now also available in native and exiqgrep formats)
1940 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1941 with very large, slow to parse queues
1943 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1945 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1947 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1948 responses to authentication challenges, though it was showing the
1949 challenges; (ii) I've removed the CR characters from the debug output for
1952 PH/58 Allow for the insertion of a newline as well as a space when a string
1953 is turned into more than one encoded-word during RFC 2047 encoding. The
1954 Sieve code now uses this.
1956 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1957 data_4xx, lost_connection, tls_required.
1959 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1960 message was being sent as an SMTP response.
1962 PH/61 Add -l and -k options to exicyclog.
1964 PH/62 When verifying, if an address was redirected to one new address, so that
1965 verification continued, and the new address failed or deferred after
1966 having set something in $address_data, the value of $address_data was not
1967 passed back to the ACL. This was different to the case when no
1968 redirection occurred. The value is now passed back in both cases.
1970 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1971 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1972 use login_cap.h, so on its own it isn't the distinguishing feature. The
1973 new name refers directly to the setclassresources() function.
1975 PH/65 Added configuration files for NetBSD3.
1977 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1979 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1980 is preferred over IPv4.
1982 PH/68 The bounce_return_message and bounce_return_body options were not being
1983 honoured for bounces generated during the reception of non-SMTP messages.
1984 In particular, this applied to messages rejected by the ACL. This bug has
1985 been fixed. However, if bounce_return_message is true and bounce_return_
1986 body is false, the headers that are returned for a non-SMTP message
1987 include only those that have been read before the error was detected.
1988 (In the case of an ACL rejection, they have all been read.)
1990 PH/69 The HTML version of the specification is now built in a directory called
1991 spec_html instead of spec.html, because the latter looks like a path with
1992 a MIME-type, and this confuses some software.
1994 PH/70 Catch two compiler warnings in sieve.c.
1996 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1997 function verify_get_ident() calls ip_connect() to connect a socket, but
1998 if the "connect()" function timed out, ip_connect() used to close the
1999 socket. However, verify_get_ident() also closes the socket later, and in
2000 between Exim writes to the log, which may get opened at this point. When
2001 the socket was closed in ip_connect(), the log could get the same file
2002 descriptor number as the socket. This naturally causes chaos. The fix is
2003 not to close the socket in ip_connect(); the socket should be closed by
2004 the function that creates it. There was only one place in the code where
2005 this was missing, in the iplookup router, which I don't think anybody now
2006 uses, but I've fixed it anyway.
2008 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2009 well as to direct DNS lookups. Otherwise the handling of names in host
2010 lists is inconsistent and therefore confusing.
2016 PH/01 Two changes to the default runtime configuration:
2018 (1) Move the checks for relay_from_hosts and authenticated clients from
2019 after to before the (commented out) DNS black list checks.
2021 (2) Add control=submission to the relay_from_hosts and authenticated
2022 clients checks, on the grounds that messages accepted by these
2023 statements are most likely to be submissions.
2025 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2027 (1) Generate an error if the third argument for the ${prvs expansion is
2030 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2033 (3) Reset the variables that are obtained from the first argument of
2034 ${prvscheck and used in the second argument before leaving the code,
2035 because their memory is reclaimed, so using them afterwards may do
2038 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2039 one (it's much easier than Tom thought :-).
2041 (5) Because of (4), we can now allow for the use of $prvscheck_result
2042 inside the third argument.
2044 PH/03 For some reason, the default setting of PATH when running a command from
2045 a pipe transport was just "/usr/bin". I have changed it to
2048 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2049 anything to be listed in the output from -bV.
2051 PH/05 When a filter generated an autoreply, the entire To: header line was
2052 quoted in the delivery log line, like this:
2054 => >A.N.Other <ano@some.domain> <original@ddress> ...
2056 This has been changed so that it extracts the operative address. There
2057 may be more than one such address. If so, they are comma-separated, like
2060 => >ano@some.domain,ona@other.domain <original@ddress> ...
2062 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2063 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2064 not being looked up in the rDNS to get a host name, Exim was showing the
2065 IP address twice in Received: lines, even though the IP addresses were
2066 identical. For example:
2068 Received: from [1.2.3.4] (helo=[1.2.3.4])
2070 However, if the real host name was known, it was omitting the HELO data
2071 if it matched the actual IP address. This has been tidied up so that it
2072 doesn't show the same IP address twice.
2074 PH/07 When both +timestamp and +memory debugging was on, the value given by
2075 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2076 called by the debug printing, thereby overwriting the timestamp buffer.
2077 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2079 PH/08 When the original message was included in an autoreply transport, it
2080 always said "this is a copy of the message, including all the headers",
2081 even if body_only or headers_only was set. It now gives an appropriate
2084 PH/09 Applied a patch from the Sieve maintainer which:
2086 o fixes some comments
2087 o adds the (disabled) notify extension core
2088 o adds some debug output for the result of if/elsif tests
2089 o points to the current vacation draft in the documentation
2090 and documents the missing references header update
2094 o fixes a bug in processing the envelope test (when testing
2095 multiple envelope elements, the last element determinted the
2098 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2099 Electronic Mail") by including:
2101 Auto-submitted: auto-generated
2103 in the messages that it generates (bounce messages and others, such as
2104 warnings). In the case of bounce messages for non-SMTP mesages, there was
2105 also a typo: it was using "Auto_submitted" (underscore instead of
2106 hyphen). Since every message generated by Exim is necessarily in response
2107 to another message, thes have all been changed to:
2109 Auto-Submitted: auto-replied
2111 in accordance with these statements in the RFC:
2113 The auto-replied keyword:
2115 - SHOULD be used on messages sent in direct response to another
2116 message by an automatic process,
2118 - MUST NOT be used on manually-generated messages,
2120 - MAY be used on Delivery Status Notifications (DSNs) and Message
2121 Disposition Notifications (MDNs),
2123 - MUST NOT be used on messages generated by automatic or periodic
2124 processes, except for messages which are automatic responses to
2127 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2128 to the default Received: header definition.
2130 PH/12 Added log selector acl_warn_skipped (default on).
2132 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2133 variables because (a) they are in the wrong storage pool and (b) even if
2134 they were copied, it wouldn't work properly because of the caching.
2136 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2137 checking when decoding. Apparently there are clients that generate
2138 overlong encoded strings. Why am I not surprised?
2140 PH/15 If the first argument of "${if match_address" was not empty, but did not
2141 contain an "@" character, Exim crashed. Now it writes a panic log message
2142 and treats the condition as false.
2144 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2146 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2147 extension "envelope-auth". The code is finished and in agreement with
2148 other implementations, but there is no documentation so far and in fact,
2149 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2150 not changing the active code.
2152 Print executed "if" and "elsif" statements when debugging is used. This
2153 helps a great deal to understand what a filter does.
2155 Document more things not specified clearly in RFC3028. I had all this
2156 sorted out, when out of a sudden new issues came to my mind. Oops."
2158 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2161 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2162 canonical form (as documented). However, after a host name lookup from
2163 the IP address, check_host() was doing a simple string comparison with
2164 addresses acquired from the DNS when checking that the found name did
2165 have the original IP as one of its addresses. Since any found IPv6
2166 addresses are likely to be in abbreviated form, the comparison could
2167 fail. Luckily, there already exists a function for doing the comparison
2168 by converting both addresses to binary, so now that is used instead of
2169 the text comparison.
2171 PH/20 There was another similar case to PH/19, when a complete host name was
2172 given in a host list; looking up its IP address could give an abbreviated
2173 form, whereas the current host's name might or might not be abbreviated.
2174 The same fix has been applied.
2180 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2181 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2184 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2185 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2187 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2189 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2190 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2191 Also, for "Auto-Submitted", treat anything other than "no" as
2192 non-personal, in accordance with RFC 3834. (Previously it treated
2193 anything starting "auto-" as non-personal.)
2195 TF/01 The control=submission/name=... option had a problem with syntax
2196 errors if the name included a slash character. The /name= option
2197 now slurps the rest of the string, so it can include any characters
2198 but it must come last in the list of options (after /sender_retain
2201 PH/05 Some modifications to the interface to the fake nameserver for the new
2209 TK/01 Added the "success_on_redirect" address verification option. See
2210 NewStuff for rationale and an example.
2212 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2214 PH/02 Patch to exigrep to allow it to work on syslog lines.
2216 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2217 fread() to skip over the body file's header line, because in Cygwin the
2218 header line is locked and is inaccessible.
2220 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2221 co-exist for some time) to make it clear that it is the Exim ID that is
2222 referenced, not the Message-ID: header line.
2224 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2225 string_format() function, because snprintf() does not exist on all
2228 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2229 ${stat: expansion item.
2231 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2232 protocol synchronization error", to keep the pedants happy.
2234 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2235 well as for IRIX systems, when gcc is being used. See the host.c source
2238 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2240 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2243 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2244 transport and to the smtp transport in LMTP mode.
2246 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2248 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2249 run a filter in a subprocess. This could lead to confusion in subsequent
2250 lookups in the parent process. There should also be a search_tidyup() at
2251 the end of the subprocess.
2253 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2254 only if the host matched helo_try_verify_hosts, which caused the
2255 verification to occur when the EHLO/HELO command was issued. The ACL just
2256 tested the remembered result. Now, if a previous verification attempt has
2257 not happened, "verify = helo" does it there and then.
2259 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2261 TK/03 Fix log output including CR from clamd.
2263 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2264 exist provoked a memory error which could cause a segfault.
2266 PH/15 Installed PCRE 6.2
2268 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2270 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2271 of the problem. Specifically, suggested +O2 rather than +O1 for the
2274 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2276 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2277 record, the port was not correctly logged when the outgoing_port log
2278 selector was set (it logged the transort's default port).
2280 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2281 fallback_hosts, and "hosts" in the smtp transport.
2283 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2284 host errors such as "Connection refused".
2286 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2287 authentication with radiusclient 0.4.9:
2289 - Error returned from rc_read_config was caught wrongly
2290 - Username/password not passed on to radius server due to wrong length.
2292 The presumption is that some radiusclient API changes for 4.51/PH/17
2293 were not taken care of correctly. The code is still untested by me (my
2294 Linux distribution still has 0.3.2 of radiusclient), but it was
2295 contributed by a Radius user.
2297 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2298 expanding the "port" option of the smtp transport.
2300 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2301 while reading a MIME header. Thanks to Tom Hughes for a patch.
2303 PH/24 Include config.h inside local_scan.h so that configuration settings are
2306 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2307 that in an ACL for RCPT (for example), you can examine exactly what was
2310 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2311 commands, but it was not correctly comparing the address with the actual
2312 client host address. Thus, it would show the EHLO address in Received:
2313 header lines when this was not necessary.
2315 PH/27 Added the % operator to ${eval:}.
2317 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2318 it should be ignoring failures (because with -C, for example, it has lost
2319 privilege). It wasn't ignoring creation failures other than "already
2322 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2325 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2326 because an input line was too long, either on its own, or by virtue of
2327 too many continuations, the temporary file was not being removed, and the
2328 return code was incorrect.
2330 PH/31 Missing "BOOL" in function definition in filtertest.c.
2332 PH/32 Applied Sieve patches from the maintainer.
2334 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2336 PH/33 Added "verify = not_blind".
2338 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2339 Local/Makefile (with some defaults set). These are used in built scripts
2340 such as exicyclog, but they have never been used in the exim_install
2341 script (though there are many overriding facilities there). I have
2342 arranged that the exim_install script now takes note of these two
2345 PH/35 Installed configuration files for Dragonfly.
2347 PH/36 When a locally submitted message by a trusted user did not contain a
2348 From: header, and the sender address was obtained from -f or from an SMTP
2349 MAIL command, and the trusted user did not use -F to supply a sender
2350 name, $originator_name was incorrectly used when constructing a From:
2351 header. Furthermore, $originator_name was used for submission mode
2352 messages from external hosts without From: headers in a similar way,
2353 which is clearly wrong.
2355 PH/37 Added control=suppress_local_fixups.
2357 PH/38 When log_selector = +received_sender was set, and the addition of the
2358 sender made the log line's construction buffer exactly full, or one byte
2359 less than full, an overflow happened when the terminating "\n" was
2362 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2363 when the result of a list match is failure because a DNS lookup failed.
2365 PH/40 RM_COMMAND is now used in the building process.
2367 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2368 the "build-* directories that it finds.
2370 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2371 address in a domain literal was a prefix of an interface address.
2373 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2374 when verifying a sender address, unless rewrite_headers is false.
2376 PH/44 Wrote a long comment about why errors_to addresses are verified as
2377 recipients, not senders.
2379 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2380 the ratelimit ACL was added.
2382 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2384 PH/46 Added extra information about PostgreSQL errors to the error string.
2386 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2387 avoiding the need to install special zones in a real server. This is
2388 backwards compatible; if it can't find the fake resolver, it drops back.
2389 Thus, both old and new test suites can be run.
2391 TF/02 Added util/ratelimit.pl
2393 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2394 clock is set back in time.
2396 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2397 Candler <B.Candler@pobox.com>.
2399 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2400 OK for addresses that are the result of redirections.
2402 PH/48 A number of further additions for the benefit of the new test suite,
2403 including a fake gethostbyname() that interfaces to the fake DNS resolver
2406 TF/06 The fix for widen_domains has also been applied to qualify_single and
2407 search_parents which are the other dnslookup options that can cause
2410 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2413 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2415 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2416 service" if it encountered an MX record with an empty target hostname.
2417 The message is now "an MX or SRV record indicated no SMTP service".
2419 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2420 if the DNS of the sending site is misconfigured. This is quite a
2421 common situation. This change restores the behaviour of treating a
2422 helo verification defer as a failure.
2424 PH/51 If self=fail was set on a router, the bounce message did not include the
2425 actual error message.
2431 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2433 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2434 command itself ended in error, the underlying message about the transport
2435 filter timeout was being overwritten with the pipe command error. Now the
2436 underlying error message should be appended to the second error message.
2438 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2440 PH/02 Reduce the amount of output that "make" produces by default. Full output
2441 can still be requested.
2443 PH/03 The warning log line about a condition test deferring for a "warn" verb
2444 was being output only once per connection, rather than after each
2445 occurrence (because it was using the same function as for successful
2446 "warn" verbs). This seems wrong, so I have changed it.
2448 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2449 it should not have, which might have caused a crash in the right
2450 circumstances, but probably never did.
2452 PH/04 Installed a modified version of Tony Finch's patch to make submission
2453 mode fix the return path as well as the Sender: header line, and to
2454 add a /name= option so that you can make the user's friendly name appear
2457 TF/03 Added the control = fakedefer ACL modifier.
2459 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2460 Mark Lowes for thorough testing.
2462 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2464 TK/03 Merged latest SRS patch from Miles Wilton.
2466 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2467 with the definition in sysexits.h (which is #included earlier).
2468 Fortunately, Exim does not actually use EX_OK. The code used to try to
2469 preserve the sysexits.h value, by assumimg that macro definitions were
2470 scanned for macro replacements. I have been disabused of this notion,
2471 so now the code just undefines EX_OK before #including unistd.h.
2473 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2474 in the smtp transport. When a block could not be written in a single
2475 write() function, the timeout was being re-applied to each part-write.
2476 This seems wrong - if the receiver was accepting one byte at a time it
2477 would take for ever. The timeout is now adjusted when this happens. It
2478 doesn't have to be particularly precise.
2480 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2481 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2483 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2484 without a domain if the check to <postmaster@domain> fails.
2486 SC/01 Eximstats: added -xls and the ability to specify output files
2487 (patch written by Frank Heydlauf).
2489 SC/02 Eximstats: use FileHandles for outputing results.
2491 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2493 SC/04 Eximstats: fixed display of large numbers with -nvr option
2495 SC/05 Eximstats: fixed merging of reports with empty tables.
2497 SC/06 Eximstats: added the -include_original_destination flag
2499 SC/07 Eximstats: removed tabs and trailing whitespace.
2501 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2503 TK/06 MBOX spool code: Add real "From " MBOX separator line
2504 so the .eml file is really in mbox format (even though
2505 most programs do not really care). Patch from Alex Miller.
2507 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2508 The latter is generated from $received_to and is only set if the
2509 message has one envelope recipient. SA can use these headers,
2510 obviously out-of-the-box. Patch from Alex Miller.
2512 PH/08 The ${def test on a variable was returning false if the variable's
2513 value was "0", contrary to what the specification has always said!
2514 The result should be true unless the variable is empty.
2516 PH/09 The syntax error of a character other than { following "${if
2517 def:variable_name" (after optional whitespace) was not being diagnosed.
2518 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2519 accidental colon was present, for example, could give incorrect results.
2521 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2522 result is used (not including appendfile, where other changes are about
2525 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2526 This involved changing a lot of size variables from int to off_t. It
2527 should work with maildirs and everything.
2529 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2530 spamd dying while we are connected to it.
2532 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2535 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2536 function for BDB 4.3.
2538 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2540 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2541 the sources for PCRE have been re-arranged and more files are now
2544 PH/15 The code I had for printing potentially long long variables in PH/11
2545 above was not the best (it lost precision). The length of off_t variables
2546 is now inspected at build time, and an appropriate printing format (%ld
2547 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2548 to be "long long int" or "long int". This is needed for the internal
2549 formatting function string_vformat().
2551 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2552 the configuration file to be ":syslog", then the script "guesses" where
2553 the logs files are, rather than using the compiled in default. In our
2554 case the guess is not the same as the compiled default, so the script
2555 suddenly stopped working when I started to use syslog. The patch checks
2556 to see if log_file_path is "". If so, it attempts to read it from exim
2557 with no configuration file to get the compiled in version, before it
2558 falls back to the previous guessing code."
2560 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2561 implementing BATV in an Exim configuration. See NewStuff for the gory
2564 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2565 Makefile that are specific to HP-UX.
2567 PH/18 If the "use_postmaster" option was set for a recipient callout together
2568 with the "random" option, the postmaster address was used as the MAIL
2569 FROM address for the random test, but not for the subsequent recipient
2570 test. It is now used for both.
2572 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2573 patch removes a few documentation additions to RFC 3028, because the
2574 latest draft now contains them. It adds the new en;ascii-case comparator
2575 and a new error check for 8bit text in MIME parts. Comparator and
2576 require names are now matched exactly. I enabled the subaddress
2577 extension, but it is not well tested yet (read: it works for me)."
2579 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2580 rework some of the code of TK/09 above to avoid the hardwired use of
2581 "%lld" and "long long". Replaced the call to snprintf() with a call to
2584 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2585 records point to non-existent hosts", "retry timeout exceeded", and
2586 "retry time not reached for any host after a long failure period".
2588 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2589 experimental DomainKeys support:
2591 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2592 (2) On an error such as an illegally used "control", the wrong name for
2593 the control was given.
2595 These problems did NOT occur unless DomainKeys support was compiled.
2597 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2599 PH/24 Added ${if match_ip condition.
2601 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2602 databases so that it will be absolutely obvious if a crash occurs in the
2603 DB library. This is a regular occurrence (often caused by mis-matched
2606 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2607 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2608 picked up on a user's system that detects such things. There doesn't seem
2609 to be a gcc warning option for this - only an attribute that has to be
2610 put on the function's prototype. It seems that in Fedora Core 4 they have
2611 set this on a number of new functions. No doubt there will be more in due
2614 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2615 specify a transport. However, if an address that was verified by such a
2616 router was the subject of a callout, Exim crashed because it tried to
2617 read the rcpt_include_affixes from the non-existent transport. Now it
2618 just assumes that the setting of that option is false. This bug was
2619 introduced by 4.51/PH/31.
2621 PH/28 Changed -d+all to exclude +memory, because that information is very
2622 rarely of interest, but it makes the output a lot bigger. People tend to
2623 do -d+all out of habit.
2625 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2626 code in os-type was giving problems when libc.so lives in lib64, like on
2629 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2630 aren't the modern standard, and it seems that some systems' include files
2631 don't always have them. Exim was already checking for some of the newer
2632 ones like T_AAAA, and defining it itself. I've added checks for all the
2633 record types that Exim uses.
2635 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2636 not automatically generating a new one, as it is supposed to. This
2637 prevented TLS from working. If the file did exist, but contained invalid
2638 data, a new version was generated, as expected. It was only the case of a
2639 non-existent file that was broken.
2641 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2642 with a change in libdomainkeys > 0.64.
2644 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2645 from DNS. If the selector record carries the flag, it now has
2646 precedence over the domain-wide flag.
2648 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2650 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2651 the use of a port name that isn't defined in /etc/services) occurred, the
2652 message was deferred as in a normal delivery, and thus remained on the
2653 spool, instead of being failed because of the mua_wrapper setting. This
2654 is now fixed, and I tidied up some of the mua_wrapper messages at the
2657 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2658 the messages in a hash of arrays rather than using individual hashes.
2659 This is a bit cleaner and results in dramatic memory savings, albeit
2660 at a slight CPU cost.
2662 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2663 as requested by Marc Sherman.
2665 SC/10 Eximstats: added histograms for user specified patterns as requested
2668 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2670 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2671 fopen() in the content-scanning modules that did not already have it.
2677 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2678 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2680 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2682 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2684 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2685 to test Sieve filters that use "vacation".
2687 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2688 that changes the way the GnuTLS parameters are stored in the cache file.
2689 The new format can be generated externally. For backward compatibility,
2690 if the data in the cache doesn't make sense, Exim assumes it has read an
2691 old-format file, and it generates new data and writes a new file. This
2692 means that you can't go back to an older release without removing the
2695 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2696 work if there are any delivery delays because "one_time" forces the
2697 parent to be marked "delivered", so its unseen clone is never tried
2698 again. For this reason, Exim now forbids the simultaneous setting of
2701 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2702 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2703 entry if you want to know the details. The fix, however, neglected to
2704 consider the case when local delivery batching is involved. The test for
2705 "previously delivered" was not happening when checking to see if an
2706 address could be batched with a previous (undelivered) one; under
2707 certain circumstances this could lead to multiple deliveries to the same
2710 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2711 in its include files, and this causes problems building Exim.
2713 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2714 header_syntax) but Exim was just ignoring anything given after a slash.
2715 In particular, this caused confusion with an attempt to use "verify =
2716 reverse_host_lookup/defer_ok". An error is now given when options are
2717 supplied for verify items that do not have them. (Maybe reverse_host_
2718 lookup should have a defer_ok option, but that's a different point.)
2720 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2721 defined by RFC 821) to 2048, because there were problems with some AUTH
2722 commands, and RFC 1869 says the size should be increased for extended
2723 SMTP commands that take arguments.
2725 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2728 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2729 "unknown" error; now it says that the functionality isn't in the binary.
2731 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2732 an address' error message when a string expansion fails (syntax or
2733 whatever). Otherwise the password may appear in the log. Following change
2734 PH/42 below, there is no longer a chance of it appearing in a bounce
2737 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2739 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2740 of its addresses was ever tried. (Bugzilla bug #2.)
2742 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2743 the result incorrectly in the debug output. (It correctly added a newline
2744 to what was transported.)
2746 TF/01 Added $received_time.
2748 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2749 commented out examples of how to interface to a virus scanner and to
2750 SpamAssassin. Also added commented examples of av_scanner and
2751 spamd_address settings.
2753 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2754 and controls are allowed in which ACLs. There were a couple of minor
2755 errors. Some of the entries in the conditions table (which is a table of
2756 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2757 negation of where the condition IS allowed.
2759 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2761 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2762 header file does not have a version number, so I've had to invent a new
2763 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2764 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2765 radiusclient), but it was contributed by a Radius user.
2767 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2768 files or directories, for OpenSSL.
2770 PH/19 When an Exim process that is running as root has to create an Exim log
2771 file, it does so in a subprocess that runs as exim:exim so as to get the
2772 ownership right at creation (otherwise, other Exim processes might see
2773 the file with the wrong ownership). There was no test for failure of this
2774 fork() call, which would lead to the process getting stuck as it waited
2775 for a non-existent subprocess. Forks do occasionally fail when resources
2776 run out. I reviewed all the other calls to fork(); they all seem to check
2779 PH/20 When checking for unexpected SMTP input at connect time (before writing
2780 the banner), Exim was not dealing correctly with a non-positive return
2781 from the read() function. If the client had disconnected by this time,
2782 the result was a log entry for a synchronization error with an empty
2783 string after "input=" when read() returned zero. If read() returned -1
2784 (an event I could not check), uninitialized data bytes were printed.
2785 There were reports of junk text (parts of files, etc) appearing after
2788 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2790 PH/22 Added support for macro redefinition, and (re)definition in between
2791 driver and ACL definitions.
2793 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2794 forgetting to use the resulting value; it was using the unexpanded value.
2796 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2797 hadn't been configured. The fix is from Juergen Kreileder, who
2798 understands it better than I do:
2800 "Here's what I see happening with three configured cyrus_sasl
2801 authenticators configured (plain, login, cram-md5):
2803 On startup auth_cyrus_sasl_init() gets called for each of these.
2804 This means three calls to sasl_listmech() without a specified mech_list.
2805 => SASL tests which mechs of all available mechs actually work
2806 => three warnings about OTP not working
2807 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2809 With the patch, sasl_listmech() also gets called three times. But now
2810 SASL's mech_list option is set to the server_mech specified in the the
2811 authenticator. Or in other words, the answer from sasl_listmech()
2812 gets limited to just the mech you're testing for (which is different
2814 => the return list contains just 'plain' or 'login', 'cram-md5' or
2815 nothing depending on the value of ob->server_mech.
2817 I've just tested the patch: Authentication still works fine,
2818 unavailable mechs specified in the exim configuration are still
2819 caught, and the auth.log warnings about OTP are gone."
2821 PH/25 When debugging is enabled, the contents of the command line are added
2822 to the debugging output, even when log_selector=+arguments is not
2825 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2826 answer is "GNU", and only if the return is "GNU/something" is the answer
2829 PH/27 $acl_verify_message is now set immediately after the failure of a
2830 verification in an ACL, and so is available in subsequent modifiers. In
2831 particular, the message can be preserved by coding like this:
2833 warn !verify = sender
2834 set acl_m0 = $acl_verify_message
2836 Previously, $acl_verify_message was set only while expanding "message"
2837 and "log_message" when a very denied access.
2839 PH/28 Modified OS/os.c-Linux with
2841 -#ifndef OS_LOAD_AVERAGE
2842 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2844 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2845 nomenclature these days.)
2847 PH/29 Installed patch from the Sieve maintainer that adds the options
2848 sieve_useraddress and sieve_subaddress to the redirect router.
2850 PH/30 In these circumstances:
2851 . Two addresses routed to the same list of hosts;
2852 . First host does not offer TLS;
2853 . First host accepts first address;
2854 . First host gives temporary error to second address;
2855 . Second host offers TLS and a TLS session is established;
2856 . Second host accepts second address.
2857 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2858 and peerdn, if requested) that were in fact used only for the second
2861 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2862 attention to any local part prefix or suffix that was matched by the
2863 router that accepted the address. It now behaves in the same way as it
2864 does for delivery: the affixes are removed from the local part unless
2865 rcpt_include_affixes is set on the transport.
2867 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2868 timeout during the DATA phase of an incoming message.
2870 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2871 applied a patch sanctioned by the Sieve maintainer.
2873 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2874 the uid or gid is negative. A case of a negative gid caused this to be
2875 noticed. The fix allows for either to be negative.
2877 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2878 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2881 PH/36 The change PH/12 above was broken. Fixed it.
2883 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2884 the grounds that routing the same address twice would always produce the
2885 same answer. This might have been true once, but it is certainly no
2886 longer true now. Routing a child address may depend on the previous
2887 routing that produced that child. Some complicated redirection strategies
2888 went wrong when messages had multiple recipients, and made Exim's
2889 behaviour dependent on the order in which the addresses were given.
2891 I have moved the duplicate checking until after the routing is complete.
2892 Exim scans the addresses that are assigned to local and remote
2893 transports, and removes any duplicates. This means that more work will be
2894 done, as duplicates will always all be routed, but duplicates are
2895 presumably rare, so I don't expect this is of any significance.
2897 For deliveries to pipes, files, and autoreplies, the duplicate checking
2898 still happens during the routing process, since they are not going to be
2901 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2902 It corrects a timeout issue with spamd. This is Ian's comment: "The
2903 background is that sometimes spamd either never reads data from a
2904 connection it has accepted, or it never writes response data. The exiscan
2905 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2906 blindly assumes that writes won't block so it may never time out."
2908 PH/39 Allow G after quota size as well as K and M.
2910 PH/40 The value set for $authenticated_id in an authenticator may not contain
2911 binary zeroes or newlines because the value is written to log lines and
2912 to spool files. There was no check on this. Now the value is run through
2913 the string_printing() function so that such characters are converted to
2914 printable escape sequences.
2916 PH/41 $message_linecount is a new variable that contains the total number of
2917 lines in the message. Compare $body_linecount, which is the count for the
2920 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2921 bounce and delay warning messages, except in certain special cases, which
2924 (a) An SMTP error message from a remote host;
2925 (b) A message specified in a :fail: redirection;
2926 (c) A message specified in a "fail" command in a system filter;
2927 (d) A message specified in a FAIL return from the queryprogram router;
2928 (e) A message specified by the cannot_route_message router option.
2930 In these cases only, Exim does include the error details in bounce and
2931 warning messages. There are also a few cases where bland messages such
2932 as "unrouteable address" or "local delivery error" are given.
2934 PH/43 $value is now also set for the "else" part of a ${run expansion.
2936 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2937 being worked on, but at least Exim now implements the latest version to
2940 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2941 process to complete was treated as a delivery failure, a timeout while
2942 writing the message to the pipe was logged, but erroneously treated as a
2943 successful delivery. Such timeouts include transport filter timeouts. For
2944 consistency with the overall process timeout, these timeouts are now
2945 treated as errors, giving rise to delivery failures by default. However,
2946 there is now a new Boolean option for the pipe transport called
2947 timeout_defer, which, if set TRUE, converts the failures into defers for
2948 both kinds of timeout. A transport filter timeout is now identified in
2951 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2952 systems where "make" and "gmake" are different, calling "gmake" at top
2953 level broke things. I've arranged for the value of $(MAKE) to be passed
2954 from the Makefile to this script so that it can call the same version of
2958 A note about Exim versions 4.44 and 4.50
2959 ----------------------------------------
2961 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2962 changes of various kinds. As a consequence, a big documentation update was
2963 needed. This delayed the release for rather longer than seemed good, especially
2964 in the light of a couple of (minor) security issues. Therefore, the changes
2965 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2966 release. So 4.44 and 4.50 are in effect two different branches that both start
2969 I have left the 4.50 change log unchanged; it contains all the changes since
2970 4.43. The change log for 4.44 is below; many of its items are identical to
2971 those for 4.50. This seems to be the most sensible way to preserve the
2972 historical information.
2978 1. Minor wording change to the doc/README.SIEVE file.
2980 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2981 computation of the current number of files was incorrect.
2983 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2984 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2987 4. Give more explanation in the error message when the command for a transport
2988 filter fails to execute.
2990 5. There are several places where Exim runs a non-Exim command in a
2991 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2992 was being done only for the command run by the queryprogram router. It is
2993 now done for all such subprocesses. The other cases are: ${run, transport
2994 filters, and the commands run by the lmtp and pipe transports.
2996 6. Added CONFIGURE_GROUP build-time option.
2998 7. Some older OS have a limit of 256 on the maximum number of file
2999 descriptors. Exim was using setrlimit() to set 1000 as a large value
3000 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3001 systems. I've change it so that if it can't get 1000, it tries for 256.
3003 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3004 was an oversight, and furthermore, ever since the addition of extra
3005 controls (e.g. 4.43/32), the checks on when to allow different forms of
3006 "control" were broken. There should now be diagnostics for all cases when a
3007 control that does not make sense is encountered.
3009 9. Added the /retain_sender option to "control=submission".
3011 10. $recipients is now available in the predata ACL (oversight).
3013 11. Tidy the search cache before the fork to do a delivery from a message
3014 received from the command line. Otherwise the child will trigger a lookup
3015 failure and thereby defer the delivery if it tries to use (for example) a
3016 cached ldap connection that the parent has called unbind on.
3018 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3019 of $address_data from the recipient verification was clobbered by the
3020 sender verification.
3022 13. The value of address_data from a sender verification is now available in
3023 $sender_address_data in subsequent conditions in the ACL statement.
3025 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3027 15. Added a new option "connect=<time>" to callout options, to set a different
3030 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3031 was its contents. (It was OK if the option was not defined at all.)
3033 17. A "Completed" log line is now written for messages that are removed from
3034 the spool by the -Mrm option.
3036 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3037 information about exactly what failed.
3039 19. Added -dd to debug only the daemon process.
3041 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3042 handles timeouts, both on the server side and network timeouts. Renamed the
3043 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3045 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3046 It is now set to "smtps".
3048 22. $host_address is now set to the target address during the checking of
3049 ignore_target_hosts.
3051 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3052 being passed; this would have caused $sender_host_name to have been used if
3053 matching the list had actually called for a host name (not very likely,
3054 since this list is usually IP addresses). A host name is now passed as
3057 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3058 code that specifies a non-restarting handler (typically sigaction() in
3059 modern systems) in an attempt to fix a rare and obscure crash bug.
3061 25. Narrowed the window for a race in the daemon that could cause it to ignore
3062 SIGCHLD signals. This is not a major problem, because they are used only to
3063 wake it up if nothing else does.
3065 26. A malformed maildirsize file could cause Exim to calculate negative values
3066 for the mailbox size or file count. Odd effects could occur as a result.
3067 The maildirsize information is now recalculated if the size or filecount
3070 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3071 support for a long time. Removed HAVE_SYS_VFS_H.
3073 28. Installed the latest version of exipick from John Jetmore.
3075 29. In an address list, if the pattern was not a regular expression, an empty
3076 subject address (from a bounce message) matched only if the pattern was an
3077 empty string. Non-empty patterns were not even tested. This was the wrong
3078 because it is perfectly reasonable to use an empty address as part of a
3079 database query. An empty address is now tested by patterns that are
3080 lookups. However, all the other forms of pattern expect the subject to
3081 contain a local part and a domain, and therefore, for them, an empty
3082 address still always fails if the pattern is not itself empty.
3084 30. Exim went into a mad DNS loop when attempting to do a callout where the
3085 host was specified on an smtp transport, and looking it up yielded more
3086 than one IP address.
3088 31. Re-factored the code for checking spool and log partition space into a
3089 function that finds that data and another that does the check. The former
3090 is then used to implement four new variables: $spool_space, $log_space,
3091 $spool_inodes, and $log_inodes.
3093 32. The RFC2047 encoding function was originally intended for short strings
3094 such as real names; it was not keeping to the 75-character limit for
3095 encoded words that the RFC imposes. It now respects the limit, and
3096 generates multiple encoded words if necessary. To be on the safe side, I
3097 have increased the buffer size for the ${rfc2047: expansion operator from
3100 33. It is now permitted to omit both strings after an "if" condition; if the
3101 condition is true, the result is "true". As before, when the second string
3102 is omitted, a false condition yields an empty string. This makes it less
3103 cumbersome to write custom ACL and router conditions.
3105 34. Failure to deliver a bounce message always caused it to be frozen, even if
3106 there was an errors_to setting on the router. The errors_to setting is now
3109 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3110 canonical form (fully expanded) before being placed in
3111 $sender_host_address.
3113 36. The table in the code that translates DNS record types into text (T_A to
3114 "A" for instance) was missing entries for NS and CNAME. It is just possible
3115 that this could have caused confusion if both these types were looked up
3116 for the same domain, because the text type is used as part of Exim's
3117 per-process caching. But the chance of anyone hitting this buglet seems
3120 37. The dnsdb lookup has been extended in a number of ways.
3122 (1) There is a new type, "zns", which walks up the domain tree until it
3123 finds some nameserver records. It should be used with care.
3125 (2) There is a new type, "mxh", which is like "mx" except that it returns
3126 just the host names, not the priorities.
3128 (3) It is now possible to give a list of domains (or IP addresses) to be
3129 looked up. The behaviour when one of the lookups defers can be
3130 controlled by a keyword.
3132 (4) It is now possible to specify the separator character for use when
3133 multiple records are returned.
3135 38. The dnslists ACL condition has been extended: it is now possible to supply
3136 a list of IP addresses and/or domains to be looked up in a particular DNS
3139 39. Added log_selector=+queue_time_overall.
3141 40. When running the queue in the test harness, wait just a tad after forking a
3142 delivery process, to get repeatability of debugging output.
3144 41. Include certificate and key file names in error message when GnuTLS fails
3145 to set them up, because the GnuTLS error message doesn't include the name
3146 of the failing file when there is a problem reading it.
3148 42. Allow both -bf and -bF in the same test run.
3150 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3152 44. The "Exiscan patch" is now merged into the mainline Exim source.
3154 45. Sometimes the final signoff response after QUIT could fail to get
3155 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3156 < 0 before doing a fflush(). This bug looks as though it goes back to the
3157 introduction of TLS in release 3.20, but "sometimes" must have been rare
3158 because the tests only now provoked it.
3160 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3161 (this can affect the format of dates).
3163 47. exim_tidydb, when checking for the continued existence of a message for
3164 which it has found a message-specific retry record, was not finding
3165 messages that were in split spool directories. Consequently, it was
3166 deleting retry records that should have stayed in existence.
3168 48. Steve fixed some bugs in eximstats.
3170 49. The SPA authentication driver was not abandoning authentication and moving
3171 on to the next authenticator when an expansion was forced to fail,
3172 contradicting the general specification for all authenticators. Instead it
3173 was generating a temporary error. It now behaves as specified.
3175 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3176 (the order specifies the preference for clients). The order is now AES256,
3177 AES128, 3DES, ARCFOUR128.
3179 51. Small patch to Sieve code - explicitly set From: when generating an
3182 52. Exim crashed if a remote delivery caused a very long error message to be
3183 recorded - for instance if somebody sent an entire SpamAssassin report back
3184 as a large number of 550 error lines. This bug was coincidentally fixed by
3185 increasing the size of one of Exim's internal buffers (big_buffer) that
3186 happened as part of the Exiscan merge. However, to be on the safe side, I
3187 have made the code more robust (and fixed the comments that describe what
3190 53. Now that there can be additional text after "Completed" in log lines (if
3191 the queue_time_overall log selector is set), a one-byte patch to exigrep
3192 was needed to allow it to recognize "Completed" as not the last thing in
3195 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3196 patch that reportedly fixes this has been added. I am not expert enough to
3197 create a test for it. This is what the patch creator wrote:
3199 "I found a little strange behaviour of ldap code when working with
3200 Windows 2003 AD Domain, where users was placed in more than one
3201 Organization Units. When I tried to give exim partial DN, the exit code
3202 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3203 But simultaneously result of request was absolutely normal ldap result,
3204 so I produce this patch..."
3206 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3207 REFERENCE, so I have modified the code to exclude the patch when that macro
3210 55. Some experimental protocols are using DNS PTR records for new purposes. The
3211 keys for these records are domain names, not reversed IP addresses. The
3212 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3213 leaves it alone. Component reversal etc. now happens only for IP addresses.
3216 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3218 57. Double the size of the debug message buffer (to 2048) so that more of very
3219 long debug lines gets shown.
3221 58. The exicyclog utility now does better if the number of log files to keep
3222 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3224 59. Two changes related to the smtp_active_hostname option:
3226 (1) $smtp_active_hostname is now available as a variable.
3227 (2) The default for smtp_banner uses $smtp_active_hostname instead
3228 of $primary_hostname.
3230 60. The host_aton() function is supposed to be passed a string that is known
3231 to be a valid IP address. However, in the case of IPv6 addresses, it was
3232 not checking this. This is a hostage to fortune. Exim now panics and dies
3233 if the condition is not met. A case was found where this could be provoked
3234 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3235 components; fortuitously, this particular loophole had already been fixed
3236 by change 4.50/55 above.
3238 If there are any other similar loopholes, the new check in host_aton()
3239 itself should stop them being exploited. The report I received stated that
3240 data on the command line could provoke the exploit when Exim was running as
3241 exim, but did not say which command line option was involved. All I could
3242 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3243 running as the user.
3246 61. There was a buffer overflow vulnerability in the SPA authentication code
3247 (which came originally from the Samba project). I have added a test to the
3248 spa_base64_to_bits() function which I hope fixes it.
3251 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3252 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3254 63. The daemon start-up calls getloadavg() while still root for those OS that
3255 need the first call to be done as root, but it missed one case: when
3256 deliver_queue_load_max is set with deliver_drop_privilege. This is
3257 necessary for the benefit of the queue runner, because there is no re-exec
3258 when deliver_drop_privilege is set.
3260 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3261 This has been fixed.
3263 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3264 was in use, was not putting the data itself into the right store pool;
3265 consequently, it could be overwritten for a subsequent message in the same
3266 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3269 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3271 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3272 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3273 the function were treating the return as a boolean value, which happened to
3274 work because 0=false and not-0=true, but is not correct code.
3276 68. The host_aton() function was not handling scoped IPv6 addresses (those
3277 with, for example, "%eth0" on the end) correctly.
3279 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3280 negated items (that is, ~something) in unsigned ints. Some compilers
3281 apparently mutter when there is no cast.
3283 70. If an address verification called from an ACL failed, and did not produce a
3284 user-specific message (i.e. there was only a "system" message), nothing was
3285 put in $acl_verify_message. In this situation, it now puts the system
3288 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3289 session; change 4.31/43 added the unwanted input to the log line - except
3290 that it did not do this in the start of session case. It now does.
3292 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3293 This is wrong and can cause the other end to generate a synchronization
3294 error if it is another Exim or anything else that does the synchronization
3295 check. A QUIT command is no longer sent after a timeout.
3297 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3298 during host lookups.
3300 74. The defer_ok option of callout verification was not working if it was used
3301 when verifying addresses in header lines, that is, for this case:
3303 verify = header_sender/callout=defer_ok
3305 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3306 those file descriptors could be used for SMTP connections. If anything
3307 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3308 could be sent to the SMTP client, causing chaos. The daemon now opens
3309 stdin, stdout, and stderr to /dev/null when it puts itself into the
3312 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3313 log by default. The user can override this with suitable Perl magic.
3315 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3316 the log message when discard triggers, was not working for the DATA ACL or
3317 for the non-SMTP ACL.
3319 78. Error message wording change in sieve.c.
3321 79. If smtp_accept_max_per_host was set, the number of connections could be
3322 restricted to fewer than expected, because the daemon was trying to set up
3323 a new connection before checking whether the processes handling previous
3324 connections had finished. The check for completed processes is now done
3325 earlier. On busy systems, this bug wouldn't be noticed because something
3326 else would have woken the daemon, and it would have reaped the completed
3329 80. If a message was submitted locally by a user whose login name contained one
3330 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3331 It caused a spool format error. I have fixed the spool reading code. A
3332 related problem was that the "from" clause in the Received: line became
3333 illegal because of the space(s). It is now covered by ${quote_local_part.
3335 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3338 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3340 83. Patch from Sieve maintainer to fix unterminated string problem in
3341 "vacation" handling.
3343 84. Some minor changes to the Linux configuration files to help with other
3344 OS variants using glibc.
3346 85. One more patch for Sieve to update vacation handling to latest spec.
3349 ----------------------------------------------------
3350 See the note above about the 4.44 and 4.50 releases.
3351 ----------------------------------------------------
3357 1. Change 4.43/35 introduced a bug that caused file counts to be
3358 incorrectly computed when quota_filecount was set in an appendfile
3361 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3362 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3365 3. Give more explanation in the error message when the command for a transport
3366 filter fails to execute.
3368 4. There are several places where Exim runs a non-Exim command in a
3369 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3370 was being done only for the command run by the queryprogram router. It is
3371 now done for all such subprocesses. The other cases are: ${run, transport
3372 filters, and the commands run by the lmtp and pipe transports.
3374 5. Some older OS have a limit of 256 on the maximum number of file
3375 descriptors. Exim was using setrlimit() to set 1000 as a large value
3376 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3377 systems. I've change it so that if it can't get 1000, it tries for 256.
3379 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3380 was an oversight, and furthermore, ever since the addition of extra
3381 controls (e.g. 4.43/32), the checks on when to allow different forms of
3382 "control" were broken. There should now be diagnostics for all cases when a
3383 control that does not make sense is encountered.
3385 7. $recipients is now available in the predata ACL (oversight).
3387 8. Tidy the search cache before the fork to do a delivery from a message
3388 received from the command line. Otherwise the child will trigger a lookup
3389 failure and thereby defer the delivery if it tries to use (for example) a
3390 cached ldap connection that the parent has called unbind on.
3392 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3393 of $address_data from the recipient verification was clobbered by the
3394 sender verification.
3396 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3397 was its contents. (It was OK if the option was not defined at all.)
3399 11. A "Completed" log line is now written for messages that are removed from
3400 the spool by the -Mrm option.
3402 12. $host_address is now set to the target address during the checking of
3403 ignore_target_hosts.
3405 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3406 being passed; this would have caused $sender_host_name to have been used if
3407 matching the list had actually called for a host name (not very likely,
3408 since this list is usually IP addresses). A host name is now passed as
3411 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3412 code that specifies a non-restarting handler (typically sigaction() in
3413 modern systems) in an attempt to fix a rare and obscure crash bug.
3415 15. Narrowed the window for a race in the daemon that could cause it to ignore
3416 SIGCHLD signals. This is not a major problem, because they are used only to
3417 wake it up if nothing else does.
3419 16. A malformed maildirsize file could cause Exim to calculate negative values
3420 for the mailbox size or file count. Odd effects could occur as a result.
3421 The maildirsize information is now recalculated if the size or filecount
3424 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3425 support for a long time. Removed HAVE_SYS_VFS_H.
3427 18. Updated exipick to current release from John Jetmore.
3429 19. Allow an empty sender to be matched against a lookup in an address list.
3430 Previously the only cases considered were a regular expression, or an
3433 20. Exim went into a mad DNS lookup loop when doing a callout where the
3434 host was specified on the transport, if the DNS lookup yielded more than
3437 21. The RFC2047 encoding function was originally intended for short strings
3438 such as real names; it was not keeping to the 75-character limit for
3439 encoded words that the RFC imposes. It now respects the limit, and
3440 generates multiple encoded words if necessary. To be on the safe side, I
3441 have increased the buffer size for the ${rfc2047: expansion operator from
3444 22. Failure to deliver a bounce message always caused it to be frozen, even if
3445 there was an errors_to setting on the router. The errors_to setting is now
3448 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3449 canonical form (fully expanded) before being placed in
3450 $sender_host_address.
3452 24. Updated eximstats to version 1.33
3454 25. Include certificate and key file names in error message when GnuTLS fails
3455 to set them up, because the GnuTLS error message doesn't include the name
3456 of the failing file when there is a problem reading it.
3458 26. Expand error message when OpenSSL has problems setting up cert/key files.
3461 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3462 (this can affect the format of dates).
3464 28. exim_tidydb, when checking for the continued existence of a message for
3465 which it has found a message-specific retry record, was not finding
3466 messages that were in split spool directories. Consequently, it was
3467 deleting retry records that should have stayed in existence.
3469 29. eximstats updated to version 1.35
3470 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3471 1.35 - bugfix such that pie charts by volume are generated correctly
3473 30. The SPA authentication driver was not abandoning authentication and moving
3474 on to the next authenticator when an expansion was forced to fail,
3475 contradicting the general specification for all authenticators. Instead it
3476 was generating a temporary error. It now behaves as specified.
3478 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3479 (the order specifies the preference for clients). The order is now AES256,
3480 AES128, 3DES, ARCFOUR128.
3482 31. Small patch to Sieve code - explicitly set From: when generating an
3485 32. Exim crashed if a remote delivery caused a very long error message to be
3486 recorded - for instance if somebody sent an entire SpamAssassin report back
3487 as a large number of 550 error lines. This bug was coincidentally fixed by
3488 increasing the size of one of Exim's internal buffers (big_buffer) that
3489 happened as part of the Exiscan merge. However, to be on the safe side, I
3490 have made the code more robust (and fixed the comments that describe what
3493 33. Some experimental protocols are using DNS PTR records for new purposes. The
3494 keys for these records are domain names, not reversed IP addresses. The
3495 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3496 leaves it alone. Component reversal etc. now happens only for IP addresses.
3499 34. The host_aton() function is supposed to be passed a string that is known
3500 to be a valid IP address. However, in the case of IPv6 addresses, it was
3501 not checking this. This is a hostage to fortune. Exim now panics and dies
3502 if the condition is not met. A case was found where this could be provoked
3503 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3504 components; fortuitously, this particular loophole had already been fixed
3505 by change 4.50/55 or 4.44/33 above.
3507 If there are any other similar loopholes, the new check in host_aton()
3508 itself should stop them being exploited. The report I received stated that
3509 data on the command line could provoke the exploit when Exim was running as
3510 exim, but did not say which command line option was involved. All I could
3511 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3512 running as the user.
3515 35. There was a buffer overflow vulnerability in the SPA authentication code
3516 (which came originally from the Samba project). I have added a test to the
3517 spa_base64_to_bits() function which I hope fixes it.
3520 36. The daemon start-up calls getloadavg() while still root for those OS that
3521 need the first call to be done as root, but it missed one case: when
3522 deliver_queue_load_max is set with deliver_drop_privilege. This is
3523 necessary for the benefit of the queue runner, because there is no re-exec
3524 when deliver_drop_privilege is set.
3526 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3527 was in use, was not putting the data itself into the right store pool;
3528 consequently, it could be overwritten for a subsequent message in the same
3529 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3532 38. Sometimes the final signoff response after QUIT could fail to get
3533 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3534 < 0 before doing a fflush(). This bug looks as though it goes back to the
3535 introduction of TLS in release 3.20, but "sometimes" must have been rare
3536 because the tests only now provoked it.
3542 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3543 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3544 Now it returns an integer. A number of calls were still expecting a T/F
3545 return. Fortuitously, in all cases, the tests worked in OK situations,
3546 which is the norm. However, things would have gone wrong on any write
3547 failures on the smtp file descriptor. This function is used when sending
3548 messages over SMTP and also when doing verify callouts.
3550 2. When Exim is called to do synchronous delivery of a locally submitted
3551 message (the -odf or -odi options), it no longer closes stderr before doing
3554 3. Implemented the mua_wrapper option.
3556 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3558 5. Implemented the functions header_remove(), header_testname(),
3559 header_add_at_position(), and receive_remove_recipient(), and exported them
3562 6. If an ACL "warn" statement specified the addition of headers, Exim already
3563 inserted X-ACL-Warn: at the start if there was no header name. However, it
3564 was not making this test for the second and subsequent header lines if
3565 there were newlines in the string. This meant that an invalid header could
3566 be inserted if Exim was badly configured.
3568 7. Allow an ACL "warn" statement to add header lines at the start or after all
3569 the Received: headers, as well as at the end.
3571 8. Added the rcpt_4xx retry error code.
3573 9. Added postmaster_mailfrom=xxx to callout verification option.
3575 10. Added mailfrom=xxxx to the callout verification option, for verify=
3578 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3579 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3581 12. Inserted some casts to stop certain compilers warning when using pointer
3582 differences as field lengths or precisions in printf-type calls (mostly
3583 affecting debugging statements).
3585 13. Added optional readline() support for -be (dynamically loaded).
3587 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3588 same clock tick as a message's arrival, so that its received time was the
3589 same as the "first fail" time on the retry record, and that message
3590 remained on the queue past the ultimate address timeout, every queue runner
3591 would try a delivery (because it was past the ultimate address timeout) but
3592 after another failure, the ultimate address timeout, which should have then
3593 bounced the address, did not kick in. This was a "< instead of <=" error;
3594 in most cases the first failure would have been in the next clock tick
3595 after the received time, and all would be well.
3597 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3598 being recognized when the domain list was tested by the match_domain
3599 condition in an expansion string.
3601 16. Added the ${str2b64: operator.
3603 17. Exim was always calling setrlimit() to set a large limit for the number of
3604 processes, without checking whether the existing limit was already
3605 adequate. (It did check for the limit on file descriptors.) Furthermore,
3606 errors from getrlimit() and setrlimit() were being ignored. Now they are
3607 logged to the main and panic logs, but Exim does carry on, to try to do its
3608 job under whatever limits there are.
3610 18. Imported PCRE 5.0.
3612 19. Trivial typo in log message " temporarily refused connection" (the leading
3615 20. If the log selector return_path_on_delivery was set and an address was
3616 redirected to /dev/null, the delivery process crashed because it assumed
3617 that a return path would always be set for a "successful" delivery. In this
3618 case, the whole delivery is bypassed as an optimization, and therefore no
3621 21. Internal re-arrangement: the function for sending a challenge and reading
3622 a response while authentication was assuming a zero-terminated challenge
3623 string. It's now changed to take a pointer and a length, to allow for
3624 binary data in such strings.
3626 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3628 23. Exim was not respecting finduser_retries when seeking the login of the
3629 uid under which it was called; it was always trying 10 times. (The default
3630 setting of finduser_retries is zero.) Also, it was sleeping after the final
3631 failure, which is pointless.
3633 24. Implemented tls_on_connect_ports.
3635 25. Implemented acl_smtp_predata.
3637 26. If the domain in control=submission is set empty, Exim assumes that the
3638 authenticated id is a complete email address when it generates From: or
3639 Sender: header lines.
3641 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3642 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3643 chown and chgrp in /bin and hostname in /usr/bin.
3645 28. Exim was keeping the "process log" file open after each use, just as it
3646 does for the main log. This opens the possibility of it remaining open for
3647 long periods when the USR1 signal hits a daemon. Occasional processlog
3648 errors were reported, that could have been caused by this. Anyway, it seems
3649 much more sensible not to leave this file open at all, so that is what now
3652 29. The long-running daemon process does not normally write to the log once it
3653 has entered its main loop, and it closes the log before doing so. This is
3654 so that log files can straightforwardly be renamed and moved. However,
3655 there are a couple of unusual error situations where the daemon does write
3656 log entries, and I had neglected to close the log afterwards.
3658 30. The text of an SMTP error response that was received during a remote
3659 delivery was being truncated at 512 bytes. This is too short for some of
3660 the long messages that one sometimes sees. I've increased the limit to
3663 31. It is now possible to make retry rules that apply only when a message has a
3664 specific sender, in particular, an empty sender.
3666 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3667 it possible to be selective about when SMTP synchronization is enforced.
3669 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3671 32. Implemented hosts_connection_nolog.
3673 33. Added an ACL for QUIT.
3675 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3678 35. Added mailbox_size and mailbox_filecount to appendfile.
3680 36. Added control = no_multiline_responses to ACLs.
3682 37. There was a bug in the logic of the code that waits for the clock to tick
3683 in the case where the clock went backwards by a substantial amount such
3684 that the microsecond fraction of "now" was more than the microsecond
3685 fraction of "then" (but the whole seconds number was less).
3687 38. Added support for the libradius Radius client library this is found on
3688 FreeBSD (previously only the radiusclient library was supported).
3694 1. When certain lookups returned multiple values in the form name=value, the
3695 quoting of the values was not always being done properly. Specifically:
3696 (a) If the value started with a double quote, but contained no whitespace,
3698 (b) If the value contained whitespace other than a space character (i.e.
3699 tabs or newlines or carriage returns) it was not quoted.
3700 This fix has been applied to the mysql and pgsql lookups by writing a
3701 separate quoting function and calling it from the lookup code. The fix
3702 should probably also be applied to nisplus, ibase and oracle lookups, but
3703 since I cannot test any of those, I have not disturbed their existing code.
3705 2. A hit in the callout cache for a specific address caused a log line with no
3706 reason for rejecting RCPT. Now it says "Previous (cached) callout
3707 verification failure".
3709 3. There was an off-by-one bug in the queryprogram router. An over-long
3710 return line was truncated at 256 instead of 255 characters, thereby
3711 overflowing its buffer with the terminating zero. As well as fixing this, I
3712 have increased the buffer size to 1024 (and made a note to document this).
3714 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3715 when Exim is waiting for an SMTP response from a remote server, Exim
3716 restarts its select() call on the socket, thereby resetting its timeout.
3717 This is not a problem when such interrupts are rare. Somebody set up a cron
3718 job to run exiwhat every 2 minutes, which is less than the normal select()
3719 timeout (5 or 10 minutes). This meant that the select() timeout never
3720 kicked in because it was always reset. I have fixed this by comparing the
3721 time when an interrupt arrives with the time at the start of the first call
3722 to select(). If more time than the timeout has elapsed, the interrupt is
3723 treated as a timeout.
3725 5. Some internal re-factoring in preparation for the addition of Sieve
3726 extensions (by MH). In particular, the "personal" test is moved to a
3727 separate function, and given an option for scanning Cc: and Bcc: (which is
3728 not set for Exim filters).
3730 6. When Exim created an email address using the login of the caller as the
3731 local part (e.g. when creating a From: or Sender: header line), it was not
3732 quoting the local part when it contained special characters such as @.
3734 7. Installed new OpenBSD configuration files.
3736 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3737 try to make them clearer.
3739 9. Callout options, other than the timeout value, were being ignored when
3740 verifying sender addresses in header lines. For example, when using
3742 verify = header_sender/callout=no_cache
3744 the cache was (incorrectly) being used.
3746 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3747 only the Cygwin environment.
3749 11. When return_path_on_delivery was set as a log selector, if different remote
3750 addresses in the same message used different return paths and parallel
3751 remote delivery occurred, the wrong values would sometimes be logged.
3752 (Whenever a remote delivery process finished, the return path value from
3753 the most recently started remote delivery process was logged.)
3755 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3756 lines when AUTH and/or TLS are in use. This is the "received protocol"
3757 field. Exim used to use "asmtp" for authenticated SMTP, without any
3758 indication (in the protocol name) for TLS use. Now it follows the RFC and
3759 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3760 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3761 names appear in log lines as well as in Received: header lines.
3763 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3764 extensions, and comparison tests, and to fix some bugs.
3766 14. Changes to the "personal" filter test:
3768 (1) The test was buggy in that it was just doing the equivalent of
3769 "contains" tests on header lines. For example, if a user's address was
3770 anne@some.where, the "personal" test would incorrectly be true for
3772 To: susanne@some.where
3774 This test is now done by extracting each address from the header in turn,
3775 and checking the entire address. Other tests that are part of "personal"
3776 are now done using regular expressions (for example, to check local parts
3777 of addresses in From: header lines).
3779 (2) The list of non-personal local parts in From: addresses has been
3780 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3781 taken from the Sieve specification recommendations.
3783 (3) If the message contains any header line starting with "List-" it is
3784 treated as non-personal.
3786 (4) The test for "circular" in the Subject: header line has been removed
3787 because it now seems ill-conceived.
3789 15. Minor typos in src/EDITME comments corrected.
3791 16. Installed latest exipick from John Jetmore.
3793 17. If headers_add on a router specified a text string that was too long for
3794 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3795 of string_sprintf() is now avoided.
3797 18. $message_body_size was not set (it was always zero) when running the DATA
3798 ACL and the local_scan() function.
3800 19. For the "mail" command in an Exim filter, no default was being set for
3801 the once_repeat time, causing a random time value to be used if "once" was
3802 specified. (If the value happened to be <= 0, no repeat happened.) The
3803 default is now 0s, meaning "never repeat". The "vacation" command was OK
3804 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3805 (I found it when inspecting the code).
3807 20. There is now an overall timeout for performing a callout verification. It
3808 defaults to 4 times the callout timeout, which applies to individual SMTP
3809 commands during the callout. The overall timeout applies when there is more
3810 than one host that can be tried. The timeout is checked before trying the
3811 next host. This prevents very long delays if there are a large number of
3812 hosts and all are timing out (e.g. when the network connections are timing
3813 out). The value of the overall timeout can be changed by specifying an
3814 additional sub-option for "callout", called "maxwait". For example:
3816 verify = sender/callout=5s,maxwait=20s
3818 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3819 to the end before writing, but this should make it even safer).
3821 22. Exim was forgetting that it had advertised PIPELINING for the second and
3822 subsequent messages on an SMTP connection. It was also not resetting its
3823 memory on STARTTLS and an internal HELO.
3825 23. When Exim logs an SMTP synchronization error within a session, it now
3826 records whether PIPELINING has been advertised or not.
3828 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3829 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3830 rather than long int.
3832 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3834 26. Added the never_mail option to autoreply.
3840 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3841 crash if the getsockname() call failed; this can happen if a connection is
3842 closed very soon after it is established. The problem was simply in the
3843 order in which certain operations were done, causing Exim to try to write
3844 to the SMTP stream before it had set up the file descriptor. The bug has
3845 been fixed by making things happen in the correct order.
3851 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3852 before the connection was closed, thus losing the rejection response.
3854 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3855 some early Solaris releases, but causes trouble in current releases where
3856 socklen_t is defined.
3858 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3861 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3864 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3865 that releasing the top of it at the end releases what was used for sub-
3866 expansions (unless the block got too big). However, discard this block if
3867 the first thing is a variable or header, so that we can use its block when
3868 it is dynamic (useful for very large $message_headers, for example).
3870 6. Lookups now cache *every* query, not just the most recent. A new, separate
3871 store pool is used for this. It can be recovered when all lookup caches are
3872 flushed. Lookups now release memory at the end of their result strings.
3873 This has involved some general refactoring of the lookup sources.
3875 7. Some code has been added to the store_xxx() functions to reduce the amount
3876 of flapping under certain conditions.
3878 8. log_incoming_interface used to affect only the <= reception log lines. Now
3879 it causes the local interface and port to be added to several more SMTP log
3880 lines, for example "SMTP connection from", and rejection lines.
3882 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3884 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3886 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3887 could be overwritten at the end of the current message (or the start of a
3888 new message if it was set in a HELO ACL). The value is now preserved for
3889 the duration of the SMTP connection.
3891 12. If a transport had a headers_rewrite setting, and a matching header line
3892 contained an unqualified address, that address was qualified, even if it
3893 did not match any rewriting rules. The underlying bug was that the values
3894 of the flags that permit the existence of unqualified sender and recipient
3895 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3896 non-local messages, and by -bnq for local messages) were not being
3897 preserved with the message after it was received.
3899 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3900 "next input=" as part of the text comprising the host identity instead of
3901 the correct text. The code was using the same buffer for two different
3902 strings. However, depending on which order the printing function evaluated
3903 its arguments, the bug did not always show up. Under Linux, for example, my
3904 test suite worked just fine.
3906 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3907 doesn't work with some older versions of Perl. It has been changed to "my",
3908 which in any case is probably the better facility to use.
3910 15. A really picky compiler found some instances of statements for creating
3911 error messages that either had too many or two few arguments for the format
3914 16. The size of the buffer for calls to the DNS resolver has been increased
3915 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3916 for addresses that have a lot of PTR records. This alleviates a problem; it
3917 does not fully solve it.
3919 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3920 buffer now truncates the list and logs the incident, which is the same
3921 action as happens when Exim is looking up a host name and its aliases.
3922 Previously in this situation something unpredictable would happen;
3923 sometimes it was "internal error: store_reset failed".
3925 18. If a server dropped the connection unexpectedly when an Exim client was
3926 using GnuTLS and trying to read a response, the client delivery process
3927 crashed while trying to generate an error log message.
3929 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3930 string, for example:
3932 warn message = H1: something\nH2: something
3934 the text was added as a single header line from Exim's point of view
3935 though it ended up OK in the delivered message. However, searching for the
3936 second and subsequent header lines using $h_h2: did not work. This has been
3937 fixed. Similarly, if a system filter added multiple headers in this way,
3938 the routers could not see them.
3940 20. Expanded the error message when iplsearch is called with an invalid key to
3941 suggest using net-iplsearch in a host list.
3943 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3944 ACLs are no longer actually imposed (and a message to that effect is
3947 22. If a "gecos" field in a passwd entry contained escaped characters, in
3948 particular, if it contained a \" sequence, Exim got it wrong when building
3949 a From: or a Sender: header from that name. A second bug also caused
3950 incorrect handling when an unquoted " was present following a character
3951 that needed quoting.
3953 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3954 was not being matched caselessly.
3956 24. Arranged for all hyphens in the exim.8 source to be escaped with
3959 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3960 properly. Recipient callout cache records were still being keyed to include
3961 the sender, even when use_sender was set false. This led to far more
3962 callouts that were necessary. The sender is no longer included in the key
3963 when use_sender is false.
3965 26. Added "control = submission" modifier to ACLs.
3967 27. Added the ${base62d: operator to decode base 62 numbers.
3969 28. dnsdb lookups can now access SRV records.
3971 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3972 the configuration file.
3974 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3975 (-v) mode. This makes the output for a verbose queue run more intelligible.
3977 31. Added a use_postmaster feature to recipient callouts.
3979 32. Added the $body_zerocount variable, containing the number of binary zero
3980 bytes in the message body.
3982 33. The time of last modification of the "new" subdirectory is now used as the
3983 "mailbox time last read" when there is a quota error for a maildir
3986 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3988 35. Added +ignore_unknown as a special item in host lists.
3990 36. Code for decoding IPv6 addresses in host lists is now included, even if
3991 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3992 address was recognized as an IP address, but was then not correctly decoded
3993 into binary, causing unexpected and incorrect effects when compared with
4000 1. Very minor rewording of debugging text in manualroute to say "list of
4001 hosts" instead of "hostlist".
4003 2. If verify=header_syntax was set, and a header line with an unqualified
4004 address (no domain) and a large number of spaces between the end of the
4005 name and the colon was received, the reception process suffered a buffer
4006 overflow, and (when I tested it) crashed. This was caused by some obsolete
4007 code that should have been removed. The fix is to remove it!
4009 3. When running in the test harness, delay a bit after writing a bounce
4010 message to get a bit more predictability in the log output.
4012 4. Added a call to search_tidyup() just before forking a reception process. In
4013 theory, someone could use a lookup in the expansion of smtp_accept_max_
4014 per_host which, without the tidyup, could leave open a database connection.
4016 5. Added the variables $recipient_data and $sender_data which get set from a
4017 lookup success in an ACL "recipients" or "senders" condition, or a router
4018 "senders" option, similar to $domain_data and $local_part_data.
4020 6. Moved the writing of debug_print from before to after the "senders" test
4023 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4024 problems for message scanning, either using a data ACL, or using
4025 local_scan() because the Received: header was not generated till after they
4026 were called (in order to set the time as the time of reception completion).
4027 I have revised the way this works. The header is now generated after the
4028 body is received, but before the ACL or local_scan() are called. After they
4029 are run, the timestamp in the header is updated.
4035 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4036 before starting a queue runner without re-exec. This happened only when
4037 deliver_drop_privilege was set or when the Exim user was set to root. The
4038 effect of the bug was that timeouts during subsequent deliveries caused
4039 crashes instead of being properly handled. The handler is now left at its
4040 default (and expected) setting.
4042 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4043 message, again when deliver_drop_privilege is set or Exim is run as root.
4044 The bug described in (1) was not present in this case, but the tidying up
4045 of the other signals was missing. I have made the two cases consistent.
4047 3. The ignore_target_hosts setting on a manualroute router was being ignored
4048 for hosts that were looked up using the /MX notation.
4050 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4053 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4054 operated on the sender address. After changing the $sender_address to <>
4055 for the sender address verify, Exim was re-instated it as the original
4056 (before rewriting) address, but remembering that it had rewritten it, so it
4057 wasn't rewriting it again. This bug also had the effect of breaking the
4058 sender address verification caching when the sender address was rewritten.
4060 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4061 This has been changed so that if the ip literal address matches
4062 ignore_target_hosts, the router declines.
4064 7. Added expansion conditions match_domain, match_address, and match_local_
4065 part (NOT match_host).
4067 8. The placeholder for the Received: header didn't have a length field set.
4069 9. Added code to Exim itself and to exim_lock to test for a specific race
4070 condition that could lead to file corruption when using MBX delivery. The
4071 issue is with the lockfile that is created in /tmp. If this file is removed
4072 after a process has opened it but before that process has acquired a lock,
4073 there is the potential for a second process to recreate the file and also
4074 acquire a lock. This could lead to two Exim processes writing to the file
4075 at the same time. The added code performs the same test as UW imapd; it
4076 checks after acquiring the lock that its file descriptor still refers to
4077 the same named file.
4079 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4080 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4083 11. Added the smtp_active_hostname option. If used, this will typically be made
4084 to depend on the incoming interface address. Because $interface_address is
4085 not set up until the daemon has forked a reception process, error responses
4086 that can happen earlier (such as "too many connections") no longer contain
4089 12. If an expansion in a condition on a "warn" statement fails because a lookup
4090 defers, the "warn" statement is abandoned, and the next ACL statement is
4091 processed. Previously this caused the whole ACL to be aborted.
4093 13. Added the iplsearch lookup type.
4095 14. Added ident_timeout as a log selector.
4097 15. Added tls_certificate_verified as a log selector.
4099 16. Added a global option tls_require_ciphers (compare the smtp transport
4100 option of the same name). This controls incoming TLS connections.
4102 17. I finally figured out how to make tls_require_ciphers do a similar thing
4103 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4104 before starting the TLS session.
4106 18. Tabs are now shown as \t in -bP output.
4108 19. If the log selector return_path_on_delivery was set, Exim crashed when
4109 bouncing a message because it had too many Received: header lines.
4111 20. If two routers both had headers_remove settings, and the first one included
4112 a superfluous trailing colon, the final name in the first list and the
4113 first name in the second list were incorrectly joined into one item (with a
4114 colon in the middle).
4120 1. Added -C and -D options to the exinext utility, mainly to make it easier
4121 to include in the automated testing, but these could be helpful when
4122 multiple configurations are in use.
4124 2. The exinext utility was not formatting the output nicely when there was
4125 an alternate port involved in the retry record key, nor when there was a
4126 message id as well (for retries that were specific to a specific message
4127 and a specific host). It was also confused by IPv6 addresses, because of
4128 the additional colons they contain. I have fixed the IPv4 problem, and
4129 patched it up to do a reasonable job for IPv6.
4131 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4132 delivery, the log line now contains "pipelined" if PIPELINING was used.
4134 4. An SMTP transport process used to panic and die if the bind() call to set
4135 an explicit outgoing interface failed. This has been changed; it is now
4136 treated in the same way as a connect() failure.
4138 5. A reference to $sender_host_name in the part of a conditional expansion
4139 that was being skipped was still causing a DNS lookup. This no longer
4142 6. The def: expansion condition was not recognizing references to header lines
4143 that used bh_ and bheader_.
4145 7. Added the _cache feature to named lists.
4147 8. The code for checking quota_filecount in the appendfile transport was
4148 allowing one more file than it should have been.
4150 9. For compatibility with Sendmail, the command line option
4158 and sets the incoming protocol and host name (for trusted callers). The
4159 host name and its colon can be omitted when only the protocol is to be set.
4160 Note the Exim already has two private options, -pd and -ps, that refer to
4161 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4162 "s", but I don't think that's a major issue.
4164 10. A number of refactoring changes to the code, none of which should affect
4167 (a) The number of logging options was getting close to filling up the
4168 32-bit word that was used as a bit map. I have split them into two classes:
4169 those that are passed in the argument to log_write(), and those that are
4170 only ever tested independently outside of that function. These are now in
4171 separate 32-bit words, so there is plenty of room for expansion again.
4172 There is no change in the user interface or the logging behaviour.
4174 (b) When building, for example, log lines, the code previously used a
4175 macro that called string_cat() twice, in order to add two strings. This is
4176 not really sufficiently general. Furthermore, there was one instance where
4177 it was actually wrong because one of the argument was used twice, and in
4178 one call a function was used. (As it happened, calling the function twice
4179 did not affect the overall behaviour.) The macro has been replaced by a
4180 function that can join an arbitrary number of extra strings onto a growing
4183 (c) The code for expansion conditions now uses a table and a binary chop
4184 instead of a serial search (which was left over from when there were very
4185 few conditions). Also, it now recognizes conditions like "pam" even when
4186 the relevant support is not compiled in: a suitably worded error message is
4187 given if an attempt is made to use such a condition.
4189 11. Added ${time_interval:xxxxx}.
4191 12. A bug was causing one of the ddress fields not to be passed back correctly
4192 from remote delivery subprocesses. The field in question was not being
4193 subsequently used, so this caused to problems in practice.
4195 13. Added new log selectors queue_time and deliver_time.
4197 14. Might have fixed a bug in maildirsizefile handling that threw up
4198 "unexpected character" debug warnings, and recalculated the data
4199 unnecessarily. In any case, I expanded the warning message to give more
4202 15. Added the message "Restricted characters in address" to the statements in
4203 the default ACL that block characters like @ and % in local parts.
4205 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4206 Three changes have been made:
4208 (a) There was a serious bug; a negative response to MAIL caused the whole
4209 recipient domain to be cached as invalid, thereby blocking all messages
4210 to all local parts at the same domain, from all senders. This bug has
4211 been fixed. The domain is no longer cached after a negative response to
4212 MAIL if the sender used is not empty.
4214 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4217 (c) A new callout option, "use_sender" has been added for people who want
4218 the modified behaviour.
4224 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4227 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4228 indeed breaks things for older releases.
4230 3. Added additional logging to the case where there is a problem reading data
4231 from a filter that is running in a subprocess using a pipe, in order to
4232 try to track down a specific problem.
4234 4. Testing facility fudge: when running in the test harness and attempting
4235 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4236 getting "No route to host". Convert this to a timeout.
4238 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4241 6. Some OS don't have socklen_t but use size_t instead. This affects the
4242 fifth argument of getsockopt() amongst other things. This is now
4243 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4244 can be set for individual OS. I have set it for SunOS5, OSF1, and
4245 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4246 some earlier ones do not.
4248 7. Change 4.30/15 was not doing the test caselessly.
4250 8. The standard form for an IPv6 address literal was being rejected by address
4251 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4252 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4253 this, as well as the form without the "IPv6" on the front (but only when
4254 address literals are enabled, of course).
4256 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4258 10. Exim crashed if a message with an empty sender address specified by -f
4259 encountered a router with an errors_to setting. This could be provoked only
4260 by a command such as
4264 where an empty string was supplied; "<>" did not hit this bug.
4266 11. Installed PCRE release 4.5.
4268 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4269 remained set. It is now erased.
4271 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4272 times from message ids (which are base 36 rather than the normal 62).
4274 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4275 were being counted as actual protocol errors, and logged if the log
4276 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4277 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4278 and DATA following a set of rejected RCPTs do not count as protocol errors.
4279 In other words, Exim assumes they were pipelined, though this may not
4280 actually be the case. Of course, in all cases the client gets an
4281 appropriate error code.
4283 15. If a lookup fails in an ACL condition, a message about the failure may
4284 be available; it is used if testing the ACL cannot continue, because most
4285 such messages specify what the cause of the deferral is. However, some
4286 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4287 that caused an old message to be retained and used if a later statement
4288 caused a defer, replacing the real cause of the deferral.
4290 16. If an IP address had so many PTR records that the DNS lookup buffer
4291 was not large enough to hold them, Exim could crash while trying to process
4292 the truncated data. It now detects and logs this case.
4294 17. Further to 4.21/58, another change has been made: if (and only if) the
4295 first line of a message (the first header line) ends with CRLF, a bare LF
4296 in a subsequent header line has a space inserted after it, so as not to
4297 terminate the header.
4299 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4300 unnecessarily, used atoi() instead of strtol(), and didn't check the
4301 termination when getting file sizes from file names by regex.
4303 19. Completely re-implemented the support for maildirsize files, in the light
4304 of a number of problems with the previous contributed implementation
4305 (4.30/29). In particular:
4307 . If the quota is zero, the maildirsize file is maintained, but no quota is
4310 . If the maildir directory does not exist, it is created before any attempt
4311 to write a maildirsize file.
4313 . The quota value in the file is just a cache; if the quota is changed in
4314 the transport, the new value overrides.
4316 . A regular expression is available for excluding directories from the
4319 20. The autoreply transport checks the characters in options that define the
4320 message's headers; it allows continued headers, but it was checking with
4321 isspace() after an embedded newline instead of explicitly looking for a
4324 21. If all the "regular" hosts to which an address was routed had passed their
4325 expiry times, and had not reached their retry times, the address was
4326 bounced, even if fallback hosts were defined. Now Exim should go on to try
4329 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4330 equivalent code in the SMTP transport. Some hosts send humungous responses
4331 to HELO/EHLO, more than 1024 it seems.
4333 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4334 gives compiler warnings in some environments. I've now done it "properly",
4337 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4338 (because of problems with the built-in one) was declared to return uschar *
4339 instead of char *, causing compiler failure.
4341 25. Fixed a file descriptor leak when processing alias/forward files.
4343 26. Fixed a minor format string issue in dbfn.c.
4345 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4347 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4348 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4351 29. When a sender address is verified, it is cached, to save repeating the test
4352 when there is more than one recipient in a message. However, when the
4353 verification involves a callout, it is possible for different callout
4354 options to be set for different recipients. It is too complicated to keep
4355 track of this in the cache, so now Exim always runs a verification when a
4356 callout is required, relying on the callout cache for the optimization.
4357 The overhead is duplication of the address routing, but this should not be
4360 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4361 to be verified with callout=postmaster, and the main callout worked but the
4362 postmaster check failed, the verification correctly failed. However, if a
4363 subsequent RCPT command asked for sender verification *without* the
4364 postmaster check, incorrect caching caused this verification also to fail,
4367 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4368 it was not caching the DNS options (qualify_single, search_parents) that
4369 were used when the lookup failed. A subsequent lookup with different
4370 options therefore always gave the same answer, though there were cases
4371 where it should not have. (Example: a "domains = !$mx_any" option on a
4372 dnslookup router: the "domains" option is always processed without any
4373 widening, but the router might have qualify_single set.) Now Exim uses the
4374 cached value only when the same options are set.
4376 32. Added John Jetmore's "exipick" utility to the distribution.
4378 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4379 than a timeout (e.g. a certificate is required, and is not provided), an
4380 Exim server now closes the connection immediately. Previously it waited for
4381 the client to close - but if the client is SSL, it seems that they each
4382 wait for each other, leading to a delay before one of them times out.
4384 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4385 maintained 0.8.x compatibility because I don't think many are using it, and
4386 it is clearly obsolete.
4388 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4391 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4392 client certificate was expired. A simple patch fixes this, though I don't
4393 understand the full logic of why the verify callback is called multiple
4396 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4397 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4398 which causes problems with some clients (such as the Certicom SSL Plus
4399 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4400 disables the coutermeasure allowing Eudora to connect."
4402 38. Exim was not checking that a write() to a log file succeeded. This could
4403 lead to Bad Things if a log got too big, in particular if it hit a file
4404 size limit. Exim now panics and dies if it cannot write to a log file, just
4405 as it does if it cannot open a log file.
4407 39. Modified OS/Makefile-Linux so that it now contains
4409 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4411 The two -D definitions ensure that Exim is compiled with large file
4412 support, which makes it possible to handle log files that are bigger than
4415 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4416 instance) a domain was checked against a named list that involved a lookup,
4417 causing $domain_data to be set, then another domain was checked against the
4418 same list, then the first domain was re-checked, the value of $domain_data
4419 after the final check could be wrong. In particular, if the second check
4420 failed, it could be set empty. This bug probably also applied to
4423 41. The strip_trailing_dot option was not being applied to the address given
4424 with the -f command-line option.
4426 42. The code for reading a message's header from the spool was incrementing
4427 $received_count, but never initializing it. This meant that the value was
4428 incorrect (doubled) while delivering a message in the same process in which
4429 it was received. In the most common configuration of Exim, this never
4430 happens - a fresh exec is done - but it can happen when
4431 deliver_drop_privilege is set.
4433 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4434 it now includes up to 150 characters of the unexpected data in the log
4437 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4438 and building data strings. The size of both of these buffers was 10 000
4439 bytes - far larger than anybody would *ever* want, thought I. Needless to
4440 say, somebody hit the limit. I have increased the maximum line length to
4441 20 000 and the maximum data length of concatenated lines to 100 000. I have
4442 also fixed two bugs, because there was no checking on these buffers. Tsk,
4443 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4444 buffer is too small.
4446 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4447 lsearch lookups. Now it does.
4449 46. When parsing a route_list item in a manualroute router, a fixed-length
4450 buffer was used for the list of hosts. I made this 1024 bytes long,
4451 thinking that nobody would ever have a list of hosts that long. Wrong.
4452 Somebody had a whole pile of complicated expansion conditions, and the
4453 string was silently truncated, leading to an expansion error. It turns out
4454 that it is easier to change to an unlimited length (owing to other changes
4455 that have happened since this code was originally written) than to build
4456 structure for giving a limitation error. The length of the item that
4457 expands into the list of hosts is now unlimited.
4459 47. The lsearch lookup could not handle data where the length of text line was
4460 more than 4095 characters. Such lines were truncated, leading to shortened
4461 data being returned. It should now handle lines of any length.
4463 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4464 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4467 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4468 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4469 no longer happens in comment lines. A list of replacements is now placed
4470 at the head of all of the source files, except those whose only change is
4471 to replace PERL_COMMAND in the very first #! line.
4473 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4474 messages on the queue, with a bottom-up merge sort, using code contributed
4475 by Michael Haardt. This should make operations like -bp somewhat faster on
4476 large queues. It won't affect queue runners, except when queue_run_in_order
4479 51. Installed eximstats 1.31 in the distribution.
4481 52. Added support for SRV lookups to the dnslookup router.
4483 53. If an ACL referred to $message_body or $message_body_end, the value was not
4484 reset for any messages that followed in the same SMTP session.
4486 54. The store-handling optimization for building very long strings was not
4487 differentiating between the different store pools. I don't think this
4488 actually made any difference in practice, but I've tidied it.
4490 55. While running the routers to verify a sender address, $sender_address
4491 was still set to the sender address. This is wrong, because when routing to
4492 send a bounce to the sender, it would be empty. Therefore, I have changed
4493 it so that, while verifying a sender address, $sender_address is set to <>.
4494 (There is no change to what happens when verifying a recipient address.)
4496 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4497 target A or AAAA records (if not already returned) without resetting the
4498 qualify_single or search_parents options of the DNS resolver. These are
4499 inappropriate in this case because the targets of MX and SRV records must
4500 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4501 target that, when qualified, matched something in the local domain. These
4502 two options are now turned off when doing these lookups.
4504 57. It seems that at least some releases of Reiserfs (which does not have the
4505 concept of a fixed number of inodes) returns zero and not -1 for the
4506 number of available inodes. This interacted badly with check_spool_inodes,
4507 which assumed that -1 was the "no such thing" setting. What I have done is
4508 to check that the total number of inodes is greater than zero before doing
4509 the test of how many are available.
4511 58. When a "warn" ACL statement has a log_message modifier, the message is
4512 remembered, and not repeated. This is to avoid a lot of repetition when a
4513 message has many recipients that cause the same warning to be written.
4514 Howewer, Exim was preserving the list of already written lines for an
4515 entire SMTP session, which doesn't seem right. The memory is now reset if a
4516 new message is started.
4518 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4519 debugging output unless log_rewrite was also set.
4521 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4522 of (int)(handle) when we know that handle contains (void *)(-1).
4524 61. The Exim daemon panic-logs an error return when it closes the incoming
4525 connection. However "connection reset by peer" seems to be common, and
4526 isn't really an error worthy of noting specially, so that particular error
4529 62. When Exim is trying to find all the local interfaces, it used to panic and
4530 die if the ioctl to get the interface flags failed. However, it seems that
4531 on at least one OS (Solaris 9) it is possible to have an interface that is
4532 included in the list of interfaces, but for which you get a failure error
4533 for this call. This happens when the interface is not "plumbed" into a
4534 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4535 failure of the "get flags" call assumes that the interface is down.
4537 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4538 makes life easier for people who are doing arithmetic on fields extracted
4539 from dates, where you often get leading zeros that should not be
4540 interpreted as octal.
4542 64. Added qualify_domain to the redirect router, to override the global
4545 65. If a pathologically long header line contained very many addresses (the
4546 report of this problem mentioned 10 000) and each of them was rewritten,
4547 Exim could use up a very large amount of memory. (It kept on making new
4548 copies of the header line as it rewrote, and never released the old ones.)
4549 At the expense of a bit more processing, the header rewriting function has
4550 been changed so that it no longer eats memory in this way.
4552 66. The generation of the Received: header has been moved from the time that a
4553 message starts to be received, to the time that it finishes. The timestamp
4554 in the Received: header should now be very close to that of the <= log
4555 line. There are two side-effects of this change:
4557 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4558 logged header lines no longer include the local Received: line, because
4559 it has not yet been created. The same applies to a copy of the message
4560 that is returned to a non-SMTP sender when a message is rejected.
4562 (b) When a filter file is tested using -bf, no additional Received: header
4563 is added to the test message. After some thought, I decided that this
4566 This change does not affect the value of $received_for. It is still set
4567 after address rewriting, but before local_scan() is called.
4569 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4571 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4572 gave an unhelpful panic error message, and a defer error. I have managed to
4573 change this behaviour so that it now rejects any supplied certificate,
4574 which seems right, as the list of acceptable certificates is empty.
4576 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4577 gave an unhelpful defer error. I have not managed to make this reject any
4578 supplied certificates, but the error message it gives is "no certificate
4579 supplied", which is not helpful.
4581 70. exigrep's output now also includes lines that are not associated with any
4582 message, but which match the given pattern. Implemented by a patch from
4583 Martin Sluka, which also tidied up the Perl a bit.
4585 71. Recipient callout verification, like sender verification, was using <> in
4586 the MAIL FROM command. This isn't really the right thing, since the actual
4587 sender may affect whether the remote host accepts the recipient or not. I
4588 have changed it to use the actual sender in the callout; this means that
4589 the cache record is now keyed on a recipient/sender pair, not just the
4590 recipient address. There doesn't seem to be a real danger of callout loops,
4591 since a callout by the remote host to check the sender would use <>.
4592 [SEE ABOVE: changed after hitting problems.]
4594 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4595 temporary errors. However, in the case of such a code being given after
4596 the end of a data transmission (i.e. after ".") Exim was failing to write
4597 a retry record for the message. (Yes, there was some broken host that was
4598 actually sending 8xx at this point.)
4600 73. An unknown lookup type in a host list could cause Exim to panic-die when
4601 the list was checked. (An example that provoked this was putting <; in the
4602 middle of a list instead of at the start.) If this happened during a DATA
4603 ACL check, a -D file could be left lying around. This kind of configuration
4604 error no longer causes Exim to die; instead it causes a defer errror. The
4605 incident is still logged to the main and panic logs.
4607 74. Buglet left over from Exim 3 conversion. The message "too many messages
4608 in one connection" was written to the rejectlog but not the mainlog, except
4609 when address rewriting (yes!) was being logged.
4611 75. Added write_rejectlog option.
4613 76. When a system filter was run not as root (that is, when system_filter_user
4614 was set), the values of the $n variables were not being returned to the
4615 main process; thus, they were not subsequently available in the $sn
4618 77. Added +return_path_on_delivery log selector.
4620 78. A connection timeout was being treated differently from recipients deferred
4621 when testing hosts_max_try with a message that was older than the host's
4622 retry timeout. (The host should not be counted, thus allowing all hosts to
4623 be tried at least once before bouncing.) This may have been the cause of an
4624 occasionally reported bug whereby a message would remain on the queue
4625 longer than the retry timeout, but would be bounced if a delivery was
4626 forced. I say "may" because I never totally pinned down the problem;
4627 setting up timeout/retry tests is difficult. See also the next item.
4629 79. The ultimate address timeout was not being applied to errors that involved
4630 a combination of host plus message (for example, a timeout on a MAIL
4631 command). When an address resolved to a number of possible hosts, and they
4632 were not all tried for each delivery (e.g. because of hosts_max_try), a
4633 message could remain on the queue longer than the retry timeout.
4635 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4638 81. Fixed an obscure SMTP outgoing bug which required at least the following
4639 conditions: (a) there was another message waiting for the same server;
4640 (b) the server returned 5xx to all RCPT commands in the first message so
4641 that the message was not completed; (c) the server dropped the connection
4642 or gave a negative response to the RSET that Exim sends to abort the
4643 transaction. The observed case was a dropped connection after DATA that had
4644 been sent in pipelining mode. That is, the server had advertised PIPELINING
4645 but was not implementing it correctly. The effect of the bug was incorrect
4646 behaviour, such as trying another host, and this could lead to a crash.
4652 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4653 and daemon.c were passed as pointers to ints; they should have been
4654 pointers to socklen_t variables (which are typically unsigned ints).
4656 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4659 3. Fixed a really odd bug that affected only the testing scheme; patching a
4660 certain fixed string in the binary changed the value of another string that
4661 happened to be identical to the end of the original first string.
4663 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4664 name", it returns that address as the IP address. On some operating
4665 systems (e.g. Solaris), it also passes back the IP address string as the
4666 "host name". However, on others (e.g. Linux), it passes back an empty
4667 string. Exim wasn't checking for this, and was changing the host name to an
4668 empty string, assuming it had been canonicized.
4670 5. Although rare, it is permitted to have more than one PTR record for a given
4671 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4672 all the names associated with an address, because they do in Solaris.
4673 However, it seems that they do not in Linux for data that comes from the
4674 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4675 I found this out when I moved to a new Linux workstation and tried to run
4676 the Exim test suite.
4678 To get round this problem I have changed the code so that it now does its
4679 own call to the DNS to look up PTR records when searching for a host name.
4680 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4681 addresses that are only in /etc/hosts are still found.
4683 This behaviour is, however, controlled by an option called host_lookup_
4684 order, which defaults to "bydns:byaddr". If people want to use the other
4685 order, or indeed, just use one or the other means of lookup, they can
4686 specify it in this variable.
4688 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4689 some operating systems, this comes back from gethostbyaddr() as an empty
4690 string, and this is what Exim used to test for. However, it seems that in
4691 other systems, "." is yielded. Exim now tests for this case too.
4693 7. The values of check_spool_space and check_log_space are now held internally
4694 as a number of kilobytes instead of an absolute number of bytes. If a
4695 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4696 kilobyte. This means that much larger values can be stored.
4698 8. Exim monitor: an attempt to get the action menu when not actually pointing
4699 at a message produces an empty menu entitled "No message selected". This
4700 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4701 no entries in it ("Shell widget menu has zero width and/or height"). So I
4702 have added a single, blank menu entry in this case.
4704 9. Added ${quote_local_part.
4706 10. MIME decoding is now applied to the contents of Subject: header lines when
4709 11. Now that a reference to $sender_host_address automatically causes a reverse
4710 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4711 host lookup before query-style lookups in lists that might use this
4712 variable. This has therefore been abolished, and the "net-" prefix is no
4713 longer necessary for query-style lookups.
4715 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4716 have been a typo for LFLAGS, so it has been changed.
4718 13. The install script calls Exim with "-C /dev/null" in order to find the
4719 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4720 to be output. Howeve, since Exim outputs its version number before the
4721 error, it didn't break the script. It just looked ugly. I fixed this by
4722 always allowing "-C /dev/null" if the caller is root.
4724 14. Ignore overlarge ACL variable number when reading spool file - insurance
4725 against a later release with more variables having written the file.
4727 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4728 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4729 this, as well as the form without the "IPv6" on the front.
4731 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4732 OS/Makefile-Darwin file.
4734 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4735 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4737 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4738 variable. However, some OS use ints for inodes. I've added cast to long int
4739 to get rid of the compiler warning.
4741 19. I had forgotten to lock out "/../" in configuration file names when
4742 ALT_CONFIG_PREFIX was set.
4744 20. Routers used for verification do not need to specify transports. However,
4745 if such a router generated a host list, and callout was configured, Exim
4746 crashed, because it could not find a port number from the (non-existent)
4747 transport. It now assumes port 25 in this circumstance.
4749 21. Added the -t option to exigrep.
4751 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4752 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4753 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4754 them is set, LOOKUP_LSEARCH is forced.
4756 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4758 24. Added sender and host information to the "rejected by local_scan()" log
4759 line; previously there was no indication of these.
4761 25. Added .include_if_exists.
4763 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4764 Linux. It turns out that not all file systems support this. Apparently some
4765 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4766 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4767 not supported on the file descriptor, is now ignored when Exim is trying to
4768 sync a directory. This applies only to Linux.
4770 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4772 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4773 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4776 29. Added support for maildirsize files from supplied patch (modified a bit).
4778 30. The use of :fail: followed by an empty string could lead Exim to respond to
4779 sender verification failures with (e.g.):
4781 550 Verification failed for <xxx>
4782 550 Sender verify failed
4784 where the first response line was missing the '-' that indicates it is not
4785 the final line of the response.
4787 31. The loop for finding the name of the user that called Exim had a hardwired
4788 limit of 10; it now uses the value of finduser_retries, which is used for
4789 all other user lookups.
4791 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4794 33. Exim was neglecting to zero errno before one call of strtol() when
4795 expanding a string and expecting an integer value. On some systems this
4796 resulted in spurious "integer overflow" errors. Also, it was casting the
4797 result into an int without checking.
4799 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4800 did not work. The code looks as if it has *never* worked, though it appears
4801 to have been documented since at least releast 1.62. I have made it work.
4803 35. The "timeout_DNS" error in retry rules, also documented since at least
4804 1.62, also never worked. As it isn't clear exactly what this means, and
4805 clearly it isn't a major issue, I have abolished the feature by treating it
4806 as "timeout", and writing a warning to the main and panic logs.
4808 36. The display of retry rules for -brt wasn't always showing the error code
4811 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4812 timeout_connect_A, timeout_connect_MX.
4814 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4815 to the empty sender.
4817 39. The daemon was not analysing the content of -oX till after it had closed
4818 stderr and disconnected from the controlling terminal. This meant that any
4819 syntax errors were only noted on the panic log, and the return code from
4820 the command was 0. By re-arranging the code a little, I've made the
4821 decoding happen first, so such errors now appear on stderr, and the return
4822 code is 1. However, the actual setting up of the sockets still happens in
4823 the disconnected process, so errors there are still only recorded on the
4826 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4827 connections (as happens on some IP stacks) was logged at start up time as
4828 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4829 it from "IPv6 and IPv4", which means that two separate sockets are being
4832 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4833 says whether AF_INET or AF_INET6 was passed as an argument.
4835 42. Exiwhat output was messed up when time zones were included in log
4838 43. Exiwhat now gives more information about the daemon's listening ports,
4839 and whether -tls-on-connect was used.
4841 44. The "port" option of the smtp transport is now expanded.
4843 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4844 silently ignored. Now an error message is written to the main and panic
4847 46. There's a new ACL modifier called "logwrite" which writes to a log file
4848 as soon as it is encountered.
4850 47. Added $local_user_uid and $local_user_gid at routing time.
4852 48. Exim crashed when trying to verify a sender address that was being
4855 49. Exim was recognizing only a space character after ".include". It now also
4856 recognizes a tab character.
4858 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4859 extracting the relevant information from the specification. The man page no
4860 longer contains scrambled data for the -d option, and I've added a section
4861 at the front about calling Exim under different names.
4863 51. Added "extra_headers" argument to the "mail" command in filter files.
4865 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4868 53. Installed eximstats 1.29.
4870 54. Added transport_filter_timeout as a generic transport option.
4872 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4873 Cc: header lines. This was required by RFC 822, but it not required by RFC
4876 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4877 incoming messages that did not have them. Now it does so only if the
4878 message originates locally, that is, if there is no associated remote host
4879 address. When Resent- header lines are present, this applies to the Resent-
4880 lines rather than the non-Resent- lines.
4882 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4883 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4885 58. Messages for configuration errors now include the name of the main
4886 configuration file - useful now that there may be more than one file in a
4887 list (.included file names were always shown).
4889 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4890 for those rare installations that do not start the daemon as root or run it
4891 setuid root. I've cut out the call to initgroups() if the daemon is not
4894 60. The Exim user and group can now be bound into the binary as text strings
4895 that are looked up at the start of Exim's processing.
4897 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4899 62. Added $mailstore_basename variable.
4901 63. Installed patch to sieve.c from Michael Haardt.
4903 64. When Exim failed to open the panic log after failing to open the main log,
4904 the original message it was trying to log was written to stderr and debug
4905 output, but if they were not available (the usual case in production), it
4906 was lost. Now it is written to syslog before the two lines that record the
4907 failures to open the logs.
4909 65. Users' Exim filters run in subprocesses under the user's uid. It is
4910 possible for a "deliver" command or an alias in a "personal" command to
4911 provoke an address rewrite. If logging of address rewriting is configured,
4912 this fails because the process is not running as root or exim. There may be
4913 a better way of dealing with this, but for the moment (because 4.30 needs
4914 to be released), I have disabled address rewrite logging when running a
4915 filter in a non-root, non-exim process.
4921 1. The buildconfig auxiliary program wasn't quoting the value set for
4922 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4923 not defined. This bug was masked in 4.22 by the effect that was fixed in
4926 2. Some messages that were rejected after a message id was allocated were
4927 shown as "incomplete" by exigrep. It no longer does this for messages that
4928 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4930 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4931 have allow_domain_literals set, the ID did not get logged in the <= line.
4932 Domain literals are now always recognized in Message-ID: header lines.
4934 4. The first argument for a ${extract expansion item is the key name or field
4935 number. Leading and trailing spaces in this item were not being ignored,
4936 causing some misleading effects.
4938 5. When deliver_drop_privilege was set, single queue runner processes started
4939 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4940 same command in the process it spins off) were not dropping privilege.
4942 6. When the daemon running as "exim" started a queue runner, it always
4943 re-executed Exim in the spun-off process. This is a waste of effort when
4944 deliver_drop_privilege is set. The new process now just calls the
4945 queue-runner function directly.
4951 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4954 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4955 ignored. Though the use of -oP was forcing the writing of a pid file, it
4956 was always written to the default place.
4958 3. If the message "no IP address found for host xxxx" is generated during
4959 incoming verification, it is now followed by identification of the incoming
4960 connection (so you can more easily find what provoked it).
4962 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4964 5. Added some features to "harden" Exim a bit more against certain attacks:
4966 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4967 be put in Local/Makefile. This is like the never_users runtime option,
4968 but it cannot be overridden. The default setting is "root".
4970 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4971 prefix string with which any file named in a -C command line option
4974 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4975 is retained for -C and -D only if the caller of Exim is root. Without
4976 it, the exim user may also use -C and -D and retain privilege.
4978 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4979 command line option is disabled.
4981 6. Macro names set by the -D option must start with an upper case letter, just
4982 like macro names defined in the configuration file.
4984 7. Added "dereference=" facility to LDAP.
4986 8. Two instances of the typo "uknown" in the source files are fixed.
4988 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4989 the Configure-Makefile script screwed up while processing it.
4991 10. Incorporated PCRE 4.4.
4993 11. The SMTP synchronization check was not operating right at the start of an
4994 SMTP session. For example, it could not catch a HELO sent before the client
4995 waited for the greeting. There is now a check for outstanding input at the
4996 point when the greeting is written. Because of the duplex, asynchronous
4997 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4998 way, but not yet received, when the check is performed.
5000 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5001 on TCP/IP sockets, because this apparently causes some broken clients to
5004 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5005 unchanged) from the Cygwin maintainer.
5007 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5008 was supported instead of testing for mbx. Effectively a typo.
5010 15. The spa authenticator server code was not checking that the input it
5011 received was valid base64.
5013 16. The debug output line for the "set" modifier in ACLs was not showing the
5014 name of the variable that was being set.
5016 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5018 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5019 lookup on its own. Something else was needed to trigger the lookup. For
5020 example, a match in host_lookup or the need for a host name in a host list.
5021 Now, if $sender_host_name is referenced and the host name has not yet been
5022 looked up, a lookup is performed. If the lookup fails, the variable remains
5023 empty, and $host_lookup_failed is set to "1".
5025 19. Added "eqi" as a case-independent comparison operator.
5027 20. The saslauthd authentication condition could segfault if neither service
5028 nor realm was specified.
5030 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5031 error message that was logged was misleading, and incoming SMTP
5032 connections were dropped. The message is now more accurate, and temporary
5033 errors are given to SMTP connections.
5035 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5036 (except RSET) with a 421 error, until QUIT is received. However, it was
5037 failing to send a response to QUIT.
5039 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5040 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5041 ACL is now run after the helo_try_verify_hosts code.
5043 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5044 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5045 case-independent, so other case variants are also recognized). Apparently
5046 some systems use these upper case variants.
5048 25. If more than two messages were waiting for the same host, and a transport
5049 filter was specified for the transport, Exim sent two messages over the
5050 same TCP/IP connection, and then failed with "socket operation on non-
5051 socket" when it tried to send the third.
5053 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5055 27. The extern definition of crypt16() in expand.c was not being excluded when
5056 the OS had its own crypt16() function.
5058 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5059 as a preferred synonym for return_size_limit, both as an option and as an
5062 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5064 30. Changed the default configuration ACL to relax the local part checking rule
5065 for addresses that are not in any local domains. For these addresses,
5066 slashes and pipe symbols are allowed within local parts, but the sequence
5067 /../ is explicitly forbidden.
5069 31. SPA server authentication was not clearing the challenge buffer before
5072 32. log_message in a "warn" ACL statement was writing to the reject log as
5073 well as to the main log, which contradicts the documentation and doesn't
5074 seem right (because no rejection is happening). So I have stopped it.
5076 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5077 However, I am unable to do any testing of this.
5079 34. Fixed an infelicity in the appendfile transport. When checking directories
5080 for a mailbox, to see if any needed to be created, it was accidentally
5081 using path names with one or more superfluous leading slashes; tracing
5082 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5084 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5085 contents are added to the log line that is written for every discarded
5086 recipient. (Previously a log_message setting was ignored.)
5088 36. The ${quote: operator now quotes the string if it is empty.
5090 37. The install script runs exim in order to find its version number. If for
5091 some reason other than non-existence or emptiness, which it checks, it
5092 could not run './exim', it was installing it with an empty version number,
5093 i.e. as "exim-". This error state is now caught, and the installation is
5096 38. An argument was missing from the function that creates an error message
5097 when Exim fails to connect to the socket for saslauthd authentication.
5098 This could cause Exim to crash, or give a corrupted message.
5100 39. Added isip, isip4, and isip6 to ${if conditions.
5102 40. The ACL variables $acl_xx are now saved with the message, and can be
5103 accessed later in routers, transports, and filters.
5105 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5106 strings in the file are not string-expanded.
5108 42. If a MAIL command specified a SIZE value that was too large to fit into an
5109 int variable, the check against message_size_limit failed. Such values are
5110 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5111 day this will have to be increased, but I don't think I want to be around
5112 when emails are that large.
5119 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5120 iconv() is not standard in FreeBSD.
5122 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5123 IPv6 enabled. The observed symptom was a segmentation fault on return from
5124 the function os_common_find_running_interfaces() in src/os.c.
5126 3. In the check_special_case() function in daemon.c I had used "errno" as an
5127 argument name, which causes warnings on some systems. This was basically a
5128 typo, since it was named "eno" in the comments!
5130 4. The code that waits for the clock to tick (at a resolution of some fraction
5131 of a second) so as to ensure message-id uniqueness was always waiting for
5132 at least one whole tick, when it could have waited for less. [This is
5133 almost certainly not relevant at current processor speeds, where it is
5134 unlikely to ever wait at all. But we try to future-proof.]
5136 5. The function that sleeps for a time interval that includes fractions of a
5137 second contained a race. It did not block SIGALRM between setting the
5138 timer, and suspending (a couple of lines later). If the interval was short
5139 and the sigsuspend() was delayed until after it had expired, the suspension
5140 never ended. On busy systems this could lead to processes getting stuck for
5143 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5144 process, before it forks any delivery processes. The open lookup caching
5145 mechanism meant that the open file or database connection was passed into
5146 the delivery process. The problem was that delivery processes always tidy
5147 up cached lookup data. This could cause a problem for the next delivery
5148 process started by the queue runner, because the external queue runner
5149 process does not know about the closure. So the next delivery process
5150 still has data in the lookup cache. In the case of a file lookup, there was
5151 no problem because closing a file descriptor in a subprocess doesn't affect
5152 the parent. However, if the lookup was caching a connection to a database,
5153 the connection was closed, and the second delivery process was likely to
5154 see errors such as "PGSQL: query failed: server closed the connection
5155 unexpectedly". The problem has been fixed by closing all cached lookups
5156 in a queue runner before running a delivery process.
5158 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5159 seem to have the "const" qualifier which it has on other OS. I've
5162 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5163 *of the same type* (client or server) with the same public name that an
5164 error should be diagnosed.
5166 9. When Exim looked up a host name for an IP address, but failed to find the
5167 original IP address when looking up the host name (a safety check), it
5168 output the message "<ip address> does not match any IP for NULL", which was
5169 confusing, to say the least. The bug was that the host name should have
5170 appeared instead of "NULL".
5172 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5173 user that is built into the binary, and the -C or -D options is used, root
5174 privilege is dropped before the configuration file is read. In addition,
5175 logging is switched to stderr instead of the normal log files. If the
5176 configuration then re-defines the Exim user, the unprivileged environment
5177 is probably not what is expected, so Exim logs a panic warning message (but
5180 However, if deliver_drop_privilege is set, the unprivileged state may well
5181 be exactly what is intended, so the warning has been cut out in that case,
5182 and Exim is allowed to try to write to its normal log files.
5188 1. smtp_return_error_details was not giving details for temporary sender
5189 or receiver verification errors.
5191 2. Diagnose a configuration error if two authenticators have the same public
5194 3. Exim used not to create the message log file for a message until the first
5195 delivery attempt. This could be confusing when incoming messages were held
5196 for policy or load reasons. The message log file is now created at the time
5197 the message is received, and an initial "Received" line is written to it.
5199 4. The automatically generated man page for command line options had a minor
5200 bug that caused no ill effects; however, a more serious problem was that
5201 the procedure for building the man page automatically didn't always
5202 operate. Consequently, release 4.20 contains an out-of-date version. This
5203 shouldn't happen again.
5205 5. When building Exim with embedded Perl support, the script that builds the
5206 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5207 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5209 6. The freeze_tell option was not being used for messages that were frozen on
5210 arrival, either by an ACL or by local_scan().
5212 7. Added the smtp_incomplete_transaction log selector.
5214 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5215 was accepting AUTH without a new EHLO.
5217 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5218 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5221 10. Logging of TCP/IP connections (when configured) now happens in the main
5222 daemon process instead of the child process, so that the TCP/IP connection
5223 count is more accurate (but it can never be perfect).
5225 11. The use of "drop" in a nested ACL was not being handled correctly in the
5226 outer ACL. Now, if condition failure induced by the nested "drop" causes
5227 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5228 or "require"), the connection is dropped.
5230 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5231 that yield "discard" can now be used with an "accept" or a "discard" verb,
5232 but an error is generated for any others (because I can't see a useful way
5233 to define what should happen).
5235 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5236 are now processed in the same way as lines in the Exim configuration file.
5237 In particular, continuation lines are supported.
5239 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5241 15. Added -ti meaning -t -i.
5243 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5244 domains, and warn by logging if others are found.
5246 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5247 returned from GIFCONF when Exim is trying to find the list of interfaces on
5248 a host. The code in os.c has been modified to copy each ifreq to an aligned
5249 structure in all cases.
5251 Also, in some cases, the returned ifreq's were being copied to a 'struct
5252 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5253 means the last couple of bytes of an IPv6 address could be chopped if the
5254 ifreq contained only a normal sockaddr (14 bytes storage).
5256 18. Named domain lists were not supported in the hosts_treat_as_local option.
5257 An entry such as +xxxx was not recognized, and was treated as a literal
5260 19. Ensure that header lines added by a DATA ACL are included in the reject log
5261 if the ACL subsequently rejects the message.
5263 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5264 MD5 (which is deprecated).
5266 21. When testing a filter file using -bf, Exim was writing a message when it
5267 took the sender from a "From " line in the message, but it was not doing so
5268 when it took $return_path from a Return-Path: header line. It now does.
5270 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5271 with a valid header line field name (a series of printing characters
5272 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5274 23. Changed "disc" in the source to "disk" to conform to the documentation and
5275 the book and for uniformity.
5277 24. Ignore Sendmail's -Ooption=value command line item.
5279 25. When execve() failed while trying to run a command in a pipe transport,
5280 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5281 could be confused with a return value of 69 from the command itself. This
5282 has been changed to 127, the value the shell returns if it is asked to run
5283 a non-existent command. The wording for the related log line suggests a
5284 non-existent command as the problem.
5286 26. If received_header_text expands to an empty string, do not add a Received:
5287 header line to the message. (Well, it adds a token one on the spool, but
5288 marks it "old" so that it doesn't get used or transmitted.)
5290 27. Installed eximstats 1.28 (addition of -nt option).
5292 28. There was no check for failure on the call to getsockname() in the daemon
5293 code. This can fail if there is a shortage of resources on the system, with
5294 ENOMEM, for example. A temporary error is now given on failure.
5296 29. Contrary to the C standard, it seems that in some environments, the
5297 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5298 program. Exim now does this explicitly; it affects the formatting of
5299 timestamps using strftime().
5301 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5302 complaints. I've now initialized all the variables, to avoid this.
5304 32. Header lines added by a system filter were not being "seen" during
5305 transport-time rewrites.
5307 33. The info_callback() function passed to OpenSSL is set up with type void
5308 (*)(SSL *, int, int), as described somewhere. However, when calling the
5309 function (actually a macro) that sets it up, the type void(*)() is
5310 expected. I've put in a cast to prevent warnings from picky compilers.
5312 34. If a DNS black list lookup found a CNAME record, but there were no A
5313 records associated with the domain it pointed at, Exim crashed.
5315 35. If a DNS black list lookup returned more than one A record, Exim ignored
5316 all but the first. It now scans all returned addresses if a particular IP
5317 value is being sought. In this situation, the contents of the
5318 $dnslist_value variable are a list of all the addresses, separated by a
5321 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5322 to accept a host name and all its aliases if the forward lookup for any of
5323 them yielded the IP address of the incoming connection. Now it accepts only
5324 those names whose forward lookup yields the correct IP address. Any other
5325 names are discarded. This closes a loophole whereby a rogue DNS
5326 administrator could create reverse DNS records to break through a
5327 wildcarded host restriction in an ACL.
5329 37. If a user filter or a system filter that ran in a subprocess used any of
5330 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5331 the wrong values were passed to the pipe command ($thisaddress had the
5332 value of $0, $0 had the value of $1, etc). This bug was introduced by
5333 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5335 38. Improved the line breaking for long SMTP error messages from ACLs.
5336 Previously, if there was no break point between 40 and 75 characters, Exim
5337 left the rest of the message alone. Two changes have been made: (a) I've
5338 reduced the minimum length to 35 characters; (b) if it can't find a break
5339 point between 35 and 75 characters, it looks ahead and uses the first one
5340 that it finds. This may give the occasional overlong line, but at least the
5341 remaining text gets split now.
5343 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5344 file descriptors might be low, and that setting 1000 would always raise it.
5345 It turns out that in some environments, the limit is already over 1000 and
5346 that lowering it causes trouble. So now Exim takes care not to decrease it.
5348 40. When delivering a message, the value of $return_path is set to $sender_
5349 address at the start of routing (routers may change the value). By an
5350 oversight, this default was not being set up when an address was tested by
5351 -bt or -bv, which affected the outcome if any router or filter referred to
5354 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5355 log only when "message" or "log_message" are set. However, if one of the
5356 conditions was an address verification, or a call to a nested ACL, the
5357 messages generated by the underlying test were being passed through. This
5358 no longer happens. The underlying message is available in $acl_verify_
5359 message for both "message" and "log_message" expansions, so it can be
5360 passed through if needed.
5362 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5363 new expansion $bh_ to give the encoded byte string without charset
5364 translation. Translation happens only if iconv() is available; HAVE_ICONV
5365 indicates this at build time. HEADERS_CHARSET gives the charset to
5366 translate to; headers_charset can change it in the configuration, and
5367 "headers charset" can change it in an individual filter file.
5369 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5370 that creates RFC 2047 encoded "words" labels them as that charset instead
5371 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5372 expansion operator; (ii) when Exim creates a From: line for a local
5373 message; (iii) when a header line is rewritten to include a "phrase" part.
5375 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5376 buggy, causing it to skip the first lines of messages whose message ID
5377 ended in 'D'. This would not have bitten before Exim release 4.14, because
5378 message IDs were unlikely to end in 'D' before then. The effect was to have
5379 incorrect size information for certain domains.
5381 45. #include "config.h" was missing at the start of the crypt16.c module. This
5382 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5385 46. If there was a timeout during a "random" callout check, Exim treated it as
5386 a failure of the random address, and carried on sending RSET and the real
5387 address. If the delay was just some slowness somewhere, the response to the
5388 original RCPT would be taken as a response to RSET and so on, causing
5389 mayhem of various kinds.
5391 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5392 when I implemented it. It didn't allow for the fact that some option values
5393 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5394 the right test for positive values.
5396 48. Domain names in DNS records are case-independent. Exim always looks them up
5397 in lower case. Some resolvers return domain names in exactly the case they
5398 appear in the zone file, that is, they may contain uppercase letters. Not
5399 all resolvers do this - some return always lower case. Exim was treating a
5400 change of case by a resolver as a change of domain, similar to a widening
5401 of a domain abbreviation. This triggered its re-routing code and so it was
5402 trying to route what was effectively the same domain again. This normally
5403 caused routing to fail (because the router wouldn't handle the domain
5404 twice). Now Exim checks for this case specially, and just changes the
5405 casing of the domain that it ultimately uses when it transmits the message
5408 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5411 50. If a filter generated a file delivery with a non-absolute name (possible if
5412 no home directory exists for the router), the forbid_file option was not
5415 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5416 the existing equality matching.
5418 52. Exim was using ints instead of ino_t variables in some places where it was
5419 dealing with inode numbers.
5421 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5422 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5423 called TMPDIR, and if it finds it is different, it changes its value.
5425 54. The smtp_printf() function is now made available to local_scan() so
5426 additional output lines can be written before returning. There is also an
5427 smtp_fflush() function to enable the detection of a dropped connection.
5428 The variables smtp_input and smtp_batched_input are exported to
5431 55. Changed the default runtime configuration: the message "Unknown user"
5432 has been removed from the ACL, and instead placed on the localuser router,
5433 using the cannot_route_message feature. This means that any verification
5434 failures that generate their own messages won't get overridden. Similarly,
5435 the "Unrouteable address" message that was in the ACL for unverifiable
5436 relay addresses has also been removed.
5438 56. Added hosts_avoid_esmtp to the smtp transport.
5440 57. The exicyclog script was not checking for the esoteric option
5441 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5442 will work only if exicyclog is run under the appropriate euid.
5444 58. Following a discussion on the list, the rules by which Exim recognises line
5445 endings on incoming messages have been changed. The -dropcr and drop_cr
5446 options are now no-ops, retained only for backwards compatibility. The
5447 following line terminators are recognized: LF CRLF CR. However, special
5448 processing applies to CR:
5450 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5451 nor a local message in the state where . is a terminator.
5453 (ii) If a bare CR is encountered in a header line, an extra space is added
5454 after the line terminator so as not to end the header. The reasoning
5455 behind this is that bare CRs in header lines are most likely either
5456 to be mistakes, or people trying to play silly games.
5458 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5459 was being incorrectly given as 18 bytes larger than it should have been.
5460 This is a VOB (very old bug).
5462 60. This may never have affected anything current, but just in case it has:
5463 When the local host is found other than at the start of a list of hosts,
5464 the local host, those with the same MX, and any that follow, are discarded.
5465 When the list in question was part of a longer list of hosts, the following
5466 hosts (not currently being processed) were also being discarded. This no
5467 longer happens. I'm not sure if this situation could ever has previously
5470 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5473 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5474 header. This is something that is recommended in a new Internet Draft, and
5475 is something that is documented as being done by Sendmail. There are two
5476 possible values. For messages generated by the autoreply transport, Exim
5479 Auto-Submitted: auto-replied
5481 whereas for all other generated messages (e.g. bounces) it adds
5483 Auto-Submitted: auto-generated
5485 63. The "personal" condition in filters now includes a test for the
5486 Auto-Submitted: header. If it contains the string "auto-" the message it
5487 not considered personal.
5489 64. Added rcpt_include_affixes as a generic transport option.
5491 65. Added queue_only_override (default true).
5493 66. Added the syslog_duplication option.
5495 67. If what should have been the first header line of a message consisted of
5496 a space followed by a colon, Exim was mis-interpreting it as a header line.
5497 It isn't of course - it is syntactically invalid and should therefore be
5498 treated as the start of the message body. The misbehaviour could have
5499 caused a number of strange effects, including loss of data in subsequent
5500 header lines, and spool format errors.
5502 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5503 client host had authenticated. This control can now be exercised by an ACL
5504 for more flexibility.
5506 69. By default, callouts do not happen when testing with -bh. There is now a
5507 variant, -bhc, which does actually run the callout code, including
5508 consulting and updating the callout cache.
5510 70. Added support for saslauthd authentication, courtesy of Alexander
5513 71. If statvfs() failed on the spool or log directories while checking their
5514 size for availability, Exim confusingly gave the error "space shortage".
5515 Furthermore, in debugging mode it crashed with a floating point exception.
5516 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5517 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5518 serious problem, Exim now writes to the main and panic logs when this
5519 happens, with details of the failure. It then refuses to accept the
5520 incoming message, giving the message "spool directory problem" or "log
5521 directory problem" with a 421 code for SMTP messages.
5523 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5524 0, 1, and 2 exist, because some OS complain for execs without them (see
5525 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5526 descriptors. However, the code omitted to check that the open succeeded,
5527 causing mysterious errors if for some reason the permissions on /dev/null
5528 got screwed. Now Exim writes a message to the main and panic logs, and
5529 bombs out if it can't open /dev/null.
5531 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5532 interact so that it is all more flexible. It is supposed to remain
5533 backwards compatible. Also added extra_local_interfaces.
5535 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5536 to bomb out with an assertion failure - to the client this appears as a
5537 connection drop. This problem occurs in the part of the code that was taken
5538 from the Samba project. Fortunately, the assertion is in a very simple
5539 function, so I have fixed this by reproducing the function inline in the
5540 one place where it is called, and arranging for authentication to fail
5541 instead of killing the process with assert().
5543 75. The SPA client code was not working when the server requested OEM rather
5544 than Unicode encoding.
5546 76. Added code to make require_files with a specific uid setting more usable in
5547 the case where statting the file as root fails - usually a non-root-mounted
5548 NFS file system. When this happens and the failure is EACCES, Exim now
5549 forks a subprocess and does the per-uid checking as the relevant uid.
5551 77. Added process_log_path.
5553 78. If log_file_path was not explicitly set, a setting of check_log_space or
5554 check_log_inodes was ignored.
5556 79. If a space check for the spool or log partitions fails, the incident is now
5557 logged. Of course, in the latter case the data may get lost...
5559 80. Added the %p formatting code to string_format() so that it can be used to
5560 print addresses in debug_print(). Adjusted all the address printing in the
5561 debugging in store.c to use %p rather than %d.
5563 81. There was a concern that a line of code in smtp_in.c could overflow a
5564 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5565 initially expressed, the concern was not well-founded, because trailing
5566 spaces are removed early. However, if the trailing spaces were followed by
5567 a NULL, they did not get removed, so the overflow was possible. Two fixes
5570 (a) I re-wrote the offending code in a cleaner fashion.
5571 (b) If an incoming SMTP command contains a NULL character, it is rejected
5574 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5575 runs initgroups(), so that if the Exim user is in any additional groups,
5576 they will be used during message reception.
5582 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob