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 not 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 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
97 locate the relevant includes and libraries. Made this the default.
99 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
100 Bugzilla 1246, report and most of solution from Tomasz Kusy.
102 JH/02 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
103 This may cause build issues on older platforms.
105 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
106 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
107 gnutls_require_mac & gnutls_require_protocols (no longer supported).
108 Added SNI support via GnuTLS too.
110 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
112 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
118 PP/01 Solaris build fix for Oracle's LDAP libraries.
119 Bugzilla 1109, patch from Stephen Usher.
121 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
123 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
126 TF/02 Fix a couple more cases where we did not log the error message
127 when unlink() failed. See also change 4.74-TF/03.
129 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
130 lock up or crash if it happened to be inside a call to libc when it
131 got a SIGUSR1 from exiwhat.
133 The SIGUSR1 handler appends the current process status to the process
134 log which is later printed by exiwhat. It used to use the general
135 purpose logging code to do this, but several functions it calls are
136 not safe for signals.
138 The new output code in the SIGUSR1 handler is specific to the process
139 log, and simple enough that it's easy to inspect for signal safety.
140 Removing some special cases also simplifies the general logging code.
141 Removing the spurious timestamps from the process log simplifies
144 TF/04 Improved ratelimit ACL condition.
146 The /noupdate option has been deprecated in favour of /readonly which
147 has clearer semantics. The /leaky, /strict, and /readonly update modes
148 are mutually exclusive. The update mode is no longer included in the
149 database key; it just determines when the database is updated. (This
150 means that when you upgrde Exim will forget old rate measurements.)
152 Exim now checks that the per_* options are used with an update mode that
153 makes sense for the current ACL. For example, when Exim is processing a
154 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
155 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
156 must specify per_mail/readonly. If you omit the update mode it defaults to
157 /leaky where that makes sense (as before) or /readonly where required.
159 The /noupdate option is now undocumented but still supported for
160 backwards compatibility. It is equivalent to /readonly except that in
161 ACLs where /readonly is required you may specify /leaky/noupdate or
162 /strict/noupdate which are treated the same as /readonly.
164 A useful new feature is the /count= option. This is a generalization
165 of the per_byte option, so that you can measure the throughput of other
166 aggregate values. For example, the per_byte option is now equivalent
167 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
169 The per_rcpt option has been generalized using the /count= mechanism
170 (though it's more complicated than the per_byte equivalence). When it is
171 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
172 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
173 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
174 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
175 Note that using per_rcpt with a non-readonly update mode in more than
176 one ACL will cause the recipients to be double-counted. (The per_mail
177 and per_byte options don't have this problem.)
179 The handling of very low rates has changed slightly. If the computed rate
180 is less than the event's count (usually one) then this event is the first
181 after a long gap. In this case the rate is set to the same as this event's
182 count, so that the first message of a spam run is counted properly.
184 The major new feature is a mechanism for counting the rate of unique
185 events. The new per_addr option counts the number of different
186 recipients that someone has sent messages to in the last time period. It
187 behaves like per_rcpt if all the recipient addresses are different, but
188 duplicate recipient addresses do not increase the measured rate. Like
189 the /count= option this is a general mechanism, so the per_addr option
190 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
191 example, measure the rate that a client uses different sender addresses
192 with the options per_mail/unique=$sender_address. There are further
193 details in the main documentation.
195 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
197 TF/06 Removed a few PCRE remnants.
199 TF/07 Automatically extract Exim's version number from tags in the git
200 repository when doing development or release builds.
202 PP/02 Raise smtp_cmd_buffer_size to 16kB.
203 Bugzilla 879. Patch from Paul Fisher.
205 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
206 Heavily based on revision 40f9a89a from Simon Arlott's tree.
209 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
211 PP/05 Variable $av_failed, true if the AV scanner deferred.
212 Bugzilla 1078. Patch from John Horne.
214 PP/06 Stop make process more reliably on build failure.
215 Bugzilla 1087. Patch from Heiko Schlittermann.
217 PP/07 Make maildir_use_size_file an _expandable_ boolean.
218 Bugzilla 1089. Patch from Heiko Schlittermann.
220 PP/08 Handle ${run} returning more data than OS pipe buffer size.
221 Bugzilla 1131. Patch from Holger Weiß.
223 PP/09 Handle IPv6 addresses with SPF.
224 Bugzilla 860. Patch from Wolfgang Breyha.
226 PP/10 GnuTLS: support TLS 1.2 & 1.1.
228 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
231 PP/11 match_* no longer expand right-hand-side by default.
232 New compile-time build option, EXPAND_LISTMATCH_RHS.
233 New expansion conditions, "inlist", "inlisti".
235 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
237 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
239 PP/14 fix log_write() format string regression from TF/03.
240 Bugzilla 1152. Patch from Dmitry Isaikin.
246 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
248 PP/02 Harmonised TLS library version reporting; only show if debugging.
249 Layout now matches that introduced for other libraries in 4.74 PP/03.
251 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
253 PP/04 New "dns_use_edns0" global option.
255 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
258 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
259 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
261 TK/01 Updated PolarSSL code to 0.14.2.
262 Bugzilla 1097. Patch from Andreas Metzler.
264 PP/07 Catch divide-by-zero in ${eval:...}.
267 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
270 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
271 format-string attack -- SECURITY: remote arbitrary code execution.
273 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
274 time unintentionally subject to list matching rules, letting the header
275 cause arbitrary Exim lookups (of items which can occur in lists, *not*
276 arbitrary string expansion). This allowed for information disclosure.
278 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
279 INT_MIN/-1 -- value coerced to INT_MAX.
285 NM/01 Workround for PCRE version dependency in version reporting
288 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
289 This fixes portability to compilers other than gcc, notably
290 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
292 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
293 makefiles for portability to HP-UX and POSIX correctness.
295 PP/01 Permit LOOKUP_foo enabling on the make command-line.
296 Also via indented variable definition in the Makefile.
297 (Debugging by Oliver Heesakkers).
299 PP/02 Restore caching of spamd results with expanded spamd_address.
300 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
302 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
303 Improves build reliability. Fix from: Frank Elsner
305 NM/02 Fix wide character breakage in the rfc2047 coding
306 Fixes bug 1064. Patch from Andrey N. Oktyabrski
308 NM/03 Allow underscore in dnslist lookups
309 Fixes bug 1026. Patch from Graeme Fowler
311 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
312 Code patches from Adam Ciarcinski of NetBSD.
314 NM/04 Fixed exiqgrep to cope with mailq missing size issue
317 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
318 is logged, to avoid truncation. Patch from John Horne.
320 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
321 Patch from Jakob Hirsch.
323 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
324 SQL string expansion failure details.
325 Patch from Andrey Oktyabrski.
327 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
328 Patch from Simon Arlott.
330 PP/09 New lookups functionality failed to compile on old gcc which rejects
331 extern declarations in function scope.
332 Patch from Oliver Fleischmann
334 PP/10 Use sig_atomic_t for flags set from signal handlers.
335 Check getgroups() return and improve debugging.
336 Fixed developed for diagnosis in bug 927 (which turned out to be
339 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
340 Patch from Mark Zealey.
342 PP/12 Bugzilla 1056: Improved spamd server selection.
343 Patch from Mark Zealey.
345 PP/13 Bugzilla 1086: Deal with maildir quota file races.
346 Based on patch from Heiko Schlittermann.
348 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
349 Patch from Uwe Doering, sign-off by Michael Haardt.
351 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
352 variable declaration deep within a block. Bug and patch from
355 PP/15 lookups-Makefile IRIX compatibilty coercion.
357 PP/16 Make DISABLE_DKIM build knob functional.
359 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
360 Patch by Simon Arlott
362 TF/03 Fix valgrind.h portability to C89 compilers that do not support
363 variable argument macros. Our copy now differs from upstream.
369 TF/01 Failure to get a lock on a hints database can have serious
370 consequences so log it to the panic log.
372 TF/02 Log LMTP confirmation messages in the same way as SMTP,
373 controlled using the smtp_confirmation log selector.
375 TF/03 Include the error message when we fail to unlink a spool file.
377 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
378 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
379 for maintaining out-of-tree patches for some time.
381 PP/01 Bugzilla 139: Documentation and portability issues.
382 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
383 Handle per-OS dynamic-module compilation flags.
385 PP/02 Let /dev/null have normal permissions.
386 The 4.73 fixes were a little too stringent and complained about the
387 permissions on /dev/null. Exempt it from some checks.
388 Reported by Andreas M. Kirchwitz.
390 PP/03 Report version information for many libraries, including
391 Exim version information for dynamically loaded libraries. Created
392 version.h, now support a version extension string for distributors
393 who patch heavily. Dynamic module ABI change.
395 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
396 privilege escalation vulnerability whereby the Exim run-time user
397 can cause root to append content of the attacker's choosing to
400 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
403 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
404 If dropping privileges for untrusted macros, we disabled normal logging
405 on the basis that it would fail; for the Exim run-time user, this is not
406 the case, and it resulted in successful deliveries going unlogged.
407 Fixed. Reported by Andreas Metzler.
413 PP/01 Date: & Message-Id: revert to normally being appended to a message,
414 only prepend for the Resent-* case. Fixes regression introduced in
415 Exim 4.70 by NM/22 for Bugzilla 607.
417 PP/02 Include check_rfc2047_length in configure.default because we're seeing
418 increasing numbers of administrators be bitten by this.
420 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
422 PP/03 Bugzilla 994: added openssl_options main configuration option.
424 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
426 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
428 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
430 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
431 without a peer certificate, leading to a segfault because of an
432 assumption that peers always have certificates. Be a little more
433 paranoid. Problem reported by Martin Tscholak.
435 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
436 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
437 NB: ClamAV planning to remove STREAM in "middle of 2010".
438 CL also introduces -bmalware, various -d+acl logging additions and
439 more caution in buffer sizes.
441 PP/09 Implemented reverse_ip expansion operator.
443 PP/10 Bugzilla 937: provide a "debug" ACL control.
445 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
447 PP/12 Bugzilla 973: Implement --version.
449 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
451 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
453 PP/15 Bugzilla 816: support multiple condition rules on Routers.
455 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
456 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
457 ignore trailing whitespace.
459 JJ/02 prevent non-panic DKIM error from being sent to paniclog
461 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
464 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
465 Notification from Dr Andrew Aitchison.
467 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
468 ExtendedDetectionInfo response format.
469 Notification from John Horne.
471 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
474 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
475 XSL and documented dependency on system catalogs, with examples of how
478 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
481 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
482 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
483 configuration file which is writeable by the Exim user or group.
485 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
486 of configuration files to cover files specified with the -C option if
487 they are going to be used with root privileges, not just the default
490 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
491 option (effectively making it always true).
493 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
494 files to be used while preserving root privileges.
496 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
497 that rogue child processes cannot use them.
499 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
500 run-time user, instead of root.
502 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
503 Exim run-time user without dropping privileges.
505 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
506 result string, instead of calling string_vformat() twice with the same
509 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
510 for other users. Others should always drop root privileges if they use
511 -C on the command line, even for a whitelisted configure file.
513 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
515 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
521 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
522 $data_path, and $header_path variables; fixed documentation bugs and
525 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
526 exipick to access non-standard spools, including the "frozen" queue
529 NM/01 Bugzilla 965: Support mysql stored procedures.
530 Patch from Alain Williams
532 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
534 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
535 Patch from Andreas Metzler
537 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
538 Patch from Kirill Miazine
540 NM/05 Bugzilla 671: Added umask to procmail example.
542 JJ/03 installed exipick 20100323.0, fixing doc bug
544 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
545 directory. Notification and patch from Dan Rosenberg.
547 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
549 TK/02 Improve log output when DKIM signing operation fails.
551 MH/01 Treat the transport option dkim_domain as a colon separated
552 list, not as a single string, and sign the message with each element,
553 omitting multiple occurences of the same signer.
555 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
556 Bugzilla 985, 986. Patch by Simon Arlott
558 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
559 Patch by Simon Arlott
561 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
562 MBX locking. Notification from Dan Rosenberg.
568 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
570 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
572 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
574 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
576 NM/04 Bugzilla 915: Flag broken perl installation during build.
582 TK/01 Added patch by Johannes Berg that expands the main option
583 "spamd_address" if it starts with a dollar sign.
585 TK/02 Write list of recipients to X-Envelope-Sender header when building
586 the mbox-format spool file for content scanning (suggested by Jakob
589 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
590 (http://www.dcc-servers.net/) support via dccifd. Activated by
591 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
593 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
594 by Mark Daniel Reidel <mr@df.eu>.
596 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
597 When building exim an external PCRE library is now needed -
598 PCRE is a system library on the majority of modern systems.
599 See entry on PCRE_LIBS in EDITME file.
601 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
602 conversation. Added nologin parameter to request.
603 Patch contributed by Kirill Miazine.
605 TF/01 Do not log submission mode rewrites if they do not change the address.
607 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
609 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
610 log files in place. Contributed by Roberto Lima.
612 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
614 TF/03 Bugzilla 615: When checking the local_parts router precondition
615 after a local_part_suffix or local_part_prefix option, Exim now
616 does not use the address's named list lookup cache, since this
617 contains cached lookups for the whole local part.
619 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
620 Robert Millan. Documentation is in experimental-spec.txt.
622 TF/04 Bugzilla 668: Fix parallel build (make -j).
624 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
626 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
627 Patch provided by Jan Srzednicki.
629 TF/05 Leading white space used to be stripped from $spam_report which
630 wrecked the formatting. Now it is preserved.
632 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
633 that they are available at delivery time.
635 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
637 TF/08 TLS error reporting now respects the incoming_interface and
638 incoming_port log selectors.
640 TF/09 Produce a more useful error message if an SMTP transport's hosts
641 setting expands to an empty string.
643 NM/06 Bugzilla 744: EXPN did not work under TLS.
644 Patch provided by Phil Pennock.
646 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
647 Patch provided by Richard Godbee.
649 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
650 acl_smtp_notquit, added index entry.
652 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
653 Patch provided by Eugene Bujak.
655 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
656 accept(). Patch provided by Maxim Dounin.
658 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
659 Patch provided by Phil Pennock.
661 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
663 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
664 Patch provided by Brad "anomie" Jorsch.
666 NM/14 Bugzilla 622: Added timeout setting to transport filter.
667 Patch provided by Dean Brooks.
669 TK/05 Add native DKIM support (does not depend on external libraries).
671 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
672 Patch provided by Graeme Fowler.
674 NM/16 Bugzilla 851: Documentation example syntax fix.
676 NM/17 Changed NOTICE file to remove references to embedded PCRE.
678 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
681 NM/19 Bugzilla 745: TLS version reporting.
682 Patch provided by Phil Pennock.
684 NM/20 Bugzilla 167: bool: condition support.
685 Patch provided by Phil Pennock.
687 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
688 clients. Patch provided by Phil Pennock.
690 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
691 Patch provided by Brad "anomie" Jorsch.
693 NM/23 Bugzilla 687: Fix misparses in eximstats.
694 Patch provided by Heiko Schlittermann.
696 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
697 Patch provided by Heiko Schlittermann.
699 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
700 plus update to original patch.
702 NM/26 Bugzilla 799: Documentation correction for ratelimit.
704 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
705 Patch provided by David Brownlee.
707 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
709 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
711 NM/30 Bugzilla 888: TLS documentation bugfixes.
713 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
715 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
716 Unlike the original bugzilla I have changed all shell scripts in src tree.
718 NM/33 Bugzilla 898: Transport filter timeout fix.
719 Patch by Todd Rinaldo.
721 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
722 Patch by Serge Demonchaux.
724 NM/35 Bugzilla 39: Base64 decode bug fixes.
725 Patch by Jakob Hirsch.
727 NM/36 Bugzilla 909: Correct connect() call in dcc code.
729 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
731 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
733 NM/39 Bugzilla 911: Fixed MakeLinks build script.
739 TK/01 Add preliminary DKIM support. Currently requires a forked version of
740 ALT-N's libdkim that I have put here:
741 http://duncanthrax.net/exim-experimental/
743 Note to Michael Haardt: I had to rename some vars in sieve.c. They
744 were called 'true' and it seems that C99 defines that as a reserved
745 keyword to be used with 'bool' variable types. That means you could
746 not include C99-style headers which use bools without triggering
747 build errors in sieve.c.
749 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
750 as mailq or other aliases. Changed the --help handling significantly
751 to do whats expected. exim_usage() emits usage/help information.
753 SC/01 Added the -bylocaldomain option to eximstats.
755 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
757 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
759 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
765 PH/01 Another patch from the Sieve maintainer.
767 PH/02 When an IPv6 address is converted to a string for single-key lookup
768 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
769 dots are used instead of colons so that keys in lsearch files need not
770 contain colons. This was done some time before quoting was made available
771 in lsearch files. However, iplsearch files do require colons in IPv6 keys
772 (notated using the quote facility) so as to distinguish them from IPv4
773 keys. This meant that lookups for IP addresses in host lists did not work
774 for iplsearch lookups.
776 This has been fixed by arranging for IPv6 addresses to be expressed with
777 colons if the lookup type is iplsearch. This is not incompatible, because
778 previously such lookups could never work.
780 The situation is now rather anomolous, since one *can* have colons in
781 ordinary lsearch keys. However, making the change in all cases is
782 incompatible and would probably break a number of configurations.
784 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
787 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
788 conversion specification without a maximum field width, thereby enabling
789 a rogue spamd server to cause a buffer overflow. While nobody in their
790 right mind would setup Exim to query an untrusted spamd server, an
791 attacker that gains access to a server running spamd could potentially
792 exploit this vulnerability to run arbitrary code as the Exim user.
794 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
795 $primary_hostname instead of what libspf2 thinks the hosts name is.
797 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
798 a directory entry by the name of the lookup key. Previously, if a
799 symlink pointed to a non-existing file or a file in a directory that
800 Exim lacked permissions to read, a lookup for a key matching that
801 symlink would fail. Now it is enough that a matching directory entry
802 exists, symlink or not. (Bugzilla 503.)
804 PH/03 The body_linecount and body_zerocount variables are now exported in the
807 PH/04 Added the $dnslist_matched variable.
809 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
810 This means they are set thereafter only if the connection becomes
813 PH/06 Added the client_condition to authenticators so that some can be skipped
814 by clients under certain conditions.
816 PH/07 The error message for a badly-placed control=no_multiline_responses left
817 "_responses" off the end of the name.
819 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
821 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
822 (without spaces) instead of just copying the configuration text.
824 PH/10 Added the /noupdate option to the ratelimit ACL condition.
826 PH/11 Added $max_received_linelength.
828 PH/12 Added +ignore_defer and +include_defer to host lists.
830 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
831 way in which PCRE > 7.0 is built.
833 PH/14 Implemented queue_only_load_latch.
835 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
836 MAIL command. The effect was to mangle the value on 64-bit systems.
838 PH/16 Another patch from the Sieve maintainer.
840 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
842 PH/18 If a system quota error occurred while trying to create the file for
843 a maildir delivery, the message "Mailbox is full" was not appended to the
844 bounce if the delivery eventually timed out. Change 4.67/27 below applied
845 only to a quota excession during the actual writing of the file.
847 PH/19 It seems that peer DN values may contain newlines (and other non-printing
848 characters?) which causes problems in log lines. The DN values are now
849 passed through string_printing() before being added to log lines.
851 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
852 and InterBase are left for another time.)
854 PH/21 Added message_body_newlines option.
856 PH/22 Guard against possible overflow in moan_check_errorcopy().
858 PH/23 POSIX allows open() to be a macro; guard against that.
860 PH/24 If the recipient of an error message contained an @ in the local part
861 (suitably quoted, of course), incorrect values were put in $domain and
862 $local_part during the evaluation of errors_copy.
868 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
869 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
872 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
873 issue a MAIL command.
875 PH/02 In an ACL statement such as
877 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
879 if a client was not listed at all, or was listed with a value other than
880 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
881 the condition was not true (as it should be), so access was not denied.
882 The bug was that the ! inversion was incorrectly passed on to the second
883 item. This has been fixed.
885 PH/03 Added additional dnslists conditions == and =& which are different from
886 = and & when the dns lookup returns more than one IP address.
888 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
889 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
891 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
892 FSYNC, which compiles an option called disable_fsync that allows for
893 bypassing fsync(). The documentation is heavily laced with warnings.
895 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
897 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
898 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
899 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
900 including adding "make clean"; (3) Added -fPIC when compiling the test
901 dynamically loaded module, to get rid of a warning.
903 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
904 message fails, move_frozen_messages = true and ignore_bounce_errors_after
905 = 0s. The bug is otherwise harmless.
907 PH/07 There was a bug in the dovecot authenticator such that the value of
908 $auth1 could be overwritten, and so not correctly preserved, after a
909 successful authentication. This usually meant that the value preserved by
910 the server_setid option was incorrect.
912 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
914 PH/09 Installed PCRE release 7.0.
916 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
917 run for batched SMTP input. It is now run at the start of every message
918 in the batch. While fixing this I discovered that the process information
919 (output by running exiwhat) was not always getting set for -bs and -bS
920 input. This is fixed, and it now also says "batched" for BSMTP.
922 PH/11 Added control=no_pipelining.
924 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
925 patch, slightly modified), and move the expansion of helo_data till after
926 the connection is made in the smtp transport (so it can use these
929 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
931 PH/14 Added log_selector = +pid.
933 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
935 PH/16 Add ${if forany and ${if forall.
937 PH/17 Added dsn_from option to vary the From: line in DSNs.
939 PH/18 Flush SMTP output before performing a callout, unless control =
940 no_callout_flush is set.
942 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
943 was true (the default) a successful delivery failed to delete the retry
944 item, thus causing premature timeout of the address. The bug is now
947 PH/20 Added hosts_avoid_pipelining to the smtp transport.
949 PH/21 Long custom messages for fakedefer and fakereject are now split up
950 into multiline reponses in the same way that messages for "deny" and
951 other ACL rejections are.
953 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
954 with slight modification.
956 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
957 draft, changing the syntax and factoring some duplicate code".
959 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
960 for deliveries of the second and subsequent messages over the same SMTP
963 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
964 ${reduce, with only minor "tidies".
966 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
968 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
969 expansion side effects.
971 PH/27 When a message times out after an over-quota error from an Exim-imposed
972 quota, the bounce message says "mailbox is full". This message was not
973 being given when it was a system quota that was exceeded. It now should
976 MH/03 Made $recipients available in local_scan(). local_scan() already has
977 better access to the recipient list through recipients_list[], but
978 $recipients can be useful in postmaster-provided expansion strings.
980 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
981 in the case of a MAIL command with additional options following the
982 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
983 were accidentally chopped off.
985 PH/29 SMTP synchronization checks are implemented when a command is read -
986 there is a check that no more input is waiting when there shouldn't be
987 any. However, for some commands, a delay in an ACL can mean that it is
988 some time before the response is written. In this time, more input might
989 arrive, invalidly. So now there are extra checks after an ACL has run for
990 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
991 pipelining has not been advertised.
993 PH/30 MH's patch to allow iscntrl() characters to be list separators.
995 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
996 returned in the SMTP response when smtp_return_error_details was false.
999 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1000 instead of the C I/O that was originally supplied, because problems were
1001 reported on Solaris.
1003 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1004 Exim which did not show up earlier: it was assuming that a call to
1005 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1006 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1007 was a macro that became an assignment, so it seemed to work. This has
1008 changed to a proper function call with a void return, hence the compile
1009 error. Exim's code has been fixed.
1011 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1014 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1016 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1018 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1019 because it turns out to be too big or there is a timeout) there is no
1020 "Completed" line in the log. When some messages of this type were
1021 selected by exigrep, they were listed as "not completed". Others were
1022 picked up by some special patterns. I have improved the selection
1023 criteria to be more general.
1025 PH/38 The host_find_failed option in the manualroute router can now be set
1026 to "ignore", to completely ignore a host whose IP address cannot be
1027 found. If all hosts are ignored, the behaviour is controlled by the new
1028 host_all_ignored option.
1030 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1031 homing or because of multiple MX records with /mx) generated more than
1032 one IP address, and the following item turned out to be the local host,
1033 all the secondary addresses of the first item were incorrectly removed
1034 from the list, along with the local host and any following hosts (which
1035 is what is supposed to happen).
1037 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1038 whoever called Exim into the -H file. In the case of the daemon it was
1039 behaving confusingly. When first started, it used values for whoever
1040 started the daemon, but after a SIGHUP it used the Exim user (because it
1041 calls itself on a restart). I have changed the code so that it now always
1044 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1045 message are rejected with the same error (e.g. no authentication or bad
1046 sender address), and a DATA command is nevertheless sent (as can happen
1047 with PIPELINING or a stupid MUA), the error message that was given to the
1048 RCPT commands is included in the rejection of the DATA command. This is
1049 intended to be helpful for MUAs that show only the final error to their
1052 PH/42 Another patch from the Sieve maintainer.
1054 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1055 Eximstats - Fixed some broken HTML links and added missing column headers
1057 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1058 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1060 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1062 PH/43 Yet another patch from the Sieve maintainer.
1064 PH/44 I found a way to check for a TCP/IP connection going away before sending
1065 the response to the final '.' that terminates a message, but only in the
1066 case where the client has not sent further data following the '.'
1067 (unfortunately, this is allowed). However, in many cases there won't be
1068 any further data because there won't be any more messages to send. A call
1069 to select() can be used: if it shows that the input is "ready", there is
1070 either input waiting, or the socket has been closed. An attempt to read
1071 the next input character can distinguish the two cases. Previously, Exim
1072 would have sent an OK response which the client would never have see.
1073 This could lead to message repetition. This fix should cure that, at
1074 least in a lot of common cases.
1076 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1077 advertised in response to EHLO.
1083 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1084 fixed by 4.65/MH/01 (is this a record?) are fixed:
1086 (i) An empty string was always treated as zero by the numeric comparison
1087 operators. This behaviour has been restored.
1089 (ii) It is documented that the numeric comparison operators always treat
1090 their arguments as decimal numbers. This was broken in that numbers
1091 starting with 0 were being interpreted as octal.
1093 While fixing these problems I realized that there was another issue that
1094 hadn't been noticed. Values of message_size_limit (both the global option
1095 and the transport option) were treated as octal if they started with 0.
1096 The documentation was vague. These values are now always treated as
1097 decimal, and I will make that clear in the documentation.
1103 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1104 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1107 MH/01 Don't check that the operands of numeric comparison operators are
1108 integers when their expansion is in "skipping" mode (fixes bug
1109 introduced by 4.64-PH/07).
1111 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1112 child addresses, Exim now panics and dies. Previously, because the count
1113 is held in a short int, deliveries were likely to be lost. As such a
1114 large number of recipients for a single message is ridiculous
1115 (performance will be very, very poor), I have chosen to impose a limit
1116 rather than extend the field.
1122 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1123 leftover -K file (the existence of which was triggered by #402).
1124 While we were at it, introduced process PID as part of the -K
1125 filename. This should rule out race conditions when creating
1128 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1129 processing considerably. Previous code took too long for large mails,
1130 triggering a timeout which in turn triggers #401.
1132 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1133 in the DK code in transports.c. sendfile() is not really portable,
1134 hence the _LINUX specificness.
1136 TF/01 In the add_headers option to the mail command in an Exim filter,
1137 there was a bug that Exim would claim a syntax error in any
1138 header after the first one which had an odd number of characters
1141 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1142 callout verification, Exim cached a "reject" for the entire domain. This
1143 is correct for most verifications, but it is not correct for a recipient
1144 verification with use_sender or use_postmaster set, because in that case
1145 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1146 case of MAIL FROM:<> rejection from other early rejections (e.g.
1147 rejection of HELO). When verifying a recipient using a non-null MAIL
1148 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1149 Whatever the result of the callout, the value of the domain cache is
1150 left unchanged (for any other kind of callout, getting as far as trying
1151 RCPT means that the domain itself is ok).
1153 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1156 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1157 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1158 session, the daemon ignores ECONNECT errors and logs others; it now
1159 ignores EPIPE as well.
1161 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1162 (quoted-printable decoding).
1164 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1165 later the small subsequent patch to fix an introduced bug.
1167 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1169 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1171 PH/08 An error is now given if message_size_limit is specified negative.
1173 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1174 to be given (somewhat) arbitrary names.
1176 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1179 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1180 miscellaneous code fixes
1182 PH/10 Added the log_reject_target ACL modifier to specify where to log
1185 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1186 hostname. This is wrong, because it relates to the incoming message (and
1187 probably the interface on which it is arriving) and not to the outgoing
1188 callout (which could be using a different interface). This has been
1189 changed to use the value of the helo_data option from the smtp transport
1190 instead - this is what is used when a message is actually being sent. If
1191 there is no remote transport (possible with a router that sets up host
1192 addresses), $smtp_active_hostname is used.
1194 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1195 tweaks were necessary in order to get it to work (see also 21 below):
1196 (a) The code assumed that strncpy() returns a negative number on buffer
1197 overflow, which isn't the case. Replaced with Exim's string_format()
1199 (b) There were several signed/unsigned issues. I just did the minimum
1200 hacking in of casts. There is scope for a larger refactoring.
1201 (c) The code used strcasecmp() which is not a standard C function.
1202 Replaced with Exim's strcmpic() function.
1203 (d) The code set only $1; it now sets $auth1 as well.
1204 (e) A simple test gave the error "authentication client didn't specify
1205 service in request". It would seem that Dovecot has changed its
1206 interface. Fortunately there's a specification; I followed it and
1207 changed what the client sends and it appears to be working now.
1209 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1212 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1213 address A is aliased to B and C, where B exists and C does not. Without
1214 -v the output is "A verified" because verification stops after a
1215 successful redirection if more than one address is generated. However,
1216 with -v the child addresses are also verified. Exim was outputting "A
1217 failed to verify" and then showing the successful verification for C,
1218 with its parentage. It now outputs "B failed to verify", showing B's
1219 parentage before showing the successful verification of C.
1221 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1222 look up a TXT record in a specific list after matching in a combined
1225 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1226 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1227 they consult the DNS. I had assumed they would set it the way they
1228 wanted; and indeed my experiments on Linux seem to show that in some
1229 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1230 To be on the safe side, however, I have now made the interface to
1231 host_find_byname() similar to host_find_bydns(), with an argument
1232 containing the DNS resolver options. The host_find_byname() function now
1233 sets these options at its start, just as host_find_bydns() does. The smtp
1234 transport options dns_qualify_single and dns_search_parents are passed to
1235 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1236 of host_find_byname() use the default settings of RES_DEFNAMES
1237 (qualify_single) but not RES_DNSRCH (search_parents).
1239 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1240 spool_read_header() do less string testing, by means of a preliminary
1241 switch on the second character of optional "-foo" lines. (This is
1242 overdue, caused by the large number of possibilities that now exist.
1243 Originally there were few.) While I was there, I also converted the
1244 str(n)cmp tests so they don't re-test the leading "-" and the first
1245 character, in the hope this might squeeze out yet more improvement.
1247 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1248 flag allowing group syntax was set by the header_syntax check but not
1249 turned off, possible causing trouble later; (2) The flag was not being
1250 set at all for the header_verify test, causing "group"-style headers to
1251 be rejected. I have now set it in this case, and also caused header_
1252 verify to ignore an empty address taken from a group. While doing this, I
1253 came across some other cases where the code for allowing group syntax
1254 while scanning a header line wasn't quite right (mostly, not resetting
1255 the flag correctly in the right place). These bugs could have caused
1256 trouble for malformed header lines. I hope it is now all correct.
1258 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1259 with the "reply" argument non-NULL. The code, however (which originally
1260 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1261 but it didn't always do it. This confused somebody who was copying the
1262 code for some other use. I have removed all the tests.
1264 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1265 feature that was used to support insecure browsers during the U.S. crypto
1266 embargo. It requires special client support, and Exim is probably the
1267 only MTA that supported it -- and would never use it because real RSA is
1268 always available. This code has been removed, because it had the bad
1269 effect of slowing Exim down by computing (never used) parameters for the
1270 RSA_EXPORT functionality.
1272 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1273 authenticator to fail if there's a tab character in the incoming data
1274 (there should never be unless someone is messing about, as it's supposed
1275 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1276 if the connection is using TLS or if the remote IP is the same as the
1277 local IP, and the "valid-client-cert option" if a client certificate has
1280 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1281 authenticators. This can be used for authorization after authentication
1282 succeeds. (In the case of plaintext, it servers for both authentication
1285 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1286 if any retry times were supplied.
1288 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1289 connection, where there is no client IP address to check. In this
1290 situation, the verify now always succeeds.
1292 PH/25 Applied John Jetmore's -Mset patch.
1294 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1296 PH/27 In a string expansion for a processed (not raw) header when multiple
1297 headers of the same name were present, leading whitespace was being
1298 removed from all of them, but trailing whitespace was being removed only
1299 from the last one. Now trailing whitespace is removed from each header
1300 before concatenation. Completely empty headers in a concatenation (as
1301 before) are ignored.
1303 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1304 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1306 PH/29 [Removed. This was a change that I later backed out, and forgot to
1307 correct the ChangeLog entry (that I had efficiently created) before
1308 committing the later change.]
1310 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1311 address errors (4xx response to RCPT) over the same connection as other
1312 messages routed to the same hosts. Such deliveries are always "forced",
1313 so retry times are not inspected. This resulted in far too many retries
1314 for the affected addresses. The effect occurred only when there were more
1315 hosts than the hosts_max_try setting in the smtp transport when it had
1316 the 4xx errors. Those hosts that it had tried were not added to the list
1317 of hosts for which the message was waiting, so if all were tried, there
1318 was no problem. Two fixes have been applied:
1320 (i) If there are any address or message errors in an SMTP delivery, none
1321 of the hosts (tried or untried) are now added to the list of hosts
1322 for which the message is waiting, so the message should not be a
1323 candidate for sending over the same connection that was used for a
1324 successful delivery of some other message. This seems entirely
1325 reasonable: after all the message is NOT "waiting for some host".
1326 This is so "obvious" that I'm not sure why it wasn't done
1327 previously. Hope I haven't missed anything, but it can't do any
1328 harm, as the worst effect is to miss an optimization.
1330 (ii) If, despite (i), such a delivery is accidentally attempted, the
1331 routing retry time is respected, so at least it doesn't keep
1332 hammering the server.
1334 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1335 in ${readsocket because some servers need this prod.
1337 PH/32 Added some extra debug output when updating a wait-xxx database.
1339 PH/33 The hint "could be header name not terminated by colon", which has been
1340 given for certain expansion errors for a long time, was not being given
1341 for the ${if def:h_colon_omitted{... case.
1343 PH/34 The spec says: "With one important exception, whenever a domain list is
1344 being scanned, $domain contains the subject domain." There was at least
1345 one case where this was not true.
1347 PH/35 The error "getsockname() failed: connection reset by peer" was being
1348 written to the panic log as well as the main log, but it isn't really
1349 panic-worthy as it just means the connection died rather early on. I have
1350 removed the panic log writing for the ECONNRESET error when getsockname()
1353 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1354 runs only) independently of the message's sender address. This meant
1355 that, if the 4xx error was in fact related to the sender, a different
1356 message to the same recipient with a different sender could confuse
1357 things. In particualar, this can happen when sending to a greylisting
1358 server, but other circumstances could also provoke similar problems.
1359 I have changed the default so that the retry time for these errors is now
1360 based a combination of the sender and recipient addresses. This change
1361 can be overridden by setting address_retry_include_sender=false in the
1364 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1365 remote server are returned as part of bounce messages. This was not
1366 happening for LMTP over a pipe (the lmtp transport), but now it is the
1367 same for both kinds of LMTP.
1369 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1370 in header lines that were in fact CNAMEs. This is no longer the case.
1372 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1373 and queue runs started by the daemon processed all messages. This has
1374 been fixed so that -R and -S can now usefully be given with -q<time>.
1376 PH/40 Import PCRE release 6.7 (fixes some bugs).
1378 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1380 PH/42 Give an error if -q is specified more than once.
1382 PH/43 Renamed the variables $interface_address and $interface_port as
1383 $received_ip_address and $received_port, to make it clear that these
1384 values apply to message reception, and not to the outgoing interface when
1385 a message is delivered. (The old names remain recognized, of course.)
1387 PH/44 There was no timeout on the connect() call when using a Unix domain
1388 socket in the ${readsocket expansion. There now is.
1390 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1391 be meaningful with "accept".
1393 SC/01 Eximstats V1.43
1394 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1396 SC/02 Eximstats V1.44
1397 Use a glob alias rather than an array ref in the generated
1398 parser. This improves both readability and performance.
1400 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1401 Collect SpamAssassin and rejection statistics.
1402 Don't display local sender or destination tables unless
1403 there is data to show.
1404 Added average volumes into the top table text output.
1406 SC/04 Eximstats V1.46
1407 Collect data on the number of addresses (recipients)
1408 as well as the number of messages.
1410 SC/05 Eximstats V1.47
1411 Added 'Message too big' to the list of mail rejection
1412 reasons (thanks to Marco Gaiarin).
1414 SC/06 Eximstats V1.48
1415 Mainlog lines which have GMT offsets and are too short to
1416 have a flag are now skipped.
1418 SC/07 Eximstats V1.49 (Alain Williams)
1419 Added the -emptyok flag.
1421 SC/08 Eximstats V1.50
1422 Fixes for obtaining the IP address from reject messages.
1424 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1425 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1426 whitesspace changes from 4.64-PH/27
1428 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1431 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1432 are found, allow negative numbers in numeric criteria)
1434 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1436 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1437 to match changes made in 4.64-PH/43
1439 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1441 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1442 "domain" is actually an IP address, and force a failure. This locks out
1443 those revolvers/nameservers that support "A-for-A" lookups, in
1444 contravention of the specifications.
1446 PH/48 When a host name was looked up from an IP address, and the subsequent
1447 forward lookup of the name timed out, the host name was left in
1448 $sender_host_name, contrary to the specification.
1450 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1451 restricted to single-key lookups, Exim was not diagnosing an error if
1452 * or *@ was used with a query-style lookup.
1454 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1456 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1457 long ago, but noone interested enough thought of it. Let's just say that
1458 the "1.1" means that there are some new functions that weren't there at
1459 some point in the past.
1461 PH/51 Error processing for expansion failure of helo_data from an smtp
1462 transport during callout processing was broken.
1464 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1465 tested/used via the -bh/-bhc/-bs options.
1467 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1468 bug, fixed in subsequent PCRE releases).
1470 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1471 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1473 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1479 SC/01 Use a glob alias rather than an array ref in eximstats generated
1480 parser. This improves both readability and performance.
1482 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1483 Don't display local sender or destination tables in eximstats unless
1484 there is data to show.
1485 Added average volumes into the eximstats top table text output.
1487 SC/03 Collect data on the number of addresses (recipients) as well
1488 as the number of messages in eximstats.
1490 TF/01 Correct an error in the documentation for the redirect router. Exim
1491 does (usually) call initgroups() when daemonizing.
1493 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1494 with consistent privilege compared to when running as a daemon.
1496 TF/03 Note in the spec that $authenticated_id is not set for local
1497 submissions from trusted users.
1499 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1500 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1502 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1503 by adding some example configuration directives to the default
1504 configuration file. A little bit of work is required to uncomment the
1505 directives and define how usernames and passwords are checked, but
1506 there is now a framework to start from.
1508 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1509 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1510 without this. I don't know how relevant this is to other LDAP libraries.
1512 PH/02 Add the verb name to the "unknown ACL verb" error.
1514 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1516 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1518 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1519 directory not expanded when it should be if an expanded home directory
1520 was set for the address (which is overridden by the transport).
1522 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1525 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1526 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1527 because it is too late at that time, and has no effect.
1529 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1530 security issue with \' (bugzilla #107). I could not use the
1531 PQescapeStringConn() function, because it needs a PGconn value as one of
1534 PH/08 When testing addresses using -bt, indicate those final addresses that
1535 are duplicates that would not cause an additional delivery. At least one
1536 person was confused, thinking that -bt output corresponded to deliveries.
1537 (Suppressing duplicates isn't a good idea as you lose the information
1538 about possibly different redirections that led to the duplicates.)
1540 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1541 systems where poll() doesn't work, in particular OS X.
1543 PH/10 Added more information to debugging output for retry time not reached.
1545 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1546 operations in malware.c.
1548 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1551 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1552 syslog_duplication set false, log lines that would normally be written
1553 both the the main log and to the reject log were not written to syslog at
1556 PH/14 In the default configuration, change the use of "message" in ACL warn
1557 statements to "add_header".
1559 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1560 not followed by a command (e.g. "seen endif").
1562 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1563 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1566 PH/17 Added extra conditions to the default value of delay_warning_condition
1570 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1571 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1572 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1575 The Auto-Submitted: and various List- headers are standardised, whereas I
1576 don't think Precedence: ever was.
1578 PH/18 Refactored debugging code in route_finduser() to show more information,
1579 in particular, the error code if getpwnam() issues one.
1581 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1582 This is apparently needed in addition to the PH/07 change above to avoid
1583 any possible encoding problems.
1585 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1586 but not after initializing Perl.
1588 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1589 output them only if debugging. By default they are written stderr,
1590 apparently, which is not desirable.
1592 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1595 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1598 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1600 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1601 authenticated or an ident call has been made. Suppress the default
1602 values for $authenticated_id and $authenticated_sender (but permit -oMai
1603 and -oMas) when testing with -bh.
1605 PH/24 Re-jigged the order of the tests in the default configuration so that the
1606 tests for valid domains and recipients precede the DNS black list and CSA
1607 tests, on the grounds that those ones are more expensive.
1609 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1610 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1611 This bug exists in every version of Exim that I still have, right back to
1614 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1615 However, an attempt to turn on case-sensitivity in a regex key by
1616 including (?-i) didn't work because the subject string was already
1617 lowercased, and the effects were non-intuitive. It turns out that a
1618 one-line patch can be used to allow (?-i) to work as expected.
1624 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1625 other effects) broke the use of negated acl sub-conditions.
1627 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1630 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1631 "Deny" causes Exim to reject the incoming connection with a 554 error.
1632 Unfortunately, if there is a major crisis, such as a disk failure,
1633 tcp-wrappers gives "deny", whereas what one would like would be some
1634 kind of temporary error. A kludge has been added to help with this.
1635 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1636 554 error is used if errno is still zero or contains ENOENT (which occurs
1637 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1640 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1642 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1643 errors. Otherwise a message that provokes a temporary error (when other
1644 messages do not) can cause a whole host to time out.
1646 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1647 addresses were routed directly to files or pipes from a redirect router.
1648 File deliveries just didn't batch; pipe deliveries might have suffered
1651 PH/06 A failure to get a lock for a hints database would erroneously always say
1652 "Failed to get write lock", even when it was really a read lock.
1654 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1655 of 0600. This has been changed to use the value of the lockfile_mode
1656 option (which defaults to 0600).
1658 PH/08 Applied small patch from the Sieve maintainer.
1660 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1661 folder from quota calculations, a direct delivery into this folder messed
1662 up the contents of the maildirsize file. This was because the regex was
1663 used only to exclude .Trash (or whatever) when the size of the mailbox
1664 was calculated. There was no check that a delivery was happening into an
1665 excluded directory. This bug has been fixed by ignoring all quota
1666 processing for deliveries into excluded directories.
1668 PH/10 Added the maildirfolder_create_regex option to appendfile.
1674 PH/01 The code for finding all the local interface addresses on a FreeBSD
1675 system running IPv6 was broken. This may well have applied to all BSD
1676 systems, as well as to others that have similar system calls. The broken
1677 code found IPv4 interfaces correctly, but gave incorrect values for the
1678 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1679 that it would not match correctly against @[] and not recognize the IPv6
1682 PH/02 The ipliteral router was not recognizing addresses of the form user@
1683 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1685 PH/03 Added disable_ipv6.
1687 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1688 decoded form, because it is most often used to construct To: headers
1689 lines in autoreplies, and the decoded form may well be syntactically
1690 invalid. However, $reply_address has leading white space removed, and all
1691 newlines turned into spaces so that the autoreply transport does not
1694 PH/05 If group was specified without a user on a router, and no group or user
1695 was specified on a transport, the group from the router was ignored.
1697 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1698 for visible compile-time settings that can be used to change these
1699 numbers, for those that want even more. Backwards compatibility with old
1700 spool files has been maintained. However, going back to a previous Exim
1701 release will lost any variables that are in spool files.
1703 PH/07 Two small changes when running in the test harness: increase delay when
1704 passing a TCP/IP connection to a new process, in case the original
1705 process has to generate a bounce, and remove special handling of
1706 127.0.0.2 (sic), which is no longer necessary.
1708 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1709 be the same on different OS.
1711 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1714 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1715 whether --show-vars was specified or not
1717 JJ/02 exipick: Added support for new ACL variable spool format introduced
1720 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1721 syntactically invalid From: or Reply-to: line, and a filter used this to
1722 generate an autoreply, and therefore failed to obtain an address for the
1723 autoreply, Exim could try to deliver to a non-existent relative file
1724 name, causing unrelated and misleading errors. What now happens is that
1725 it logs this as a hard delivery error, but does not attempt to create a
1728 PH/11 The exinext utility has a -C option for testing purposes, but although
1729 the given file was scanned by exinext itself; it wasn't being passed on
1730 when Exim was called.
1732 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1733 an end-of-file indication when reading a command response.
1735 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1736 compiled. In many other places in Exim, IPv6 addresses are always
1737 recognized, so I have changed this. It also means that IPv4 domain
1738 literals of the form [IPV4:n.n.n.n] are now always recognized.
1740 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1741 used if the router is not running as root, for example, when verifying at
1742 ACL time, or when using -bh. The debugging output from this situation was
1743 non-existent - all you got was a failure to exec. I have made two
1746 (a) Failures to set uid/gid, the current directory, or a process leader
1747 in a subprocess such as that created by queryprogram now generate
1748 suitable debugging ouput when -d is set.
1750 (b) The queryprogram router detects when it is not running as root,
1751 outputs suitable debugging information if -d is set, and then runs
1752 the subprocess without attempting to change uid/gid.
1754 PH/15 Minor change to Makefile for building test_host (undocumented testing
1757 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1758 additional section of a DNS packet that returns MX or SRV records.
1759 Instead, it always explicitly searches for A/AAAA records. This avoids
1760 major problems that occur when a DNS server includes only records of one
1761 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1762 fixed another bug: if SRV records were looked up and the corresponding
1763 address records were *not* found in the additional section, the port
1764 values from the SRV records were lost.
1766 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1767 using the correct key (the original address) when searching the retry
1768 rules in order to find which one to use for generating the retry hint.
1770 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1771 adding the default one. Similarly, if it contains a Reply-To: header, the
1772 errors_reply_to option, if set, is not used.
1774 PH/19 When calculating a retry time, Exim used to measure the "time since
1775 failure" by looking at the "first failed" field in the retry record. Now
1776 it does not use this if it is later than than the arrival time of the
1777 message. Instead it uses the arrival time. This makes for better
1778 behaviour in cases where some deliveries succeed, thus re-setting the
1779 "first failed" field. An example is a quota failure for a huge message
1780 when small messages continue to be delivered. Without this change, the
1781 "time since failure" will always be short, possible causing more frequent
1782 delivery attempts for the huge message than are intended.
1783 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1785 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1786 $1, $2, $3) because the numerical variables can be reset during some
1787 expansion items (e.g. "match"), thereby losing the authentication data.
1789 PH/21 Make -bV show the size of off_t variables so that the test suite can
1790 decide whether to run tests for quotas > 2G.
1792 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1793 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1794 filecount value is greater than 2G or if a quota value is greater than 2G
1795 on a system where the size of off_t is not greater than 4, a panic error
1798 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1799 never match. The debug and -bh output now contains an explicit error
1800 message indicating a malformed IPv4 address or mask.
1802 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1803 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1804 PH/23 above applies.
1806 PH/25 Do not write to syslog when running in the test harness. The only
1807 occasion when this arises is a failure to open the main or panic logs
1808 (for which there is an explicit test).
1810 PH/26 Added the /no_tell option to "control=freeze".
1812 PH/27 If a host name lookup failed very early in a connection, for example, if
1813 the IP address matched host_lookup and the reverse lookup yielded a name
1814 that did not have a forward lookup, an error message of the form "no IP
1815 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1816 could be logged. Now it outputs the IP address instead of "NULL".
1818 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1819 allows the sender and the authenticated sender to be set when
1820 submitting a message from within Exim. Since child_open_exim() is
1821 documented for local_scan(), the new function should be too.
1823 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1824 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1825 results in an empty string is now treated as unset.
1827 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1829 PH/31 Added sender_verify_fail logging option.
1831 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1832 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1833 tidied the source and removed it altogether.
1835 PH/33 When a queue run was abandoned because the load average was too high, a
1836 log line was always written; now it is written only if the queue_run log
1837 selector is set. In addition, the log line for abandonment now contains
1838 information about the queue run such as the pid. This is always present
1839 in "start" and "stop" lines but was omitted from the "abandon" line.
1841 PH/34 Omit spaces between a header name and the colon in the error message that
1842 is given when verify = headers_syntax fails (if there are lots of them,
1843 the message gets confusing).
1845 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1846 names, as there are now some PTR records that contain slashes. This check
1847 is only to protect against broken name servers that fall over on strange
1848 characters, so the fact that it applies to all lookups doesn't matter.
1850 PH/36 Now that the new test suite is complete, we can remove some of the
1851 special code in Exim that was needed for the old test suite. For example,
1852 sorting DNS records because real resolvers return them in an arbitrary
1853 order. The new test suite's fake resolver always returns records in the
1856 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1857 bounces) except when queue_only is set, to avoid logging races between
1858 the different processes.
1860 PH/38 Panic-die if .include specifies a non-absolute path.
1862 PH/39 A tweak to the "H" retry rule from its user.
1864 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1865 a label. They prevented compilation on older perls.
1867 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1868 a warning to be raised on newish perls.
1870 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1871 on queue. Changes to match documented behaviour of showing count of
1872 messages matching specified criteria.
1874 PH/40 Changed the default ident timeout from 30s to 5s.
1876 PH/41 Added support for the use of login_cap features, on those BSD systems
1877 that have them, for controlling the resources used by pipe deliveries.
1879 PH/42 The content-scanning code uses fopen() to create files in which to put
1880 message data. Previously it was not paying any attention to the mode of
1881 the files. Exim runs with umask(0) because the rest of the code creates
1882 files with open(), and sets the required mode explicitly. Thus, these
1883 files were ending up world-writeable. This was not a big issue, because,
1884 being within the spool directory, they were not world-accessible. I have
1885 created a function called modefopen, which takes an additional mode
1886 argument. It sets umask(777), creates the file, chmods it to the required
1887 mode, then resets the umask. All the relevant calls to fopen() in the
1888 content scanning code have been changed to use this function.
1890 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1891 to 24 hours. This avoids potential overflow problems when processing G
1892 and H retry rules. I suspect nobody ever tinkers with this value.
1894 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1896 PH/45 When the plaintext authenticator is running as a client, the server's
1897 challenges are checked to ensure they are valid base64 strings. By
1898 default, the authentication attempt is cancelled if an invalid string is
1899 received. Setting client_ignore_invalid_base64 true ignores these errors.
1900 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1901 they are received. Thus, the responses can be made to depend on the
1902 challenges. If an invalid string is ignored, an empty string is placed in
1905 PH/46 Messages that are created by the autoreply transport now contains a
1906 References: header, in accordance with RFCs 2822 and 3834.
1908 PH/47 Added authenticated_sender_force to the smtp transport.
1910 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1912 PH/49 Installed latest patch from the Sieve maintainer.
1914 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1915 was also set, the appendfile transport was unnecessarily scanning a
1916 directory of message files (e.g. for maildir delivery) to find the count
1917 of files (along with the size), even though it did not need this
1918 information. It now does the scan only if it needs to find either the
1919 size of the count of files.
1921 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1923 PH/52 Two bugs concerned with error handling when the smtp transport is
1926 (i) Exim was not creating retry information for temporary errors given
1927 for individual recipients after the DATA command when the smtp transport
1928 was used in LMTP mode. This meant that they could be retried too
1929 frequently, and not timed out correctly.
1931 (ii) Exim was setting the flag that allows error details to be returned
1932 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1933 recipients that were returned after the DATA command.
1935 PH/53 This is related to PH/52, but is more general: for any failing address,
1936 when detailed error information was permitted to be returned to the
1937 sender, but the error was temporary, then after the final timeout, only
1938 "retry timeout exceeded" was returned. Now it returns the full error as
1939 well as "retry timeout exceeded".
1941 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1942 do this, and (what is worse) MTAs that accept it.
1944 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1945 will now be deprecated.
1947 PH/56 New os.c-cygwin from the Cygwin maintainer.
1949 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1950 formats (previously only available in exim formats via -bpr, -bpru,
1951 and -bpra. Now also available in native and exiqgrep formats)
1953 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1954 with very large, slow to parse queues
1956 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1958 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1960 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1961 responses to authentication challenges, though it was showing the
1962 challenges; (ii) I've removed the CR characters from the debug output for
1965 PH/58 Allow for the insertion of a newline as well as a space when a string
1966 is turned into more than one encoded-word during RFC 2047 encoding. The
1967 Sieve code now uses this.
1969 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1970 data_4xx, lost_connection, tls_required.
1972 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1973 message was being sent as an SMTP response.
1975 PH/61 Add -l and -k options to exicyclog.
1977 PH/62 When verifying, if an address was redirected to one new address, so that
1978 verification continued, and the new address failed or deferred after
1979 having set something in $address_data, the value of $address_data was not
1980 passed back to the ACL. This was different to the case when no
1981 redirection occurred. The value is now passed back in both cases.
1983 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1984 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1985 use login_cap.h, so on its own it isn't the distinguishing feature. The
1986 new name refers directly to the setclassresources() function.
1988 PH/65 Added configuration files for NetBSD3.
1990 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1992 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1993 is preferred over IPv4.
1995 PH/68 The bounce_return_message and bounce_return_body options were not being
1996 honoured for bounces generated during the reception of non-SMTP messages.
1997 In particular, this applied to messages rejected by the ACL. This bug has
1998 been fixed. However, if bounce_return_message is true and bounce_return_
1999 body is false, the headers that are returned for a non-SMTP message
2000 include only those that have been read before the error was detected.
2001 (In the case of an ACL rejection, they have all been read.)
2003 PH/69 The HTML version of the specification is now built in a directory called
2004 spec_html instead of spec.html, because the latter looks like a path with
2005 a MIME-type, and this confuses some software.
2007 PH/70 Catch two compiler warnings in sieve.c.
2009 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2010 function verify_get_ident() calls ip_connect() to connect a socket, but
2011 if the "connect()" function timed out, ip_connect() used to close the
2012 socket. However, verify_get_ident() also closes the socket later, and in
2013 between Exim writes to the log, which may get opened at this point. When
2014 the socket was closed in ip_connect(), the log could get the same file
2015 descriptor number as the socket. This naturally causes chaos. The fix is
2016 not to close the socket in ip_connect(); the socket should be closed by
2017 the function that creates it. There was only one place in the code where
2018 this was missing, in the iplookup router, which I don't think anybody now
2019 uses, but I've fixed it anyway.
2021 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2022 well as to direct DNS lookups. Otherwise the handling of names in host
2023 lists is inconsistent and therefore confusing.
2029 PH/01 Two changes to the default runtime configuration:
2031 (1) Move the checks for relay_from_hosts and authenticated clients from
2032 after to before the (commented out) DNS black list checks.
2034 (2) Add control=submission to the relay_from_hosts and authenticated
2035 clients checks, on the grounds that messages accepted by these
2036 statements are most likely to be submissions.
2038 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2040 (1) Generate an error if the third argument for the ${prvs expansion is
2043 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2046 (3) Reset the variables that are obtained from the first argument of
2047 ${prvscheck and used in the second argument before leaving the code,
2048 because their memory is reclaimed, so using them afterwards may do
2051 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2052 one (it's much easier than Tom thought :-).
2054 (5) Because of (4), we can now allow for the use of $prvscheck_result
2055 inside the third argument.
2057 PH/03 For some reason, the default setting of PATH when running a command from
2058 a pipe transport was just "/usr/bin". I have changed it to
2061 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2062 anything to be listed in the output from -bV.
2064 PH/05 When a filter generated an autoreply, the entire To: header line was
2065 quoted in the delivery log line, like this:
2067 => >A.N.Other <ano@some.domain> <original@ddress> ...
2069 This has been changed so that it extracts the operative address. There
2070 may be more than one such address. If so, they are comma-separated, like
2073 => >ano@some.domain,ona@other.domain <original@ddress> ...
2075 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2076 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2077 not being looked up in the rDNS to get a host name, Exim was showing the
2078 IP address twice in Received: lines, even though the IP addresses were
2079 identical. For example:
2081 Received: from [1.2.3.4] (helo=[1.2.3.4])
2083 However, if the real host name was known, it was omitting the HELO data
2084 if it matched the actual IP address. This has been tidied up so that it
2085 doesn't show the same IP address twice.
2087 PH/07 When both +timestamp and +memory debugging was on, the value given by
2088 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2089 called by the debug printing, thereby overwriting the timestamp buffer.
2090 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2092 PH/08 When the original message was included in an autoreply transport, it
2093 always said "this is a copy of the message, including all the headers",
2094 even if body_only or headers_only was set. It now gives an appropriate
2097 PH/09 Applied a patch from the Sieve maintainer which:
2099 o fixes some comments
2100 o adds the (disabled) notify extension core
2101 o adds some debug output for the result of if/elsif tests
2102 o points to the current vacation draft in the documentation
2103 and documents the missing references header update
2107 o fixes a bug in processing the envelope test (when testing
2108 multiple envelope elements, the last element determinted the
2111 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2112 Electronic Mail") by including:
2114 Auto-submitted: auto-generated
2116 in the messages that it generates (bounce messages and others, such as
2117 warnings). In the case of bounce messages for non-SMTP mesages, there was
2118 also a typo: it was using "Auto_submitted" (underscore instead of
2119 hyphen). Since every message generated by Exim is necessarily in response
2120 to another message, thes have all been changed to:
2122 Auto-Submitted: auto-replied
2124 in accordance with these statements in the RFC:
2126 The auto-replied keyword:
2128 - SHOULD be used on messages sent in direct response to another
2129 message by an automatic process,
2131 - MUST NOT be used on manually-generated messages,
2133 - MAY be used on Delivery Status Notifications (DSNs) and Message
2134 Disposition Notifications (MDNs),
2136 - MUST NOT be used on messages generated by automatic or periodic
2137 processes, except for messages which are automatic responses to
2140 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2141 to the default Received: header definition.
2143 PH/12 Added log selector acl_warn_skipped (default on).
2145 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2146 variables because (a) they are in the wrong storage pool and (b) even if
2147 they were copied, it wouldn't work properly because of the caching.
2149 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2150 checking when decoding. Apparently there are clients that generate
2151 overlong encoded strings. Why am I not surprised?
2153 PH/15 If the first argument of "${if match_address" was not empty, but did not
2154 contain an "@" character, Exim crashed. Now it writes a panic log message
2155 and treats the condition as false.
2157 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2159 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2160 extension "envelope-auth". The code is finished and in agreement with
2161 other implementations, but there is no documentation so far and in fact,
2162 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2163 not changing the active code.
2165 Print executed "if" and "elsif" statements when debugging is used. This
2166 helps a great deal to understand what a filter does.
2168 Document more things not specified clearly in RFC3028. I had all this
2169 sorted out, when out of a sudden new issues came to my mind. Oops."
2171 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2174 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2175 canonical form (as documented). However, after a host name lookup from
2176 the IP address, check_host() was doing a simple string comparison with
2177 addresses acquired from the DNS when checking that the found name did
2178 have the original IP as one of its addresses. Since any found IPv6
2179 addresses are likely to be in abbreviated form, the comparison could
2180 fail. Luckily, there already exists a function for doing the comparison
2181 by converting both addresses to binary, so now that is used instead of
2182 the text comparison.
2184 PH/20 There was another similar case to PH/19, when a complete host name was
2185 given in a host list; looking up its IP address could give an abbreviated
2186 form, whereas the current host's name might or might not be abbreviated.
2187 The same fix has been applied.
2193 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2194 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2197 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2198 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2200 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2202 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2203 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2204 Also, for "Auto-Submitted", treat anything other than "no" as
2205 non-personal, in accordance with RFC 3834. (Previously it treated
2206 anything starting "auto-" as non-personal.)
2208 TF/01 The control=submission/name=... option had a problem with syntax
2209 errors if the name included a slash character. The /name= option
2210 now slurps the rest of the string, so it can include any characters
2211 but it must come last in the list of options (after /sender_retain
2214 PH/05 Some modifications to the interface to the fake nameserver for the new
2222 TK/01 Added the "success_on_redirect" address verification option. See
2223 NewStuff for rationale and an example.
2225 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2227 PH/02 Patch to exigrep to allow it to work on syslog lines.
2229 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2230 fread() to skip over the body file's header line, because in Cygwin the
2231 header line is locked and is inaccessible.
2233 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2234 co-exist for some time) to make it clear that it is the Exim ID that is
2235 referenced, not the Message-ID: header line.
2237 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2238 string_format() function, because snprintf() does not exist on all
2241 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2242 ${stat: expansion item.
2244 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2245 protocol synchronization error", to keep the pedants happy.
2247 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2248 well as for IRIX systems, when gcc is being used. See the host.c source
2251 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2253 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2256 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2257 transport and to the smtp transport in LMTP mode.
2259 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2261 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2262 run a filter in a subprocess. This could lead to confusion in subsequent
2263 lookups in the parent process. There should also be a search_tidyup() at
2264 the end of the subprocess.
2266 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2267 only if the host matched helo_try_verify_hosts, which caused the
2268 verification to occur when the EHLO/HELO command was issued. The ACL just
2269 tested the remembered result. Now, if a previous verification attempt has
2270 not happened, "verify = helo" does it there and then.
2272 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2274 TK/03 Fix log output including CR from clamd.
2276 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2277 exist provoked a memory error which could cause a segfault.
2279 PH/15 Installed PCRE 6.2
2281 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2283 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2284 of the problem. Specifically, suggested +O2 rather than +O1 for the
2287 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2289 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2290 record, the port was not correctly logged when the outgoing_port log
2291 selector was set (it logged the transort's default port).
2293 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2294 fallback_hosts, and "hosts" in the smtp transport.
2296 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2297 host errors such as "Connection refused".
2299 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2300 authentication with radiusclient 0.4.9:
2302 - Error returned from rc_read_config was caught wrongly
2303 - Username/password not passed on to radius server due to wrong length.
2305 The presumption is that some radiusclient API changes for 4.51/PH/17
2306 were not taken care of correctly. The code is still untested by me (my
2307 Linux distribution still has 0.3.2 of radiusclient), but it was
2308 contributed by a Radius user.
2310 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2311 expanding the "port" option of the smtp transport.
2313 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2314 while reading a MIME header. Thanks to Tom Hughes for a patch.
2316 PH/24 Include config.h inside local_scan.h so that configuration settings are
2319 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2320 that in an ACL for RCPT (for example), you can examine exactly what was
2323 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2324 commands, but it was not correctly comparing the address with the actual
2325 client host address. Thus, it would show the EHLO address in Received:
2326 header lines when this was not necessary.
2328 PH/27 Added the % operator to ${eval:}.
2330 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2331 it should be ignoring failures (because with -C, for example, it has lost
2332 privilege). It wasn't ignoring creation failures other than "already
2335 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2338 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2339 because an input line was too long, either on its own, or by virtue of
2340 too many continuations, the temporary file was not being removed, and the
2341 return code was incorrect.
2343 PH/31 Missing "BOOL" in function definition in filtertest.c.
2345 PH/32 Applied Sieve patches from the maintainer.
2347 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2349 PH/33 Added "verify = not_blind".
2351 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2352 Local/Makefile (with some defaults set). These are used in built scripts
2353 such as exicyclog, but they have never been used in the exim_install
2354 script (though there are many overriding facilities there). I have
2355 arranged that the exim_install script now takes note of these two
2358 PH/35 Installed configuration files for Dragonfly.
2360 PH/36 When a locally submitted message by a trusted user did not contain a
2361 From: header, and the sender address was obtained from -f or from an SMTP
2362 MAIL command, and the trusted user did not use -F to supply a sender
2363 name, $originator_name was incorrectly used when constructing a From:
2364 header. Furthermore, $originator_name was used for submission mode
2365 messages from external hosts without From: headers in a similar way,
2366 which is clearly wrong.
2368 PH/37 Added control=suppress_local_fixups.
2370 PH/38 When log_selector = +received_sender was set, and the addition of the
2371 sender made the log line's construction buffer exactly full, or one byte
2372 less than full, an overflow happened when the terminating "\n" was
2375 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2376 when the result of a list match is failure because a DNS lookup failed.
2378 PH/40 RM_COMMAND is now used in the building process.
2380 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2381 the "build-* directories that it finds.
2383 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2384 address in a domain literal was a prefix of an interface address.
2386 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2387 when verifying a sender address, unless rewrite_headers is false.
2389 PH/44 Wrote a long comment about why errors_to addresses are verified as
2390 recipients, not senders.
2392 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2393 the ratelimit ACL was added.
2395 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2397 PH/46 Added extra information about PostgreSQL errors to the error string.
2399 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2400 avoiding the need to install special zones in a real server. This is
2401 backwards compatible; if it can't find the fake resolver, it drops back.
2402 Thus, both old and new test suites can be run.
2404 TF/02 Added util/ratelimit.pl
2406 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2407 clock is set back in time.
2409 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2410 Candler <B.Candler@pobox.com>.
2412 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2413 OK for addresses that are the result of redirections.
2415 PH/48 A number of further additions for the benefit of the new test suite,
2416 including a fake gethostbyname() that interfaces to the fake DNS resolver
2419 TF/06 The fix for widen_domains has also been applied to qualify_single and
2420 search_parents which are the other dnslookup options that can cause
2423 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2426 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2428 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2429 service" if it encountered an MX record with an empty target hostname.
2430 The message is now "an MX or SRV record indicated no SMTP service".
2432 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2433 if the DNS of the sending site is misconfigured. This is quite a
2434 common situation. This change restores the behaviour of treating a
2435 helo verification defer as a failure.
2437 PH/51 If self=fail was set on a router, the bounce message did not include the
2438 actual error message.
2444 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2446 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2447 command itself ended in error, the underlying message about the transport
2448 filter timeout was being overwritten with the pipe command error. Now the
2449 underlying error message should be appended to the second error message.
2451 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2453 PH/02 Reduce the amount of output that "make" produces by default. Full output
2454 can still be requested.
2456 PH/03 The warning log line about a condition test deferring for a "warn" verb
2457 was being output only once per connection, rather than after each
2458 occurrence (because it was using the same function as for successful
2459 "warn" verbs). This seems wrong, so I have changed it.
2461 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2462 it should not have, which might have caused a crash in the right
2463 circumstances, but probably never did.
2465 PH/04 Installed a modified version of Tony Finch's patch to make submission
2466 mode fix the return path as well as the Sender: header line, and to
2467 add a /name= option so that you can make the user's friendly name appear
2470 TF/03 Added the control = fakedefer ACL modifier.
2472 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2473 Mark Lowes for thorough testing.
2475 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2477 TK/03 Merged latest SRS patch from Miles Wilton.
2479 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2480 with the definition in sysexits.h (which is #included earlier).
2481 Fortunately, Exim does not actually use EX_OK. The code used to try to
2482 preserve the sysexits.h value, by assumimg that macro definitions were
2483 scanned for macro replacements. I have been disabused of this notion,
2484 so now the code just undefines EX_OK before #including unistd.h.
2486 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2487 in the smtp transport. When a block could not be written in a single
2488 write() function, the timeout was being re-applied to each part-write.
2489 This seems wrong - if the receiver was accepting one byte at a time it
2490 would take for ever. The timeout is now adjusted when this happens. It
2491 doesn't have to be particularly precise.
2493 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2494 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2496 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2497 without a domain if the check to <postmaster@domain> fails.
2499 SC/01 Eximstats: added -xls and the ability to specify output files
2500 (patch written by Frank Heydlauf).
2502 SC/02 Eximstats: use FileHandles for outputing results.
2504 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2506 SC/04 Eximstats: fixed display of large numbers with -nvr option
2508 SC/05 Eximstats: fixed merging of reports with empty tables.
2510 SC/06 Eximstats: added the -include_original_destination flag
2512 SC/07 Eximstats: removed tabs and trailing whitespace.
2514 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2516 TK/06 MBOX spool code: Add real "From " MBOX separator line
2517 so the .eml file is really in mbox format (even though
2518 most programs do not really care). Patch from Alex Miller.
2520 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2521 The latter is generated from $received_to and is only set if the
2522 message has one envelope recipient. SA can use these headers,
2523 obviously out-of-the-box. Patch from Alex Miller.
2525 PH/08 The ${def test on a variable was returning false if the variable's
2526 value was "0", contrary to what the specification has always said!
2527 The result should be true unless the variable is empty.
2529 PH/09 The syntax error of a character other than { following "${if
2530 def:variable_name" (after optional whitespace) was not being diagnosed.
2531 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2532 accidental colon was present, for example, could give incorrect results.
2534 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2535 result is used (not including appendfile, where other changes are about
2538 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2539 This involved changing a lot of size variables from int to off_t. It
2540 should work with maildirs and everything.
2542 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2543 spamd dying while we are connected to it.
2545 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2548 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2549 function for BDB 4.3.
2551 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2553 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2554 the sources for PCRE have been re-arranged and more files are now
2557 PH/15 The code I had for printing potentially long long variables in PH/11
2558 above was not the best (it lost precision). The length of off_t variables
2559 is now inspected at build time, and an appropriate printing format (%ld
2560 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2561 to be "long long int" or "long int". This is needed for the internal
2562 formatting function string_vformat().
2564 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2565 the configuration file to be ":syslog", then the script "guesses" where
2566 the logs files are, rather than using the compiled in default. In our
2567 case the guess is not the same as the compiled default, so the script
2568 suddenly stopped working when I started to use syslog. The patch checks
2569 to see if log_file_path is "". If so, it attempts to read it from exim
2570 with no configuration file to get the compiled in version, before it
2571 falls back to the previous guessing code."
2573 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2574 implementing BATV in an Exim configuration. See NewStuff for the gory
2577 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2578 Makefile that are specific to HP-UX.
2580 PH/18 If the "use_postmaster" option was set for a recipient callout together
2581 with the "random" option, the postmaster address was used as the MAIL
2582 FROM address for the random test, but not for the subsequent recipient
2583 test. It is now used for both.
2585 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2586 patch removes a few documentation additions to RFC 3028, because the
2587 latest draft now contains them. It adds the new en;ascii-case comparator
2588 and a new error check for 8bit text in MIME parts. Comparator and
2589 require names are now matched exactly. I enabled the subaddress
2590 extension, but it is not well tested yet (read: it works for me)."
2592 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2593 rework some of the code of TK/09 above to avoid the hardwired use of
2594 "%lld" and "long long". Replaced the call to snprintf() with a call to
2597 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2598 records point to non-existent hosts", "retry timeout exceeded", and
2599 "retry time not reached for any host after a long failure period".
2601 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2602 experimental DomainKeys support:
2604 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2605 (2) On an error such as an illegally used "control", the wrong name for
2606 the control was given.
2608 These problems did NOT occur unless DomainKeys support was compiled.
2610 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2612 PH/24 Added ${if match_ip condition.
2614 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2615 databases so that it will be absolutely obvious if a crash occurs in the
2616 DB library. This is a regular occurrence (often caused by mis-matched
2619 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2620 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2621 picked up on a user's system that detects such things. There doesn't seem
2622 to be a gcc warning option for this - only an attribute that has to be
2623 put on the function's prototype. It seems that in Fedora Core 4 they have
2624 set this on a number of new functions. No doubt there will be more in due
2627 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2628 specify a transport. However, if an address that was verified by such a
2629 router was the subject of a callout, Exim crashed because it tried to
2630 read the rcpt_include_affixes from the non-existent transport. Now it
2631 just assumes that the setting of that option is false. This bug was
2632 introduced by 4.51/PH/31.
2634 PH/28 Changed -d+all to exclude +memory, because that information is very
2635 rarely of interest, but it makes the output a lot bigger. People tend to
2636 do -d+all out of habit.
2638 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2639 code in os-type was giving problems when libc.so lives in lib64, like on
2642 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2643 aren't the modern standard, and it seems that some systems' include files
2644 don't always have them. Exim was already checking for some of the newer
2645 ones like T_AAAA, and defining it itself. I've added checks for all the
2646 record types that Exim uses.
2648 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2649 not automatically generating a new one, as it is supposed to. This
2650 prevented TLS from working. If the file did exist, but contained invalid
2651 data, a new version was generated, as expected. It was only the case of a
2652 non-existent file that was broken.
2654 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2655 with a change in libdomainkeys > 0.64.
2657 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2658 from DNS. If the selector record carries the flag, it now has
2659 precedence over the domain-wide flag.
2661 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2663 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2664 the use of a port name that isn't defined in /etc/services) occurred, the
2665 message was deferred as in a normal delivery, and thus remained on the
2666 spool, instead of being failed because of the mua_wrapper setting. This
2667 is now fixed, and I tidied up some of the mua_wrapper messages at the
2670 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2671 the messages in a hash of arrays rather than using individual hashes.
2672 This is a bit cleaner and results in dramatic memory savings, albeit
2673 at a slight CPU cost.
2675 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2676 as requested by Marc Sherman.
2678 SC/10 Eximstats: added histograms for user specified patterns as requested
2681 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2683 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2684 fopen() in the content-scanning modules that did not already have it.
2690 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2691 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2693 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2695 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2697 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2698 to test Sieve filters that use "vacation".
2700 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2701 that changes the way the GnuTLS parameters are stored in the cache file.
2702 The new format can be generated externally. For backward compatibility,
2703 if the data in the cache doesn't make sense, Exim assumes it has read an
2704 old-format file, and it generates new data and writes a new file. This
2705 means that you can't go back to an older release without removing the
2708 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2709 work if there are any delivery delays because "one_time" forces the
2710 parent to be marked "delivered", so its unseen clone is never tried
2711 again. For this reason, Exim now forbids the simultaneous setting of
2714 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2715 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2716 entry if you want to know the details. The fix, however, neglected to
2717 consider the case when local delivery batching is involved. The test for
2718 "previously delivered" was not happening when checking to see if an
2719 address could be batched with a previous (undelivered) one; under
2720 certain circumstances this could lead to multiple deliveries to the same
2723 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2724 in its include files, and this causes problems building Exim.
2726 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2727 header_syntax) but Exim was just ignoring anything given after a slash.
2728 In particular, this caused confusion with an attempt to use "verify =
2729 reverse_host_lookup/defer_ok". An error is now given when options are
2730 supplied for verify items that do not have them. (Maybe reverse_host_
2731 lookup should have a defer_ok option, but that's a different point.)
2733 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2734 defined by RFC 821) to 2048, because there were problems with some AUTH
2735 commands, and RFC 1869 says the size should be increased for extended
2736 SMTP commands that take arguments.
2738 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2741 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2742 "unknown" error; now it says that the functionality isn't in the binary.
2744 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2745 an address' error message when a string expansion fails (syntax or
2746 whatever). Otherwise the password may appear in the log. Following change
2747 PH/42 below, there is no longer a chance of it appearing in a bounce
2750 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2752 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2753 of its addresses was ever tried. (Bugzilla bug #2.)
2755 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2756 the result incorrectly in the debug output. (It correctly added a newline
2757 to what was transported.)
2759 TF/01 Added $received_time.
2761 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2762 commented out examples of how to interface to a virus scanner and to
2763 SpamAssassin. Also added commented examples of av_scanner and
2764 spamd_address settings.
2766 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2767 and controls are allowed in which ACLs. There were a couple of minor
2768 errors. Some of the entries in the conditions table (which is a table of
2769 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2770 negation of where the condition IS allowed.
2772 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2774 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2775 header file does not have a version number, so I've had to invent a new
2776 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2777 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2778 radiusclient), but it was contributed by a Radius user.
2780 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2781 files or directories, for OpenSSL.
2783 PH/19 When an Exim process that is running as root has to create an Exim log
2784 file, it does so in a subprocess that runs as exim:exim so as to get the
2785 ownership right at creation (otherwise, other Exim processes might see
2786 the file with the wrong ownership). There was no test for failure of this
2787 fork() call, which would lead to the process getting stuck as it waited
2788 for a non-existent subprocess. Forks do occasionally fail when resources
2789 run out. I reviewed all the other calls to fork(); they all seem to check
2792 PH/20 When checking for unexpected SMTP input at connect time (before writing
2793 the banner), Exim was not dealing correctly with a non-positive return
2794 from the read() function. If the client had disconnected by this time,
2795 the result was a log entry for a synchronization error with an empty
2796 string after "input=" when read() returned zero. If read() returned -1
2797 (an event I could not check), uninitialized data bytes were printed.
2798 There were reports of junk text (parts of files, etc) appearing after
2801 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2803 PH/22 Added support for macro redefinition, and (re)definition in between
2804 driver and ACL definitions.
2806 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2807 forgetting to use the resulting value; it was using the unexpanded value.
2809 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2810 hadn't been configured. The fix is from Juergen Kreileder, who
2811 understands it better than I do:
2813 "Here's what I see happening with three configured cyrus_sasl
2814 authenticators configured (plain, login, cram-md5):
2816 On startup auth_cyrus_sasl_init() gets called for each of these.
2817 This means three calls to sasl_listmech() without a specified mech_list.
2818 => SASL tests which mechs of all available mechs actually work
2819 => three warnings about OTP not working
2820 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2822 With the patch, sasl_listmech() also gets called three times. But now
2823 SASL's mech_list option is set to the server_mech specified in the the
2824 authenticator. Or in other words, the answer from sasl_listmech()
2825 gets limited to just the mech you're testing for (which is different
2827 => the return list contains just 'plain' or 'login', 'cram-md5' or
2828 nothing depending on the value of ob->server_mech.
2830 I've just tested the patch: Authentication still works fine,
2831 unavailable mechs specified in the exim configuration are still
2832 caught, and the auth.log warnings about OTP are gone."
2834 PH/25 When debugging is enabled, the contents of the command line are added
2835 to the debugging output, even when log_selector=+arguments is not
2838 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2839 answer is "GNU", and only if the return is "GNU/something" is the answer
2842 PH/27 $acl_verify_message is now set immediately after the failure of a
2843 verification in an ACL, and so is available in subsequent modifiers. In
2844 particular, the message can be preserved by coding like this:
2846 warn !verify = sender
2847 set acl_m0 = $acl_verify_message
2849 Previously, $acl_verify_message was set only while expanding "message"
2850 and "log_message" when a very denied access.
2852 PH/28 Modified OS/os.c-Linux with
2854 -#ifndef OS_LOAD_AVERAGE
2855 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2857 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2858 nomenclature these days.)
2860 PH/29 Installed patch from the Sieve maintainer that adds the options
2861 sieve_useraddress and sieve_subaddress to the redirect router.
2863 PH/30 In these circumstances:
2864 . Two addresses routed to the same list of hosts;
2865 . First host does not offer TLS;
2866 . First host accepts first address;
2867 . First host gives temporary error to second address;
2868 . Second host offers TLS and a TLS session is established;
2869 . Second host accepts second address.
2870 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2871 and peerdn, if requested) that were in fact used only for the second
2874 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2875 attention to any local part prefix or suffix that was matched by the
2876 router that accepted the address. It now behaves in the same way as it
2877 does for delivery: the affixes are removed from the local part unless
2878 rcpt_include_affixes is set on the transport.
2880 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2881 timeout during the DATA phase of an incoming message.
2883 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2884 applied a patch sanctioned by the Sieve maintainer.
2886 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2887 the uid or gid is negative. A case of a negative gid caused this to be
2888 noticed. The fix allows for either to be negative.
2890 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2891 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2894 PH/36 The change PH/12 above was broken. Fixed it.
2896 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2897 the grounds that routing the same address twice would always produce the
2898 same answer. This might have been true once, but it is certainly no
2899 longer true now. Routing a child address may depend on the previous
2900 routing that produced that child. Some complicated redirection strategies
2901 went wrong when messages had multiple recipients, and made Exim's
2902 behaviour dependent on the order in which the addresses were given.
2904 I have moved the duplicate checking until after the routing is complete.
2905 Exim scans the addresses that are assigned to local and remote
2906 transports, and removes any duplicates. This means that more work will be
2907 done, as duplicates will always all be routed, but duplicates are
2908 presumably rare, so I don't expect this is of any significance.
2910 For deliveries to pipes, files, and autoreplies, the duplicate checking
2911 still happens during the routing process, since they are not going to be
2914 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2915 It corrects a timeout issue with spamd. This is Ian's comment: "The
2916 background is that sometimes spamd either never reads data from a
2917 connection it has accepted, or it never writes response data. The exiscan
2918 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2919 blindly assumes that writes won't block so it may never time out."
2921 PH/39 Allow G after quota size as well as K and M.
2923 PH/40 The value set for $authenticated_id in an authenticator may not contain
2924 binary zeroes or newlines because the value is written to log lines and
2925 to spool files. There was no check on this. Now the value is run through
2926 the string_printing() function so that such characters are converted to
2927 printable escape sequences.
2929 PH/41 $message_linecount is a new variable that contains the total number of
2930 lines in the message. Compare $body_linecount, which is the count for the
2933 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2934 bounce and delay warning messages, except in certain special cases, which
2937 (a) An SMTP error message from a remote host;
2938 (b) A message specified in a :fail: redirection;
2939 (c) A message specified in a "fail" command in a system filter;
2940 (d) A message specified in a FAIL return from the queryprogram router;
2941 (e) A message specified by the cannot_route_message router option.
2943 In these cases only, Exim does include the error details in bounce and
2944 warning messages. There are also a few cases where bland messages such
2945 as "unrouteable address" or "local delivery error" are given.
2947 PH/43 $value is now also set for the "else" part of a ${run expansion.
2949 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2950 being worked on, but at least Exim now implements the latest version to
2953 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2954 process to complete was treated as a delivery failure, a timeout while
2955 writing the message to the pipe was logged, but erroneously treated as a
2956 successful delivery. Such timeouts include transport filter timeouts. For
2957 consistency with the overall process timeout, these timeouts are now
2958 treated as errors, giving rise to delivery failures by default. However,
2959 there is now a new Boolean option for the pipe transport called
2960 timeout_defer, which, if set TRUE, converts the failures into defers for
2961 both kinds of timeout. A transport filter timeout is now identified in
2964 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2965 systems where "make" and "gmake" are different, calling "gmake" at top
2966 level broke things. I've arranged for the value of $(MAKE) to be passed
2967 from the Makefile to this script so that it can call the same version of
2971 A note about Exim versions 4.44 and 4.50
2972 ----------------------------------------
2974 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2975 changes of various kinds. As a consequence, a big documentation update was
2976 needed. This delayed the release for rather longer than seemed good, especially
2977 in the light of a couple of (minor) security issues. Therefore, the changes
2978 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2979 release. So 4.44 and 4.50 are in effect two different branches that both start
2982 I have left the 4.50 change log unchanged; it contains all the changes since
2983 4.43. The change log for 4.44 is below; many of its items are identical to
2984 those for 4.50. This seems to be the most sensible way to preserve the
2985 historical information.
2991 1. Minor wording change to the doc/README.SIEVE file.
2993 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2994 computation of the current number of files was incorrect.
2996 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2997 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3000 4. Give more explanation in the error message when the command for a transport
3001 filter fails to execute.
3003 5. There are several places where Exim runs a non-Exim command in a
3004 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3005 was being done only for the command run by the queryprogram router. It is
3006 now done for all such subprocesses. The other cases are: ${run, transport
3007 filters, and the commands run by the lmtp and pipe transports.
3009 6. Added CONFIGURE_GROUP build-time option.
3011 7. Some older OS have a limit of 256 on the maximum number of file
3012 descriptors. Exim was using setrlimit() to set 1000 as a large value
3013 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3014 systems. I've change it so that if it can't get 1000, it tries for 256.
3016 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3017 was an oversight, and furthermore, ever since the addition of extra
3018 controls (e.g. 4.43/32), the checks on when to allow different forms of
3019 "control" were broken. There should now be diagnostics for all cases when a
3020 control that does not make sense is encountered.
3022 9. Added the /retain_sender option to "control=submission".
3024 10. $recipients is now available in the predata ACL (oversight).
3026 11. Tidy the search cache before the fork to do a delivery from a message
3027 received from the command line. Otherwise the child will trigger a lookup
3028 failure and thereby defer the delivery if it tries to use (for example) a
3029 cached ldap connection that the parent has called unbind on.
3031 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3032 of $address_data from the recipient verification was clobbered by the
3033 sender verification.
3035 13. The value of address_data from a sender verification is now available in
3036 $sender_address_data in subsequent conditions in the ACL statement.
3038 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3040 15. Added a new option "connect=<time>" to callout options, to set a different
3043 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3044 was its contents. (It was OK if the option was not defined at all.)
3046 17. A "Completed" log line is now written for messages that are removed from
3047 the spool by the -Mrm option.
3049 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3050 information about exactly what failed.
3052 19. Added -dd to debug only the daemon process.
3054 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3055 handles timeouts, both on the server side and network timeouts. Renamed the
3056 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3058 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3059 It is now set to "smtps".
3061 22. $host_address is now set to the target address during the checking of
3062 ignore_target_hosts.
3064 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3065 being passed; this would have caused $sender_host_name to have been used if
3066 matching the list had actually called for a host name (not very likely,
3067 since this list is usually IP addresses). A host name is now passed as
3070 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3071 code that specifies a non-restarting handler (typically sigaction() in
3072 modern systems) in an attempt to fix a rare and obscure crash bug.
3074 25. Narrowed the window for a race in the daemon that could cause it to ignore
3075 SIGCHLD signals. This is not a major problem, because they are used only to
3076 wake it up if nothing else does.
3078 26. A malformed maildirsize file could cause Exim to calculate negative values
3079 for the mailbox size or file count. Odd effects could occur as a result.
3080 The maildirsize information is now recalculated if the size or filecount
3083 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3084 support for a long time. Removed HAVE_SYS_VFS_H.
3086 28. Installed the latest version of exipick from John Jetmore.
3088 29. In an address list, if the pattern was not a regular expression, an empty
3089 subject address (from a bounce message) matched only if the pattern was an
3090 empty string. Non-empty patterns were not even tested. This was the wrong
3091 because it is perfectly reasonable to use an empty address as part of a
3092 database query. An empty address is now tested by patterns that are
3093 lookups. However, all the other forms of pattern expect the subject to
3094 contain a local part and a domain, and therefore, for them, an empty
3095 address still always fails if the pattern is not itself empty.
3097 30. Exim went into a mad DNS loop when attempting to do a callout where the
3098 host was specified on an smtp transport, and looking it up yielded more
3099 than one IP address.
3101 31. Re-factored the code for checking spool and log partition space into a
3102 function that finds that data and another that does the check. The former
3103 is then used to implement four new variables: $spool_space, $log_space,
3104 $spool_inodes, and $log_inodes.
3106 32. The RFC2047 encoding function was originally intended for short strings
3107 such as real names; it was not keeping to the 75-character limit for
3108 encoded words that the RFC imposes. It now respects the limit, and
3109 generates multiple encoded words if necessary. To be on the safe side, I
3110 have increased the buffer size for the ${rfc2047: expansion operator from
3113 33. It is now permitted to omit both strings after an "if" condition; if the
3114 condition is true, the result is "true". As before, when the second string
3115 is omitted, a false condition yields an empty string. This makes it less
3116 cumbersome to write custom ACL and router conditions.
3118 34. Failure to deliver a bounce message always caused it to be frozen, even if
3119 there was an errors_to setting on the router. The errors_to setting is now
3122 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3123 canonical form (fully expanded) before being placed in
3124 $sender_host_address.
3126 36. The table in the code that translates DNS record types into text (T_A to
3127 "A" for instance) was missing entries for NS and CNAME. It is just possible
3128 that this could have caused confusion if both these types were looked up
3129 for the same domain, because the text type is used as part of Exim's
3130 per-process caching. But the chance of anyone hitting this buglet seems
3133 37. The dnsdb lookup has been extended in a number of ways.
3135 (1) There is a new type, "zns", which walks up the domain tree until it
3136 finds some nameserver records. It should be used with care.
3138 (2) There is a new type, "mxh", which is like "mx" except that it returns
3139 just the host names, not the priorities.
3141 (3) It is now possible to give a list of domains (or IP addresses) to be
3142 looked up. The behaviour when one of the lookups defers can be
3143 controlled by a keyword.
3145 (4) It is now possible to specify the separator character for use when
3146 multiple records are returned.
3148 38. The dnslists ACL condition has been extended: it is now possible to supply
3149 a list of IP addresses and/or domains to be looked up in a particular DNS
3152 39. Added log_selector=+queue_time_overall.
3154 40. When running the queue in the test harness, wait just a tad after forking a
3155 delivery process, to get repeatability of debugging output.
3157 41. Include certificate and key file names in error message when GnuTLS fails
3158 to set them up, because the GnuTLS error message doesn't include the name
3159 of the failing file when there is a problem reading it.
3161 42. Allow both -bf and -bF in the same test run.
3163 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3165 44. The "Exiscan patch" is now merged into the mainline Exim source.
3167 45. Sometimes the final signoff response after QUIT could fail to get
3168 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3169 < 0 before doing a fflush(). This bug looks as though it goes back to the
3170 introduction of TLS in release 3.20, but "sometimes" must have been rare
3171 because the tests only now provoked it.
3173 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3174 (this can affect the format of dates).
3176 47. exim_tidydb, when checking for the continued existence of a message for
3177 which it has found a message-specific retry record, was not finding
3178 messages that were in split spool directories. Consequently, it was
3179 deleting retry records that should have stayed in existence.
3181 48. Steve fixed some bugs in eximstats.
3183 49. The SPA authentication driver was not abandoning authentication and moving
3184 on to the next authenticator when an expansion was forced to fail,
3185 contradicting the general specification for all authenticators. Instead it
3186 was generating a temporary error. It now behaves as specified.
3188 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3189 (the order specifies the preference for clients). The order is now AES256,
3190 AES128, 3DES, ARCFOUR128.
3192 51. Small patch to Sieve code - explicitly set From: when generating an
3195 52. Exim crashed if a remote delivery caused a very long error message to be
3196 recorded - for instance if somebody sent an entire SpamAssassin report back
3197 as a large number of 550 error lines. This bug was coincidentally fixed by
3198 increasing the size of one of Exim's internal buffers (big_buffer) that
3199 happened as part of the Exiscan merge. However, to be on the safe side, I
3200 have made the code more robust (and fixed the comments that describe what
3203 53. Now that there can be additional text after "Completed" in log lines (if
3204 the queue_time_overall log selector is set), a one-byte patch to exigrep
3205 was needed to allow it to recognize "Completed" as not the last thing in
3208 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3209 patch that reportedly fixes this has been added. I am not expert enough to
3210 create a test for it. This is what the patch creator wrote:
3212 "I found a little strange behaviour of ldap code when working with
3213 Windows 2003 AD Domain, where users was placed in more than one
3214 Organization Units. When I tried to give exim partial DN, the exit code
3215 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3216 But simultaneously result of request was absolutely normal ldap result,
3217 so I produce this patch..."
3219 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3220 REFERENCE, so I have modified the code to exclude the patch when that macro
3223 55. Some experimental protocols are using DNS PTR records for new purposes. The
3224 keys for these records are domain names, not reversed IP addresses. The
3225 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3226 leaves it alone. Component reversal etc. now happens only for IP addresses.
3229 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3231 57. Double the size of the debug message buffer (to 2048) so that more of very
3232 long debug lines gets shown.
3234 58. The exicyclog utility now does better if the number of log files to keep
3235 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3237 59. Two changes related to the smtp_active_hostname option:
3239 (1) $smtp_active_hostname is now available as a variable.
3240 (2) The default for smtp_banner uses $smtp_active_hostname instead
3241 of $primary_hostname.
3243 60. The host_aton() function is supposed to be passed a string that is known
3244 to be a valid IP address. However, in the case of IPv6 addresses, it was
3245 not checking this. This is a hostage to fortune. Exim now panics and dies
3246 if the condition is not met. A case was found where this could be provoked
3247 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3248 components; fortuitously, this particular loophole had already been fixed
3249 by change 4.50/55 above.
3251 If there are any other similar loopholes, the new check in host_aton()
3252 itself should stop them being exploited. The report I received stated that
3253 data on the command line could provoke the exploit when Exim was running as
3254 exim, but did not say which command line option was involved. All I could
3255 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3256 running as the user.
3259 61. There was a buffer overflow vulnerability in the SPA authentication code
3260 (which came originally from the Samba project). I have added a test to the
3261 spa_base64_to_bits() function which I hope fixes it.
3264 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3265 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3267 63. The daemon start-up calls getloadavg() while still root for those OS that
3268 need the first call to be done as root, but it missed one case: when
3269 deliver_queue_load_max is set with deliver_drop_privilege. This is
3270 necessary for the benefit of the queue runner, because there is no re-exec
3271 when deliver_drop_privilege is set.
3273 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3274 This has been fixed.
3276 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3277 was in use, was not putting the data itself into the right store pool;
3278 consequently, it could be overwritten for a subsequent message in the same
3279 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3282 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3284 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3285 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3286 the function were treating the return as a boolean value, which happened to
3287 work because 0=false and not-0=true, but is not correct code.
3289 68. The host_aton() function was not handling scoped IPv6 addresses (those
3290 with, for example, "%eth0" on the end) correctly.
3292 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3293 negated items (that is, ~something) in unsigned ints. Some compilers
3294 apparently mutter when there is no cast.
3296 70. If an address verification called from an ACL failed, and did not produce a
3297 user-specific message (i.e. there was only a "system" message), nothing was
3298 put in $acl_verify_message. In this situation, it now puts the system
3301 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3302 session; change 4.31/43 added the unwanted input to the log line - except
3303 that it did not do this in the start of session case. It now does.
3305 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3306 This is wrong and can cause the other end to generate a synchronization
3307 error if it is another Exim or anything else that does the synchronization
3308 check. A QUIT command is no longer sent after a timeout.
3310 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3311 during host lookups.
3313 74. The defer_ok option of callout verification was not working if it was used
3314 when verifying addresses in header lines, that is, for this case:
3316 verify = header_sender/callout=defer_ok
3318 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3319 those file descriptors could be used for SMTP connections. If anything
3320 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3321 could be sent to the SMTP client, causing chaos. The daemon now opens
3322 stdin, stdout, and stderr to /dev/null when it puts itself into the
3325 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3326 log by default. The user can override this with suitable Perl magic.
3328 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3329 the log message when discard triggers, was not working for the DATA ACL or
3330 for the non-SMTP ACL.
3332 78. Error message wording change in sieve.c.
3334 79. If smtp_accept_max_per_host was set, the number of connections could be
3335 restricted to fewer than expected, because the daemon was trying to set up
3336 a new connection before checking whether the processes handling previous
3337 connections had finished. The check for completed processes is now done
3338 earlier. On busy systems, this bug wouldn't be noticed because something
3339 else would have woken the daemon, and it would have reaped the completed
3342 80. If a message was submitted locally by a user whose login name contained one
3343 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3344 It caused a spool format error. I have fixed the spool reading code. A
3345 related problem was that the "from" clause in the Received: line became
3346 illegal because of the space(s). It is now covered by ${quote_local_part.
3348 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3351 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3353 83. Patch from Sieve maintainer to fix unterminated string problem in
3354 "vacation" handling.
3356 84. Some minor changes to the Linux configuration files to help with other
3357 OS variants using glibc.
3359 85. One more patch for Sieve to update vacation handling to latest spec.
3362 ----------------------------------------------------
3363 See the note above about the 4.44 and 4.50 releases.
3364 ----------------------------------------------------
3370 1. Change 4.43/35 introduced a bug that caused file counts to be
3371 incorrectly computed when quota_filecount was set in an appendfile
3374 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3375 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3378 3. Give more explanation in the error message when the command for a transport
3379 filter fails to execute.
3381 4. There are several places where Exim runs a non-Exim command in a
3382 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3383 was being done only for the command run by the queryprogram router. It is
3384 now done for all such subprocesses. The other cases are: ${run, transport
3385 filters, and the commands run by the lmtp and pipe transports.
3387 5. Some older OS have a limit of 256 on the maximum number of file
3388 descriptors. Exim was using setrlimit() to set 1000 as a large value
3389 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3390 systems. I've change it so that if it can't get 1000, it tries for 256.
3392 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3393 was an oversight, and furthermore, ever since the addition of extra
3394 controls (e.g. 4.43/32), the checks on when to allow different forms of
3395 "control" were broken. There should now be diagnostics for all cases when a
3396 control that does not make sense is encountered.
3398 7. $recipients is now available in the predata ACL (oversight).
3400 8. Tidy the search cache before the fork to do a delivery from a message
3401 received from the command line. Otherwise the child will trigger a lookup
3402 failure and thereby defer the delivery if it tries to use (for example) a
3403 cached ldap connection that the parent has called unbind on.
3405 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3406 of $address_data from the recipient verification was clobbered by the
3407 sender verification.
3409 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3410 was its contents. (It was OK if the option was not defined at all.)
3412 11. A "Completed" log line is now written for messages that are removed from
3413 the spool by the -Mrm option.
3415 12. $host_address is now set to the target address during the checking of
3416 ignore_target_hosts.
3418 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3419 being passed; this would have caused $sender_host_name to have been used if
3420 matching the list had actually called for a host name (not very likely,
3421 since this list is usually IP addresses). A host name is now passed as
3424 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3425 code that specifies a non-restarting handler (typically sigaction() in
3426 modern systems) in an attempt to fix a rare and obscure crash bug.
3428 15. Narrowed the window for a race in the daemon that could cause it to ignore
3429 SIGCHLD signals. This is not a major problem, because they are used only to
3430 wake it up if nothing else does.
3432 16. A malformed maildirsize file could cause Exim to calculate negative values
3433 for the mailbox size or file count. Odd effects could occur as a result.
3434 The maildirsize information is now recalculated if the size or filecount
3437 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3438 support for a long time. Removed HAVE_SYS_VFS_H.
3440 18. Updated exipick to current release from John Jetmore.
3442 19. Allow an empty sender to be matched against a lookup in an address list.
3443 Previously the only cases considered were a regular expression, or an
3446 20. Exim went into a mad DNS lookup loop when doing a callout where the
3447 host was specified on the transport, if the DNS lookup yielded more than
3450 21. The RFC2047 encoding function was originally intended for short strings
3451 such as real names; it was not keeping to the 75-character limit for
3452 encoded words that the RFC imposes. It now respects the limit, and
3453 generates multiple encoded words if necessary. To be on the safe side, I
3454 have increased the buffer size for the ${rfc2047: expansion operator from
3457 22. Failure to deliver a bounce message always caused it to be frozen, even if
3458 there was an errors_to setting on the router. The errors_to setting is now
3461 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3462 canonical form (fully expanded) before being placed in
3463 $sender_host_address.
3465 24. Updated eximstats to version 1.33
3467 25. Include certificate and key file names in error message when GnuTLS fails
3468 to set them up, because the GnuTLS error message doesn't include the name
3469 of the failing file when there is a problem reading it.
3471 26. Expand error message when OpenSSL has problems setting up cert/key files.
3474 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3475 (this can affect the format of dates).
3477 28. exim_tidydb, when checking for the continued existence of a message for
3478 which it has found a message-specific retry record, was not finding
3479 messages that were in split spool directories. Consequently, it was
3480 deleting retry records that should have stayed in existence.
3482 29. eximstats updated to version 1.35
3483 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3484 1.35 - bugfix such that pie charts by volume are generated correctly
3486 30. The SPA authentication driver was not abandoning authentication and moving
3487 on to the next authenticator when an expansion was forced to fail,
3488 contradicting the general specification for all authenticators. Instead it
3489 was generating a temporary error. It now behaves as specified.
3491 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3492 (the order specifies the preference for clients). The order is now AES256,
3493 AES128, 3DES, ARCFOUR128.
3495 31. Small patch to Sieve code - explicitly set From: when generating an
3498 32. Exim crashed if a remote delivery caused a very long error message to be
3499 recorded - for instance if somebody sent an entire SpamAssassin report back
3500 as a large number of 550 error lines. This bug was coincidentally fixed by
3501 increasing the size of one of Exim's internal buffers (big_buffer) that
3502 happened as part of the Exiscan merge. However, to be on the safe side, I
3503 have made the code more robust (and fixed the comments that describe what
3506 33. Some experimental protocols are using DNS PTR records for new purposes. The
3507 keys for these records are domain names, not reversed IP addresses. The
3508 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3509 leaves it alone. Component reversal etc. now happens only for IP addresses.
3512 34. The host_aton() function is supposed to be passed a string that is known
3513 to be a valid IP address. However, in the case of IPv6 addresses, it was
3514 not checking this. This is a hostage to fortune. Exim now panics and dies
3515 if the condition is not met. A case was found where this could be provoked
3516 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3517 components; fortuitously, this particular loophole had already been fixed
3518 by change 4.50/55 or 4.44/33 above.
3520 If there are any other similar loopholes, the new check in host_aton()
3521 itself should stop them being exploited. The report I received stated that
3522 data on the command line could provoke the exploit when Exim was running as
3523 exim, but did not say which command line option was involved. All I could
3524 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3525 running as the user.
3528 35. There was a buffer overflow vulnerability in the SPA authentication code
3529 (which came originally from the Samba project). I have added a test to the
3530 spa_base64_to_bits() function which I hope fixes it.
3533 36. The daemon start-up calls getloadavg() while still root for those OS that
3534 need the first call to be done as root, but it missed one case: when
3535 deliver_queue_load_max is set with deliver_drop_privilege. This is
3536 necessary for the benefit of the queue runner, because there is no re-exec
3537 when deliver_drop_privilege is set.
3539 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3540 was in use, was not putting the data itself into the right store pool;
3541 consequently, it could be overwritten for a subsequent message in the same
3542 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3545 38. Sometimes the final signoff response after QUIT could fail to get
3546 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3547 < 0 before doing a fflush(). This bug looks as though it goes back to the
3548 introduction of TLS in release 3.20, but "sometimes" must have been rare
3549 because the tests only now provoked it.
3555 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3556 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3557 Now it returns an integer. A number of calls were still expecting a T/F
3558 return. Fortuitously, in all cases, the tests worked in OK situations,
3559 which is the norm. However, things would have gone wrong on any write
3560 failures on the smtp file descriptor. This function is used when sending
3561 messages over SMTP and also when doing verify callouts.
3563 2. When Exim is called to do synchronous delivery of a locally submitted
3564 message (the -odf or -odi options), it no longer closes stderr before doing
3567 3. Implemented the mua_wrapper option.
3569 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3571 5. Implemented the functions header_remove(), header_testname(),
3572 header_add_at_position(), and receive_remove_recipient(), and exported them
3575 6. If an ACL "warn" statement specified the addition of headers, Exim already
3576 inserted X-ACL-Warn: at the start if there was no header name. However, it
3577 was not making this test for the second and subsequent header lines if
3578 there were newlines in the string. This meant that an invalid header could
3579 be inserted if Exim was badly configured.
3581 7. Allow an ACL "warn" statement to add header lines at the start or after all
3582 the Received: headers, as well as at the end.
3584 8. Added the rcpt_4xx retry error code.
3586 9. Added postmaster_mailfrom=xxx to callout verification option.
3588 10. Added mailfrom=xxxx to the callout verification option, for verify=
3591 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3592 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3594 12. Inserted some casts to stop certain compilers warning when using pointer
3595 differences as field lengths or precisions in printf-type calls (mostly
3596 affecting debugging statements).
3598 13. Added optional readline() support for -be (dynamically loaded).
3600 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3601 same clock tick as a message's arrival, so that its received time was the
3602 same as the "first fail" time on the retry record, and that message
3603 remained on the queue past the ultimate address timeout, every queue runner
3604 would try a delivery (because it was past the ultimate address timeout) but
3605 after another failure, the ultimate address timeout, which should have then
3606 bounced the address, did not kick in. This was a "< instead of <=" error;
3607 in most cases the first failure would have been in the next clock tick
3608 after the received time, and all would be well.
3610 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3611 being recognized when the domain list was tested by the match_domain
3612 condition in an expansion string.
3614 16. Added the ${str2b64: operator.
3616 17. Exim was always calling setrlimit() to set a large limit for the number of
3617 processes, without checking whether the existing limit was already
3618 adequate. (It did check for the limit on file descriptors.) Furthermore,
3619 errors from getrlimit() and setrlimit() were being ignored. Now they are
3620 logged to the main and panic logs, but Exim does carry on, to try to do its
3621 job under whatever limits there are.
3623 18. Imported PCRE 5.0.
3625 19. Trivial typo in log message " temporarily refused connection" (the leading
3628 20. If the log selector return_path_on_delivery was set and an address was
3629 redirected to /dev/null, the delivery process crashed because it assumed
3630 that a return path would always be set for a "successful" delivery. In this
3631 case, the whole delivery is bypassed as an optimization, and therefore no
3634 21. Internal re-arrangement: the function for sending a challenge and reading
3635 a response while authentication was assuming a zero-terminated challenge
3636 string. It's now changed to take a pointer and a length, to allow for
3637 binary data in such strings.
3639 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3641 23. Exim was not respecting finduser_retries when seeking the login of the
3642 uid under which it was called; it was always trying 10 times. (The default
3643 setting of finduser_retries is zero.) Also, it was sleeping after the final
3644 failure, which is pointless.
3646 24. Implemented tls_on_connect_ports.
3648 25. Implemented acl_smtp_predata.
3650 26. If the domain in control=submission is set empty, Exim assumes that the
3651 authenticated id is a complete email address when it generates From: or
3652 Sender: header lines.
3654 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3655 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3656 chown and chgrp in /bin and hostname in /usr/bin.
3658 28. Exim was keeping the "process log" file open after each use, just as it
3659 does for the main log. This opens the possibility of it remaining open for
3660 long periods when the USR1 signal hits a daemon. Occasional processlog
3661 errors were reported, that could have been caused by this. Anyway, it seems
3662 much more sensible not to leave this file open at all, so that is what now
3665 29. The long-running daemon process does not normally write to the log once it
3666 has entered its main loop, and it closes the log before doing so. This is
3667 so that log files can straightforwardly be renamed and moved. However,
3668 there are a couple of unusual error situations where the daemon does write
3669 log entries, and I had neglected to close the log afterwards.
3671 30. The text of an SMTP error response that was received during a remote
3672 delivery was being truncated at 512 bytes. This is too short for some of
3673 the long messages that one sometimes sees. I've increased the limit to
3676 31. It is now possible to make retry rules that apply only when a message has a
3677 specific sender, in particular, an empty sender.
3679 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3680 it possible to be selective about when SMTP synchronization is enforced.
3682 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3684 32. Implemented hosts_connection_nolog.
3686 33. Added an ACL for QUIT.
3688 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3691 35. Added mailbox_size and mailbox_filecount to appendfile.
3693 36. Added control = no_multiline_responses to ACLs.
3695 37. There was a bug in the logic of the code that waits for the clock to tick
3696 in the case where the clock went backwards by a substantial amount such
3697 that the microsecond fraction of "now" was more than the microsecond
3698 fraction of "then" (but the whole seconds number was less).
3700 38. Added support for the libradius Radius client library this is found on
3701 FreeBSD (previously only the radiusclient library was supported).
3707 1. When certain lookups returned multiple values in the form name=value, the
3708 quoting of the values was not always being done properly. Specifically:
3709 (a) If the value started with a double quote, but contained no whitespace,
3711 (b) If the value contained whitespace other than a space character (i.e.
3712 tabs or newlines or carriage returns) it was not quoted.
3713 This fix has been applied to the mysql and pgsql lookups by writing a
3714 separate quoting function and calling it from the lookup code. The fix
3715 should probably also be applied to nisplus, ibase and oracle lookups, but
3716 since I cannot test any of those, I have not disturbed their existing code.
3718 2. A hit in the callout cache for a specific address caused a log line with no
3719 reason for rejecting RCPT. Now it says "Previous (cached) callout
3720 verification failure".
3722 3. There was an off-by-one bug in the queryprogram router. An over-long
3723 return line was truncated at 256 instead of 255 characters, thereby
3724 overflowing its buffer with the terminating zero. As well as fixing this, I
3725 have increased the buffer size to 1024 (and made a note to document this).
3727 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3728 when Exim is waiting for an SMTP response from a remote server, Exim
3729 restarts its select() call on the socket, thereby resetting its timeout.
3730 This is not a problem when such interrupts are rare. Somebody set up a cron
3731 job to run exiwhat every 2 minutes, which is less than the normal select()
3732 timeout (5 or 10 minutes). This meant that the select() timeout never
3733 kicked in because it was always reset. I have fixed this by comparing the
3734 time when an interrupt arrives with the time at the start of the first call
3735 to select(). If more time than the timeout has elapsed, the interrupt is
3736 treated as a timeout.
3738 5. Some internal re-factoring in preparation for the addition of Sieve
3739 extensions (by MH). In particular, the "personal" test is moved to a
3740 separate function, and given an option for scanning Cc: and Bcc: (which is
3741 not set for Exim filters).
3743 6. When Exim created an email address using the login of the caller as the
3744 local part (e.g. when creating a From: or Sender: header line), it was not
3745 quoting the local part when it contained special characters such as @.
3747 7. Installed new OpenBSD configuration files.
3749 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3750 try to make them clearer.
3752 9. Callout options, other than the timeout value, were being ignored when
3753 verifying sender addresses in header lines. For example, when using
3755 verify = header_sender/callout=no_cache
3757 the cache was (incorrectly) being used.
3759 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3760 only the Cygwin environment.
3762 11. When return_path_on_delivery was set as a log selector, if different remote
3763 addresses in the same message used different return paths and parallel
3764 remote delivery occurred, the wrong values would sometimes be logged.
3765 (Whenever a remote delivery process finished, the return path value from
3766 the most recently started remote delivery process was logged.)
3768 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3769 lines when AUTH and/or TLS are in use. This is the "received protocol"
3770 field. Exim used to use "asmtp" for authenticated SMTP, without any
3771 indication (in the protocol name) for TLS use. Now it follows the RFC and
3772 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3773 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3774 names appear in log lines as well as in Received: header lines.
3776 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3777 extensions, and comparison tests, and to fix some bugs.
3779 14. Changes to the "personal" filter test:
3781 (1) The test was buggy in that it was just doing the equivalent of
3782 "contains" tests on header lines. For example, if a user's address was
3783 anne@some.where, the "personal" test would incorrectly be true for
3785 To: susanne@some.where
3787 This test is now done by extracting each address from the header in turn,
3788 and checking the entire address. Other tests that are part of "personal"
3789 are now done using regular expressions (for example, to check local parts
3790 of addresses in From: header lines).
3792 (2) The list of non-personal local parts in From: addresses has been
3793 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3794 taken from the Sieve specification recommendations.
3796 (3) If the message contains any header line starting with "List-" it is
3797 treated as non-personal.
3799 (4) The test for "circular" in the Subject: header line has been removed
3800 because it now seems ill-conceived.
3802 15. Minor typos in src/EDITME comments corrected.
3804 16. Installed latest exipick from John Jetmore.
3806 17. If headers_add on a router specified a text string that was too long for
3807 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3808 of string_sprintf() is now avoided.
3810 18. $message_body_size was not set (it was always zero) when running the DATA
3811 ACL and the local_scan() function.
3813 19. For the "mail" command in an Exim filter, no default was being set for
3814 the once_repeat time, causing a random time value to be used if "once" was
3815 specified. (If the value happened to be <= 0, no repeat happened.) The
3816 default is now 0s, meaning "never repeat". The "vacation" command was OK
3817 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3818 (I found it when inspecting the code).
3820 20. There is now an overall timeout for performing a callout verification. It
3821 defaults to 4 times the callout timeout, which applies to individual SMTP
3822 commands during the callout. The overall timeout applies when there is more
3823 than one host that can be tried. The timeout is checked before trying the
3824 next host. This prevents very long delays if there are a large number of
3825 hosts and all are timing out (e.g. when the network connections are timing
3826 out). The value of the overall timeout can be changed by specifying an
3827 additional sub-option for "callout", called "maxwait". For example:
3829 verify = sender/callout=5s,maxwait=20s
3831 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3832 to the end before writing, but this should make it even safer).
3834 22. Exim was forgetting that it had advertised PIPELINING for the second and
3835 subsequent messages on an SMTP connection. It was also not resetting its
3836 memory on STARTTLS and an internal HELO.
3838 23. When Exim logs an SMTP synchronization error within a session, it now
3839 records whether PIPELINING has been advertised or not.
3841 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3842 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3843 rather than long int.
3845 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3847 26. Added the never_mail option to autoreply.
3853 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3854 crash if the getsockname() call failed; this can happen if a connection is
3855 closed very soon after it is established. The problem was simply in the
3856 order in which certain operations were done, causing Exim to try to write
3857 to the SMTP stream before it had set up the file descriptor. The bug has
3858 been fixed by making things happen in the correct order.
3864 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3865 before the connection was closed, thus losing the rejection response.
3867 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3868 some early Solaris releases, but causes trouble in current releases where
3869 socklen_t is defined.
3871 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3874 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3877 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3878 that releasing the top of it at the end releases what was used for sub-
3879 expansions (unless the block got too big). However, discard this block if
3880 the first thing is a variable or header, so that we can use its block when
3881 it is dynamic (useful for very large $message_headers, for example).
3883 6. Lookups now cache *every* query, not just the most recent. A new, separate
3884 store pool is used for this. It can be recovered when all lookup caches are
3885 flushed. Lookups now release memory at the end of their result strings.
3886 This has involved some general refactoring of the lookup sources.
3888 7. Some code has been added to the store_xxx() functions to reduce the amount
3889 of flapping under certain conditions.
3891 8. log_incoming_interface used to affect only the <= reception log lines. Now
3892 it causes the local interface and port to be added to several more SMTP log
3893 lines, for example "SMTP connection from", and rejection lines.
3895 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3897 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3899 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3900 could be overwritten at the end of the current message (or the start of a
3901 new message if it was set in a HELO ACL). The value is now preserved for
3902 the duration of the SMTP connection.
3904 12. If a transport had a headers_rewrite setting, and a matching header line
3905 contained an unqualified address, that address was qualified, even if it
3906 did not match any rewriting rules. The underlying bug was that the values
3907 of the flags that permit the existence of unqualified sender and recipient
3908 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3909 non-local messages, and by -bnq for local messages) were not being
3910 preserved with the message after it was received.
3912 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3913 "next input=" as part of the text comprising the host identity instead of
3914 the correct text. The code was using the same buffer for two different
3915 strings. However, depending on which order the printing function evaluated
3916 its arguments, the bug did not always show up. Under Linux, for example, my
3917 test suite worked just fine.
3919 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3920 doesn't work with some older versions of Perl. It has been changed to "my",
3921 which in any case is probably the better facility to use.
3923 15. A really picky compiler found some instances of statements for creating
3924 error messages that either had too many or two few arguments for the format
3927 16. The size of the buffer for calls to the DNS resolver has been increased
3928 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3929 for addresses that have a lot of PTR records. This alleviates a problem; it
3930 does not fully solve it.
3932 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3933 buffer now truncates the list and logs the incident, which is the same
3934 action as happens when Exim is looking up a host name and its aliases.
3935 Previously in this situation something unpredictable would happen;
3936 sometimes it was "internal error: store_reset failed".
3938 18. If a server dropped the connection unexpectedly when an Exim client was
3939 using GnuTLS and trying to read a response, the client delivery process
3940 crashed while trying to generate an error log message.
3942 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3943 string, for example:
3945 warn message = H1: something\nH2: something
3947 the text was added as a single header line from Exim's point of view
3948 though it ended up OK in the delivered message. However, searching for the
3949 second and subsequent header lines using $h_h2: did not work. This has been
3950 fixed. Similarly, if a system filter added multiple headers in this way,
3951 the routers could not see them.
3953 20. Expanded the error message when iplsearch is called with an invalid key to
3954 suggest using net-iplsearch in a host list.
3956 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3957 ACLs are no longer actually imposed (and a message to that effect is
3960 22. If a "gecos" field in a passwd entry contained escaped characters, in
3961 particular, if it contained a \" sequence, Exim got it wrong when building
3962 a From: or a Sender: header from that name. A second bug also caused
3963 incorrect handling when an unquoted " was present following a character
3964 that needed quoting.
3966 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3967 was not being matched caselessly.
3969 24. Arranged for all hyphens in the exim.8 source to be escaped with
3972 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3973 properly. Recipient callout cache records were still being keyed to include
3974 the sender, even when use_sender was set false. This led to far more
3975 callouts that were necessary. The sender is no longer included in the key
3976 when use_sender is false.
3978 26. Added "control = submission" modifier to ACLs.
3980 27. Added the ${base62d: operator to decode base 62 numbers.
3982 28. dnsdb lookups can now access SRV records.
3984 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3985 the configuration file.
3987 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3988 (-v) mode. This makes the output for a verbose queue run more intelligible.
3990 31. Added a use_postmaster feature to recipient callouts.
3992 32. Added the $body_zerocount variable, containing the number of binary zero
3993 bytes in the message body.
3995 33. The time of last modification of the "new" subdirectory is now used as the
3996 "mailbox time last read" when there is a quota error for a maildir
3999 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4001 35. Added +ignore_unknown as a special item in host lists.
4003 36. Code for decoding IPv6 addresses in host lists is now included, even if
4004 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4005 address was recognized as an IP address, but was then not correctly decoded
4006 into binary, causing unexpected and incorrect effects when compared with
4013 1. Very minor rewording of debugging text in manualroute to say "list of
4014 hosts" instead of "hostlist".
4016 2. If verify=header_syntax was set, and a header line with an unqualified
4017 address (no domain) and a large number of spaces between the end of the
4018 name and the colon was received, the reception process suffered a buffer
4019 overflow, and (when I tested it) crashed. This was caused by some obsolete
4020 code that should have been removed. The fix is to remove it!
4022 3. When running in the test harness, delay a bit after writing a bounce
4023 message to get a bit more predictability in the log output.
4025 4. Added a call to search_tidyup() just before forking a reception process. In
4026 theory, someone could use a lookup in the expansion of smtp_accept_max_
4027 per_host which, without the tidyup, could leave open a database connection.
4029 5. Added the variables $recipient_data and $sender_data which get set from a
4030 lookup success in an ACL "recipients" or "senders" condition, or a router
4031 "senders" option, similar to $domain_data and $local_part_data.
4033 6. Moved the writing of debug_print from before to after the "senders" test
4036 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4037 problems for message scanning, either using a data ACL, or using
4038 local_scan() because the Received: header was not generated till after they
4039 were called (in order to set the time as the time of reception completion).
4040 I have revised the way this works. The header is now generated after the
4041 body is received, but before the ACL or local_scan() are called. After they
4042 are run, the timestamp in the header is updated.
4048 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4049 before starting a queue runner without re-exec. This happened only when
4050 deliver_drop_privilege was set or when the Exim user was set to root. The
4051 effect of the bug was that timeouts during subsequent deliveries caused
4052 crashes instead of being properly handled. The handler is now left at its
4053 default (and expected) setting.
4055 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4056 message, again when deliver_drop_privilege is set or Exim is run as root.
4057 The bug described in (1) was not present in this case, but the tidying up
4058 of the other signals was missing. I have made the two cases consistent.
4060 3. The ignore_target_hosts setting on a manualroute router was being ignored
4061 for hosts that were looked up using the /MX notation.
4063 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4066 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4067 operated on the sender address. After changing the $sender_address to <>
4068 for the sender address verify, Exim was re-instated it as the original
4069 (before rewriting) address, but remembering that it had rewritten it, so it
4070 wasn't rewriting it again. This bug also had the effect of breaking the
4071 sender address verification caching when the sender address was rewritten.
4073 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4074 This has been changed so that if the ip literal address matches
4075 ignore_target_hosts, the router declines.
4077 7. Added expansion conditions match_domain, match_address, and match_local_
4078 part (NOT match_host).
4080 8. The placeholder for the Received: header didn't have a length field set.
4082 9. Added code to Exim itself and to exim_lock to test for a specific race
4083 condition that could lead to file corruption when using MBX delivery. The
4084 issue is with the lockfile that is created in /tmp. If this file is removed
4085 after a process has opened it but before that process has acquired a lock,
4086 there is the potential for a second process to recreate the file and also
4087 acquire a lock. This could lead to two Exim processes writing to the file
4088 at the same time. The added code performs the same test as UW imapd; it
4089 checks after acquiring the lock that its file descriptor still refers to
4090 the same named file.
4092 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4093 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4096 11. Added the smtp_active_hostname option. If used, this will typically be made
4097 to depend on the incoming interface address. Because $interface_address is
4098 not set up until the daemon has forked a reception process, error responses
4099 that can happen earlier (such as "too many connections") no longer contain
4102 12. If an expansion in a condition on a "warn" statement fails because a lookup
4103 defers, the "warn" statement is abandoned, and the next ACL statement is
4104 processed. Previously this caused the whole ACL to be aborted.
4106 13. Added the iplsearch lookup type.
4108 14. Added ident_timeout as a log selector.
4110 15. Added tls_certificate_verified as a log selector.
4112 16. Added a global option tls_require_ciphers (compare the smtp transport
4113 option of the same name). This controls incoming TLS connections.
4115 17. I finally figured out how to make tls_require_ciphers do a similar thing
4116 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4117 before starting the TLS session.
4119 18. Tabs are now shown as \t in -bP output.
4121 19. If the log selector return_path_on_delivery was set, Exim crashed when
4122 bouncing a message because it had too many Received: header lines.
4124 20. If two routers both had headers_remove settings, and the first one included
4125 a superfluous trailing colon, the final name in the first list and the
4126 first name in the second list were incorrectly joined into one item (with a
4127 colon in the middle).
4133 1. Added -C and -D options to the exinext utility, mainly to make it easier
4134 to include in the automated testing, but these could be helpful when
4135 multiple configurations are in use.
4137 2. The exinext utility was not formatting the output nicely when there was
4138 an alternate port involved in the retry record key, nor when there was a
4139 message id as well (for retries that were specific to a specific message
4140 and a specific host). It was also confused by IPv6 addresses, because of
4141 the additional colons they contain. I have fixed the IPv4 problem, and
4142 patched it up to do a reasonable job for IPv6.
4144 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4145 delivery, the log line now contains "pipelined" if PIPELINING was used.
4147 4. An SMTP transport process used to panic and die if the bind() call to set
4148 an explicit outgoing interface failed. This has been changed; it is now
4149 treated in the same way as a connect() failure.
4151 5. A reference to $sender_host_name in the part of a conditional expansion
4152 that was being skipped was still causing a DNS lookup. This no longer
4155 6. The def: expansion condition was not recognizing references to header lines
4156 that used bh_ and bheader_.
4158 7. Added the _cache feature to named lists.
4160 8. The code for checking quota_filecount in the appendfile transport was
4161 allowing one more file than it should have been.
4163 9. For compatibility with Sendmail, the command line option
4171 and sets the incoming protocol and host name (for trusted callers). The
4172 host name and its colon can be omitted when only the protocol is to be set.
4173 Note the Exim already has two private options, -pd and -ps, that refer to
4174 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4175 "s", but I don't think that's a major issue.
4177 10. A number of refactoring changes to the code, none of which should affect
4180 (a) The number of logging options was getting close to filling up the
4181 32-bit word that was used as a bit map. I have split them into two classes:
4182 those that are passed in the argument to log_write(), and those that are
4183 only ever tested independently outside of that function. These are now in
4184 separate 32-bit words, so there is plenty of room for expansion again.
4185 There is no change in the user interface or the logging behaviour.
4187 (b) When building, for example, log lines, the code previously used a
4188 macro that called string_cat() twice, in order to add two strings. This is
4189 not really sufficiently general. Furthermore, there was one instance where
4190 it was actually wrong because one of the argument was used twice, and in
4191 one call a function was used. (As it happened, calling the function twice
4192 did not affect the overall behaviour.) The macro has been replaced by a
4193 function that can join an arbitrary number of extra strings onto a growing
4196 (c) The code for expansion conditions now uses a table and a binary chop
4197 instead of a serial search (which was left over from when there were very
4198 few conditions). Also, it now recognizes conditions like "pam" even when
4199 the relevant support is not compiled in: a suitably worded error message is
4200 given if an attempt is made to use such a condition.
4202 11. Added ${time_interval:xxxxx}.
4204 12. A bug was causing one of the ddress fields not to be passed back correctly
4205 from remote delivery subprocesses. The field in question was not being
4206 subsequently used, so this caused to problems in practice.
4208 13. Added new log selectors queue_time and deliver_time.
4210 14. Might have fixed a bug in maildirsizefile handling that threw up
4211 "unexpected character" debug warnings, and recalculated the data
4212 unnecessarily. In any case, I expanded the warning message to give more
4215 15. Added the message "Restricted characters in address" to the statements in
4216 the default ACL that block characters like @ and % in local parts.
4218 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4219 Three changes have been made:
4221 (a) There was a serious bug; a negative response to MAIL caused the whole
4222 recipient domain to be cached as invalid, thereby blocking all messages
4223 to all local parts at the same domain, from all senders. This bug has
4224 been fixed. The domain is no longer cached after a negative response to
4225 MAIL if the sender used is not empty.
4227 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4230 (c) A new callout option, "use_sender" has been added for people who want
4231 the modified behaviour.
4237 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4240 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4241 indeed breaks things for older releases.
4243 3. Added additional logging to the case where there is a problem reading data
4244 from a filter that is running in a subprocess using a pipe, in order to
4245 try to track down a specific problem.
4247 4. Testing facility fudge: when running in the test harness and attempting
4248 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4249 getting "No route to host". Convert this to a timeout.
4251 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4254 6. Some OS don't have socklen_t but use size_t instead. This affects the
4255 fifth argument of getsockopt() amongst other things. This is now
4256 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4257 can be set for individual OS. I have set it for SunOS5, OSF1, and
4258 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4259 some earlier ones do not.
4261 7. Change 4.30/15 was not doing the test caselessly.
4263 8. The standard form for an IPv6 address literal was being rejected by address
4264 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4265 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4266 this, as well as the form without the "IPv6" on the front (but only when
4267 address literals are enabled, of course).
4269 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4271 10. Exim crashed if a message with an empty sender address specified by -f
4272 encountered a router with an errors_to setting. This could be provoked only
4273 by a command such as
4277 where an empty string was supplied; "<>" did not hit this bug.
4279 11. Installed PCRE release 4.5.
4281 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4282 remained set. It is now erased.
4284 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4285 times from message ids (which are base 36 rather than the normal 62).
4287 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4288 were being counted as actual protocol errors, and logged if the log
4289 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4290 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4291 and DATA following a set of rejected RCPTs do not count as protocol errors.
4292 In other words, Exim assumes they were pipelined, though this may not
4293 actually be the case. Of course, in all cases the client gets an
4294 appropriate error code.
4296 15. If a lookup fails in an ACL condition, a message about the failure may
4297 be available; it is used if testing the ACL cannot continue, because most
4298 such messages specify what the cause of the deferral is. However, some
4299 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4300 that caused an old message to be retained and used if a later statement
4301 caused a defer, replacing the real cause of the deferral.
4303 16. If an IP address had so many PTR records that the DNS lookup buffer
4304 was not large enough to hold them, Exim could crash while trying to process
4305 the truncated data. It now detects and logs this case.
4307 17. Further to 4.21/58, another change has been made: if (and only if) the
4308 first line of a message (the first header line) ends with CRLF, a bare LF
4309 in a subsequent header line has a space inserted after it, so as not to
4310 terminate the header.
4312 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4313 unnecessarily, used atoi() instead of strtol(), and didn't check the
4314 termination when getting file sizes from file names by regex.
4316 19. Completely re-implemented the support for maildirsize files, in the light
4317 of a number of problems with the previous contributed implementation
4318 (4.30/29). In particular:
4320 . If the quota is zero, the maildirsize file is maintained, but no quota is
4323 . If the maildir directory does not exist, it is created before any attempt
4324 to write a maildirsize file.
4326 . The quota value in the file is just a cache; if the quota is changed in
4327 the transport, the new value overrides.
4329 . A regular expression is available for excluding directories from the
4332 20. The autoreply transport checks the characters in options that define the
4333 message's headers; it allows continued headers, but it was checking with
4334 isspace() after an embedded newline instead of explicitly looking for a
4337 21. If all the "regular" hosts to which an address was routed had passed their
4338 expiry times, and had not reached their retry times, the address was
4339 bounced, even if fallback hosts were defined. Now Exim should go on to try
4342 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4343 equivalent code in the SMTP transport. Some hosts send humungous responses
4344 to HELO/EHLO, more than 1024 it seems.
4346 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4347 gives compiler warnings in some environments. I've now done it "properly",
4350 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4351 (because of problems with the built-in one) was declared to return uschar *
4352 instead of char *, causing compiler failure.
4354 25. Fixed a file descriptor leak when processing alias/forward files.
4356 26. Fixed a minor format string issue in dbfn.c.
4358 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4360 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4361 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4364 29. When a sender address is verified, it is cached, to save repeating the test
4365 when there is more than one recipient in a message. However, when the
4366 verification involves a callout, it is possible for different callout
4367 options to be set for different recipients. It is too complicated to keep
4368 track of this in the cache, so now Exim always runs a verification when a
4369 callout is required, relying on the callout cache for the optimization.
4370 The overhead is duplication of the address routing, but this should not be
4373 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4374 to be verified with callout=postmaster, and the main callout worked but the
4375 postmaster check failed, the verification correctly failed. However, if a
4376 subsequent RCPT command asked for sender verification *without* the
4377 postmaster check, incorrect caching caused this verification also to fail,
4380 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4381 it was not caching the DNS options (qualify_single, search_parents) that
4382 were used when the lookup failed. A subsequent lookup with different
4383 options therefore always gave the same answer, though there were cases
4384 where it should not have. (Example: a "domains = !$mx_any" option on a
4385 dnslookup router: the "domains" option is always processed without any
4386 widening, but the router might have qualify_single set.) Now Exim uses the
4387 cached value only when the same options are set.
4389 32. Added John Jetmore's "exipick" utility to the distribution.
4391 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4392 than a timeout (e.g. a certificate is required, and is not provided), an
4393 Exim server now closes the connection immediately. Previously it waited for
4394 the client to close - but if the client is SSL, it seems that they each
4395 wait for each other, leading to a delay before one of them times out.
4397 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4398 maintained 0.8.x compatibility because I don't think many are using it, and
4399 it is clearly obsolete.
4401 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4404 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4405 client certificate was expired. A simple patch fixes this, though I don't
4406 understand the full logic of why the verify callback is called multiple
4409 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4410 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4411 which causes problems with some clients (such as the Certicom SSL Plus
4412 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4413 disables the coutermeasure allowing Eudora to connect."
4415 38. Exim was not checking that a write() to a log file succeeded. This could
4416 lead to Bad Things if a log got too big, in particular if it hit a file
4417 size limit. Exim now panics and dies if it cannot write to a log file, just
4418 as it does if it cannot open a log file.
4420 39. Modified OS/Makefile-Linux so that it now contains
4422 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4424 The two -D definitions ensure that Exim is compiled with large file
4425 support, which makes it possible to handle log files that are bigger than
4428 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4429 instance) a domain was checked against a named list that involved a lookup,
4430 causing $domain_data to be set, then another domain was checked against the
4431 same list, then the first domain was re-checked, the value of $domain_data
4432 after the final check could be wrong. In particular, if the second check
4433 failed, it could be set empty. This bug probably also applied to
4436 41. The strip_trailing_dot option was not being applied to the address given
4437 with the -f command-line option.
4439 42. The code for reading a message's header from the spool was incrementing
4440 $received_count, but never initializing it. This meant that the value was
4441 incorrect (doubled) while delivering a message in the same process in which
4442 it was received. In the most common configuration of Exim, this never
4443 happens - a fresh exec is done - but it can happen when
4444 deliver_drop_privilege is set.
4446 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4447 it now includes up to 150 characters of the unexpected data in the log
4450 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4451 and building data strings. The size of both of these buffers was 10 000
4452 bytes - far larger than anybody would *ever* want, thought I. Needless to
4453 say, somebody hit the limit. I have increased the maximum line length to
4454 20 000 and the maximum data length of concatenated lines to 100 000. I have
4455 also fixed two bugs, because there was no checking on these buffers. Tsk,
4456 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4457 buffer is too small.
4459 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4460 lsearch lookups. Now it does.
4462 46. When parsing a route_list item in a manualroute router, a fixed-length
4463 buffer was used for the list of hosts. I made this 1024 bytes long,
4464 thinking that nobody would ever have a list of hosts that long. Wrong.
4465 Somebody had a whole pile of complicated expansion conditions, and the
4466 string was silently truncated, leading to an expansion error. It turns out
4467 that it is easier to change to an unlimited length (owing to other changes
4468 that have happened since this code was originally written) than to build
4469 structure for giving a limitation error. The length of the item that
4470 expands into the list of hosts is now unlimited.
4472 47. The lsearch lookup could not handle data where the length of text line was
4473 more than 4095 characters. Such lines were truncated, leading to shortened
4474 data being returned. It should now handle lines of any length.
4476 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4477 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4480 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4481 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4482 no longer happens in comment lines. A list of replacements is now placed
4483 at the head of all of the source files, except those whose only change is
4484 to replace PERL_COMMAND in the very first #! line.
4486 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4487 messages on the queue, with a bottom-up merge sort, using code contributed
4488 by Michael Haardt. This should make operations like -bp somewhat faster on
4489 large queues. It won't affect queue runners, except when queue_run_in_order
4492 51. Installed eximstats 1.31 in the distribution.
4494 52. Added support for SRV lookups to the dnslookup router.
4496 53. If an ACL referred to $message_body or $message_body_end, the value was not
4497 reset for any messages that followed in the same SMTP session.
4499 54. The store-handling optimization for building very long strings was not
4500 differentiating between the different store pools. I don't think this
4501 actually made any difference in practice, but I've tidied it.
4503 55. While running the routers to verify a sender address, $sender_address
4504 was still set to the sender address. This is wrong, because when routing to
4505 send a bounce to the sender, it would be empty. Therefore, I have changed
4506 it so that, while verifying a sender address, $sender_address is set to <>.
4507 (There is no change to what happens when verifying a recipient address.)
4509 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4510 target A or AAAA records (if not already returned) without resetting the
4511 qualify_single or search_parents options of the DNS resolver. These are
4512 inappropriate in this case because the targets of MX and SRV records must
4513 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4514 target that, when qualified, matched something in the local domain. These
4515 two options are now turned off when doing these lookups.
4517 57. It seems that at least some releases of Reiserfs (which does not have the
4518 concept of a fixed number of inodes) returns zero and not -1 for the
4519 number of available inodes. This interacted badly with check_spool_inodes,
4520 which assumed that -1 was the "no such thing" setting. What I have done is
4521 to check that the total number of inodes is greater than zero before doing
4522 the test of how many are available.
4524 58. When a "warn" ACL statement has a log_message modifier, the message is
4525 remembered, and not repeated. This is to avoid a lot of repetition when a
4526 message has many recipients that cause the same warning to be written.
4527 Howewer, Exim was preserving the list of already written lines for an
4528 entire SMTP session, which doesn't seem right. The memory is now reset if a
4529 new message is started.
4531 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4532 debugging output unless log_rewrite was also set.
4534 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4535 of (int)(handle) when we know that handle contains (void *)(-1).
4537 61. The Exim daemon panic-logs an error return when it closes the incoming
4538 connection. However "connection reset by peer" seems to be common, and
4539 isn't really an error worthy of noting specially, so that particular error
4542 62. When Exim is trying to find all the local interfaces, it used to panic and
4543 die if the ioctl to get the interface flags failed. However, it seems that
4544 on at least one OS (Solaris 9) it is possible to have an interface that is
4545 included in the list of interfaces, but for which you get a failure error
4546 for this call. This happens when the interface is not "plumbed" into a
4547 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4548 failure of the "get flags" call assumes that the interface is down.
4550 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4551 makes life easier for people who are doing arithmetic on fields extracted
4552 from dates, where you often get leading zeros that should not be
4553 interpreted as octal.
4555 64. Added qualify_domain to the redirect router, to override the global
4558 65. If a pathologically long header line contained very many addresses (the
4559 report of this problem mentioned 10 000) and each of them was rewritten,
4560 Exim could use up a very large amount of memory. (It kept on making new
4561 copies of the header line as it rewrote, and never released the old ones.)
4562 At the expense of a bit more processing, the header rewriting function has
4563 been changed so that it no longer eats memory in this way.
4565 66. The generation of the Received: header has been moved from the time that a
4566 message starts to be received, to the time that it finishes. The timestamp
4567 in the Received: header should now be very close to that of the <= log
4568 line. There are two side-effects of this change:
4570 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4571 logged header lines no longer include the local Received: line, because
4572 it has not yet been created. The same applies to a copy of the message
4573 that is returned to a non-SMTP sender when a message is rejected.
4575 (b) When a filter file is tested using -bf, no additional Received: header
4576 is added to the test message. After some thought, I decided that this
4579 This change does not affect the value of $received_for. It is still set
4580 after address rewriting, but before local_scan() is called.
4582 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4584 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4585 gave an unhelpful panic error message, and a defer error. I have managed to
4586 change this behaviour so that it now rejects any supplied certificate,
4587 which seems right, as the list of acceptable certificates is empty.
4589 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4590 gave an unhelpful defer error. I have not managed to make this reject any
4591 supplied certificates, but the error message it gives is "no certificate
4592 supplied", which is not helpful.
4594 70. exigrep's output now also includes lines that are not associated with any
4595 message, but which match the given pattern. Implemented by a patch from
4596 Martin Sluka, which also tidied up the Perl a bit.
4598 71. Recipient callout verification, like sender verification, was using <> in
4599 the MAIL FROM command. This isn't really the right thing, since the actual
4600 sender may affect whether the remote host accepts the recipient or not. I
4601 have changed it to use the actual sender in the callout; this means that
4602 the cache record is now keyed on a recipient/sender pair, not just the
4603 recipient address. There doesn't seem to be a real danger of callout loops,
4604 since a callout by the remote host to check the sender would use <>.
4605 [SEE ABOVE: changed after hitting problems.]
4607 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4608 temporary errors. However, in the case of such a code being given after
4609 the end of a data transmission (i.e. after ".") Exim was failing to write
4610 a retry record for the message. (Yes, there was some broken host that was
4611 actually sending 8xx at this point.)
4613 73. An unknown lookup type in a host list could cause Exim to panic-die when
4614 the list was checked. (An example that provoked this was putting <; in the
4615 middle of a list instead of at the start.) If this happened during a DATA
4616 ACL check, a -D file could be left lying around. This kind of configuration
4617 error no longer causes Exim to die; instead it causes a defer errror. The
4618 incident is still logged to the main and panic logs.
4620 74. Buglet left over from Exim 3 conversion. The message "too many messages
4621 in one connection" was written to the rejectlog but not the mainlog, except
4622 when address rewriting (yes!) was being logged.
4624 75. Added write_rejectlog option.
4626 76. When a system filter was run not as root (that is, when system_filter_user
4627 was set), the values of the $n variables were not being returned to the
4628 main process; thus, they were not subsequently available in the $sn
4631 77. Added +return_path_on_delivery log selector.
4633 78. A connection timeout was being treated differently from recipients deferred
4634 when testing hosts_max_try with a message that was older than the host's
4635 retry timeout. (The host should not be counted, thus allowing all hosts to
4636 be tried at least once before bouncing.) This may have been the cause of an
4637 occasionally reported bug whereby a message would remain on the queue
4638 longer than the retry timeout, but would be bounced if a delivery was
4639 forced. I say "may" because I never totally pinned down the problem;
4640 setting up timeout/retry tests is difficult. See also the next item.
4642 79. The ultimate address timeout was not being applied to errors that involved
4643 a combination of host plus message (for example, a timeout on a MAIL
4644 command). When an address resolved to a number of possible hosts, and they
4645 were not all tried for each delivery (e.g. because of hosts_max_try), a
4646 message could remain on the queue longer than the retry timeout.
4648 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4651 81. Fixed an obscure SMTP outgoing bug which required at least the following
4652 conditions: (a) there was another message waiting for the same server;
4653 (b) the server returned 5xx to all RCPT commands in the first message so
4654 that the message was not completed; (c) the server dropped the connection
4655 or gave a negative response to the RSET that Exim sends to abort the
4656 transaction. The observed case was a dropped connection after DATA that had
4657 been sent in pipelining mode. That is, the server had advertised PIPELINING
4658 but was not implementing it correctly. The effect of the bug was incorrect
4659 behaviour, such as trying another host, and this could lead to a crash.
4665 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4666 and daemon.c were passed as pointers to ints; they should have been
4667 pointers to socklen_t variables (which are typically unsigned ints).
4669 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4672 3. Fixed a really odd bug that affected only the testing scheme; patching a
4673 certain fixed string in the binary changed the value of another string that
4674 happened to be identical to the end of the original first string.
4676 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4677 name", it returns that address as the IP address. On some operating
4678 systems (e.g. Solaris), it also passes back the IP address string as the
4679 "host name". However, on others (e.g. Linux), it passes back an empty
4680 string. Exim wasn't checking for this, and was changing the host name to an
4681 empty string, assuming it had been canonicized.
4683 5. Although rare, it is permitted to have more than one PTR record for a given
4684 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4685 all the names associated with an address, because they do in Solaris.
4686 However, it seems that they do not in Linux for data that comes from the
4687 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4688 I found this out when I moved to a new Linux workstation and tried to run
4689 the Exim test suite.
4691 To get round this problem I have changed the code so that it now does its
4692 own call to the DNS to look up PTR records when searching for a host name.
4693 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4694 addresses that are only in /etc/hosts are still found.
4696 This behaviour is, however, controlled by an option called host_lookup_
4697 order, which defaults to "bydns:byaddr". If people want to use the other
4698 order, or indeed, just use one or the other means of lookup, they can
4699 specify it in this variable.
4701 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4702 some operating systems, this comes back from gethostbyaddr() as an empty
4703 string, and this is what Exim used to test for. However, it seems that in
4704 other systems, "." is yielded. Exim now tests for this case too.
4706 7. The values of check_spool_space and check_log_space are now held internally
4707 as a number of kilobytes instead of an absolute number of bytes. If a
4708 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4709 kilobyte. This means that much larger values can be stored.
4711 8. Exim monitor: an attempt to get the action menu when not actually pointing
4712 at a message produces an empty menu entitled "No message selected". This
4713 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4714 no entries in it ("Shell widget menu has zero width and/or height"). So I
4715 have added a single, blank menu entry in this case.
4717 9. Added ${quote_local_part.
4719 10. MIME decoding is now applied to the contents of Subject: header lines when
4722 11. Now that a reference to $sender_host_address automatically causes a reverse
4723 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4724 host lookup before query-style lookups in lists that might use this
4725 variable. This has therefore been abolished, and the "net-" prefix is no
4726 longer necessary for query-style lookups.
4728 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4729 have been a typo for LFLAGS, so it has been changed.
4731 13. The install script calls Exim with "-C /dev/null" in order to find the
4732 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4733 to be output. Howeve, since Exim outputs its version number before the
4734 error, it didn't break the script. It just looked ugly. I fixed this by
4735 always allowing "-C /dev/null" if the caller is root.
4737 14. Ignore overlarge ACL variable number when reading spool file - insurance
4738 against a later release with more variables having written the file.
4740 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4741 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4742 this, as well as the form without the "IPv6" on the front.
4744 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4745 OS/Makefile-Darwin file.
4747 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4748 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4750 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4751 variable. However, some OS use ints for inodes. I've added cast to long int
4752 to get rid of the compiler warning.
4754 19. I had forgotten to lock out "/../" in configuration file names when
4755 ALT_CONFIG_PREFIX was set.
4757 20. Routers used for verification do not need to specify transports. However,
4758 if such a router generated a host list, and callout was configured, Exim
4759 crashed, because it could not find a port number from the (non-existent)
4760 transport. It now assumes port 25 in this circumstance.
4762 21. Added the -t option to exigrep.
4764 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4765 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4766 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4767 them is set, LOOKUP_LSEARCH is forced.
4769 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4771 24. Added sender and host information to the "rejected by local_scan()" log
4772 line; previously there was no indication of these.
4774 25. Added .include_if_exists.
4776 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4777 Linux. It turns out that not all file systems support this. Apparently some
4778 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4779 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4780 not supported on the file descriptor, is now ignored when Exim is trying to
4781 sync a directory. This applies only to Linux.
4783 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4785 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4786 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4789 29. Added support for maildirsize files from supplied patch (modified a bit).
4791 30. The use of :fail: followed by an empty string could lead Exim to respond to
4792 sender verification failures with (e.g.):
4794 550 Verification failed for <xxx>
4795 550 Sender verify failed
4797 where the first response line was missing the '-' that indicates it is not
4798 the final line of the response.
4800 31. The loop for finding the name of the user that called Exim had a hardwired
4801 limit of 10; it now uses the value of finduser_retries, which is used for
4802 all other user lookups.
4804 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4807 33. Exim was neglecting to zero errno before one call of strtol() when
4808 expanding a string and expecting an integer value. On some systems this
4809 resulted in spurious "integer overflow" errors. Also, it was casting the
4810 result into an int without checking.
4812 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4813 did not work. The code looks as if it has *never* worked, though it appears
4814 to have been documented since at least releast 1.62. I have made it work.
4816 35. The "timeout_DNS" error in retry rules, also documented since at least
4817 1.62, also never worked. As it isn't clear exactly what this means, and
4818 clearly it isn't a major issue, I have abolished the feature by treating it
4819 as "timeout", and writing a warning to the main and panic logs.
4821 36. The display of retry rules for -brt wasn't always showing the error code
4824 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4825 timeout_connect_A, timeout_connect_MX.
4827 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4828 to the empty sender.
4830 39. The daemon was not analysing the content of -oX till after it had closed
4831 stderr and disconnected from the controlling terminal. This meant that any
4832 syntax errors were only noted on the panic log, and the return code from
4833 the command was 0. By re-arranging the code a little, I've made the
4834 decoding happen first, so such errors now appear on stderr, and the return
4835 code is 1. However, the actual setting up of the sockets still happens in
4836 the disconnected process, so errors there are still only recorded on the
4839 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4840 connections (as happens on some IP stacks) was logged at start up time as
4841 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4842 it from "IPv6 and IPv4", which means that two separate sockets are being
4845 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4846 says whether AF_INET or AF_INET6 was passed as an argument.
4848 42. Exiwhat output was messed up when time zones were included in log
4851 43. Exiwhat now gives more information about the daemon's listening ports,
4852 and whether -tls-on-connect was used.
4854 44. The "port" option of the smtp transport is now expanded.
4856 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4857 silently ignored. Now an error message is written to the main and panic
4860 46. There's a new ACL modifier called "logwrite" which writes to a log file
4861 as soon as it is encountered.
4863 47. Added $local_user_uid and $local_user_gid at routing time.
4865 48. Exim crashed when trying to verify a sender address that was being
4868 49. Exim was recognizing only a space character after ".include". It now also
4869 recognizes a tab character.
4871 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4872 extracting the relevant information from the specification. The man page no
4873 longer contains scrambled data for the -d option, and I've added a section
4874 at the front about calling Exim under different names.
4876 51. Added "extra_headers" argument to the "mail" command in filter files.
4878 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4881 53. Installed eximstats 1.29.
4883 54. Added transport_filter_timeout as a generic transport option.
4885 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4886 Cc: header lines. This was required by RFC 822, but it not required by RFC
4889 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4890 incoming messages that did not have them. Now it does so only if the
4891 message originates locally, that is, if there is no associated remote host
4892 address. When Resent- header lines are present, this applies to the Resent-
4893 lines rather than the non-Resent- lines.
4895 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4896 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4898 58. Messages for configuration errors now include the name of the main
4899 configuration file - useful now that there may be more than one file in a
4900 list (.included file names were always shown).
4902 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4903 for those rare installations that do not start the daemon as root or run it
4904 setuid root. I've cut out the call to initgroups() if the daemon is not
4907 60. The Exim user and group can now be bound into the binary as text strings
4908 that are looked up at the start of Exim's processing.
4910 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4912 62. Added $mailstore_basename variable.
4914 63. Installed patch to sieve.c from Michael Haardt.
4916 64. When Exim failed to open the panic log after failing to open the main log,
4917 the original message it was trying to log was written to stderr and debug
4918 output, but if they were not available (the usual case in production), it
4919 was lost. Now it is written to syslog before the two lines that record the
4920 failures to open the logs.
4922 65. Users' Exim filters run in subprocesses under the user's uid. It is
4923 possible for a "deliver" command or an alias in a "personal" command to
4924 provoke an address rewrite. If logging of address rewriting is configured,
4925 this fails because the process is not running as root or exim. There may be
4926 a better way of dealing with this, but for the moment (because 4.30 needs
4927 to be released), I have disabled address rewrite logging when running a
4928 filter in a non-root, non-exim process.
4934 1. The buildconfig auxiliary program wasn't quoting the value set for
4935 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4936 not defined. This bug was masked in 4.22 by the effect that was fixed in
4939 2. Some messages that were rejected after a message id was allocated were
4940 shown as "incomplete" by exigrep. It no longer does this for messages that
4941 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4943 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4944 have allow_domain_literals set, the ID did not get logged in the <= line.
4945 Domain literals are now always recognized in Message-ID: header lines.
4947 4. The first argument for a ${extract expansion item is the key name or field
4948 number. Leading and trailing spaces in this item were not being ignored,
4949 causing some misleading effects.
4951 5. When deliver_drop_privilege was set, single queue runner processes started
4952 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4953 same command in the process it spins off) were not dropping privilege.
4955 6. When the daemon running as "exim" started a queue runner, it always
4956 re-executed Exim in the spun-off process. This is a waste of effort when
4957 deliver_drop_privilege is set. The new process now just calls the
4958 queue-runner function directly.
4964 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4967 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4968 ignored. Though the use of -oP was forcing the writing of a pid file, it
4969 was always written to the default place.
4971 3. If the message "no IP address found for host xxxx" is generated during
4972 incoming verification, it is now followed by identification of the incoming
4973 connection (so you can more easily find what provoked it).
4975 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4977 5. Added some features to "harden" Exim a bit more against certain attacks:
4979 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4980 be put in Local/Makefile. This is like the never_users runtime option,
4981 but it cannot be overridden. The default setting is "root".
4983 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4984 prefix string with which any file named in a -C command line option
4987 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4988 is retained for -C and -D only if the caller of Exim is root. Without
4989 it, the exim user may also use -C and -D and retain privilege.
4991 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4992 command line option is disabled.
4994 6. Macro names set by the -D option must start with an upper case letter, just
4995 like macro names defined in the configuration file.
4997 7. Added "dereference=" facility to LDAP.
4999 8. Two instances of the typo "uknown" in the source files are fixed.
5001 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5002 the Configure-Makefile script screwed up while processing it.
5004 10. Incorporated PCRE 4.4.
5006 11. The SMTP synchronization check was not operating right at the start of an
5007 SMTP session. For example, it could not catch a HELO sent before the client
5008 waited for the greeting. There is now a check for outstanding input at the
5009 point when the greeting is written. Because of the duplex, asynchronous
5010 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5011 way, but not yet received, when the check is performed.
5013 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5014 on TCP/IP sockets, because this apparently causes some broken clients to
5017 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5018 unchanged) from the Cygwin maintainer.
5020 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5021 was supported instead of testing for mbx. Effectively a typo.
5023 15. The spa authenticator server code was not checking that the input it
5024 received was valid base64.
5026 16. The debug output line for the "set" modifier in ACLs was not showing the
5027 name of the variable that was being set.
5029 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5031 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5032 lookup on its own. Something else was needed to trigger the lookup. For
5033 example, a match in host_lookup or the need for a host name in a host list.
5034 Now, if $sender_host_name is referenced and the host name has not yet been
5035 looked up, a lookup is performed. If the lookup fails, the variable remains
5036 empty, and $host_lookup_failed is set to "1".
5038 19. Added "eqi" as a case-independent comparison operator.
5040 20. The saslauthd authentication condition could segfault if neither service
5041 nor realm was specified.
5043 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5044 error message that was logged was misleading, and incoming SMTP
5045 connections were dropped. The message is now more accurate, and temporary
5046 errors are given to SMTP connections.
5048 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5049 (except RSET) with a 421 error, until QUIT is received. However, it was
5050 failing to send a response to QUIT.
5052 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5053 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5054 ACL is now run after the helo_try_verify_hosts code.
5056 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5057 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5058 case-independent, so other case variants are also recognized). Apparently
5059 some systems use these upper case variants.
5061 25. If more than two messages were waiting for the same host, and a transport
5062 filter was specified for the transport, Exim sent two messages over the
5063 same TCP/IP connection, and then failed with "socket operation on non-
5064 socket" when it tried to send the third.
5066 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5068 27. The extern definition of crypt16() in expand.c was not being excluded when
5069 the OS had its own crypt16() function.
5071 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5072 as a preferred synonym for return_size_limit, both as an option and as an
5075 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5077 30. Changed the default configuration ACL to relax the local part checking rule
5078 for addresses that are not in any local domains. For these addresses,
5079 slashes and pipe symbols are allowed within local parts, but the sequence
5080 /../ is explicitly forbidden.
5082 31. SPA server authentication was not clearing the challenge buffer before
5085 32. log_message in a "warn" ACL statement was writing to the reject log as
5086 well as to the main log, which contradicts the documentation and doesn't
5087 seem right (because no rejection is happening). So I have stopped it.
5089 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5090 However, I am unable to do any testing of this.
5092 34. Fixed an infelicity in the appendfile transport. When checking directories
5093 for a mailbox, to see if any needed to be created, it was accidentally
5094 using path names with one or more superfluous leading slashes; tracing
5095 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5097 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5098 contents are added to the log line that is written for every discarded
5099 recipient. (Previously a log_message setting was ignored.)
5101 36. The ${quote: operator now quotes the string if it is empty.
5103 37. The install script runs exim in order to find its version number. If for
5104 some reason other than non-existence or emptiness, which it checks, it
5105 could not run './exim', it was installing it with an empty version number,
5106 i.e. as "exim-". This error state is now caught, and the installation is
5109 38. An argument was missing from the function that creates an error message
5110 when Exim fails to connect to the socket for saslauthd authentication.
5111 This could cause Exim to crash, or give a corrupted message.
5113 39. Added isip, isip4, and isip6 to ${if conditions.
5115 40. The ACL variables $acl_xx are now saved with the message, and can be
5116 accessed later in routers, transports, and filters.
5118 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5119 strings in the file are not string-expanded.
5121 42. If a MAIL command specified a SIZE value that was too large to fit into an
5122 int variable, the check against message_size_limit failed. Such values are
5123 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5124 day this will have to be increased, but I don't think I want to be around
5125 when emails are that large.
5132 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5133 iconv() is not standard in FreeBSD.
5135 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5136 IPv6 enabled. The observed symptom was a segmentation fault on return from
5137 the function os_common_find_running_interfaces() in src/os.c.
5139 3. In the check_special_case() function in daemon.c I had used "errno" as an
5140 argument name, which causes warnings on some systems. This was basically a
5141 typo, since it was named "eno" in the comments!
5143 4. The code that waits for the clock to tick (at a resolution of some fraction
5144 of a second) so as to ensure message-id uniqueness was always waiting for
5145 at least one whole tick, when it could have waited for less. [This is
5146 almost certainly not relevant at current processor speeds, where it is
5147 unlikely to ever wait at all. But we try to future-proof.]
5149 5. The function that sleeps for a time interval that includes fractions of a
5150 second contained a race. It did not block SIGALRM between setting the
5151 timer, and suspending (a couple of lines later). If the interval was short
5152 and the sigsuspend() was delayed until after it had expired, the suspension
5153 never ended. On busy systems this could lead to processes getting stuck for
5156 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5157 process, before it forks any delivery processes. The open lookup caching
5158 mechanism meant that the open file or database connection was passed into
5159 the delivery process. The problem was that delivery processes always tidy
5160 up cached lookup data. This could cause a problem for the next delivery
5161 process started by the queue runner, because the external queue runner
5162 process does not know about the closure. So the next delivery process
5163 still has data in the lookup cache. In the case of a file lookup, there was
5164 no problem because closing a file descriptor in a subprocess doesn't affect
5165 the parent. However, if the lookup was caching a connection to a database,
5166 the connection was closed, and the second delivery process was likely to
5167 see errors such as "PGSQL: query failed: server closed the connection
5168 unexpectedly". The problem has been fixed by closing all cached lookups
5169 in a queue runner before running a delivery process.
5171 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5172 seem to have the "const" qualifier which it has on other OS. I've
5175 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5176 *of the same type* (client or server) with the same public name that an
5177 error should be diagnosed.
5179 9. When Exim looked up a host name for an IP address, but failed to find the
5180 original IP address when looking up the host name (a safety check), it
5181 output the message "<ip address> does not match any IP for NULL", which was
5182 confusing, to say the least. The bug was that the host name should have
5183 appeared instead of "NULL".
5185 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5186 user that is built into the binary, and the -C or -D options is used, root
5187 privilege is dropped before the configuration file is read. In addition,
5188 logging is switched to stderr instead of the normal log files. If the
5189 configuration then re-defines the Exim user, the unprivileged environment
5190 is probably not what is expected, so Exim logs a panic warning message (but
5193 However, if deliver_drop_privilege is set, the unprivileged state may well
5194 be exactly what is intended, so the warning has been cut out in that case,
5195 and Exim is allowed to try to write to its normal log files.
5201 1. smtp_return_error_details was not giving details for temporary sender
5202 or receiver verification errors.
5204 2. Diagnose a configuration error if two authenticators have the same public
5207 3. Exim used not to create the message log file for a message until the first
5208 delivery attempt. This could be confusing when incoming messages were held
5209 for policy or load reasons. The message log file is now created at the time
5210 the message is received, and an initial "Received" line is written to it.
5212 4. The automatically generated man page for command line options had a minor
5213 bug that caused no ill effects; however, a more serious problem was that
5214 the procedure for building the man page automatically didn't always
5215 operate. Consequently, release 4.20 contains an out-of-date version. This
5216 shouldn't happen again.
5218 5. When building Exim with embedded Perl support, the script that builds the
5219 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5220 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5222 6. The freeze_tell option was not being used for messages that were frozen on
5223 arrival, either by an ACL or by local_scan().
5225 7. Added the smtp_incomplete_transaction log selector.
5227 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5228 was accepting AUTH without a new EHLO.
5230 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5231 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5234 10. Logging of TCP/IP connections (when configured) now happens in the main
5235 daemon process instead of the child process, so that the TCP/IP connection
5236 count is more accurate (but it can never be perfect).
5238 11. The use of "drop" in a nested ACL was not being handled correctly in the
5239 outer ACL. Now, if condition failure induced by the nested "drop" causes
5240 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5241 or "require"), the connection is dropped.
5243 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5244 that yield "discard" can now be used with an "accept" or a "discard" verb,
5245 but an error is generated for any others (because I can't see a useful way
5246 to define what should happen).
5248 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5249 are now processed in the same way as lines in the Exim configuration file.
5250 In particular, continuation lines are supported.
5252 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5254 15. Added -ti meaning -t -i.
5256 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5257 domains, and warn by logging if others are found.
5259 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5260 returned from GIFCONF when Exim is trying to find the list of interfaces on
5261 a host. The code in os.c has been modified to copy each ifreq to an aligned
5262 structure in all cases.
5264 Also, in some cases, the returned ifreq's were being copied to a 'struct
5265 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5266 means the last couple of bytes of an IPv6 address could be chopped if the
5267 ifreq contained only a normal sockaddr (14 bytes storage).
5269 18. Named domain lists were not supported in the hosts_treat_as_local option.
5270 An entry such as +xxxx was not recognized, and was treated as a literal
5273 19. Ensure that header lines added by a DATA ACL are included in the reject log
5274 if the ACL subsequently rejects the message.
5276 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5277 MD5 (which is deprecated).
5279 21. When testing a filter file using -bf, Exim was writing a message when it
5280 took the sender from a "From " line in the message, but it was not doing so
5281 when it took $return_path from a Return-Path: header line. It now does.
5283 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5284 with a valid header line field name (a series of printing characters
5285 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5287 23. Changed "disc" in the source to "disk" to conform to the documentation and
5288 the book and for uniformity.
5290 24. Ignore Sendmail's -Ooption=value command line item.
5292 25. When execve() failed while trying to run a command in a pipe transport,
5293 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5294 could be confused with a return value of 69 from the command itself. This
5295 has been changed to 127, the value the shell returns if it is asked to run
5296 a non-existent command. The wording for the related log line suggests a
5297 non-existent command as the problem.
5299 26. If received_header_text expands to an empty string, do not add a Received:
5300 header line to the message. (Well, it adds a token one on the spool, but
5301 marks it "old" so that it doesn't get used or transmitted.)
5303 27. Installed eximstats 1.28 (addition of -nt option).
5305 28. There was no check for failure on the call to getsockname() in the daemon
5306 code. This can fail if there is a shortage of resources on the system, with
5307 ENOMEM, for example. A temporary error is now given on failure.
5309 29. Contrary to the C standard, it seems that in some environments, the
5310 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5311 program. Exim now does this explicitly; it affects the formatting of
5312 timestamps using strftime().
5314 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5315 complaints. I've now initialized all the variables, to avoid this.
5317 32. Header lines added by a system filter were not being "seen" during
5318 transport-time rewrites.
5320 33. The info_callback() function passed to OpenSSL is set up with type void
5321 (*)(SSL *, int, int), as described somewhere. However, when calling the
5322 function (actually a macro) that sets it up, the type void(*)() is
5323 expected. I've put in a cast to prevent warnings from picky compilers.
5325 34. If a DNS black list lookup found a CNAME record, but there were no A
5326 records associated with the domain it pointed at, Exim crashed.
5328 35. If a DNS black list lookup returned more than one A record, Exim ignored
5329 all but the first. It now scans all returned addresses if a particular IP
5330 value is being sought. In this situation, the contents of the
5331 $dnslist_value variable are a list of all the addresses, separated by a
5334 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5335 to accept a host name and all its aliases if the forward lookup for any of
5336 them yielded the IP address of the incoming connection. Now it accepts only
5337 those names whose forward lookup yields the correct IP address. Any other
5338 names are discarded. This closes a loophole whereby a rogue DNS
5339 administrator could create reverse DNS records to break through a
5340 wildcarded host restriction in an ACL.
5342 37. If a user filter or a system filter that ran in a subprocess used any of
5343 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5344 the wrong values were passed to the pipe command ($thisaddress had the
5345 value of $0, $0 had the value of $1, etc). This bug was introduced by
5346 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5348 38. Improved the line breaking for long SMTP error messages from ACLs.
5349 Previously, if there was no break point between 40 and 75 characters, Exim
5350 left the rest of the message alone. Two changes have been made: (a) I've
5351 reduced the minimum length to 35 characters; (b) if it can't find a break
5352 point between 35 and 75 characters, it looks ahead and uses the first one
5353 that it finds. This may give the occasional overlong line, but at least the
5354 remaining text gets split now.
5356 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5357 file descriptors might be low, and that setting 1000 would always raise it.
5358 It turns out that in some environments, the limit is already over 1000 and
5359 that lowering it causes trouble. So now Exim takes care not to decrease it.
5361 40. When delivering a message, the value of $return_path is set to $sender_
5362 address at the start of routing (routers may change the value). By an
5363 oversight, this default was not being set up when an address was tested by
5364 -bt or -bv, which affected the outcome if any router or filter referred to
5367 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5368 log only when "message" or "log_message" are set. However, if one of the
5369 conditions was an address verification, or a call to a nested ACL, the
5370 messages generated by the underlying test were being passed through. This
5371 no longer happens. The underlying message is available in $acl_verify_
5372 message for both "message" and "log_message" expansions, so it can be
5373 passed through if needed.
5375 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5376 new expansion $bh_ to give the encoded byte string without charset
5377 translation. Translation happens only if iconv() is available; HAVE_ICONV
5378 indicates this at build time. HEADERS_CHARSET gives the charset to
5379 translate to; headers_charset can change it in the configuration, and
5380 "headers charset" can change it in an individual filter file.
5382 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5383 that creates RFC 2047 encoded "words" labels them as that charset instead
5384 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5385 expansion operator; (ii) when Exim creates a From: line for a local
5386 message; (iii) when a header line is rewritten to include a "phrase" part.
5388 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5389 buggy, causing it to skip the first lines of messages whose message ID
5390 ended in 'D'. This would not have bitten before Exim release 4.14, because
5391 message IDs were unlikely to end in 'D' before then. The effect was to have
5392 incorrect size information for certain domains.
5394 45. #include "config.h" was missing at the start of the crypt16.c module. This
5395 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5398 46. If there was a timeout during a "random" callout check, Exim treated it as
5399 a failure of the random address, and carried on sending RSET and the real
5400 address. If the delay was just some slowness somewhere, the response to the
5401 original RCPT would be taken as a response to RSET and so on, causing
5402 mayhem of various kinds.
5404 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5405 when I implemented it. It didn't allow for the fact that some option values
5406 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5407 the right test for positive values.
5409 48. Domain names in DNS records are case-independent. Exim always looks them up
5410 in lower case. Some resolvers return domain names in exactly the case they
5411 appear in the zone file, that is, they may contain uppercase letters. Not
5412 all resolvers do this - some return always lower case. Exim was treating a
5413 change of case by a resolver as a change of domain, similar to a widening
5414 of a domain abbreviation. This triggered its re-routing code and so it was
5415 trying to route what was effectively the same domain again. This normally
5416 caused routing to fail (because the router wouldn't handle the domain
5417 twice). Now Exim checks for this case specially, and just changes the
5418 casing of the domain that it ultimately uses when it transmits the message
5421 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5424 50. If a filter generated a file delivery with a non-absolute name (possible if
5425 no home directory exists for the router), the forbid_file option was not
5428 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5429 the existing equality matching.
5431 52. Exim was using ints instead of ino_t variables in some places where it was
5432 dealing with inode numbers.
5434 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5435 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5436 called TMPDIR, and if it finds it is different, it changes its value.
5438 54. The smtp_printf() function is now made available to local_scan() so
5439 additional output lines can be written before returning. There is also an
5440 smtp_fflush() function to enable the detection of a dropped connection.
5441 The variables smtp_input and smtp_batched_input are exported to
5444 55. Changed the default runtime configuration: the message "Unknown user"
5445 has been removed from the ACL, and instead placed on the localuser router,
5446 using the cannot_route_message feature. This means that any verification
5447 failures that generate their own messages won't get overridden. Similarly,
5448 the "Unrouteable address" message that was in the ACL for unverifiable
5449 relay addresses has also been removed.
5451 56. Added hosts_avoid_esmtp to the smtp transport.
5453 57. The exicyclog script was not checking for the esoteric option
5454 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5455 will work only if exicyclog is run under the appropriate euid.
5457 58. Following a discussion on the list, the rules by which Exim recognises line
5458 endings on incoming messages have been changed. The -dropcr and drop_cr
5459 options are now no-ops, retained only for backwards compatibility. The
5460 following line terminators are recognized: LF CRLF CR. However, special
5461 processing applies to CR:
5463 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5464 nor a local message in the state where . is a terminator.
5466 (ii) If a bare CR is encountered in a header line, an extra space is added
5467 after the line terminator so as not to end the header. The reasoning
5468 behind this is that bare CRs in header lines are most likely either
5469 to be mistakes, or people trying to play silly games.
5471 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5472 was being incorrectly given as 18 bytes larger than it should have been.
5473 This is a VOB (very old bug).
5475 60. This may never have affected anything current, but just in case it has:
5476 When the local host is found other than at the start of a list of hosts,
5477 the local host, those with the same MX, and any that follow, are discarded.
5478 When the list in question was part of a longer list of hosts, the following
5479 hosts (not currently being processed) were also being discarded. This no
5480 longer happens. I'm not sure if this situation could ever has previously
5483 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5486 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5487 header. This is something that is recommended in a new Internet Draft, and
5488 is something that is documented as being done by Sendmail. There are two
5489 possible values. For messages generated by the autoreply transport, Exim
5492 Auto-Submitted: auto-replied
5494 whereas for all other generated messages (e.g. bounces) it adds
5496 Auto-Submitted: auto-generated
5498 63. The "personal" condition in filters now includes a test for the
5499 Auto-Submitted: header. If it contains the string "auto-" the message it
5500 not considered personal.
5502 64. Added rcpt_include_affixes as a generic transport option.
5504 65. Added queue_only_override (default true).
5506 66. Added the syslog_duplication option.
5508 67. If what should have been the first header line of a message consisted of
5509 a space followed by a colon, Exim was mis-interpreting it as a header line.
5510 It isn't of course - it is syntactically invalid and should therefore be
5511 treated as the start of the message body. The misbehaviour could have
5512 caused a number of strange effects, including loss of data in subsequent
5513 header lines, and spool format errors.
5515 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5516 client host had authenticated. This control can now be exercised by an ACL
5517 for more flexibility.
5519 69. By default, callouts do not happen when testing with -bh. There is now a
5520 variant, -bhc, which does actually run the callout code, including
5521 consulting and updating the callout cache.
5523 70. Added support for saslauthd authentication, courtesy of Alexander
5526 71. If statvfs() failed on the spool or log directories while checking their
5527 size for availability, Exim confusingly gave the error "space shortage".
5528 Furthermore, in debugging mode it crashed with a floating point exception.
5529 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5530 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5531 serious problem, Exim now writes to the main and panic logs when this
5532 happens, with details of the failure. It then refuses to accept the
5533 incoming message, giving the message "spool directory problem" or "log
5534 directory problem" with a 421 code for SMTP messages.
5536 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5537 0, 1, and 2 exist, because some OS complain for execs without them (see
5538 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5539 descriptors. However, the code omitted to check that the open succeeded,
5540 causing mysterious errors if for some reason the permissions on /dev/null
5541 got screwed. Now Exim writes a message to the main and panic logs, and
5542 bombs out if it can't open /dev/null.
5544 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5545 interact so that it is all more flexible. It is supposed to remain
5546 backwards compatible. Also added extra_local_interfaces.
5548 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5549 to bomb out with an assertion failure - to the client this appears as a
5550 connection drop. This problem occurs in the part of the code that was taken
5551 from the Samba project. Fortunately, the assertion is in a very simple
5552 function, so I have fixed this by reproducing the function inline in the
5553 one place where it is called, and arranging for authentication to fail
5554 instead of killing the process with assert().
5556 75. The SPA client code was not working when the server requested OEM rather
5557 than Unicode encoding.
5559 76. Added code to make require_files with a specific uid setting more usable in
5560 the case where statting the file as root fails - usually a non-root-mounted
5561 NFS file system. When this happens and the failure is EACCES, Exim now
5562 forks a subprocess and does the per-uid checking as the relevant uid.
5564 77. Added process_log_path.
5566 78. If log_file_path was not explicitly set, a setting of check_log_space or
5567 check_log_inodes was ignored.
5569 79. If a space check for the spool or log partitions fails, the incident is now
5570 logged. Of course, in the latter case the data may get lost...
5572 80. Added the %p formatting code to string_format() so that it can be used to
5573 print addresses in debug_print(). Adjusted all the address printing in the
5574 debugging in store.c to use %p rather than %d.
5576 81. There was a concern that a line of code in smtp_in.c could overflow a
5577 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5578 initially expressed, the concern was not well-founded, because trailing
5579 spaces are removed early. However, if the trailing spaces were followed by
5580 a NULL, they did not get removed, so the overflow was possible. Two fixes
5583 (a) I re-wrote the offending code in a cleaner fashion.
5584 (b) If an incoming SMTP command contains a NULL character, it is rejected
5587 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5588 runs initgroups(), so that if the Exim user is in any additional groups,
5589 they will be used during message reception.
5595 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob