1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
81 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
82 NULL dereference. Report and patch from Alun Jones.
84 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
85 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
86 Not seeing resolver debug output on NetBSD, but suspect this is a
87 resolver implementation change.
93 PP/01 Solaris build fix for Oracle's LDAP libraries.
94 Bugzilla 1109, patch from Stephen Usher.
96 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
98 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
101 TF/02 Fix a couple more cases where we did not log the error message
102 when unlink() failed. See also change 4.74-TF/03.
104 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
105 lock up or crash if it happened to be inside a call to libc when it
106 got a SIGUSR1 from exiwhat.
108 The SIGUSR1 handler appends the current process status to the process
109 log which is later printed by exiwhat. It used to use the general
110 purpose logging code to do this, but several functions it calls are
111 not safe for signals.
113 The new output code in the SIGUSR1 handler is specific to the process
114 log, and simple enough that it's easy to inspect for signal safety.
115 Removing some special cases also simplifies the general logging code.
116 Removing the spurious timestamps from the process log simplifies
119 TF/04 Improved ratelimit ACL condition.
121 The /noupdate option has been deprecated in favour of /readonly which
122 has clearer semantics. The /leaky, /strict, and /readonly update modes
123 are mutually exclusive. The update mode is no longer included in the
124 database key; it just determines when the database is updated. (This
125 means that when you upgrde Exim will forget old rate measurements.)
127 Exim now checks that the per_* options are used with an update mode that
128 makes sense for the current ACL. For example, when Exim is processing a
129 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
130 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
131 must specify per_mail/readonly. If you omit the update mode it defaults to
132 /leaky where that makes sense (as before) or /readonly where required.
134 The /noupdate option is now undocumented but still supported for
135 backwards compatibility. It is equivalent to /readonly except that in
136 ACLs where /readonly is required you may specify /leaky/noupdate or
137 /strict/noupdate which are treated the same as /readonly.
139 A useful new feature is the /count= option. This is a generalization
140 of the per_byte option, so that you can measure the throughput of other
141 aggregate values. For example, the per_byte option is now equivalent
142 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
144 The per_rcpt option has been generalized using the /count= mechanism
145 (though it's more complicated than the per_byte equivalence). When it is
146 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
147 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
148 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
149 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
150 Note that using per_rcpt with a non-readonly update mode in more than
151 one ACL will cause the recipients to be double-counted. (The per_mail
152 and per_byte options don't have this problem.)
154 The handling of very low rates has changed slightly. If the computed rate
155 is less than the event's count (usually one) then this event is the first
156 after a long gap. In this case the rate is set to the same as this event's
157 count, so that the first message of a spam run is counted properly.
159 The major new feature is a mechanism for counting the rate of unique
160 events. The new per_addr option counts the number of different
161 recipients that someone has sent messages to in the last time period. It
162 behaves like per_rcpt if all the recipient addresses are different, but
163 duplicate recipient addresses do not increase the measured rate. Like
164 the /count= option this is a general mechanism, so the per_addr option
165 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
166 example, measure the rate that a client uses different sender addresses
167 with the options per_mail/unique=$sender_address. There are further
168 details in the main documentation.
170 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
172 TF/06 Removed a few PCRE remnants.
174 TF/07 Automatically extract Exim's version number from tags in the git
175 repository when doing development or release builds.
177 PP/02 Raise smtp_cmd_buffer_size to 16kB.
178 Bugzilla 879. Patch from Paul Fisher.
180 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
181 Heavily based on revision 40f9a89a from Simon Arlott's tree.
184 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
186 PP/05 Variable $av_failed, true if the AV scanner deferred.
187 Bugzilla 1078. Patch from John Horne.
189 PP/06 Stop make process more reliably on build failure.
190 Bugzilla 1087. Patch from Heiko Schlittermann.
192 PP/07 Make maildir_use_size_file an _expandable_ boolean.
193 Bugzilla 1089. Patch from Heiko Schlittermann.
195 PP/08 Handle ${run} returning more data than OS pipe buffer size.
196 Bugzilla 1131. Patch from Holger Weiß.
198 PP/09 Handle IPv6 addresses with SPF.
199 Bugzilla 860. Patch from Wolfgang Breyha.
201 PP/10 GnuTLS: support TLS 1.2 & 1.1.
203 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
206 PP/11 match_* no longer expand right-hand-side by default.
207 New compile-time build option, EXPAND_LISTMATCH_RHS.
208 New expansion conditions, "inlist", "inlisti".
210 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
212 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
214 PP/14 fix log_write() format string regression from TF/03.
215 Bugzilla 1152. Patch from Dmitry Isaikin.
221 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
223 PP/02 Harmonised TLS library version reporting; only show if debugging.
224 Layout now matches that introduced for other libraries in 4.74 PP/03.
226 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
228 PP/04 New "dns_use_edns0" global option.
230 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
233 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
234 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
236 TK/01 Updated PolarSSL code to 0.14.2.
237 Bugzilla 1097. Patch from Andreas Metzler.
239 PP/07 Catch divide-by-zero in ${eval:...}.
242 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
245 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
246 format-string attack -- SECURITY: remote arbitrary code execution.
248 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
249 time unintentionally subject to list matching rules, letting the header
250 cause arbitrary Exim lookups (of items which can occur in lists, *not*
251 arbitrary string expansion). This allowed for information disclosure.
253 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
254 INT_MIN/-1 -- value coerced to INT_MAX.
260 NM/01 Workround for PCRE version dependency in version reporting
263 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
264 This fixes portability to compilers other than gcc, notably
265 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
267 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
268 makefiles for portability to HP-UX and POSIX correctness.
270 PP/01 Permit LOOKUP_foo enabling on the make command-line.
271 Also via indented variable definition in the Makefile.
272 (Debugging by Oliver Heesakkers).
274 PP/02 Restore caching of spamd results with expanded spamd_address.
275 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
277 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
278 Improves build reliability. Fix from: Frank Elsner
280 NM/02 Fix wide character breakage in the rfc2047 coding
281 Fixes bug 1064. Patch from Andrey N. Oktyabrski
283 NM/03 Allow underscore in dnslist lookups
284 Fixes bug 1026. Patch from Graeme Fowler
286 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
287 Code patches from Adam Ciarcinski of NetBSD.
289 NM/04 Fixed exiqgrep to cope with mailq missing size issue
292 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
293 is logged, to avoid truncation. Patch from John Horne.
295 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
296 Patch from Jakob Hirsch.
298 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
299 SQL string expansion failure details.
300 Patch from Andrey Oktyabrski.
302 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
303 Patch from Simon Arlott.
305 PP/09 New lookups functionality failed to compile on old gcc which rejects
306 extern declarations in function scope.
307 Patch from Oliver Fleischmann
309 PP/10 Use sig_atomic_t for flags set from signal handlers.
310 Check getgroups() return and improve debugging.
311 Fixed developed for diagnosis in bug 927 (which turned out to be
314 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
315 Patch from Mark Zealey.
317 PP/12 Bugzilla 1056: Improved spamd server selection.
318 Patch from Mark Zealey.
320 PP/13 Bugzilla 1086: Deal with maildir quota file races.
321 Based on patch from Heiko Schlittermann.
323 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
324 Patch from Uwe Doering, sign-off by Michael Haardt.
326 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
327 variable declaration deep within a block. Bug and patch from
330 PP/15 lookups-Makefile IRIX compatibilty coercion.
332 PP/16 Make DISABLE_DKIM build knob functional.
334 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
335 Patch by Simon Arlott
337 TF/03 Fix valgrind.h portability to C89 compilers that do not support
338 variable argument macros. Our copy now differs from upstream.
344 TF/01 Failure to get a lock on a hints database can have serious
345 consequences so log it to the panic log.
347 TF/02 Log LMTP confirmation messages in the same way as SMTP,
348 controlled using the smtp_confirmation log selector.
350 TF/03 Include the error message when we fail to unlink a spool file.
352 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
353 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
354 for maintaining out-of-tree patches for some time.
356 PP/01 Bugzilla 139: Documentation and portability issues.
357 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
358 Handle per-OS dynamic-module compilation flags.
360 PP/02 Let /dev/null have normal permissions.
361 The 4.73 fixes were a little too stringent and complained about the
362 permissions on /dev/null. Exempt it from some checks.
363 Reported by Andreas M. Kirchwitz.
365 PP/03 Report version information for many libraries, including
366 Exim version information for dynamically loaded libraries. Created
367 version.h, now support a version extension string for distributors
368 who patch heavily. Dynamic module ABI change.
370 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
371 privilege escalation vulnerability whereby the Exim run-time user
372 can cause root to append content of the attacker's choosing to
375 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
378 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
379 If dropping privileges for untrusted macros, we disabled normal logging
380 on the basis that it would fail; for the Exim run-time user, this is not
381 the case, and it resulted in successful deliveries going unlogged.
382 Fixed. Reported by Andreas Metzler.
388 PP/01 Date: & Message-Id: revert to normally being appended to a message,
389 only prepend for the Resent-* case. Fixes regression introduced in
390 Exim 4.70 by NM/22 for Bugzilla 607.
392 PP/02 Include check_rfc2047_length in configure.default because we're seeing
393 increasing numbers of administrators be bitten by this.
395 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
397 PP/03 Bugzilla 994: added openssl_options main configuration option.
399 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
401 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
403 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
405 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
406 without a peer certificate, leading to a segfault because of an
407 assumption that peers always have certificates. Be a little more
408 paranoid. Problem reported by Martin Tscholak.
410 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
411 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
412 NB: ClamAV planning to remove STREAM in "middle of 2010".
413 CL also introduces -bmalware, various -d+acl logging additions and
414 more caution in buffer sizes.
416 PP/09 Implemented reverse_ip expansion operator.
418 PP/10 Bugzilla 937: provide a "debug" ACL control.
420 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
422 PP/12 Bugzilla 973: Implement --version.
424 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
426 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
428 PP/15 Bugzilla 816: support multiple condition rules on Routers.
430 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
431 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
432 ignore trailing whitespace.
434 JJ/02 prevent non-panic DKIM error from being sent to paniclog
436 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
439 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
440 Notification from Dr Andrew Aitchison.
442 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
443 ExtendedDetectionInfo response format.
444 Notification from John Horne.
446 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
449 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
450 XSL and documented dependency on system catalogs, with examples of how
453 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
456 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
457 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
458 configuration file which is writeable by the Exim user or group.
460 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
461 of configuration files to cover files specified with the -C option if
462 they are going to be used with root privileges, not just the default
465 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
466 option (effectively making it always true).
468 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
469 files to be used while preserving root privileges.
471 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
472 that rogue child processes cannot use them.
474 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
475 run-time user, instead of root.
477 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
478 Exim run-time user without dropping privileges.
480 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
481 result string, instead of calling string_vformat() twice with the same
484 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
485 for other users. Others should always drop root privileges if they use
486 -C on the command line, even for a whitelisted configure file.
488 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
490 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
496 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
497 $data_path, and $header_path variables; fixed documentation bugs and
500 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
501 exipick to access non-standard spools, including the "frozen" queue
504 NM/01 Bugzilla 965: Support mysql stored procedures.
505 Patch from Alain Williams
507 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
509 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
510 Patch from Andreas Metzler
512 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
513 Patch from Kirill Miazine
515 NM/05 Bugzilla 671: Added umask to procmail example.
517 JJ/03 installed exipick 20100323.0, fixing doc bug
519 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
520 directory. Notification and patch from Dan Rosenberg.
522 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
524 TK/02 Improve log output when DKIM signing operation fails.
526 MH/01 Treat the transport option dkim_domain as a colon separated
527 list, not as a single string, and sign the message with each element,
528 omitting multiple occurences of the same signer.
530 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
531 Bugzilla 985, 986. Patch by Simon Arlott
533 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
534 Patch by Simon Arlott
536 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
537 MBX locking. Notification from Dan Rosenberg.
543 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
545 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
547 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
549 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
551 NM/04 Bugzilla 915: Flag broken perl installation during build.
557 TK/01 Added patch by Johannes Berg that expands the main option
558 "spamd_address" if it starts with a dollar sign.
560 TK/02 Write list of recipients to X-Envelope-Sender header when building
561 the mbox-format spool file for content scanning (suggested by Jakob
564 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
565 (http://www.dcc-servers.net/) support via dccifd. Activated by
566 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
568 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
569 by Mark Daniel Reidel <mr@df.eu>.
571 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
572 When building exim an external PCRE library is now needed -
573 PCRE is a system library on the majority of modern systems.
574 See entry on PCRE_LIBS in EDITME file.
576 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
577 conversation. Added nologin parameter to request.
578 Patch contributed by Kirill Miazine.
580 TF/01 Do not log submission mode rewrites if they do not change the address.
582 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
584 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
585 log files in place. Contributed by Roberto Lima.
587 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
589 TF/03 Bugzilla 615: When checking the local_parts router precondition
590 after a local_part_suffix or local_part_prefix option, Exim now
591 does not use the address's named list lookup cache, since this
592 contains cached lookups for the whole local part.
594 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
595 Robert Millan. Documentation is in experimental-spec.txt.
597 TF/04 Bugzilla 668: Fix parallel build (make -j).
599 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
601 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
602 Patch provided by Jan Srzednicki.
604 TF/05 Leading white space used to be stripped from $spam_report which
605 wrecked the formatting. Now it is preserved.
607 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
608 that they are available at delivery time.
610 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
612 TF/08 TLS error reporting now respects the incoming_interface and
613 incoming_port log selectors.
615 TF/09 Produce a more useful error message if an SMTP transport's hosts
616 setting expands to an empty string.
618 NM/06 Bugzilla 744: EXPN did not work under TLS.
619 Patch provided by Phil Pennock.
621 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
622 Patch provided by Richard Godbee.
624 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
625 acl_smtp_notquit, added index entry.
627 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
628 Patch provided by Eugene Bujak.
630 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
631 accept(). Patch provided by Maxim Dounin.
633 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
634 Patch provided by Phil Pennock.
636 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
638 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
639 Patch provided by Brad "anomie" Jorsch.
641 NM/14 Bugzilla 622: Added timeout setting to transport filter.
642 Patch provided by Dean Brooks.
644 TK/05 Add native DKIM support (does not depend on external libraries).
646 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
647 Patch provided by Graeme Fowler.
649 NM/16 Bugzilla 851: Documentation example syntax fix.
651 NM/17 Changed NOTICE file to remove references to embedded PCRE.
653 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
656 NM/19 Bugzilla 745: TLS version reporting.
657 Patch provided by Phil Pennock.
659 NM/20 Bugzilla 167: bool: condition support.
660 Patch provided by Phil Pennock.
662 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
663 clients. Patch provided by Phil Pennock.
665 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
666 Patch provided by Brad "anomie" Jorsch.
668 NM/23 Bugzilla 687: Fix misparses in eximstats.
669 Patch provided by Heiko Schlittermann.
671 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
672 Patch provided by Heiko Schlittermann.
674 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
675 plus update to original patch.
677 NM/26 Bugzilla 799: Documentation correction for ratelimit.
679 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
680 Patch provided by David Brownlee.
682 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
684 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
686 NM/30 Bugzilla 888: TLS documentation bugfixes.
688 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
690 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
691 Unlike the original bugzilla I have changed all shell scripts in src tree.
693 NM/33 Bugzilla 898: Transport filter timeout fix.
694 Patch by Todd Rinaldo.
696 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
697 Patch by Serge Demonchaux.
699 NM/35 Bugzilla 39: Base64 decode bug fixes.
700 Patch by Jakob Hirsch.
702 NM/36 Bugzilla 909: Correct connect() call in dcc code.
704 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
706 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
708 NM/39 Bugzilla 911: Fixed MakeLinks build script.
714 TK/01 Add preliminary DKIM support. Currently requires a forked version of
715 ALT-N's libdkim that I have put here:
716 http://duncanthrax.net/exim-experimental/
718 Note to Michael Haardt: I had to rename some vars in sieve.c. They
719 were called 'true' and it seems that C99 defines that as a reserved
720 keyword to be used with 'bool' variable types. That means you could
721 not include C99-style headers which use bools without triggering
722 build errors in sieve.c.
724 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
725 as mailq or other aliases. Changed the --help handling significantly
726 to do whats expected. exim_usage() emits usage/help information.
728 SC/01 Added the -bylocaldomain option to eximstats.
730 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
732 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
734 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
740 PH/01 Another patch from the Sieve maintainer.
742 PH/02 When an IPv6 address is converted to a string for single-key lookup
743 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
744 dots are used instead of colons so that keys in lsearch files need not
745 contain colons. This was done some time before quoting was made available
746 in lsearch files. However, iplsearch files do require colons in IPv6 keys
747 (notated using the quote facility) so as to distinguish them from IPv4
748 keys. This meant that lookups for IP addresses in host lists did not work
749 for iplsearch lookups.
751 This has been fixed by arranging for IPv6 addresses to be expressed with
752 colons if the lookup type is iplsearch. This is not incompatible, because
753 previously such lookups could never work.
755 The situation is now rather anomolous, since one *can* have colons in
756 ordinary lsearch keys. However, making the change in all cases is
757 incompatible and would probably break a number of configurations.
759 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
762 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
763 conversion specification without a maximum field width, thereby enabling
764 a rogue spamd server to cause a buffer overflow. While nobody in their
765 right mind would setup Exim to query an untrusted spamd server, an
766 attacker that gains access to a server running spamd could potentially
767 exploit this vulnerability to run arbitrary code as the Exim user.
769 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
770 $primary_hostname instead of what libspf2 thinks the hosts name is.
772 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
773 a directory entry by the name of the lookup key. Previously, if a
774 symlink pointed to a non-existing file or a file in a directory that
775 Exim lacked permissions to read, a lookup for a key matching that
776 symlink would fail. Now it is enough that a matching directory entry
777 exists, symlink or not. (Bugzilla 503.)
779 PH/03 The body_linecount and body_zerocount variables are now exported in the
782 PH/04 Added the $dnslist_matched variable.
784 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
785 This means they are set thereafter only if the connection becomes
788 PH/06 Added the client_condition to authenticators so that some can be skipped
789 by clients under certain conditions.
791 PH/07 The error message for a badly-placed control=no_multiline_responses left
792 "_responses" off the end of the name.
794 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
796 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
797 (without spaces) instead of just copying the configuration text.
799 PH/10 Added the /noupdate option to the ratelimit ACL condition.
801 PH/11 Added $max_received_linelength.
803 PH/12 Added +ignore_defer and +include_defer to host lists.
805 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
806 way in which PCRE > 7.0 is built.
808 PH/14 Implemented queue_only_load_latch.
810 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
811 MAIL command. The effect was to mangle the value on 64-bit systems.
813 PH/16 Another patch from the Sieve maintainer.
815 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
817 PH/18 If a system quota error occurred while trying to create the file for
818 a maildir delivery, the message "Mailbox is full" was not appended to the
819 bounce if the delivery eventually timed out. Change 4.67/27 below applied
820 only to a quota excession during the actual writing of the file.
822 PH/19 It seems that peer DN values may contain newlines (and other non-printing
823 characters?) which causes problems in log lines. The DN values are now
824 passed through string_printing() before being added to log lines.
826 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
827 and InterBase are left for another time.)
829 PH/21 Added message_body_newlines option.
831 PH/22 Guard against possible overflow in moan_check_errorcopy().
833 PH/23 POSIX allows open() to be a macro; guard against that.
835 PH/24 If the recipient of an error message contained an @ in the local part
836 (suitably quoted, of course), incorrect values were put in $domain and
837 $local_part during the evaluation of errors_copy.
843 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
844 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
847 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
848 issue a MAIL command.
850 PH/02 In an ACL statement such as
852 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
854 if a client was not listed at all, or was listed with a value other than
855 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
856 the condition was not true (as it should be), so access was not denied.
857 The bug was that the ! inversion was incorrectly passed on to the second
858 item. This has been fixed.
860 PH/03 Added additional dnslists conditions == and =& which are different from
861 = and & when the dns lookup returns more than one IP address.
863 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
864 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
866 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
867 FSYNC, which compiles an option called disable_fsync that allows for
868 bypassing fsync(). The documentation is heavily laced with warnings.
870 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
872 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
873 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
874 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
875 including adding "make clean"; (3) Added -fPIC when compiling the test
876 dynamically loaded module, to get rid of a warning.
878 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
879 message fails, move_frozen_messages = true and ignore_bounce_errors_after
880 = 0s. The bug is otherwise harmless.
882 PH/07 There was a bug in the dovecot authenticator such that the value of
883 $auth1 could be overwritten, and so not correctly preserved, after a
884 successful authentication. This usually meant that the value preserved by
885 the server_setid option was incorrect.
887 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
889 PH/09 Installed PCRE release 7.0.
891 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
892 run for batched SMTP input. It is now run at the start of every message
893 in the batch. While fixing this I discovered that the process information
894 (output by running exiwhat) was not always getting set for -bs and -bS
895 input. This is fixed, and it now also says "batched" for BSMTP.
897 PH/11 Added control=no_pipelining.
899 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
900 patch, slightly modified), and move the expansion of helo_data till after
901 the connection is made in the smtp transport (so it can use these
904 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
906 PH/14 Added log_selector = +pid.
908 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
910 PH/16 Add ${if forany and ${if forall.
912 PH/17 Added dsn_from option to vary the From: line in DSNs.
914 PH/18 Flush SMTP output before performing a callout, unless control =
915 no_callout_flush is set.
917 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
918 was true (the default) a successful delivery failed to delete the retry
919 item, thus causing premature timeout of the address. The bug is now
922 PH/20 Added hosts_avoid_pipelining to the smtp transport.
924 PH/21 Long custom messages for fakedefer and fakereject are now split up
925 into multiline reponses in the same way that messages for "deny" and
926 other ACL rejections are.
928 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
929 with slight modification.
931 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
932 draft, changing the syntax and factoring some duplicate code".
934 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
935 for deliveries of the second and subsequent messages over the same SMTP
938 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
939 ${reduce, with only minor "tidies".
941 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
943 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
944 expansion side effects.
946 PH/27 When a message times out after an over-quota error from an Exim-imposed
947 quota, the bounce message says "mailbox is full". This message was not
948 being given when it was a system quota that was exceeded. It now should
951 MH/03 Made $recipients available in local_scan(). local_scan() already has
952 better access to the recipient list through recipients_list[], but
953 $recipients can be useful in postmaster-provided expansion strings.
955 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
956 in the case of a MAIL command with additional options following the
957 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
958 were accidentally chopped off.
960 PH/29 SMTP synchronization checks are implemented when a command is read -
961 there is a check that no more input is waiting when there shouldn't be
962 any. However, for some commands, a delay in an ACL can mean that it is
963 some time before the response is written. In this time, more input might
964 arrive, invalidly. So now there are extra checks after an ACL has run for
965 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
966 pipelining has not been advertised.
968 PH/30 MH's patch to allow iscntrl() characters to be list separators.
970 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
971 returned in the SMTP response when smtp_return_error_details was false.
974 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
975 instead of the C I/O that was originally supplied, because problems were
978 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
979 Exim which did not show up earlier: it was assuming that a call to
980 SSL_CTX_set_info_callback() might give an error value. In fact, there is
981 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
982 was a macro that became an assignment, so it seemed to work. This has
983 changed to a proper function call with a void return, hence the compile
984 error. Exim's code has been fixed.
986 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
989 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
991 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
993 PH/37 If a message is not accepted after it has had an id assigned (e.g.
994 because it turns out to be too big or there is a timeout) there is no
995 "Completed" line in the log. When some messages of this type were
996 selected by exigrep, they were listed as "not completed". Others were
997 picked up by some special patterns. I have improved the selection
998 criteria to be more general.
1000 PH/38 The host_find_failed option in the manualroute router can now be set
1001 to "ignore", to completely ignore a host whose IP address cannot be
1002 found. If all hosts are ignored, the behaviour is controlled by the new
1003 host_all_ignored option.
1005 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1006 homing or because of multiple MX records with /mx) generated more than
1007 one IP address, and the following item turned out to be the local host,
1008 all the secondary addresses of the first item were incorrectly removed
1009 from the list, along with the local host and any following hosts (which
1010 is what is supposed to happen).
1012 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1013 whoever called Exim into the -H file. In the case of the daemon it was
1014 behaving confusingly. When first started, it used values for whoever
1015 started the daemon, but after a SIGHUP it used the Exim user (because it
1016 calls itself on a restart). I have changed the code so that it now always
1019 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1020 message are rejected with the same error (e.g. no authentication or bad
1021 sender address), and a DATA command is nevertheless sent (as can happen
1022 with PIPELINING or a stupid MUA), the error message that was given to the
1023 RCPT commands is included in the rejection of the DATA command. This is
1024 intended to be helpful for MUAs that show only the final error to their
1027 PH/42 Another patch from the Sieve maintainer.
1029 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1030 Eximstats - Fixed some broken HTML links and added missing column headers
1032 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1033 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1035 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1037 PH/43 Yet another patch from the Sieve maintainer.
1039 PH/44 I found a way to check for a TCP/IP connection going away before sending
1040 the response to the final '.' that terminates a message, but only in the
1041 case where the client has not sent further data following the '.'
1042 (unfortunately, this is allowed). However, in many cases there won't be
1043 any further data because there won't be any more messages to send. A call
1044 to select() can be used: if it shows that the input is "ready", there is
1045 either input waiting, or the socket has been closed. An attempt to read
1046 the next input character can distinguish the two cases. Previously, Exim
1047 would have sent an OK response which the client would never have see.
1048 This could lead to message repetition. This fix should cure that, at
1049 least in a lot of common cases.
1051 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1052 advertised in response to EHLO.
1058 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1059 fixed by 4.65/MH/01 (is this a record?) are fixed:
1061 (i) An empty string was always treated as zero by the numeric comparison
1062 operators. This behaviour has been restored.
1064 (ii) It is documented that the numeric comparison operators always treat
1065 their arguments as decimal numbers. This was broken in that numbers
1066 starting with 0 were being interpreted as octal.
1068 While fixing these problems I realized that there was another issue that
1069 hadn't been noticed. Values of message_size_limit (both the global option
1070 and the transport option) were treated as octal if they started with 0.
1071 The documentation was vague. These values are now always treated as
1072 decimal, and I will make that clear in the documentation.
1078 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1079 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1082 MH/01 Don't check that the operands of numeric comparison operators are
1083 integers when their expansion is in "skipping" mode (fixes bug
1084 introduced by 4.64-PH/07).
1086 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1087 child addresses, Exim now panics and dies. Previously, because the count
1088 is held in a short int, deliveries were likely to be lost. As such a
1089 large number of recipients for a single message is ridiculous
1090 (performance will be very, very poor), I have chosen to impose a limit
1091 rather than extend the field.
1097 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1098 leftover -K file (the existence of which was triggered by #402).
1099 While we were at it, introduced process PID as part of the -K
1100 filename. This should rule out race conditions when creating
1103 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1104 processing considerably. Previous code took too long for large mails,
1105 triggering a timeout which in turn triggers #401.
1107 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1108 in the DK code in transports.c. sendfile() is not really portable,
1109 hence the _LINUX specificness.
1111 TF/01 In the add_headers option to the mail command in an Exim filter,
1112 there was a bug that Exim would claim a syntax error in any
1113 header after the first one which had an odd number of characters
1116 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1117 callout verification, Exim cached a "reject" for the entire domain. This
1118 is correct for most verifications, but it is not correct for a recipient
1119 verification with use_sender or use_postmaster set, because in that case
1120 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1121 case of MAIL FROM:<> rejection from other early rejections (e.g.
1122 rejection of HELO). When verifying a recipient using a non-null MAIL
1123 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1124 Whatever the result of the callout, the value of the domain cache is
1125 left unchanged (for any other kind of callout, getting as far as trying
1126 RCPT means that the domain itself is ok).
1128 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1131 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1132 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1133 session, the daemon ignores ECONNECT errors and logs others; it now
1134 ignores EPIPE as well.
1136 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1137 (quoted-printable decoding).
1139 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1140 later the small subsequent patch to fix an introduced bug.
1142 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1144 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1146 PH/08 An error is now given if message_size_limit is specified negative.
1148 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1149 to be given (somewhat) arbitrary names.
1151 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1154 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1155 miscellaneous code fixes
1157 PH/10 Added the log_reject_target ACL modifier to specify where to log
1160 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1161 hostname. This is wrong, because it relates to the incoming message (and
1162 probably the interface on which it is arriving) and not to the outgoing
1163 callout (which could be using a different interface). This has been
1164 changed to use the value of the helo_data option from the smtp transport
1165 instead - this is what is used when a message is actually being sent. If
1166 there is no remote transport (possible with a router that sets up host
1167 addresses), $smtp_active_hostname is used.
1169 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1170 tweaks were necessary in order to get it to work (see also 21 below):
1171 (a) The code assumed that strncpy() returns a negative number on buffer
1172 overflow, which isn't the case. Replaced with Exim's string_format()
1174 (b) There were several signed/unsigned issues. I just did the minimum
1175 hacking in of casts. There is scope for a larger refactoring.
1176 (c) The code used strcasecmp() which is not a standard C function.
1177 Replaced with Exim's strcmpic() function.
1178 (d) The code set only $1; it now sets $auth1 as well.
1179 (e) A simple test gave the error "authentication client didn't specify
1180 service in request". It would seem that Dovecot has changed its
1181 interface. Fortunately there's a specification; I followed it and
1182 changed what the client sends and it appears to be working now.
1184 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1187 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1188 address A is aliased to B and C, where B exists and C does not. Without
1189 -v the output is "A verified" because verification stops after a
1190 successful redirection if more than one address is generated. However,
1191 with -v the child addresses are also verified. Exim was outputting "A
1192 failed to verify" and then showing the successful verification for C,
1193 with its parentage. It now outputs "B failed to verify", showing B's
1194 parentage before showing the successful verification of C.
1196 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1197 look up a TXT record in a specific list after matching in a combined
1200 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1201 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1202 they consult the DNS. I had assumed they would set it the way they
1203 wanted; and indeed my experiments on Linux seem to show that in some
1204 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1205 To be on the safe side, however, I have now made the interface to
1206 host_find_byname() similar to host_find_bydns(), with an argument
1207 containing the DNS resolver options. The host_find_byname() function now
1208 sets these options at its start, just as host_find_bydns() does. The smtp
1209 transport options dns_qualify_single and dns_search_parents are passed to
1210 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1211 of host_find_byname() use the default settings of RES_DEFNAMES
1212 (qualify_single) but not RES_DNSRCH (search_parents).
1214 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1215 spool_read_header() do less string testing, by means of a preliminary
1216 switch on the second character of optional "-foo" lines. (This is
1217 overdue, caused by the large number of possibilities that now exist.
1218 Originally there were few.) While I was there, I also converted the
1219 str(n)cmp tests so they don't re-test the leading "-" and the first
1220 character, in the hope this might squeeze out yet more improvement.
1222 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1223 flag allowing group syntax was set by the header_syntax check but not
1224 turned off, possible causing trouble later; (2) The flag was not being
1225 set at all for the header_verify test, causing "group"-style headers to
1226 be rejected. I have now set it in this case, and also caused header_
1227 verify to ignore an empty address taken from a group. While doing this, I
1228 came across some other cases where the code for allowing group syntax
1229 while scanning a header line wasn't quite right (mostly, not resetting
1230 the flag correctly in the right place). These bugs could have caused
1231 trouble for malformed header lines. I hope it is now all correct.
1233 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1234 with the "reply" argument non-NULL. The code, however (which originally
1235 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1236 but it didn't always do it. This confused somebody who was copying the
1237 code for some other use. I have removed all the tests.
1239 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1240 feature that was used to support insecure browsers during the U.S. crypto
1241 embargo. It requires special client support, and Exim is probably the
1242 only MTA that supported it -- and would never use it because real RSA is
1243 always available. This code has been removed, because it had the bad
1244 effect of slowing Exim down by computing (never used) parameters for the
1245 RSA_EXPORT functionality.
1247 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1248 authenticator to fail if there's a tab character in the incoming data
1249 (there should never be unless someone is messing about, as it's supposed
1250 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1251 if the connection is using TLS or if the remote IP is the same as the
1252 local IP, and the "valid-client-cert option" if a client certificate has
1255 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1256 authenticators. This can be used for authorization after authentication
1257 succeeds. (In the case of plaintext, it servers for both authentication
1260 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1261 if any retry times were supplied.
1263 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1264 connection, where there is no client IP address to check. In this
1265 situation, the verify now always succeeds.
1267 PH/25 Applied John Jetmore's -Mset patch.
1269 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1271 PH/27 In a string expansion for a processed (not raw) header when multiple
1272 headers of the same name were present, leading whitespace was being
1273 removed from all of them, but trailing whitespace was being removed only
1274 from the last one. Now trailing whitespace is removed from each header
1275 before concatenation. Completely empty headers in a concatenation (as
1276 before) are ignored.
1278 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1279 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1281 PH/29 [Removed. This was a change that I later backed out, and forgot to
1282 correct the ChangeLog entry (that I had efficiently created) before
1283 committing the later change.]
1285 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1286 address errors (4xx response to RCPT) over the same connection as other
1287 messages routed to the same hosts. Such deliveries are always "forced",
1288 so retry times are not inspected. This resulted in far too many retries
1289 for the affected addresses. The effect occurred only when there were more
1290 hosts than the hosts_max_try setting in the smtp transport when it had
1291 the 4xx errors. Those hosts that it had tried were not added to the list
1292 of hosts for which the message was waiting, so if all were tried, there
1293 was no problem. Two fixes have been applied:
1295 (i) If there are any address or message errors in an SMTP delivery, none
1296 of the hosts (tried or untried) are now added to the list of hosts
1297 for which the message is waiting, so the message should not be a
1298 candidate for sending over the same connection that was used for a
1299 successful delivery of some other message. This seems entirely
1300 reasonable: after all the message is NOT "waiting for some host".
1301 This is so "obvious" that I'm not sure why it wasn't done
1302 previously. Hope I haven't missed anything, but it can't do any
1303 harm, as the worst effect is to miss an optimization.
1305 (ii) If, despite (i), such a delivery is accidentally attempted, the
1306 routing retry time is respected, so at least it doesn't keep
1307 hammering the server.
1309 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1310 in ${readsocket because some servers need this prod.
1312 PH/32 Added some extra debug output when updating a wait-xxx database.
1314 PH/33 The hint "could be header name not terminated by colon", which has been
1315 given for certain expansion errors for a long time, was not being given
1316 for the ${if def:h_colon_omitted{... case.
1318 PH/34 The spec says: "With one important exception, whenever a domain list is
1319 being scanned, $domain contains the subject domain." There was at least
1320 one case where this was not true.
1322 PH/35 The error "getsockname() failed: connection reset by peer" was being
1323 written to the panic log as well as the main log, but it isn't really
1324 panic-worthy as it just means the connection died rather early on. I have
1325 removed the panic log writing for the ECONNRESET error when getsockname()
1328 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1329 runs only) independently of the message's sender address. This meant
1330 that, if the 4xx error was in fact related to the sender, a different
1331 message to the same recipient with a different sender could confuse
1332 things. In particualar, this can happen when sending to a greylisting
1333 server, but other circumstances could also provoke similar problems.
1334 I have changed the default so that the retry time for these errors is now
1335 based a combination of the sender and recipient addresses. This change
1336 can be overridden by setting address_retry_include_sender=false in the
1339 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1340 remote server are returned as part of bounce messages. This was not
1341 happening for LMTP over a pipe (the lmtp transport), but now it is the
1342 same for both kinds of LMTP.
1344 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1345 in header lines that were in fact CNAMEs. This is no longer the case.
1347 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1348 and queue runs started by the daemon processed all messages. This has
1349 been fixed so that -R and -S can now usefully be given with -q<time>.
1351 PH/40 Import PCRE release 6.7 (fixes some bugs).
1353 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1355 PH/42 Give an error if -q is specified more than once.
1357 PH/43 Renamed the variables $interface_address and $interface_port as
1358 $received_ip_address and $received_port, to make it clear that these
1359 values apply to message reception, and not to the outgoing interface when
1360 a message is delivered. (The old names remain recognized, of course.)
1362 PH/44 There was no timeout on the connect() call when using a Unix domain
1363 socket in the ${readsocket expansion. There now is.
1365 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1366 be meaningful with "accept".
1368 SC/01 Eximstats V1.43
1369 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1371 SC/02 Eximstats V1.44
1372 Use a glob alias rather than an array ref in the generated
1373 parser. This improves both readability and performance.
1375 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1376 Collect SpamAssassin and rejection statistics.
1377 Don't display local sender or destination tables unless
1378 there is data to show.
1379 Added average volumes into the top table text output.
1381 SC/04 Eximstats V1.46
1382 Collect data on the number of addresses (recipients)
1383 as well as the number of messages.
1385 SC/05 Eximstats V1.47
1386 Added 'Message too big' to the list of mail rejection
1387 reasons (thanks to Marco Gaiarin).
1389 SC/06 Eximstats V1.48
1390 Mainlog lines which have GMT offsets and are too short to
1391 have a flag are now skipped.
1393 SC/07 Eximstats V1.49 (Alain Williams)
1394 Added the -emptyok flag.
1396 SC/08 Eximstats V1.50
1397 Fixes for obtaining the IP address from reject messages.
1399 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1400 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1401 whitesspace changes from 4.64-PH/27
1403 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1406 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1407 are found, allow negative numbers in numeric criteria)
1409 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1411 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1412 to match changes made in 4.64-PH/43
1414 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1416 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1417 "domain" is actually an IP address, and force a failure. This locks out
1418 those revolvers/nameservers that support "A-for-A" lookups, in
1419 contravention of the specifications.
1421 PH/48 When a host name was looked up from an IP address, and the subsequent
1422 forward lookup of the name timed out, the host name was left in
1423 $sender_host_name, contrary to the specification.
1425 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1426 restricted to single-key lookups, Exim was not diagnosing an error if
1427 * or *@ was used with a query-style lookup.
1429 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1431 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1432 long ago, but noone interested enough thought of it. Let's just say that
1433 the "1.1" means that there are some new functions that weren't there at
1434 some point in the past.
1436 PH/51 Error processing for expansion failure of helo_data from an smtp
1437 transport during callout processing was broken.
1439 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1440 tested/used via the -bh/-bhc/-bs options.
1442 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1443 bug, fixed in subsequent PCRE releases).
1445 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1446 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1448 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1454 SC/01 Use a glob alias rather than an array ref in eximstats generated
1455 parser. This improves both readability and performance.
1457 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1458 Don't display local sender or destination tables in eximstats unless
1459 there is data to show.
1460 Added average volumes into the eximstats top table text output.
1462 SC/03 Collect data on the number of addresses (recipients) as well
1463 as the number of messages in eximstats.
1465 TF/01 Correct an error in the documentation for the redirect router. Exim
1466 does (usually) call initgroups() when daemonizing.
1468 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1469 with consistent privilege compared to when running as a daemon.
1471 TF/03 Note in the spec that $authenticated_id is not set for local
1472 submissions from trusted users.
1474 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1475 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1477 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1478 by adding some example configuration directives to the default
1479 configuration file. A little bit of work is required to uncomment the
1480 directives and define how usernames and passwords are checked, but
1481 there is now a framework to start from.
1483 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1484 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1485 without this. I don't know how relevant this is to other LDAP libraries.
1487 PH/02 Add the verb name to the "unknown ACL verb" error.
1489 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1491 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1493 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1494 directory not expanded when it should be if an expanded home directory
1495 was set for the address (which is overridden by the transport).
1497 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1500 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1501 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1502 because it is too late at that time, and has no effect.
1504 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1505 security issue with \' (bugzilla #107). I could not use the
1506 PQescapeStringConn() function, because it needs a PGconn value as one of
1509 PH/08 When testing addresses using -bt, indicate those final addresses that
1510 are duplicates that would not cause an additional delivery. At least one
1511 person was confused, thinking that -bt output corresponded to deliveries.
1512 (Suppressing duplicates isn't a good idea as you lose the information
1513 about possibly different redirections that led to the duplicates.)
1515 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1516 systems where poll() doesn't work, in particular OS X.
1518 PH/10 Added more information to debugging output for retry time not reached.
1520 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1521 operations in malware.c.
1523 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1526 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1527 syslog_duplication set false, log lines that would normally be written
1528 both the the main log and to the reject log were not written to syslog at
1531 PH/14 In the default configuration, change the use of "message" in ACL warn
1532 statements to "add_header".
1534 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1535 not followed by a command (e.g. "seen endif").
1537 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1538 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1541 PH/17 Added extra conditions to the default value of delay_warning_condition
1545 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1546 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1547 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1550 The Auto-Submitted: and various List- headers are standardised, whereas I
1551 don't think Precedence: ever was.
1553 PH/18 Refactored debugging code in route_finduser() to show more information,
1554 in particular, the error code if getpwnam() issues one.
1556 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1557 This is apparently needed in addition to the PH/07 change above to avoid
1558 any possible encoding problems.
1560 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1561 but not after initializing Perl.
1563 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1564 output them only if debugging. By default they are written stderr,
1565 apparently, which is not desirable.
1567 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1570 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1573 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1575 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1576 authenticated or an ident call has been made. Suppress the default
1577 values for $authenticated_id and $authenticated_sender (but permit -oMai
1578 and -oMas) when testing with -bh.
1580 PH/24 Re-jigged the order of the tests in the default configuration so that the
1581 tests for valid domains and recipients precede the DNS black list and CSA
1582 tests, on the grounds that those ones are more expensive.
1584 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1585 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1586 This bug exists in every version of Exim that I still have, right back to
1589 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1590 However, an attempt to turn on case-sensitivity in a regex key by
1591 including (?-i) didn't work because the subject string was already
1592 lowercased, and the effects were non-intuitive. It turns out that a
1593 one-line patch can be used to allow (?-i) to work as expected.
1599 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1600 other effects) broke the use of negated acl sub-conditions.
1602 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1605 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1606 "Deny" causes Exim to reject the incoming connection with a 554 error.
1607 Unfortunately, if there is a major crisis, such as a disk failure,
1608 tcp-wrappers gives "deny", whereas what one would like would be some
1609 kind of temporary error. A kludge has been added to help with this.
1610 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1611 554 error is used if errno is still zero or contains ENOENT (which occurs
1612 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1615 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1617 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1618 errors. Otherwise a message that provokes a temporary error (when other
1619 messages do not) can cause a whole host to time out.
1621 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1622 addresses were routed directly to files or pipes from a redirect router.
1623 File deliveries just didn't batch; pipe deliveries might have suffered
1626 PH/06 A failure to get a lock for a hints database would erroneously always say
1627 "Failed to get write lock", even when it was really a read lock.
1629 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1630 of 0600. This has been changed to use the value of the lockfile_mode
1631 option (which defaults to 0600).
1633 PH/08 Applied small patch from the Sieve maintainer.
1635 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1636 folder from quota calculations, a direct delivery into this folder messed
1637 up the contents of the maildirsize file. This was because the regex was
1638 used only to exclude .Trash (or whatever) when the size of the mailbox
1639 was calculated. There was no check that a delivery was happening into an
1640 excluded directory. This bug has been fixed by ignoring all quota
1641 processing for deliveries into excluded directories.
1643 PH/10 Added the maildirfolder_create_regex option to appendfile.
1649 PH/01 The code for finding all the local interface addresses on a FreeBSD
1650 system running IPv6 was broken. This may well have applied to all BSD
1651 systems, as well as to others that have similar system calls. The broken
1652 code found IPv4 interfaces correctly, but gave incorrect values for the
1653 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1654 that it would not match correctly against @[] and not recognize the IPv6
1657 PH/02 The ipliteral router was not recognizing addresses of the form user@
1658 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1660 PH/03 Added disable_ipv6.
1662 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1663 decoded form, because it is most often used to construct To: headers
1664 lines in autoreplies, and the decoded form may well be syntactically
1665 invalid. However, $reply_address has leading white space removed, and all
1666 newlines turned into spaces so that the autoreply transport does not
1669 PH/05 If group was specified without a user on a router, and no group or user
1670 was specified on a transport, the group from the router was ignored.
1672 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1673 for visible compile-time settings that can be used to change these
1674 numbers, for those that want even more. Backwards compatibility with old
1675 spool files has been maintained. However, going back to a previous Exim
1676 release will lost any variables that are in spool files.
1678 PH/07 Two small changes when running in the test harness: increase delay when
1679 passing a TCP/IP connection to a new process, in case the original
1680 process has to generate a bounce, and remove special handling of
1681 127.0.0.2 (sic), which is no longer necessary.
1683 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1684 be the same on different OS.
1686 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1689 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1690 whether --show-vars was specified or not
1692 JJ/02 exipick: Added support for new ACL variable spool format introduced
1695 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1696 syntactically invalid From: or Reply-to: line, and a filter used this to
1697 generate an autoreply, and therefore failed to obtain an address for the
1698 autoreply, Exim could try to deliver to a non-existent relative file
1699 name, causing unrelated and misleading errors. What now happens is that
1700 it logs this as a hard delivery error, but does not attempt to create a
1703 PH/11 The exinext utility has a -C option for testing purposes, but although
1704 the given file was scanned by exinext itself; it wasn't being passed on
1705 when Exim was called.
1707 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1708 an end-of-file indication when reading a command response.
1710 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1711 compiled. In many other places in Exim, IPv6 addresses are always
1712 recognized, so I have changed this. It also means that IPv4 domain
1713 literals of the form [IPV4:n.n.n.n] are now always recognized.
1715 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1716 used if the router is not running as root, for example, when verifying at
1717 ACL time, or when using -bh. The debugging output from this situation was
1718 non-existent - all you got was a failure to exec. I have made two
1721 (a) Failures to set uid/gid, the current directory, or a process leader
1722 in a subprocess such as that created by queryprogram now generate
1723 suitable debugging ouput when -d is set.
1725 (b) The queryprogram router detects when it is not running as root,
1726 outputs suitable debugging information if -d is set, and then runs
1727 the subprocess without attempting to change uid/gid.
1729 PH/15 Minor change to Makefile for building test_host (undocumented testing
1732 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1733 additional section of a DNS packet that returns MX or SRV records.
1734 Instead, it always explicitly searches for A/AAAA records. This avoids
1735 major problems that occur when a DNS server includes only records of one
1736 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1737 fixed another bug: if SRV records were looked up and the corresponding
1738 address records were *not* found in the additional section, the port
1739 values from the SRV records were lost.
1741 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1742 using the correct key (the original address) when searching the retry
1743 rules in order to find which one to use for generating the retry hint.
1745 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1746 adding the default one. Similarly, if it contains a Reply-To: header, the
1747 errors_reply_to option, if set, is not used.
1749 PH/19 When calculating a retry time, Exim used to measure the "time since
1750 failure" by looking at the "first failed" field in the retry record. Now
1751 it does not use this if it is later than than the arrival time of the
1752 message. Instead it uses the arrival time. This makes for better
1753 behaviour in cases where some deliveries succeed, thus re-setting the
1754 "first failed" field. An example is a quota failure for a huge message
1755 when small messages continue to be delivered. Without this change, the
1756 "time since failure" will always be short, possible causing more frequent
1757 delivery attempts for the huge message than are intended.
1758 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1760 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1761 $1, $2, $3) because the numerical variables can be reset during some
1762 expansion items (e.g. "match"), thereby losing the authentication data.
1764 PH/21 Make -bV show the size of off_t variables so that the test suite can
1765 decide whether to run tests for quotas > 2G.
1767 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1768 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1769 filecount value is greater than 2G or if a quota value is greater than 2G
1770 on a system where the size of off_t is not greater than 4, a panic error
1773 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1774 never match. The debug and -bh output now contains an explicit error
1775 message indicating a malformed IPv4 address or mask.
1777 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1778 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1779 PH/23 above applies.
1781 PH/25 Do not write to syslog when running in the test harness. The only
1782 occasion when this arises is a failure to open the main or panic logs
1783 (for which there is an explicit test).
1785 PH/26 Added the /no_tell option to "control=freeze".
1787 PH/27 If a host name lookup failed very early in a connection, for example, if
1788 the IP address matched host_lookup and the reverse lookup yielded a name
1789 that did not have a forward lookup, an error message of the form "no IP
1790 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1791 could be logged. Now it outputs the IP address instead of "NULL".
1793 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1794 allows the sender and the authenticated sender to be set when
1795 submitting a message from within Exim. Since child_open_exim() is
1796 documented for local_scan(), the new function should be too.
1798 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1799 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1800 results in an empty string is now treated as unset.
1802 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1804 PH/31 Added sender_verify_fail logging option.
1806 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1807 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1808 tidied the source and removed it altogether.
1810 PH/33 When a queue run was abandoned because the load average was too high, a
1811 log line was always written; now it is written only if the queue_run log
1812 selector is set. In addition, the log line for abandonment now contains
1813 information about the queue run such as the pid. This is always present
1814 in "start" and "stop" lines but was omitted from the "abandon" line.
1816 PH/34 Omit spaces between a header name and the colon in the error message that
1817 is given when verify = headers_syntax fails (if there are lots of them,
1818 the message gets confusing).
1820 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1821 names, as there are now some PTR records that contain slashes. This check
1822 is only to protect against broken name servers that fall over on strange
1823 characters, so the fact that it applies to all lookups doesn't matter.
1825 PH/36 Now that the new test suite is complete, we can remove some of the
1826 special code in Exim that was needed for the old test suite. For example,
1827 sorting DNS records because real resolvers return them in an arbitrary
1828 order. The new test suite's fake resolver always returns records in the
1831 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1832 bounces) except when queue_only is set, to avoid logging races between
1833 the different processes.
1835 PH/38 Panic-die if .include specifies a non-absolute path.
1837 PH/39 A tweak to the "H" retry rule from its user.
1839 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1840 a label. They prevented compilation on older perls.
1842 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1843 a warning to be raised on newish perls.
1845 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1846 on queue. Changes to match documented behaviour of showing count of
1847 messages matching specified criteria.
1849 PH/40 Changed the default ident timeout from 30s to 5s.
1851 PH/41 Added support for the use of login_cap features, on those BSD systems
1852 that have them, for controlling the resources used by pipe deliveries.
1854 PH/42 The content-scanning code uses fopen() to create files in which to put
1855 message data. Previously it was not paying any attention to the mode of
1856 the files. Exim runs with umask(0) because the rest of the code creates
1857 files with open(), and sets the required mode explicitly. Thus, these
1858 files were ending up world-writeable. This was not a big issue, because,
1859 being within the spool directory, they were not world-accessible. I have
1860 created a function called modefopen, which takes an additional mode
1861 argument. It sets umask(777), creates the file, chmods it to the required
1862 mode, then resets the umask. All the relevant calls to fopen() in the
1863 content scanning code have been changed to use this function.
1865 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1866 to 24 hours. This avoids potential overflow problems when processing G
1867 and H retry rules. I suspect nobody ever tinkers with this value.
1869 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1871 PH/45 When the plaintext authenticator is running as a client, the server's
1872 challenges are checked to ensure they are valid base64 strings. By
1873 default, the authentication attempt is cancelled if an invalid string is
1874 received. Setting client_ignore_invalid_base64 true ignores these errors.
1875 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1876 they are received. Thus, the responses can be made to depend on the
1877 challenges. If an invalid string is ignored, an empty string is placed in
1880 PH/46 Messages that are created by the autoreply transport now contains a
1881 References: header, in accordance with RFCs 2822 and 3834.
1883 PH/47 Added authenticated_sender_force to the smtp transport.
1885 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1887 PH/49 Installed latest patch from the Sieve maintainer.
1889 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1890 was also set, the appendfile transport was unnecessarily scanning a
1891 directory of message files (e.g. for maildir delivery) to find the count
1892 of files (along with the size), even though it did not need this
1893 information. It now does the scan only if it needs to find either the
1894 size of the count of files.
1896 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1898 PH/52 Two bugs concerned with error handling when the smtp transport is
1901 (i) Exim was not creating retry information for temporary errors given
1902 for individual recipients after the DATA command when the smtp transport
1903 was used in LMTP mode. This meant that they could be retried too
1904 frequently, and not timed out correctly.
1906 (ii) Exim was setting the flag that allows error details to be returned
1907 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1908 recipients that were returned after the DATA command.
1910 PH/53 This is related to PH/52, but is more general: for any failing address,
1911 when detailed error information was permitted to be returned to the
1912 sender, but the error was temporary, then after the final timeout, only
1913 "retry timeout exceeded" was returned. Now it returns the full error as
1914 well as "retry timeout exceeded".
1916 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1917 do this, and (what is worse) MTAs that accept it.
1919 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1920 will now be deprecated.
1922 PH/56 New os.c-cygwin from the Cygwin maintainer.
1924 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1925 formats (previously only available in exim formats via -bpr, -bpru,
1926 and -bpra. Now also available in native and exiqgrep formats)
1928 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1929 with very large, slow to parse queues
1931 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1933 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1935 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1936 responses to authentication challenges, though it was showing the
1937 challenges; (ii) I've removed the CR characters from the debug output for
1940 PH/58 Allow for the insertion of a newline as well as a space when a string
1941 is turned into more than one encoded-word during RFC 2047 encoding. The
1942 Sieve code now uses this.
1944 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1945 data_4xx, lost_connection, tls_required.
1947 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1948 message was being sent as an SMTP response.
1950 PH/61 Add -l and -k options to exicyclog.
1952 PH/62 When verifying, if an address was redirected to one new address, so that
1953 verification continued, and the new address failed or deferred after
1954 having set something in $address_data, the value of $address_data was not
1955 passed back to the ACL. This was different to the case when no
1956 redirection occurred. The value is now passed back in both cases.
1958 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1959 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1960 use login_cap.h, so on its own it isn't the distinguishing feature. The
1961 new name refers directly to the setclassresources() function.
1963 PH/65 Added configuration files for NetBSD3.
1965 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1967 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1968 is preferred over IPv4.
1970 PH/68 The bounce_return_message and bounce_return_body options were not being
1971 honoured for bounces generated during the reception of non-SMTP messages.
1972 In particular, this applied to messages rejected by the ACL. This bug has
1973 been fixed. However, if bounce_return_message is true and bounce_return_
1974 body is false, the headers that are returned for a non-SMTP message
1975 include only those that have been read before the error was detected.
1976 (In the case of an ACL rejection, they have all been read.)
1978 PH/69 The HTML version of the specification is now built in a directory called
1979 spec_html instead of spec.html, because the latter looks like a path with
1980 a MIME-type, and this confuses some software.
1982 PH/70 Catch two compiler warnings in sieve.c.
1984 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1985 function verify_get_ident() calls ip_connect() to connect a socket, but
1986 if the "connect()" function timed out, ip_connect() used to close the
1987 socket. However, verify_get_ident() also closes the socket later, and in
1988 between Exim writes to the log, which may get opened at this point. When
1989 the socket was closed in ip_connect(), the log could get the same file
1990 descriptor number as the socket. This naturally causes chaos. The fix is
1991 not to close the socket in ip_connect(); the socket should be closed by
1992 the function that creates it. There was only one place in the code where
1993 this was missing, in the iplookup router, which I don't think anybody now
1994 uses, but I've fixed it anyway.
1996 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1997 well as to direct DNS lookups. Otherwise the handling of names in host
1998 lists is inconsistent and therefore confusing.
2004 PH/01 Two changes to the default runtime configuration:
2006 (1) Move the checks for relay_from_hosts and authenticated clients from
2007 after to before the (commented out) DNS black list checks.
2009 (2) Add control=submission to the relay_from_hosts and authenticated
2010 clients checks, on the grounds that messages accepted by these
2011 statements are most likely to be submissions.
2013 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2015 (1) Generate an error if the third argument for the ${prvs expansion is
2018 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2021 (3) Reset the variables that are obtained from the first argument of
2022 ${prvscheck and used in the second argument before leaving the code,
2023 because their memory is reclaimed, so using them afterwards may do
2026 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2027 one (it's much easier than Tom thought :-).
2029 (5) Because of (4), we can now allow for the use of $prvscheck_result
2030 inside the third argument.
2032 PH/03 For some reason, the default setting of PATH when running a command from
2033 a pipe transport was just "/usr/bin". I have changed it to
2036 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2037 anything to be listed in the output from -bV.
2039 PH/05 When a filter generated an autoreply, the entire To: header line was
2040 quoted in the delivery log line, like this:
2042 => >A.N.Other <ano@some.domain> <original@ddress> ...
2044 This has been changed so that it extracts the operative address. There
2045 may be more than one such address. If so, they are comma-separated, like
2048 => >ano@some.domain,ona@other.domain <original@ddress> ...
2050 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2051 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2052 not being looked up in the rDNS to get a host name, Exim was showing the
2053 IP address twice in Received: lines, even though the IP addresses were
2054 identical. For example:
2056 Received: from [1.2.3.4] (helo=[1.2.3.4])
2058 However, if the real host name was known, it was omitting the HELO data
2059 if it matched the actual IP address. This has been tidied up so that it
2060 doesn't show the same IP address twice.
2062 PH/07 When both +timestamp and +memory debugging was on, the value given by
2063 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2064 called by the debug printing, thereby overwriting the timestamp buffer.
2065 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2067 PH/08 When the original message was included in an autoreply transport, it
2068 always said "this is a copy of the message, including all the headers",
2069 even if body_only or headers_only was set. It now gives an appropriate
2072 PH/09 Applied a patch from the Sieve maintainer which:
2074 o fixes some comments
2075 o adds the (disabled) notify extension core
2076 o adds some debug output for the result of if/elsif tests
2077 o points to the current vacation draft in the documentation
2078 and documents the missing references header update
2082 o fixes a bug in processing the envelope test (when testing
2083 multiple envelope elements, the last element determinted the
2086 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2087 Electronic Mail") by including:
2089 Auto-submitted: auto-generated
2091 in the messages that it generates (bounce messages and others, such as
2092 warnings). In the case of bounce messages for non-SMTP mesages, there was
2093 also a typo: it was using "Auto_submitted" (underscore instead of
2094 hyphen). Since every message generated by Exim is necessarily in response
2095 to another message, thes have all been changed to:
2097 Auto-Submitted: auto-replied
2099 in accordance with these statements in the RFC:
2101 The auto-replied keyword:
2103 - SHOULD be used on messages sent in direct response to another
2104 message by an automatic process,
2106 - MUST NOT be used on manually-generated messages,
2108 - MAY be used on Delivery Status Notifications (DSNs) and Message
2109 Disposition Notifications (MDNs),
2111 - MUST NOT be used on messages generated by automatic or periodic
2112 processes, except for messages which are automatic responses to
2115 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2116 to the default Received: header definition.
2118 PH/12 Added log selector acl_warn_skipped (default on).
2120 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2121 variables because (a) they are in the wrong storage pool and (b) even if
2122 they were copied, it wouldn't work properly because of the caching.
2124 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2125 checking when decoding. Apparently there are clients that generate
2126 overlong encoded strings. Why am I not surprised?
2128 PH/15 If the first argument of "${if match_address" was not empty, but did not
2129 contain an "@" character, Exim crashed. Now it writes a panic log message
2130 and treats the condition as false.
2132 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2134 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2135 extension "envelope-auth". The code is finished and in agreement with
2136 other implementations, but there is no documentation so far and in fact,
2137 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2138 not changing the active code.
2140 Print executed "if" and "elsif" statements when debugging is used. This
2141 helps a great deal to understand what a filter does.
2143 Document more things not specified clearly in RFC3028. I had all this
2144 sorted out, when out of a sudden new issues came to my mind. Oops."
2146 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2149 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2150 canonical form (as documented). However, after a host name lookup from
2151 the IP address, check_host() was doing a simple string comparison with
2152 addresses acquired from the DNS when checking that the found name did
2153 have the original IP as one of its addresses. Since any found IPv6
2154 addresses are likely to be in abbreviated form, the comparison could
2155 fail. Luckily, there already exists a function for doing the comparison
2156 by converting both addresses to binary, so now that is used instead of
2157 the text comparison.
2159 PH/20 There was another similar case to PH/19, when a complete host name was
2160 given in a host list; looking up its IP address could give an abbreviated
2161 form, whereas the current host's name might or might not be abbreviated.
2162 The same fix has been applied.
2168 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2169 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2172 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2173 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2175 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2177 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2178 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2179 Also, for "Auto-Submitted", treat anything other than "no" as
2180 non-personal, in accordance with RFC 3834. (Previously it treated
2181 anything starting "auto-" as non-personal.)
2183 TF/01 The control=submission/name=... option had a problem with syntax
2184 errors if the name included a slash character. The /name= option
2185 now slurps the rest of the string, so it can include any characters
2186 but it must come last in the list of options (after /sender_retain
2189 PH/05 Some modifications to the interface to the fake nameserver for the new
2197 TK/01 Added the "success_on_redirect" address verification option. See
2198 NewStuff for rationale and an example.
2200 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2202 PH/02 Patch to exigrep to allow it to work on syslog lines.
2204 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2205 fread() to skip over the body file's header line, because in Cygwin the
2206 header line is locked and is inaccessible.
2208 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2209 co-exist for some time) to make it clear that it is the Exim ID that is
2210 referenced, not the Message-ID: header line.
2212 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2213 string_format() function, because snprintf() does not exist on all
2216 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2217 ${stat: expansion item.
2219 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2220 protocol synchronization error", to keep the pedants happy.
2222 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2223 well as for IRIX systems, when gcc is being used. See the host.c source
2226 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2228 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2231 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2232 transport and to the smtp transport in LMTP mode.
2234 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2236 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2237 run a filter in a subprocess. This could lead to confusion in subsequent
2238 lookups in the parent process. There should also be a search_tidyup() at
2239 the end of the subprocess.
2241 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2242 only if the host matched helo_try_verify_hosts, which caused the
2243 verification to occur when the EHLO/HELO command was issued. The ACL just
2244 tested the remembered result. Now, if a previous verification attempt has
2245 not happened, "verify = helo" does it there and then.
2247 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2249 TK/03 Fix log output including CR from clamd.
2251 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2252 exist provoked a memory error which could cause a segfault.
2254 PH/15 Installed PCRE 6.2
2256 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2258 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2259 of the problem. Specifically, suggested +O2 rather than +O1 for the
2262 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2264 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2265 record, the port was not correctly logged when the outgoing_port log
2266 selector was set (it logged the transort's default port).
2268 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2269 fallback_hosts, and "hosts" in the smtp transport.
2271 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2272 host errors such as "Connection refused".
2274 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2275 authentication with radiusclient 0.4.9:
2277 - Error returned from rc_read_config was caught wrongly
2278 - Username/password not passed on to radius server due to wrong length.
2280 The presumption is that some radiusclient API changes for 4.51/PH/17
2281 were not taken care of correctly. The code is still untested by me (my
2282 Linux distribution still has 0.3.2 of radiusclient), but it was
2283 contributed by a Radius user.
2285 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2286 expanding the "port" option of the smtp transport.
2288 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2289 while reading a MIME header. Thanks to Tom Hughes for a patch.
2291 PH/24 Include config.h inside local_scan.h so that configuration settings are
2294 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2295 that in an ACL for RCPT (for example), you can examine exactly what was
2298 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2299 commands, but it was not correctly comparing the address with the actual
2300 client host address. Thus, it would show the EHLO address in Received:
2301 header lines when this was not necessary.
2303 PH/27 Added the % operator to ${eval:}.
2305 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2306 it should be ignoring failures (because with -C, for example, it has lost
2307 privilege). It wasn't ignoring creation failures other than "already
2310 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2313 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2314 because an input line was too long, either on its own, or by virtue of
2315 too many continuations, the temporary file was not being removed, and the
2316 return code was incorrect.
2318 PH/31 Missing "BOOL" in function definition in filtertest.c.
2320 PH/32 Applied Sieve patches from the maintainer.
2322 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2324 PH/33 Added "verify = not_blind".
2326 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2327 Local/Makefile (with some defaults set). These are used in built scripts
2328 such as exicyclog, but they have never been used in the exim_install
2329 script (though there are many overriding facilities there). I have
2330 arranged that the exim_install script now takes note of these two
2333 PH/35 Installed configuration files for Dragonfly.
2335 PH/36 When a locally submitted message by a trusted user did not contain a
2336 From: header, and the sender address was obtained from -f or from an SMTP
2337 MAIL command, and the trusted user did not use -F to supply a sender
2338 name, $originator_name was incorrectly used when constructing a From:
2339 header. Furthermore, $originator_name was used for submission mode
2340 messages from external hosts without From: headers in a similar way,
2341 which is clearly wrong.
2343 PH/37 Added control=suppress_local_fixups.
2345 PH/38 When log_selector = +received_sender was set, and the addition of the
2346 sender made the log line's construction buffer exactly full, or one byte
2347 less than full, an overflow happened when the terminating "\n" was
2350 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2351 when the result of a list match is failure because a DNS lookup failed.
2353 PH/40 RM_COMMAND is now used in the building process.
2355 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2356 the "build-* directories that it finds.
2358 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2359 address in a domain literal was a prefix of an interface address.
2361 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2362 when verifying a sender address, unless rewrite_headers is false.
2364 PH/44 Wrote a long comment about why errors_to addresses are verified as
2365 recipients, not senders.
2367 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2368 the ratelimit ACL was added.
2370 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2372 PH/46 Added extra information about PostgreSQL errors to the error string.
2374 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2375 avoiding the need to install special zones in a real server. This is
2376 backwards compatible; if it can't find the fake resolver, it drops back.
2377 Thus, both old and new test suites can be run.
2379 TF/02 Added util/ratelimit.pl
2381 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2382 clock is set back in time.
2384 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2385 Candler <B.Candler@pobox.com>.
2387 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2388 OK for addresses that are the result of redirections.
2390 PH/48 A number of further additions for the benefit of the new test suite,
2391 including a fake gethostbyname() that interfaces to the fake DNS resolver
2394 TF/06 The fix for widen_domains has also been applied to qualify_single and
2395 search_parents which are the other dnslookup options that can cause
2398 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2401 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2403 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2404 service" if it encountered an MX record with an empty target hostname.
2405 The message is now "an MX or SRV record indicated no SMTP service".
2407 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2408 if the DNS of the sending site is misconfigured. This is quite a
2409 common situation. This change restores the behaviour of treating a
2410 helo verification defer as a failure.
2412 PH/51 If self=fail was set on a router, the bounce message did not include the
2413 actual error message.
2419 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2421 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2422 command itself ended in error, the underlying message about the transport
2423 filter timeout was being overwritten with the pipe command error. Now the
2424 underlying error message should be appended to the second error message.
2426 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2428 PH/02 Reduce the amount of output that "make" produces by default. Full output
2429 can still be requested.
2431 PH/03 The warning log line about a condition test deferring for a "warn" verb
2432 was being output only once per connection, rather than after each
2433 occurrence (because it was using the same function as for successful
2434 "warn" verbs). This seems wrong, so I have changed it.
2436 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2437 it should not have, which might have caused a crash in the right
2438 circumstances, but probably never did.
2440 PH/04 Installed a modified version of Tony Finch's patch to make submission
2441 mode fix the return path as well as the Sender: header line, and to
2442 add a /name= option so that you can make the user's friendly name appear
2445 TF/03 Added the control = fakedefer ACL modifier.
2447 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2448 Mark Lowes for thorough testing.
2450 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2452 TK/03 Merged latest SRS patch from Miles Wilton.
2454 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2455 with the definition in sysexits.h (which is #included earlier).
2456 Fortunately, Exim does not actually use EX_OK. The code used to try to
2457 preserve the sysexits.h value, by assumimg that macro definitions were
2458 scanned for macro replacements. I have been disabused of this notion,
2459 so now the code just undefines EX_OK before #including unistd.h.
2461 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2462 in the smtp transport. When a block could not be written in a single
2463 write() function, the timeout was being re-applied to each part-write.
2464 This seems wrong - if the receiver was accepting one byte at a time it
2465 would take for ever. The timeout is now adjusted when this happens. It
2466 doesn't have to be particularly precise.
2468 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2469 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2471 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2472 without a domain if the check to <postmaster@domain> fails.
2474 SC/01 Eximstats: added -xls and the ability to specify output files
2475 (patch written by Frank Heydlauf).
2477 SC/02 Eximstats: use FileHandles for outputing results.
2479 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2481 SC/04 Eximstats: fixed display of large numbers with -nvr option
2483 SC/05 Eximstats: fixed merging of reports with empty tables.
2485 SC/06 Eximstats: added the -include_original_destination flag
2487 SC/07 Eximstats: removed tabs and trailing whitespace.
2489 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2491 TK/06 MBOX spool code: Add real "From " MBOX separator line
2492 so the .eml file is really in mbox format (even though
2493 most programs do not really care). Patch from Alex Miller.
2495 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2496 The latter is generated from $received_to and is only set if the
2497 message has one envelope recipient. SA can use these headers,
2498 obviously out-of-the-box. Patch from Alex Miller.
2500 PH/08 The ${def test on a variable was returning false if the variable's
2501 value was "0", contrary to what the specification has always said!
2502 The result should be true unless the variable is empty.
2504 PH/09 The syntax error of a character other than { following "${if
2505 def:variable_name" (after optional whitespace) was not being diagnosed.
2506 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2507 accidental colon was present, for example, could give incorrect results.
2509 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2510 result is used (not including appendfile, where other changes are about
2513 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2514 This involved changing a lot of size variables from int to off_t. It
2515 should work with maildirs and everything.
2517 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2518 spamd dying while we are connected to it.
2520 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2523 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2524 function for BDB 4.3.
2526 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2528 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2529 the sources for PCRE have been re-arranged and more files are now
2532 PH/15 The code I had for printing potentially long long variables in PH/11
2533 above was not the best (it lost precision). The length of off_t variables
2534 is now inspected at build time, and an appropriate printing format (%ld
2535 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2536 to be "long long int" or "long int". This is needed for the internal
2537 formatting function string_vformat().
2539 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2540 the configuration file to be ":syslog", then the script "guesses" where
2541 the logs files are, rather than using the compiled in default. In our
2542 case the guess is not the same as the compiled default, so the script
2543 suddenly stopped working when I started to use syslog. The patch checks
2544 to see if log_file_path is "". If so, it attempts to read it from exim
2545 with no configuration file to get the compiled in version, before it
2546 falls back to the previous guessing code."
2548 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2549 implementing BATV in an Exim configuration. See NewStuff for the gory
2552 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2553 Makefile that are specific to HP-UX.
2555 PH/18 If the "use_postmaster" option was set for a recipient callout together
2556 with the "random" option, the postmaster address was used as the MAIL
2557 FROM address for the random test, but not for the subsequent recipient
2558 test. It is now used for both.
2560 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2561 patch removes a few documentation additions to RFC 3028, because the
2562 latest draft now contains them. It adds the new en;ascii-case comparator
2563 and a new error check for 8bit text in MIME parts. Comparator and
2564 require names are now matched exactly. I enabled the subaddress
2565 extension, but it is not well tested yet (read: it works for me)."
2567 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2568 rework some of the code of TK/09 above to avoid the hardwired use of
2569 "%lld" and "long long". Replaced the call to snprintf() with a call to
2572 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2573 records point to non-existent hosts", "retry timeout exceeded", and
2574 "retry time not reached for any host after a long failure period".
2576 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2577 experimental DomainKeys support:
2579 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2580 (2) On an error such as an illegally used "control", the wrong name for
2581 the control was given.
2583 These problems did NOT occur unless DomainKeys support was compiled.
2585 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2587 PH/24 Added ${if match_ip condition.
2589 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2590 databases so that it will be absolutely obvious if a crash occurs in the
2591 DB library. This is a regular occurrence (often caused by mis-matched
2594 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2595 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2596 picked up on a user's system that detects such things. There doesn't seem
2597 to be a gcc warning option for this - only an attribute that has to be
2598 put on the function's prototype. It seems that in Fedora Core 4 they have
2599 set this on a number of new functions. No doubt there will be more in due
2602 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2603 specify a transport. However, if an address that was verified by such a
2604 router was the subject of a callout, Exim crashed because it tried to
2605 read the rcpt_include_affixes from the non-existent transport. Now it
2606 just assumes that the setting of that option is false. This bug was
2607 introduced by 4.51/PH/31.
2609 PH/28 Changed -d+all to exclude +memory, because that information is very
2610 rarely of interest, but it makes the output a lot bigger. People tend to
2611 do -d+all out of habit.
2613 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2614 code in os-type was giving problems when libc.so lives in lib64, like on
2617 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2618 aren't the modern standard, and it seems that some systems' include files
2619 don't always have them. Exim was already checking for some of the newer
2620 ones like T_AAAA, and defining it itself. I've added checks for all the
2621 record types that Exim uses.
2623 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2624 not automatically generating a new one, as it is supposed to. This
2625 prevented TLS from working. If the file did exist, but contained invalid
2626 data, a new version was generated, as expected. It was only the case of a
2627 non-existent file that was broken.
2629 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2630 with a change in libdomainkeys > 0.64.
2632 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2633 from DNS. If the selector record carries the flag, it now has
2634 precedence over the domain-wide flag.
2636 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2638 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2639 the use of a port name that isn't defined in /etc/services) occurred, the
2640 message was deferred as in a normal delivery, and thus remained on the
2641 spool, instead of being failed because of the mua_wrapper setting. This
2642 is now fixed, and I tidied up some of the mua_wrapper messages at the
2645 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2646 the messages in a hash of arrays rather than using individual hashes.
2647 This is a bit cleaner and results in dramatic memory savings, albeit
2648 at a slight CPU cost.
2650 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2651 as requested by Marc Sherman.
2653 SC/10 Eximstats: added histograms for user specified patterns as requested
2656 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2658 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2659 fopen() in the content-scanning modules that did not already have it.
2665 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2666 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2668 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2670 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2672 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2673 to test Sieve filters that use "vacation".
2675 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2676 that changes the way the GnuTLS parameters are stored in the cache file.
2677 The new format can be generated externally. For backward compatibility,
2678 if the data in the cache doesn't make sense, Exim assumes it has read an
2679 old-format file, and it generates new data and writes a new file. This
2680 means that you can't go back to an older release without removing the
2683 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2684 work if there are any delivery delays because "one_time" forces the
2685 parent to be marked "delivered", so its unseen clone is never tried
2686 again. For this reason, Exim now forbids the simultaneous setting of
2689 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2690 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2691 entry if you want to know the details. The fix, however, neglected to
2692 consider the case when local delivery batching is involved. The test for
2693 "previously delivered" was not happening when checking to see if an
2694 address could be batched with a previous (undelivered) one; under
2695 certain circumstances this could lead to multiple deliveries to the same
2698 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2699 in its include files, and this causes problems building Exim.
2701 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2702 header_syntax) but Exim was just ignoring anything given after a slash.
2703 In particular, this caused confusion with an attempt to use "verify =
2704 reverse_host_lookup/defer_ok". An error is now given when options are
2705 supplied for verify items that do not have them. (Maybe reverse_host_
2706 lookup should have a defer_ok option, but that's a different point.)
2708 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2709 defined by RFC 821) to 2048, because there were problems with some AUTH
2710 commands, and RFC 1869 says the size should be increased for extended
2711 SMTP commands that take arguments.
2713 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2716 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2717 "unknown" error; now it says that the functionality isn't in the binary.
2719 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2720 an address' error message when a string expansion fails (syntax or
2721 whatever). Otherwise the password may appear in the log. Following change
2722 PH/42 below, there is no longer a chance of it appearing in a bounce
2725 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2727 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2728 of its addresses was ever tried. (Bugzilla bug #2.)
2730 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2731 the result incorrectly in the debug output. (It correctly added a newline
2732 to what was transported.)
2734 TF/01 Added $received_time.
2736 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2737 commented out examples of how to interface to a virus scanner and to
2738 SpamAssassin. Also added commented examples of av_scanner and
2739 spamd_address settings.
2741 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2742 and controls are allowed in which ACLs. There were a couple of minor
2743 errors. Some of the entries in the conditions table (which is a table of
2744 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2745 negation of where the condition IS allowed.
2747 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2749 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2750 header file does not have a version number, so I've had to invent a new
2751 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2752 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2753 radiusclient), but it was contributed by a Radius user.
2755 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2756 files or directories, for OpenSSL.
2758 PH/19 When an Exim process that is running as root has to create an Exim log
2759 file, it does so in a subprocess that runs as exim:exim so as to get the
2760 ownership right at creation (otherwise, other Exim processes might see
2761 the file with the wrong ownership). There was no test for failure of this
2762 fork() call, which would lead to the process getting stuck as it waited
2763 for a non-existent subprocess. Forks do occasionally fail when resources
2764 run out. I reviewed all the other calls to fork(); they all seem to check
2767 PH/20 When checking for unexpected SMTP input at connect time (before writing
2768 the banner), Exim was not dealing correctly with a non-positive return
2769 from the read() function. If the client had disconnected by this time,
2770 the result was a log entry for a synchronization error with an empty
2771 string after "input=" when read() returned zero. If read() returned -1
2772 (an event I could not check), uninitialized data bytes were printed.
2773 There were reports of junk text (parts of files, etc) appearing after
2776 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2778 PH/22 Added support for macro redefinition, and (re)definition in between
2779 driver and ACL definitions.
2781 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2782 forgetting to use the resulting value; it was using the unexpanded value.
2784 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2785 hadn't been configured. The fix is from Juergen Kreileder, who
2786 understands it better than I do:
2788 "Here's what I see happening with three configured cyrus_sasl
2789 authenticators configured (plain, login, cram-md5):
2791 On startup auth_cyrus_sasl_init() gets called for each of these.
2792 This means three calls to sasl_listmech() without a specified mech_list.
2793 => SASL tests which mechs of all available mechs actually work
2794 => three warnings about OTP not working
2795 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2797 With the patch, sasl_listmech() also gets called three times. But now
2798 SASL's mech_list option is set to the server_mech specified in the the
2799 authenticator. Or in other words, the answer from sasl_listmech()
2800 gets limited to just the mech you're testing for (which is different
2802 => the return list contains just 'plain' or 'login', 'cram-md5' or
2803 nothing depending on the value of ob->server_mech.
2805 I've just tested the patch: Authentication still works fine,
2806 unavailable mechs specified in the exim configuration are still
2807 caught, and the auth.log warnings about OTP are gone."
2809 PH/25 When debugging is enabled, the contents of the command line are added
2810 to the debugging output, even when log_selector=+arguments is not
2813 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2814 answer is "GNU", and only if the return is "GNU/something" is the answer
2817 PH/27 $acl_verify_message is now set immediately after the failure of a
2818 verification in an ACL, and so is available in subsequent modifiers. In
2819 particular, the message can be preserved by coding like this:
2821 warn !verify = sender
2822 set acl_m0 = $acl_verify_message
2824 Previously, $acl_verify_message was set only while expanding "message"
2825 and "log_message" when a very denied access.
2827 PH/28 Modified OS/os.c-Linux with
2829 -#ifndef OS_LOAD_AVERAGE
2830 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2832 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2833 nomenclature these days.)
2835 PH/29 Installed patch from the Sieve maintainer that adds the options
2836 sieve_useraddress and sieve_subaddress to the redirect router.
2838 PH/30 In these circumstances:
2839 . Two addresses routed to the same list of hosts;
2840 . First host does not offer TLS;
2841 . First host accepts first address;
2842 . First host gives temporary error to second address;
2843 . Second host offers TLS and a TLS session is established;
2844 . Second host accepts second address.
2845 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2846 and peerdn, if requested) that were in fact used only for the second
2849 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2850 attention to any local part prefix or suffix that was matched by the
2851 router that accepted the address. It now behaves in the same way as it
2852 does for delivery: the affixes are removed from the local part unless
2853 rcpt_include_affixes is set on the transport.
2855 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2856 timeout during the DATA phase of an incoming message.
2858 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2859 applied a patch sanctioned by the Sieve maintainer.
2861 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2862 the uid or gid is negative. A case of a negative gid caused this to be
2863 noticed. The fix allows for either to be negative.
2865 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2866 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2869 PH/36 The change PH/12 above was broken. Fixed it.
2871 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2872 the grounds that routing the same address twice would always produce the
2873 same answer. This might have been true once, but it is certainly no
2874 longer true now. Routing a child address may depend on the previous
2875 routing that produced that child. Some complicated redirection strategies
2876 went wrong when messages had multiple recipients, and made Exim's
2877 behaviour dependent on the order in which the addresses were given.
2879 I have moved the duplicate checking until after the routing is complete.
2880 Exim scans the addresses that are assigned to local and remote
2881 transports, and removes any duplicates. This means that more work will be
2882 done, as duplicates will always all be routed, but duplicates are
2883 presumably rare, so I don't expect this is of any significance.
2885 For deliveries to pipes, files, and autoreplies, the duplicate checking
2886 still happens during the routing process, since they are not going to be
2889 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2890 It corrects a timeout issue with spamd. This is Ian's comment: "The
2891 background is that sometimes spamd either never reads data from a
2892 connection it has accepted, or it never writes response data. The exiscan
2893 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2894 blindly assumes that writes won't block so it may never time out."
2896 PH/39 Allow G after quota size as well as K and M.
2898 PH/40 The value set for $authenticated_id in an authenticator may not contain
2899 binary zeroes or newlines because the value is written to log lines and
2900 to spool files. There was no check on this. Now the value is run through
2901 the string_printing() function so that such characters are converted to
2902 printable escape sequences.
2904 PH/41 $message_linecount is a new variable that contains the total number of
2905 lines in the message. Compare $body_linecount, which is the count for the
2908 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2909 bounce and delay warning messages, except in certain special cases, which
2912 (a) An SMTP error message from a remote host;
2913 (b) A message specified in a :fail: redirection;
2914 (c) A message specified in a "fail" command in a system filter;
2915 (d) A message specified in a FAIL return from the queryprogram router;
2916 (e) A message specified by the cannot_route_message router option.
2918 In these cases only, Exim does include the error details in bounce and
2919 warning messages. There are also a few cases where bland messages such
2920 as "unrouteable address" or "local delivery error" are given.
2922 PH/43 $value is now also set for the "else" part of a ${run expansion.
2924 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2925 being worked on, but at least Exim now implements the latest version to
2928 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2929 process to complete was treated as a delivery failure, a timeout while
2930 writing the message to the pipe was logged, but erroneously treated as a
2931 successful delivery. Such timeouts include transport filter timeouts. For
2932 consistency with the overall process timeout, these timeouts are now
2933 treated as errors, giving rise to delivery failures by default. However,
2934 there is now a new Boolean option for the pipe transport called
2935 timeout_defer, which, if set TRUE, converts the failures into defers for
2936 both kinds of timeout. A transport filter timeout is now identified in
2939 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2940 systems where "make" and "gmake" are different, calling "gmake" at top
2941 level broke things. I've arranged for the value of $(MAKE) to be passed
2942 from the Makefile to this script so that it can call the same version of
2946 A note about Exim versions 4.44 and 4.50
2947 ----------------------------------------
2949 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2950 changes of various kinds. As a consequence, a big documentation update was
2951 needed. This delayed the release for rather longer than seemed good, especially
2952 in the light of a couple of (minor) security issues. Therefore, the changes
2953 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2954 release. So 4.44 and 4.50 are in effect two different branches that both start
2957 I have left the 4.50 change log unchanged; it contains all the changes since
2958 4.43. The change log for 4.44 is below; many of its items are identical to
2959 those for 4.50. This seems to be the most sensible way to preserve the
2960 historical information.
2966 1. Minor wording change to the doc/README.SIEVE file.
2968 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2969 computation of the current number of files was incorrect.
2971 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2972 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2975 4. Give more explanation in the error message when the command for a transport
2976 filter fails to execute.
2978 5. There are several places where Exim runs a non-Exim command in a
2979 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2980 was being done only for the command run by the queryprogram router. It is
2981 now done for all such subprocesses. The other cases are: ${run, transport
2982 filters, and the commands run by the lmtp and pipe transports.
2984 6. Added CONFIGURE_GROUP build-time option.
2986 7. Some older OS have a limit of 256 on the maximum number of file
2987 descriptors. Exim was using setrlimit() to set 1000 as a large value
2988 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2989 systems. I've change it so that if it can't get 1000, it tries for 256.
2991 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2992 was an oversight, and furthermore, ever since the addition of extra
2993 controls (e.g. 4.43/32), the checks on when to allow different forms of
2994 "control" were broken. There should now be diagnostics for all cases when a
2995 control that does not make sense is encountered.
2997 9. Added the /retain_sender option to "control=submission".
2999 10. $recipients is now available in the predata ACL (oversight).
3001 11. Tidy the search cache before the fork to do a delivery from a message
3002 received from the command line. Otherwise the child will trigger a lookup
3003 failure and thereby defer the delivery if it tries to use (for example) a
3004 cached ldap connection that the parent has called unbind on.
3006 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3007 of $address_data from the recipient verification was clobbered by the
3008 sender verification.
3010 13. The value of address_data from a sender verification is now available in
3011 $sender_address_data in subsequent conditions in the ACL statement.
3013 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3015 15. Added a new option "connect=<time>" to callout options, to set a different
3018 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3019 was its contents. (It was OK if the option was not defined at all.)
3021 17. A "Completed" log line is now written for messages that are removed from
3022 the spool by the -Mrm option.
3024 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3025 information about exactly what failed.
3027 19. Added -dd to debug only the daemon process.
3029 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3030 handles timeouts, both on the server side and network timeouts. Renamed the
3031 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3033 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3034 It is now set to "smtps".
3036 22. $host_address is now set to the target address during the checking of
3037 ignore_target_hosts.
3039 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3040 being passed; this would have caused $sender_host_name to have been used if
3041 matching the list had actually called for a host name (not very likely,
3042 since this list is usually IP addresses). A host name is now passed as
3045 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3046 code that specifies a non-restarting handler (typically sigaction() in
3047 modern systems) in an attempt to fix a rare and obscure crash bug.
3049 25. Narrowed the window for a race in the daemon that could cause it to ignore
3050 SIGCHLD signals. This is not a major problem, because they are used only to
3051 wake it up if nothing else does.
3053 26. A malformed maildirsize file could cause Exim to calculate negative values
3054 for the mailbox size or file count. Odd effects could occur as a result.
3055 The maildirsize information is now recalculated if the size or filecount
3058 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3059 support for a long time. Removed HAVE_SYS_VFS_H.
3061 28. Installed the latest version of exipick from John Jetmore.
3063 29. In an address list, if the pattern was not a regular expression, an empty
3064 subject address (from a bounce message) matched only if the pattern was an
3065 empty string. Non-empty patterns were not even tested. This was the wrong
3066 because it is perfectly reasonable to use an empty address as part of a
3067 database query. An empty address is now tested by patterns that are
3068 lookups. However, all the other forms of pattern expect the subject to
3069 contain a local part and a domain, and therefore, for them, an empty
3070 address still always fails if the pattern is not itself empty.
3072 30. Exim went into a mad DNS loop when attempting to do a callout where the
3073 host was specified on an smtp transport, and looking it up yielded more
3074 than one IP address.
3076 31. Re-factored the code for checking spool and log partition space into a
3077 function that finds that data and another that does the check. The former
3078 is then used to implement four new variables: $spool_space, $log_space,
3079 $spool_inodes, and $log_inodes.
3081 32. The RFC2047 encoding function was originally intended for short strings
3082 such as real names; it was not keeping to the 75-character limit for
3083 encoded words that the RFC imposes. It now respects the limit, and
3084 generates multiple encoded words if necessary. To be on the safe side, I
3085 have increased the buffer size for the ${rfc2047: expansion operator from
3088 33. It is now permitted to omit both strings after an "if" condition; if the
3089 condition is true, the result is "true". As before, when the second string
3090 is omitted, a false condition yields an empty string. This makes it less
3091 cumbersome to write custom ACL and router conditions.
3093 34. Failure to deliver a bounce message always caused it to be frozen, even if
3094 there was an errors_to setting on the router. The errors_to setting is now
3097 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3098 canonical form (fully expanded) before being placed in
3099 $sender_host_address.
3101 36. The table in the code that translates DNS record types into text (T_A to
3102 "A" for instance) was missing entries for NS and CNAME. It is just possible
3103 that this could have caused confusion if both these types were looked up
3104 for the same domain, because the text type is used as part of Exim's
3105 per-process caching. But the chance of anyone hitting this buglet seems
3108 37. The dnsdb lookup has been extended in a number of ways.
3110 (1) There is a new type, "zns", which walks up the domain tree until it
3111 finds some nameserver records. It should be used with care.
3113 (2) There is a new type, "mxh", which is like "mx" except that it returns
3114 just the host names, not the priorities.
3116 (3) It is now possible to give a list of domains (or IP addresses) to be
3117 looked up. The behaviour when one of the lookups defers can be
3118 controlled by a keyword.
3120 (4) It is now possible to specify the separator character for use when
3121 multiple records are returned.
3123 38. The dnslists ACL condition has been extended: it is now possible to supply
3124 a list of IP addresses and/or domains to be looked up in a particular DNS
3127 39. Added log_selector=+queue_time_overall.
3129 40. When running the queue in the test harness, wait just a tad after forking a
3130 delivery process, to get repeatability of debugging output.
3132 41. Include certificate and key file names in error message when GnuTLS fails
3133 to set them up, because the GnuTLS error message doesn't include the name
3134 of the failing file when there is a problem reading it.
3136 42. Allow both -bf and -bF in the same test run.
3138 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3140 44. The "Exiscan patch" is now merged into the mainline Exim source.
3142 45. Sometimes the final signoff response after QUIT could fail to get
3143 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3144 < 0 before doing a fflush(). This bug looks as though it goes back to the
3145 introduction of TLS in release 3.20, but "sometimes" must have been rare
3146 because the tests only now provoked it.
3148 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3149 (this can affect the format of dates).
3151 47. exim_tidydb, when checking for the continued existence of a message for
3152 which it has found a message-specific retry record, was not finding
3153 messages that were in split spool directories. Consequently, it was
3154 deleting retry records that should have stayed in existence.
3156 48. Steve fixed some bugs in eximstats.
3158 49. The SPA authentication driver was not abandoning authentication and moving
3159 on to the next authenticator when an expansion was forced to fail,
3160 contradicting the general specification for all authenticators. Instead it
3161 was generating a temporary error. It now behaves as specified.
3163 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3164 (the order specifies the preference for clients). The order is now AES256,
3165 AES128, 3DES, ARCFOUR128.
3167 51. Small patch to Sieve code - explicitly set From: when generating an
3170 52. Exim crashed if a remote delivery caused a very long error message to be
3171 recorded - for instance if somebody sent an entire SpamAssassin report back
3172 as a large number of 550 error lines. This bug was coincidentally fixed by
3173 increasing the size of one of Exim's internal buffers (big_buffer) that
3174 happened as part of the Exiscan merge. However, to be on the safe side, I
3175 have made the code more robust (and fixed the comments that describe what
3178 53. Now that there can be additional text after "Completed" in log lines (if
3179 the queue_time_overall log selector is set), a one-byte patch to exigrep
3180 was needed to allow it to recognize "Completed" as not the last thing in
3183 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3184 patch that reportedly fixes this has been added. I am not expert enough to
3185 create a test for it. This is what the patch creator wrote:
3187 "I found a little strange behaviour of ldap code when working with
3188 Windows 2003 AD Domain, where users was placed in more than one
3189 Organization Units. When I tried to give exim partial DN, the exit code
3190 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3191 But simultaneously result of request was absolutely normal ldap result,
3192 so I produce this patch..."
3194 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3195 REFERENCE, so I have modified the code to exclude the patch when that macro
3198 55. Some experimental protocols are using DNS PTR records for new purposes. The
3199 keys for these records are domain names, not reversed IP addresses. The
3200 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3201 leaves it alone. Component reversal etc. now happens only for IP addresses.
3204 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3206 57. Double the size of the debug message buffer (to 2048) so that more of very
3207 long debug lines gets shown.
3209 58. The exicyclog utility now does better if the number of log files to keep
3210 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3212 59. Two changes related to the smtp_active_hostname option:
3214 (1) $smtp_active_hostname is now available as a variable.
3215 (2) The default for smtp_banner uses $smtp_active_hostname instead
3216 of $primary_hostname.
3218 60. The host_aton() function is supposed to be passed a string that is known
3219 to be a valid IP address. However, in the case of IPv6 addresses, it was
3220 not checking this. This is a hostage to fortune. Exim now panics and dies
3221 if the condition is not met. A case was found where this could be provoked
3222 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3223 components; fortuitously, this particular loophole had already been fixed
3224 by change 4.50/55 above.
3226 If there are any other similar loopholes, the new check in host_aton()
3227 itself should stop them being exploited. The report I received stated that
3228 data on the command line could provoke the exploit when Exim was running as
3229 exim, but did not say which command line option was involved. All I could
3230 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3231 running as the user.
3234 61. There was a buffer overflow vulnerability in the SPA authentication code
3235 (which came originally from the Samba project). I have added a test to the
3236 spa_base64_to_bits() function which I hope fixes it.
3239 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3240 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3242 63. The daemon start-up calls getloadavg() while still root for those OS that
3243 need the first call to be done as root, but it missed one case: when
3244 deliver_queue_load_max is set with deliver_drop_privilege. This is
3245 necessary for the benefit of the queue runner, because there is no re-exec
3246 when deliver_drop_privilege is set.
3248 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3249 This has been fixed.
3251 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3252 was in use, was not putting the data itself into the right store pool;
3253 consequently, it could be overwritten for a subsequent message in the same
3254 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3257 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3259 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3260 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3261 the function were treating the return as a boolean value, which happened to
3262 work because 0=false and not-0=true, but is not correct code.
3264 68. The host_aton() function was not handling scoped IPv6 addresses (those
3265 with, for example, "%eth0" on the end) correctly.
3267 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3268 negated items (that is, ~something) in unsigned ints. Some compilers
3269 apparently mutter when there is no cast.
3271 70. If an address verification called from an ACL failed, and did not produce a
3272 user-specific message (i.e. there was only a "system" message), nothing was
3273 put in $acl_verify_message. In this situation, it now puts the system
3276 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3277 session; change 4.31/43 added the unwanted input to the log line - except
3278 that it did not do this in the start of session case. It now does.
3280 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3281 This is wrong and can cause the other end to generate a synchronization
3282 error if it is another Exim or anything else that does the synchronization
3283 check. A QUIT command is no longer sent after a timeout.
3285 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3286 during host lookups.
3288 74. The defer_ok option of callout verification was not working if it was used
3289 when verifying addresses in header lines, that is, for this case:
3291 verify = header_sender/callout=defer_ok
3293 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3294 those file descriptors could be used for SMTP connections. If anything
3295 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3296 could be sent to the SMTP client, causing chaos. The daemon now opens
3297 stdin, stdout, and stderr to /dev/null when it puts itself into the
3300 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3301 log by default. The user can override this with suitable Perl magic.
3303 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3304 the log message when discard triggers, was not working for the DATA ACL or
3305 for the non-SMTP ACL.
3307 78. Error message wording change in sieve.c.
3309 79. If smtp_accept_max_per_host was set, the number of connections could be
3310 restricted to fewer than expected, because the daemon was trying to set up
3311 a new connection before checking whether the processes handling previous
3312 connections had finished. The check for completed processes is now done
3313 earlier. On busy systems, this bug wouldn't be noticed because something
3314 else would have woken the daemon, and it would have reaped the completed
3317 80. If a message was submitted locally by a user whose login name contained one
3318 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3319 It caused a spool format error. I have fixed the spool reading code. A
3320 related problem was that the "from" clause in the Received: line became
3321 illegal because of the space(s). It is now covered by ${quote_local_part.
3323 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3326 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3328 83. Patch from Sieve maintainer to fix unterminated string problem in
3329 "vacation" handling.
3331 84. Some minor changes to the Linux configuration files to help with other
3332 OS variants using glibc.
3334 85. One more patch for Sieve to update vacation handling to latest spec.
3337 ----------------------------------------------------
3338 See the note above about the 4.44 and 4.50 releases.
3339 ----------------------------------------------------
3345 1. Change 4.43/35 introduced a bug that caused file counts to be
3346 incorrectly computed when quota_filecount was set in an appendfile
3349 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3350 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3353 3. Give more explanation in the error message when the command for a transport
3354 filter fails to execute.
3356 4. There are several places where Exim runs a non-Exim command in a
3357 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3358 was being done only for the command run by the queryprogram router. It is
3359 now done for all such subprocesses. The other cases are: ${run, transport
3360 filters, and the commands run by the lmtp and pipe transports.
3362 5. Some older OS have a limit of 256 on the maximum number of file
3363 descriptors. Exim was using setrlimit() to set 1000 as a large value
3364 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3365 systems. I've change it so that if it can't get 1000, it tries for 256.
3367 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3368 was an oversight, and furthermore, ever since the addition of extra
3369 controls (e.g. 4.43/32), the checks on when to allow different forms of
3370 "control" were broken. There should now be diagnostics for all cases when a
3371 control that does not make sense is encountered.
3373 7. $recipients is now available in the predata ACL (oversight).
3375 8. Tidy the search cache before the fork to do a delivery from a message
3376 received from the command line. Otherwise the child will trigger a lookup
3377 failure and thereby defer the delivery if it tries to use (for example) a
3378 cached ldap connection that the parent has called unbind on.
3380 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3381 of $address_data from the recipient verification was clobbered by the
3382 sender verification.
3384 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3385 was its contents. (It was OK if the option was not defined at all.)
3387 11. A "Completed" log line is now written for messages that are removed from
3388 the spool by the -Mrm option.
3390 12. $host_address is now set to the target address during the checking of
3391 ignore_target_hosts.
3393 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3394 being passed; this would have caused $sender_host_name to have been used if
3395 matching the list had actually called for a host name (not very likely,
3396 since this list is usually IP addresses). A host name is now passed as
3399 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3400 code that specifies a non-restarting handler (typically sigaction() in
3401 modern systems) in an attempt to fix a rare and obscure crash bug.
3403 15. Narrowed the window for a race in the daemon that could cause it to ignore
3404 SIGCHLD signals. This is not a major problem, because they are used only to
3405 wake it up if nothing else does.
3407 16. A malformed maildirsize file could cause Exim to calculate negative values
3408 for the mailbox size or file count. Odd effects could occur as a result.
3409 The maildirsize information is now recalculated if the size or filecount
3412 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3413 support for a long time. Removed HAVE_SYS_VFS_H.
3415 18. Updated exipick to current release from John Jetmore.
3417 19. Allow an empty sender to be matched against a lookup in an address list.
3418 Previously the only cases considered were a regular expression, or an
3421 20. Exim went into a mad DNS lookup loop when doing a callout where the
3422 host was specified on the transport, if the DNS lookup yielded more than
3425 21. The RFC2047 encoding function was originally intended for short strings
3426 such as real names; it was not keeping to the 75-character limit for
3427 encoded words that the RFC imposes. It now respects the limit, and
3428 generates multiple encoded words if necessary. To be on the safe side, I
3429 have increased the buffer size for the ${rfc2047: expansion operator from
3432 22. Failure to deliver a bounce message always caused it to be frozen, even if
3433 there was an errors_to setting on the router. The errors_to setting is now
3436 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3437 canonical form (fully expanded) before being placed in
3438 $sender_host_address.
3440 24. Updated eximstats to version 1.33
3442 25. Include certificate and key file names in error message when GnuTLS fails
3443 to set them up, because the GnuTLS error message doesn't include the name
3444 of the failing file when there is a problem reading it.
3446 26. Expand error message when OpenSSL has problems setting up cert/key files.
3449 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3450 (this can affect the format of dates).
3452 28. exim_tidydb, when checking for the continued existence of a message for
3453 which it has found a message-specific retry record, was not finding
3454 messages that were in split spool directories. Consequently, it was
3455 deleting retry records that should have stayed in existence.
3457 29. eximstats updated to version 1.35
3458 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3459 1.35 - bugfix such that pie charts by volume are generated correctly
3461 30. The SPA authentication driver was not abandoning authentication and moving
3462 on to the next authenticator when an expansion was forced to fail,
3463 contradicting the general specification for all authenticators. Instead it
3464 was generating a temporary error. It now behaves as specified.
3466 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3467 (the order specifies the preference for clients). The order is now AES256,
3468 AES128, 3DES, ARCFOUR128.
3470 31. Small patch to Sieve code - explicitly set From: when generating an
3473 32. Exim crashed if a remote delivery caused a very long error message to be
3474 recorded - for instance if somebody sent an entire SpamAssassin report back
3475 as a large number of 550 error lines. This bug was coincidentally fixed by
3476 increasing the size of one of Exim's internal buffers (big_buffer) that
3477 happened as part of the Exiscan merge. However, to be on the safe side, I
3478 have made the code more robust (and fixed the comments that describe what
3481 33. Some experimental protocols are using DNS PTR records for new purposes. The
3482 keys for these records are domain names, not reversed IP addresses. The
3483 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3484 leaves it alone. Component reversal etc. now happens only for IP addresses.
3487 34. The host_aton() function is supposed to be passed a string that is known
3488 to be a valid IP address. However, in the case of IPv6 addresses, it was
3489 not checking this. This is a hostage to fortune. Exim now panics and dies
3490 if the condition is not met. A case was found where this could be provoked
3491 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3492 components; fortuitously, this particular loophole had already been fixed
3493 by change 4.50/55 or 4.44/33 above.
3495 If there are any other similar loopholes, the new check in host_aton()
3496 itself should stop them being exploited. The report I received stated that
3497 data on the command line could provoke the exploit when Exim was running as
3498 exim, but did not say which command line option was involved. All I could
3499 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3500 running as the user.
3503 35. There was a buffer overflow vulnerability in the SPA authentication code
3504 (which came originally from the Samba project). I have added a test to the
3505 spa_base64_to_bits() function which I hope fixes it.
3508 36. The daemon start-up calls getloadavg() while still root for those OS that
3509 need the first call to be done as root, but it missed one case: when
3510 deliver_queue_load_max is set with deliver_drop_privilege. This is
3511 necessary for the benefit of the queue runner, because there is no re-exec
3512 when deliver_drop_privilege is set.
3514 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3515 was in use, was not putting the data itself into the right store pool;
3516 consequently, it could be overwritten for a subsequent message in the same
3517 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3520 38. Sometimes the final signoff response after QUIT could fail to get
3521 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3522 < 0 before doing a fflush(). This bug looks as though it goes back to the
3523 introduction of TLS in release 3.20, but "sometimes" must have been rare
3524 because the tests only now provoked it.
3530 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3531 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3532 Now it returns an integer. A number of calls were still expecting a T/F
3533 return. Fortuitously, in all cases, the tests worked in OK situations,
3534 which is the norm. However, things would have gone wrong on any write
3535 failures on the smtp file descriptor. This function is used when sending
3536 messages over SMTP and also when doing verify callouts.
3538 2. When Exim is called to do synchronous delivery of a locally submitted
3539 message (the -odf or -odi options), it no longer closes stderr before doing
3542 3. Implemented the mua_wrapper option.
3544 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3546 5. Implemented the functions header_remove(), header_testname(),
3547 header_add_at_position(), and receive_remove_recipient(), and exported them
3550 6. If an ACL "warn" statement specified the addition of headers, Exim already
3551 inserted X-ACL-Warn: at the start if there was no header name. However, it
3552 was not making this test for the second and subsequent header lines if
3553 there were newlines in the string. This meant that an invalid header could
3554 be inserted if Exim was badly configured.
3556 7. Allow an ACL "warn" statement to add header lines at the start or after all
3557 the Received: headers, as well as at the end.
3559 8. Added the rcpt_4xx retry error code.
3561 9. Added postmaster_mailfrom=xxx to callout verification option.
3563 10. Added mailfrom=xxxx to the callout verification option, for verify=
3566 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3567 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3569 12. Inserted some casts to stop certain compilers warning when using pointer
3570 differences as field lengths or precisions in printf-type calls (mostly
3571 affecting debugging statements).
3573 13. Added optional readline() support for -be (dynamically loaded).
3575 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3576 same clock tick as a message's arrival, so that its received time was the
3577 same as the "first fail" time on the retry record, and that message
3578 remained on the queue past the ultimate address timeout, every queue runner
3579 would try a delivery (because it was past the ultimate address timeout) but
3580 after another failure, the ultimate address timeout, which should have then
3581 bounced the address, did not kick in. This was a "< instead of <=" error;
3582 in most cases the first failure would have been in the next clock tick
3583 after the received time, and all would be well.
3585 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3586 being recognized when the domain list was tested by the match_domain
3587 condition in an expansion string.
3589 16. Added the ${str2b64: operator.
3591 17. Exim was always calling setrlimit() to set a large limit for the number of
3592 processes, without checking whether the existing limit was already
3593 adequate. (It did check for the limit on file descriptors.) Furthermore,
3594 errors from getrlimit() and setrlimit() were being ignored. Now they are
3595 logged to the main and panic logs, but Exim does carry on, to try to do its
3596 job under whatever limits there are.
3598 18. Imported PCRE 5.0.
3600 19. Trivial typo in log message " temporarily refused connection" (the leading
3603 20. If the log selector return_path_on_delivery was set and an address was
3604 redirected to /dev/null, the delivery process crashed because it assumed
3605 that a return path would always be set for a "successful" delivery. In this
3606 case, the whole delivery is bypassed as an optimization, and therefore no
3609 21. Internal re-arrangement: the function for sending a challenge and reading
3610 a response while authentication was assuming a zero-terminated challenge
3611 string. It's now changed to take a pointer and a length, to allow for
3612 binary data in such strings.
3614 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3616 23. Exim was not respecting finduser_retries when seeking the login of the
3617 uid under which it was called; it was always trying 10 times. (The default
3618 setting of finduser_retries is zero.) Also, it was sleeping after the final
3619 failure, which is pointless.
3621 24. Implemented tls_on_connect_ports.
3623 25. Implemented acl_smtp_predata.
3625 26. If the domain in control=submission is set empty, Exim assumes that the
3626 authenticated id is a complete email address when it generates From: or
3627 Sender: header lines.
3629 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3630 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3631 chown and chgrp in /bin and hostname in /usr/bin.
3633 28. Exim was keeping the "process log" file open after each use, just as it
3634 does for the main log. This opens the possibility of it remaining open for
3635 long periods when the USR1 signal hits a daemon. Occasional processlog
3636 errors were reported, that could have been caused by this. Anyway, it seems
3637 much more sensible not to leave this file open at all, so that is what now
3640 29. The long-running daemon process does not normally write to the log once it
3641 has entered its main loop, and it closes the log before doing so. This is
3642 so that log files can straightforwardly be renamed and moved. However,
3643 there are a couple of unusual error situations where the daemon does write
3644 log entries, and I had neglected to close the log afterwards.
3646 30. The text of an SMTP error response that was received during a remote
3647 delivery was being truncated at 512 bytes. This is too short for some of
3648 the long messages that one sometimes sees. I've increased the limit to
3651 31. It is now possible to make retry rules that apply only when a message has a
3652 specific sender, in particular, an empty sender.
3654 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3655 it possible to be selective about when SMTP synchronization is enforced.
3657 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3659 32. Implemented hosts_connection_nolog.
3661 33. Added an ACL for QUIT.
3663 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3666 35. Added mailbox_size and mailbox_filecount to appendfile.
3668 36. Added control = no_multiline_responses to ACLs.
3670 37. There was a bug in the logic of the code that waits for the clock to tick
3671 in the case where the clock went backwards by a substantial amount such
3672 that the microsecond fraction of "now" was more than the microsecond
3673 fraction of "then" (but the whole seconds number was less).
3675 38. Added support for the libradius Radius client library this is found on
3676 FreeBSD (previously only the radiusclient library was supported).
3682 1. When certain lookups returned multiple values in the form name=value, the
3683 quoting of the values was not always being done properly. Specifically:
3684 (a) If the value started with a double quote, but contained no whitespace,
3686 (b) If the value contained whitespace other than a space character (i.e.
3687 tabs or newlines or carriage returns) it was not quoted.
3688 This fix has been applied to the mysql and pgsql lookups by writing a
3689 separate quoting function and calling it from the lookup code. The fix
3690 should probably also be applied to nisplus, ibase and oracle lookups, but
3691 since I cannot test any of those, I have not disturbed their existing code.
3693 2. A hit in the callout cache for a specific address caused a log line with no
3694 reason for rejecting RCPT. Now it says "Previous (cached) callout
3695 verification failure".
3697 3. There was an off-by-one bug in the queryprogram router. An over-long
3698 return line was truncated at 256 instead of 255 characters, thereby
3699 overflowing its buffer with the terminating zero. As well as fixing this, I
3700 have increased the buffer size to 1024 (and made a note to document this).
3702 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3703 when Exim is waiting for an SMTP response from a remote server, Exim
3704 restarts its select() call on the socket, thereby resetting its timeout.
3705 This is not a problem when such interrupts are rare. Somebody set up a cron
3706 job to run exiwhat every 2 minutes, which is less than the normal select()
3707 timeout (5 or 10 minutes). This meant that the select() timeout never
3708 kicked in because it was always reset. I have fixed this by comparing the
3709 time when an interrupt arrives with the time at the start of the first call
3710 to select(). If more time than the timeout has elapsed, the interrupt is
3711 treated as a timeout.
3713 5. Some internal re-factoring in preparation for the addition of Sieve
3714 extensions (by MH). In particular, the "personal" test is moved to a
3715 separate function, and given an option for scanning Cc: and Bcc: (which is
3716 not set for Exim filters).
3718 6. When Exim created an email address using the login of the caller as the
3719 local part (e.g. when creating a From: or Sender: header line), it was not
3720 quoting the local part when it contained special characters such as @.
3722 7. Installed new OpenBSD configuration files.
3724 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3725 try to make them clearer.
3727 9. Callout options, other than the timeout value, were being ignored when
3728 verifying sender addresses in header lines. For example, when using
3730 verify = header_sender/callout=no_cache
3732 the cache was (incorrectly) being used.
3734 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3735 only the Cygwin environment.
3737 11. When return_path_on_delivery was set as a log selector, if different remote
3738 addresses in the same message used different return paths and parallel
3739 remote delivery occurred, the wrong values would sometimes be logged.
3740 (Whenever a remote delivery process finished, the return path value from
3741 the most recently started remote delivery process was logged.)
3743 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3744 lines when AUTH and/or TLS are in use. This is the "received protocol"
3745 field. Exim used to use "asmtp" for authenticated SMTP, without any
3746 indication (in the protocol name) for TLS use. Now it follows the RFC and
3747 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3748 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3749 names appear in log lines as well as in Received: header lines.
3751 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3752 extensions, and comparison tests, and to fix some bugs.
3754 14. Changes to the "personal" filter test:
3756 (1) The test was buggy in that it was just doing the equivalent of
3757 "contains" tests on header lines. For example, if a user's address was
3758 anne@some.where, the "personal" test would incorrectly be true for
3760 To: susanne@some.where
3762 This test is now done by extracting each address from the header in turn,
3763 and checking the entire address. Other tests that are part of "personal"
3764 are now done using regular expressions (for example, to check local parts
3765 of addresses in From: header lines).
3767 (2) The list of non-personal local parts in From: addresses has been
3768 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3769 taken from the Sieve specification recommendations.
3771 (3) If the message contains any header line starting with "List-" it is
3772 treated as non-personal.
3774 (4) The test for "circular" in the Subject: header line has been removed
3775 because it now seems ill-conceived.
3777 15. Minor typos in src/EDITME comments corrected.
3779 16. Installed latest exipick from John Jetmore.
3781 17. If headers_add on a router specified a text string that was too long for
3782 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3783 of string_sprintf() is now avoided.
3785 18. $message_body_size was not set (it was always zero) when running the DATA
3786 ACL and the local_scan() function.
3788 19. For the "mail" command in an Exim filter, no default was being set for
3789 the once_repeat time, causing a random time value to be used if "once" was
3790 specified. (If the value happened to be <= 0, no repeat happened.) The
3791 default is now 0s, meaning "never repeat". The "vacation" command was OK
3792 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3793 (I found it when inspecting the code).
3795 20. There is now an overall timeout for performing a callout verification. It
3796 defaults to 4 times the callout timeout, which applies to individual SMTP
3797 commands during the callout. The overall timeout applies when there is more
3798 than one host that can be tried. The timeout is checked before trying the
3799 next host. This prevents very long delays if there are a large number of
3800 hosts and all are timing out (e.g. when the network connections are timing
3801 out). The value of the overall timeout can be changed by specifying an
3802 additional sub-option for "callout", called "maxwait". For example:
3804 verify = sender/callout=5s,maxwait=20s
3806 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3807 to the end before writing, but this should make it even safer).
3809 22. Exim was forgetting that it had advertised PIPELINING for the second and
3810 subsequent messages on an SMTP connection. It was also not resetting its
3811 memory on STARTTLS and an internal HELO.
3813 23. When Exim logs an SMTP synchronization error within a session, it now
3814 records whether PIPELINING has been advertised or not.
3816 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3817 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3818 rather than long int.
3820 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3822 26. Added the never_mail option to autoreply.
3828 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3829 crash if the getsockname() call failed; this can happen if a connection is
3830 closed very soon after it is established. The problem was simply in the
3831 order in which certain operations were done, causing Exim to try to write
3832 to the SMTP stream before it had set up the file descriptor. The bug has
3833 been fixed by making things happen in the correct order.
3839 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3840 before the connection was closed, thus losing the rejection response.
3842 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3843 some early Solaris releases, but causes trouble in current releases where
3844 socklen_t is defined.
3846 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3849 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3852 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3853 that releasing the top of it at the end releases what was used for sub-
3854 expansions (unless the block got too big). However, discard this block if
3855 the first thing is a variable or header, so that we can use its block when
3856 it is dynamic (useful for very large $message_headers, for example).
3858 6. Lookups now cache *every* query, not just the most recent. A new, separate
3859 store pool is used for this. It can be recovered when all lookup caches are
3860 flushed. Lookups now release memory at the end of their result strings.
3861 This has involved some general refactoring of the lookup sources.
3863 7. Some code has been added to the store_xxx() functions to reduce the amount
3864 of flapping under certain conditions.
3866 8. log_incoming_interface used to affect only the <= reception log lines. Now
3867 it causes the local interface and port to be added to several more SMTP log
3868 lines, for example "SMTP connection from", and rejection lines.
3870 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3872 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3874 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3875 could be overwritten at the end of the current message (or the start of a
3876 new message if it was set in a HELO ACL). The value is now preserved for
3877 the duration of the SMTP connection.
3879 12. If a transport had a headers_rewrite setting, and a matching header line
3880 contained an unqualified address, that address was qualified, even if it
3881 did not match any rewriting rules. The underlying bug was that the values
3882 of the flags that permit the existence of unqualified sender and recipient
3883 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3884 non-local messages, and by -bnq for local messages) were not being
3885 preserved with the message after it was received.
3887 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3888 "next input=" as part of the text comprising the host identity instead of
3889 the correct text. The code was using the same buffer for two different
3890 strings. However, depending on which order the printing function evaluated
3891 its arguments, the bug did not always show up. Under Linux, for example, my
3892 test suite worked just fine.
3894 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3895 doesn't work with some older versions of Perl. It has been changed to "my",
3896 which in any case is probably the better facility to use.
3898 15. A really picky compiler found some instances of statements for creating
3899 error messages that either had too many or two few arguments for the format
3902 16. The size of the buffer for calls to the DNS resolver has been increased
3903 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3904 for addresses that have a lot of PTR records. This alleviates a problem; it
3905 does not fully solve it.
3907 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3908 buffer now truncates the list and logs the incident, which is the same
3909 action as happens when Exim is looking up a host name and its aliases.
3910 Previously in this situation something unpredictable would happen;
3911 sometimes it was "internal error: store_reset failed".
3913 18. If a server dropped the connection unexpectedly when an Exim client was
3914 using GnuTLS and trying to read a response, the client delivery process
3915 crashed while trying to generate an error log message.
3917 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3918 string, for example:
3920 warn message = H1: something\nH2: something
3922 the text was added as a single header line from Exim's point of view
3923 though it ended up OK in the delivered message. However, searching for the
3924 second and subsequent header lines using $h_h2: did not work. This has been
3925 fixed. Similarly, if a system filter added multiple headers in this way,
3926 the routers could not see them.
3928 20. Expanded the error message when iplsearch is called with an invalid key to
3929 suggest using net-iplsearch in a host list.
3931 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3932 ACLs are no longer actually imposed (and a message to that effect is
3935 22. If a "gecos" field in a passwd entry contained escaped characters, in
3936 particular, if it contained a \" sequence, Exim got it wrong when building
3937 a From: or a Sender: header from that name. A second bug also caused
3938 incorrect handling when an unquoted " was present following a character
3939 that needed quoting.
3941 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3942 was not being matched caselessly.
3944 24. Arranged for all hyphens in the exim.8 source to be escaped with
3947 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3948 properly. Recipient callout cache records were still being keyed to include
3949 the sender, even when use_sender was set false. This led to far more
3950 callouts that were necessary. The sender is no longer included in the key
3951 when use_sender is false.
3953 26. Added "control = submission" modifier to ACLs.
3955 27. Added the ${base62d: operator to decode base 62 numbers.
3957 28. dnsdb lookups can now access SRV records.
3959 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3960 the configuration file.
3962 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3963 (-v) mode. This makes the output for a verbose queue run more intelligible.
3965 31. Added a use_postmaster feature to recipient callouts.
3967 32. Added the $body_zerocount variable, containing the number of binary zero
3968 bytes in the message body.
3970 33. The time of last modification of the "new" subdirectory is now used as the
3971 "mailbox time last read" when there is a quota error for a maildir
3974 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3976 35. Added +ignore_unknown as a special item in host lists.
3978 36. Code for decoding IPv6 addresses in host lists is now included, even if
3979 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3980 address was recognized as an IP address, but was then not correctly decoded
3981 into binary, causing unexpected and incorrect effects when compared with
3988 1. Very minor rewording of debugging text in manualroute to say "list of
3989 hosts" instead of "hostlist".
3991 2. If verify=header_syntax was set, and a header line with an unqualified
3992 address (no domain) and a large number of spaces between the end of the
3993 name and the colon was received, the reception process suffered a buffer
3994 overflow, and (when I tested it) crashed. This was caused by some obsolete
3995 code that should have been removed. The fix is to remove it!
3997 3. When running in the test harness, delay a bit after writing a bounce
3998 message to get a bit more predictability in the log output.
4000 4. Added a call to search_tidyup() just before forking a reception process. In
4001 theory, someone could use a lookup in the expansion of smtp_accept_max_
4002 per_host which, without the tidyup, could leave open a database connection.
4004 5. Added the variables $recipient_data and $sender_data which get set from a
4005 lookup success in an ACL "recipients" or "senders" condition, or a router
4006 "senders" option, similar to $domain_data and $local_part_data.
4008 6. Moved the writing of debug_print from before to after the "senders" test
4011 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4012 problems for message scanning, either using a data ACL, or using
4013 local_scan() because the Received: header was not generated till after they
4014 were called (in order to set the time as the time of reception completion).
4015 I have revised the way this works. The header is now generated after the
4016 body is received, but before the ACL or local_scan() are called. After they
4017 are run, the timestamp in the header is updated.
4023 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4024 before starting a queue runner without re-exec. This happened only when
4025 deliver_drop_privilege was set or when the Exim user was set to root. The
4026 effect of the bug was that timeouts during subsequent deliveries caused
4027 crashes instead of being properly handled. The handler is now left at its
4028 default (and expected) setting.
4030 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4031 message, again when deliver_drop_privilege is set or Exim is run as root.
4032 The bug described in (1) was not present in this case, but the tidying up
4033 of the other signals was missing. I have made the two cases consistent.
4035 3. The ignore_target_hosts setting on a manualroute router was being ignored
4036 for hosts that were looked up using the /MX notation.
4038 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4041 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4042 operated on the sender address. After changing the $sender_address to <>
4043 for the sender address verify, Exim was re-instated it as the original
4044 (before rewriting) address, but remembering that it had rewritten it, so it
4045 wasn't rewriting it again. This bug also had the effect of breaking the
4046 sender address verification caching when the sender address was rewritten.
4048 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4049 This has been changed so that if the ip literal address matches
4050 ignore_target_hosts, the router declines.
4052 7. Added expansion conditions match_domain, match_address, and match_local_
4053 part (NOT match_host).
4055 8. The placeholder for the Received: header didn't have a length field set.
4057 9. Added code to Exim itself and to exim_lock to test for a specific race
4058 condition that could lead to file corruption when using MBX delivery. The
4059 issue is with the lockfile that is created in /tmp. If this file is removed
4060 after a process has opened it but before that process has acquired a lock,
4061 there is the potential for a second process to recreate the file and also
4062 acquire a lock. This could lead to two Exim processes writing to the file
4063 at the same time. The added code performs the same test as UW imapd; it
4064 checks after acquiring the lock that its file descriptor still refers to
4065 the same named file.
4067 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4068 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4071 11. Added the smtp_active_hostname option. If used, this will typically be made
4072 to depend on the incoming interface address. Because $interface_address is
4073 not set up until the daemon has forked a reception process, error responses
4074 that can happen earlier (such as "too many connections") no longer contain
4077 12. If an expansion in a condition on a "warn" statement fails because a lookup
4078 defers, the "warn" statement is abandoned, and the next ACL statement is
4079 processed. Previously this caused the whole ACL to be aborted.
4081 13. Added the iplsearch lookup type.
4083 14. Added ident_timeout as a log selector.
4085 15. Added tls_certificate_verified as a log selector.
4087 16. Added a global option tls_require_ciphers (compare the smtp transport
4088 option of the same name). This controls incoming TLS connections.
4090 17. I finally figured out how to make tls_require_ciphers do a similar thing
4091 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4092 before starting the TLS session.
4094 18. Tabs are now shown as \t in -bP output.
4096 19. If the log selector return_path_on_delivery was set, Exim crashed when
4097 bouncing a message because it had too many Received: header lines.
4099 20. If two routers both had headers_remove settings, and the first one included
4100 a superfluous trailing colon, the final name in the first list and the
4101 first name in the second list were incorrectly joined into one item (with a
4102 colon in the middle).
4108 1. Added -C and -D options to the exinext utility, mainly to make it easier
4109 to include in the automated testing, but these could be helpful when
4110 multiple configurations are in use.
4112 2. The exinext utility was not formatting the output nicely when there was
4113 an alternate port involved in the retry record key, nor when there was a
4114 message id as well (for retries that were specific to a specific message
4115 and a specific host). It was also confused by IPv6 addresses, because of
4116 the additional colons they contain. I have fixed the IPv4 problem, and
4117 patched it up to do a reasonable job for IPv6.
4119 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4120 delivery, the log line now contains "pipelined" if PIPELINING was used.
4122 4. An SMTP transport process used to panic and die if the bind() call to set
4123 an explicit outgoing interface failed. This has been changed; it is now
4124 treated in the same way as a connect() failure.
4126 5. A reference to $sender_host_name in the part of a conditional expansion
4127 that was being skipped was still causing a DNS lookup. This no longer
4130 6. The def: expansion condition was not recognizing references to header lines
4131 that used bh_ and bheader_.
4133 7. Added the _cache feature to named lists.
4135 8. The code for checking quota_filecount in the appendfile transport was
4136 allowing one more file than it should have been.
4138 9. For compatibility with Sendmail, the command line option
4146 and sets the incoming protocol and host name (for trusted callers). The
4147 host name and its colon can be omitted when only the protocol is to be set.
4148 Note the Exim already has two private options, -pd and -ps, that refer to
4149 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4150 "s", but I don't think that's a major issue.
4152 10. A number of refactoring changes to the code, none of which should affect
4155 (a) The number of logging options was getting close to filling up the
4156 32-bit word that was used as a bit map. I have split them into two classes:
4157 those that are passed in the argument to log_write(), and those that are
4158 only ever tested independently outside of that function. These are now in
4159 separate 32-bit words, so there is plenty of room for expansion again.
4160 There is no change in the user interface or the logging behaviour.
4162 (b) When building, for example, log lines, the code previously used a
4163 macro that called string_cat() twice, in order to add two strings. This is
4164 not really sufficiently general. Furthermore, there was one instance where
4165 it was actually wrong because one of the argument was used twice, and in
4166 one call a function was used. (As it happened, calling the function twice
4167 did not affect the overall behaviour.) The macro has been replaced by a
4168 function that can join an arbitrary number of extra strings onto a growing
4171 (c) The code for expansion conditions now uses a table and a binary chop
4172 instead of a serial search (which was left over from when there were very
4173 few conditions). Also, it now recognizes conditions like "pam" even when
4174 the relevant support is not compiled in: a suitably worded error message is
4175 given if an attempt is made to use such a condition.
4177 11. Added ${time_interval:xxxxx}.
4179 12. A bug was causing one of the ddress fields not to be passed back correctly
4180 from remote delivery subprocesses. The field in question was not being
4181 subsequently used, so this caused to problems in practice.
4183 13. Added new log selectors queue_time and deliver_time.
4185 14. Might have fixed a bug in maildirsizefile handling that threw up
4186 "unexpected character" debug warnings, and recalculated the data
4187 unnecessarily. In any case, I expanded the warning message to give more
4190 15. Added the message "Restricted characters in address" to the statements in
4191 the default ACL that block characters like @ and % in local parts.
4193 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4194 Three changes have been made:
4196 (a) There was a serious bug; a negative response to MAIL caused the whole
4197 recipient domain to be cached as invalid, thereby blocking all messages
4198 to all local parts at the same domain, from all senders. This bug has
4199 been fixed. The domain is no longer cached after a negative response to
4200 MAIL if the sender used is not empty.
4202 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4205 (c) A new callout option, "use_sender" has been added for people who want
4206 the modified behaviour.
4212 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4215 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4216 indeed breaks things for older releases.
4218 3. Added additional logging to the case where there is a problem reading data
4219 from a filter that is running in a subprocess using a pipe, in order to
4220 try to track down a specific problem.
4222 4. Testing facility fudge: when running in the test harness and attempting
4223 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4224 getting "No route to host". Convert this to a timeout.
4226 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4229 6. Some OS don't have socklen_t but use size_t instead. This affects the
4230 fifth argument of getsockopt() amongst other things. This is now
4231 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4232 can be set for individual OS. I have set it for SunOS5, OSF1, and
4233 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4234 some earlier ones do not.
4236 7. Change 4.30/15 was not doing the test caselessly.
4238 8. The standard form for an IPv6 address literal was being rejected by address
4239 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4240 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4241 this, as well as the form without the "IPv6" on the front (but only when
4242 address literals are enabled, of course).
4244 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4246 10. Exim crashed if a message with an empty sender address specified by -f
4247 encountered a router with an errors_to setting. This could be provoked only
4248 by a command such as
4252 where an empty string was supplied; "<>" did not hit this bug.
4254 11. Installed PCRE release 4.5.
4256 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4257 remained set. It is now erased.
4259 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4260 times from message ids (which are base 36 rather than the normal 62).
4262 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4263 were being counted as actual protocol errors, and logged if the log
4264 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4265 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4266 and DATA following a set of rejected RCPTs do not count as protocol errors.
4267 In other words, Exim assumes they were pipelined, though this may not
4268 actually be the case. Of course, in all cases the client gets an
4269 appropriate error code.
4271 15. If a lookup fails in an ACL condition, a message about the failure may
4272 be available; it is used if testing the ACL cannot continue, because most
4273 such messages specify what the cause of the deferral is. However, some
4274 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4275 that caused an old message to be retained and used if a later statement
4276 caused a defer, replacing the real cause of the deferral.
4278 16. If an IP address had so many PTR records that the DNS lookup buffer
4279 was not large enough to hold them, Exim could crash while trying to process
4280 the truncated data. It now detects and logs this case.
4282 17. Further to 4.21/58, another change has been made: if (and only if) the
4283 first line of a message (the first header line) ends with CRLF, a bare LF
4284 in a subsequent header line has a space inserted after it, so as not to
4285 terminate the header.
4287 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4288 unnecessarily, used atoi() instead of strtol(), and didn't check the
4289 termination when getting file sizes from file names by regex.
4291 19. Completely re-implemented the support for maildirsize files, in the light
4292 of a number of problems with the previous contributed implementation
4293 (4.30/29). In particular:
4295 . If the quota is zero, the maildirsize file is maintained, but no quota is
4298 . If the maildir directory does not exist, it is created before any attempt
4299 to write a maildirsize file.
4301 . The quota value in the file is just a cache; if the quota is changed in
4302 the transport, the new value overrides.
4304 . A regular expression is available for excluding directories from the
4307 20. The autoreply transport checks the characters in options that define the
4308 message's headers; it allows continued headers, but it was checking with
4309 isspace() after an embedded newline instead of explicitly looking for a
4312 21. If all the "regular" hosts to which an address was routed had passed their
4313 expiry times, and had not reached their retry times, the address was
4314 bounced, even if fallback hosts were defined. Now Exim should go on to try
4317 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4318 equivalent code in the SMTP transport. Some hosts send humungous responses
4319 to HELO/EHLO, more than 1024 it seems.
4321 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4322 gives compiler warnings in some environments. I've now done it "properly",
4325 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4326 (because of problems with the built-in one) was declared to return uschar *
4327 instead of char *, causing compiler failure.
4329 25. Fixed a file descriptor leak when processing alias/forward files.
4331 26. Fixed a minor format string issue in dbfn.c.
4333 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4335 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4336 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4339 29. When a sender address is verified, it is cached, to save repeating the test
4340 when there is more than one recipient in a message. However, when the
4341 verification involves a callout, it is possible for different callout
4342 options to be set for different recipients. It is too complicated to keep
4343 track of this in the cache, so now Exim always runs a verification when a
4344 callout is required, relying on the callout cache for the optimization.
4345 The overhead is duplication of the address routing, but this should not be
4348 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4349 to be verified with callout=postmaster, and the main callout worked but the
4350 postmaster check failed, the verification correctly failed. However, if a
4351 subsequent RCPT command asked for sender verification *without* the
4352 postmaster check, incorrect caching caused this verification also to fail,
4355 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4356 it was not caching the DNS options (qualify_single, search_parents) that
4357 were used when the lookup failed. A subsequent lookup with different
4358 options therefore always gave the same answer, though there were cases
4359 where it should not have. (Example: a "domains = !$mx_any" option on a
4360 dnslookup router: the "domains" option is always processed without any
4361 widening, but the router might have qualify_single set.) Now Exim uses the
4362 cached value only when the same options are set.
4364 32. Added John Jetmore's "exipick" utility to the distribution.
4366 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4367 than a timeout (e.g. a certificate is required, and is not provided), an
4368 Exim server now closes the connection immediately. Previously it waited for
4369 the client to close - but if the client is SSL, it seems that they each
4370 wait for each other, leading to a delay before one of them times out.
4372 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4373 maintained 0.8.x compatibility because I don't think many are using it, and
4374 it is clearly obsolete.
4376 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4379 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4380 client certificate was expired. A simple patch fixes this, though I don't
4381 understand the full logic of why the verify callback is called multiple
4384 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4385 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4386 which causes problems with some clients (such as the Certicom SSL Plus
4387 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4388 disables the coutermeasure allowing Eudora to connect."
4390 38. Exim was not checking that a write() to a log file succeeded. This could
4391 lead to Bad Things if a log got too big, in particular if it hit a file
4392 size limit. Exim now panics and dies if it cannot write to a log file, just
4393 as it does if it cannot open a log file.
4395 39. Modified OS/Makefile-Linux so that it now contains
4397 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4399 The two -D definitions ensure that Exim is compiled with large file
4400 support, which makes it possible to handle log files that are bigger than
4403 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4404 instance) a domain was checked against a named list that involved a lookup,
4405 causing $domain_data to be set, then another domain was checked against the
4406 same list, then the first domain was re-checked, the value of $domain_data
4407 after the final check could be wrong. In particular, if the second check
4408 failed, it could be set empty. This bug probably also applied to
4411 41. The strip_trailing_dot option was not being applied to the address given
4412 with the -f command-line option.
4414 42. The code for reading a message's header from the spool was incrementing
4415 $received_count, but never initializing it. This meant that the value was
4416 incorrect (doubled) while delivering a message in the same process in which
4417 it was received. In the most common configuration of Exim, this never
4418 happens - a fresh exec is done - but it can happen when
4419 deliver_drop_privilege is set.
4421 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4422 it now includes up to 150 characters of the unexpected data in the log
4425 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4426 and building data strings. The size of both of these buffers was 10 000
4427 bytes - far larger than anybody would *ever* want, thought I. Needless to
4428 say, somebody hit the limit. I have increased the maximum line length to
4429 20 000 and the maximum data length of concatenated lines to 100 000. I have
4430 also fixed two bugs, because there was no checking on these buffers. Tsk,
4431 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4432 buffer is too small.
4434 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4435 lsearch lookups. Now it does.
4437 46. When parsing a route_list item in a manualroute router, a fixed-length
4438 buffer was used for the list of hosts. I made this 1024 bytes long,
4439 thinking that nobody would ever have a list of hosts that long. Wrong.
4440 Somebody had a whole pile of complicated expansion conditions, and the
4441 string was silently truncated, leading to an expansion error. It turns out
4442 that it is easier to change to an unlimited length (owing to other changes
4443 that have happened since this code was originally written) than to build
4444 structure for giving a limitation error. The length of the item that
4445 expands into the list of hosts is now unlimited.
4447 47. The lsearch lookup could not handle data where the length of text line was
4448 more than 4095 characters. Such lines were truncated, leading to shortened
4449 data being returned. It should now handle lines of any length.
4451 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4452 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4455 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4456 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4457 no longer happens in comment lines. A list of replacements is now placed
4458 at the head of all of the source files, except those whose only change is
4459 to replace PERL_COMMAND in the very first #! line.
4461 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4462 messages on the queue, with a bottom-up merge sort, using code contributed
4463 by Michael Haardt. This should make operations like -bp somewhat faster on
4464 large queues. It won't affect queue runners, except when queue_run_in_order
4467 51. Installed eximstats 1.31 in the distribution.
4469 52. Added support for SRV lookups to the dnslookup router.
4471 53. If an ACL referred to $message_body or $message_body_end, the value was not
4472 reset for any messages that followed in the same SMTP session.
4474 54. The store-handling optimization for building very long strings was not
4475 differentiating between the different store pools. I don't think this
4476 actually made any difference in practice, but I've tidied it.
4478 55. While running the routers to verify a sender address, $sender_address
4479 was still set to the sender address. This is wrong, because when routing to
4480 send a bounce to the sender, it would be empty. Therefore, I have changed
4481 it so that, while verifying a sender address, $sender_address is set to <>.
4482 (There is no change to what happens when verifying a recipient address.)
4484 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4485 target A or AAAA records (if not already returned) without resetting the
4486 qualify_single or search_parents options of the DNS resolver. These are
4487 inappropriate in this case because the targets of MX and SRV records must
4488 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4489 target that, when qualified, matched something in the local domain. These
4490 two options are now turned off when doing these lookups.
4492 57. It seems that at least some releases of Reiserfs (which does not have the
4493 concept of a fixed number of inodes) returns zero and not -1 for the
4494 number of available inodes. This interacted badly with check_spool_inodes,
4495 which assumed that -1 was the "no such thing" setting. What I have done is
4496 to check that the total number of inodes is greater than zero before doing
4497 the test of how many are available.
4499 58. When a "warn" ACL statement has a log_message modifier, the message is
4500 remembered, and not repeated. This is to avoid a lot of repetition when a
4501 message has many recipients that cause the same warning to be written.
4502 Howewer, Exim was preserving the list of already written lines for an
4503 entire SMTP session, which doesn't seem right. The memory is now reset if a
4504 new message is started.
4506 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4507 debugging output unless log_rewrite was also set.
4509 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4510 of (int)(handle) when we know that handle contains (void *)(-1).
4512 61. The Exim daemon panic-logs an error return when it closes the incoming
4513 connection. However "connection reset by peer" seems to be common, and
4514 isn't really an error worthy of noting specially, so that particular error
4517 62. When Exim is trying to find all the local interfaces, it used to panic and
4518 die if the ioctl to get the interface flags failed. However, it seems that
4519 on at least one OS (Solaris 9) it is possible to have an interface that is
4520 included in the list of interfaces, but for which you get a failure error
4521 for this call. This happens when the interface is not "plumbed" into a
4522 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4523 failure of the "get flags" call assumes that the interface is down.
4525 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4526 makes life easier for people who are doing arithmetic on fields extracted
4527 from dates, where you often get leading zeros that should not be
4528 interpreted as octal.
4530 64. Added qualify_domain to the redirect router, to override the global
4533 65. If a pathologically long header line contained very many addresses (the
4534 report of this problem mentioned 10 000) and each of them was rewritten,
4535 Exim could use up a very large amount of memory. (It kept on making new
4536 copies of the header line as it rewrote, and never released the old ones.)
4537 At the expense of a bit more processing, the header rewriting function has
4538 been changed so that it no longer eats memory in this way.
4540 66. The generation of the Received: header has been moved from the time that a
4541 message starts to be received, to the time that it finishes. The timestamp
4542 in the Received: header should now be very close to that of the <= log
4543 line. There are two side-effects of this change:
4545 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4546 logged header lines no longer include the local Received: line, because
4547 it has not yet been created. The same applies to a copy of the message
4548 that is returned to a non-SMTP sender when a message is rejected.
4550 (b) When a filter file is tested using -bf, no additional Received: header
4551 is added to the test message. After some thought, I decided that this
4554 This change does not affect the value of $received_for. It is still set
4555 after address rewriting, but before local_scan() is called.
4557 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4559 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4560 gave an unhelpful panic error message, and a defer error. I have managed to
4561 change this behaviour so that it now rejects any supplied certificate,
4562 which seems right, as the list of acceptable certificates is empty.
4564 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4565 gave an unhelpful defer error. I have not managed to make this reject any
4566 supplied certificates, but the error message it gives is "no certificate
4567 supplied", which is not helpful.
4569 70. exigrep's output now also includes lines that are not associated with any
4570 message, but which match the given pattern. Implemented by a patch from
4571 Martin Sluka, which also tidied up the Perl a bit.
4573 71. Recipient callout verification, like sender verification, was using <> in
4574 the MAIL FROM command. This isn't really the right thing, since the actual
4575 sender may affect whether the remote host accepts the recipient or not. I
4576 have changed it to use the actual sender in the callout; this means that
4577 the cache record is now keyed on a recipient/sender pair, not just the
4578 recipient address. There doesn't seem to be a real danger of callout loops,
4579 since a callout by the remote host to check the sender would use <>.
4580 [SEE ABOVE: changed after hitting problems.]
4582 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4583 temporary errors. However, in the case of such a code being given after
4584 the end of a data transmission (i.e. after ".") Exim was failing to write
4585 a retry record for the message. (Yes, there was some broken host that was
4586 actually sending 8xx at this point.)
4588 73. An unknown lookup type in a host list could cause Exim to panic-die when
4589 the list was checked. (An example that provoked this was putting <; in the
4590 middle of a list instead of at the start.) If this happened during a DATA
4591 ACL check, a -D file could be left lying around. This kind of configuration
4592 error no longer causes Exim to die; instead it causes a defer errror. The
4593 incident is still logged to the main and panic logs.
4595 74. Buglet left over from Exim 3 conversion. The message "too many messages
4596 in one connection" was written to the rejectlog but not the mainlog, except
4597 when address rewriting (yes!) was being logged.
4599 75. Added write_rejectlog option.
4601 76. When a system filter was run not as root (that is, when system_filter_user
4602 was set), the values of the $n variables were not being returned to the
4603 main process; thus, they were not subsequently available in the $sn
4606 77. Added +return_path_on_delivery log selector.
4608 78. A connection timeout was being treated differently from recipients deferred
4609 when testing hosts_max_try with a message that was older than the host's
4610 retry timeout. (The host should not be counted, thus allowing all hosts to
4611 be tried at least once before bouncing.) This may have been the cause of an
4612 occasionally reported bug whereby a message would remain on the queue
4613 longer than the retry timeout, but would be bounced if a delivery was
4614 forced. I say "may" because I never totally pinned down the problem;
4615 setting up timeout/retry tests is difficult. See also the next item.
4617 79. The ultimate address timeout was not being applied to errors that involved
4618 a combination of host plus message (for example, a timeout on a MAIL
4619 command). When an address resolved to a number of possible hosts, and they
4620 were not all tried for each delivery (e.g. because of hosts_max_try), a
4621 message could remain on the queue longer than the retry timeout.
4623 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4626 81. Fixed an obscure SMTP outgoing bug which required at least the following
4627 conditions: (a) there was another message waiting for the same server;
4628 (b) the server returned 5xx to all RCPT commands in the first message so
4629 that the message was not completed; (c) the server dropped the connection
4630 or gave a negative response to the RSET that Exim sends to abort the
4631 transaction. The observed case was a dropped connection after DATA that had
4632 been sent in pipelining mode. That is, the server had advertised PIPELINING
4633 but was not implementing it correctly. The effect of the bug was incorrect
4634 behaviour, such as trying another host, and this could lead to a crash.
4640 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4641 and daemon.c were passed as pointers to ints; they should have been
4642 pointers to socklen_t variables (which are typically unsigned ints).
4644 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4647 3. Fixed a really odd bug that affected only the testing scheme; patching a
4648 certain fixed string in the binary changed the value of another string that
4649 happened to be identical to the end of the original first string.
4651 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4652 name", it returns that address as the IP address. On some operating
4653 systems (e.g. Solaris), it also passes back the IP address string as the
4654 "host name". However, on others (e.g. Linux), it passes back an empty
4655 string. Exim wasn't checking for this, and was changing the host name to an
4656 empty string, assuming it had been canonicized.
4658 5. Although rare, it is permitted to have more than one PTR record for a given
4659 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4660 all the names associated with an address, because they do in Solaris.
4661 However, it seems that they do not in Linux for data that comes from the
4662 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4663 I found this out when I moved to a new Linux workstation and tried to run
4664 the Exim test suite.
4666 To get round this problem I have changed the code so that it now does its
4667 own call to the DNS to look up PTR records when searching for a host name.
4668 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4669 addresses that are only in /etc/hosts are still found.
4671 This behaviour is, however, controlled by an option called host_lookup_
4672 order, which defaults to "bydns:byaddr". If people want to use the other
4673 order, or indeed, just use one or the other means of lookup, they can
4674 specify it in this variable.
4676 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4677 some operating systems, this comes back from gethostbyaddr() as an empty
4678 string, and this is what Exim used to test for. However, it seems that in
4679 other systems, "." is yielded. Exim now tests for this case too.
4681 7. The values of check_spool_space and check_log_space are now held internally
4682 as a number of kilobytes instead of an absolute number of bytes. If a
4683 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4684 kilobyte. This means that much larger values can be stored.
4686 8. Exim monitor: an attempt to get the action menu when not actually pointing
4687 at a message produces an empty menu entitled "No message selected". This
4688 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4689 no entries in it ("Shell widget menu has zero width and/or height"). So I
4690 have added a single, blank menu entry in this case.
4692 9. Added ${quote_local_part.
4694 10. MIME decoding is now applied to the contents of Subject: header lines when
4697 11. Now that a reference to $sender_host_address automatically causes a reverse
4698 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4699 host lookup before query-style lookups in lists that might use this
4700 variable. This has therefore been abolished, and the "net-" prefix is no
4701 longer necessary for query-style lookups.
4703 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4704 have been a typo for LFLAGS, so it has been changed.
4706 13. The install script calls Exim with "-C /dev/null" in order to find the
4707 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4708 to be output. Howeve, since Exim outputs its version number before the
4709 error, it didn't break the script. It just looked ugly. I fixed this by
4710 always allowing "-C /dev/null" if the caller is root.
4712 14. Ignore overlarge ACL variable number when reading spool file - insurance
4713 against a later release with more variables having written the file.
4715 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4716 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4717 this, as well as the form without the "IPv6" on the front.
4719 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4720 OS/Makefile-Darwin file.
4722 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4723 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4725 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4726 variable. However, some OS use ints for inodes. I've added cast to long int
4727 to get rid of the compiler warning.
4729 19. I had forgotten to lock out "/../" in configuration file names when
4730 ALT_CONFIG_PREFIX was set.
4732 20. Routers used for verification do not need to specify transports. However,
4733 if such a router generated a host list, and callout was configured, Exim
4734 crashed, because it could not find a port number from the (non-existent)
4735 transport. It now assumes port 25 in this circumstance.
4737 21. Added the -t option to exigrep.
4739 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4740 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4741 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4742 them is set, LOOKUP_LSEARCH is forced.
4744 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4746 24. Added sender and host information to the "rejected by local_scan()" log
4747 line; previously there was no indication of these.
4749 25. Added .include_if_exists.
4751 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4752 Linux. It turns out that not all file systems support this. Apparently some
4753 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4754 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4755 not supported on the file descriptor, is now ignored when Exim is trying to
4756 sync a directory. This applies only to Linux.
4758 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4760 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4761 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4764 29. Added support for maildirsize files from supplied patch (modified a bit).
4766 30. The use of :fail: followed by an empty string could lead Exim to respond to
4767 sender verification failures with (e.g.):
4769 550 Verification failed for <xxx>
4770 550 Sender verify failed
4772 where the first response line was missing the '-' that indicates it is not
4773 the final line of the response.
4775 31. The loop for finding the name of the user that called Exim had a hardwired
4776 limit of 10; it now uses the value of finduser_retries, which is used for
4777 all other user lookups.
4779 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4782 33. Exim was neglecting to zero errno before one call of strtol() when
4783 expanding a string and expecting an integer value. On some systems this
4784 resulted in spurious "integer overflow" errors. Also, it was casting the
4785 result into an int without checking.
4787 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4788 did not work. The code looks as if it has *never* worked, though it appears
4789 to have been documented since at least releast 1.62. I have made it work.
4791 35. The "timeout_DNS" error in retry rules, also documented since at least
4792 1.62, also never worked. As it isn't clear exactly what this means, and
4793 clearly it isn't a major issue, I have abolished the feature by treating it
4794 as "timeout", and writing a warning to the main and panic logs.
4796 36. The display of retry rules for -brt wasn't always showing the error code
4799 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4800 timeout_connect_A, timeout_connect_MX.
4802 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4803 to the empty sender.
4805 39. The daemon was not analysing the content of -oX till after it had closed
4806 stderr and disconnected from the controlling terminal. This meant that any
4807 syntax errors were only noted on the panic log, and the return code from
4808 the command was 0. By re-arranging the code a little, I've made the
4809 decoding happen first, so such errors now appear on stderr, and the return
4810 code is 1. However, the actual setting up of the sockets still happens in
4811 the disconnected process, so errors there are still only recorded on the
4814 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4815 connections (as happens on some IP stacks) was logged at start up time as
4816 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4817 it from "IPv6 and IPv4", which means that two separate sockets are being
4820 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4821 says whether AF_INET or AF_INET6 was passed as an argument.
4823 42. Exiwhat output was messed up when time zones were included in log
4826 43. Exiwhat now gives more information about the daemon's listening ports,
4827 and whether -tls-on-connect was used.
4829 44. The "port" option of the smtp transport is now expanded.
4831 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4832 silently ignored. Now an error message is written to the main and panic
4835 46. There's a new ACL modifier called "logwrite" which writes to a log file
4836 as soon as it is encountered.
4838 47. Added $local_user_uid and $local_user_gid at routing time.
4840 48. Exim crashed when trying to verify a sender address that was being
4843 49. Exim was recognizing only a space character after ".include". It now also
4844 recognizes a tab character.
4846 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4847 extracting the relevant information from the specification. The man page no
4848 longer contains scrambled data for the -d option, and I've added a section
4849 at the front about calling Exim under different names.
4851 51. Added "extra_headers" argument to the "mail" command in filter files.
4853 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4856 53. Installed eximstats 1.29.
4858 54. Added transport_filter_timeout as a generic transport option.
4860 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4861 Cc: header lines. This was required by RFC 822, but it not required by RFC
4864 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4865 incoming messages that did not have them. Now it does so only if the
4866 message originates locally, that is, if there is no associated remote host
4867 address. When Resent- header lines are present, this applies to the Resent-
4868 lines rather than the non-Resent- lines.
4870 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4871 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4873 58. Messages for configuration errors now include the name of the main
4874 configuration file - useful now that there may be more than one file in a
4875 list (.included file names were always shown).
4877 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4878 for those rare installations that do not start the daemon as root or run it
4879 setuid root. I've cut out the call to initgroups() if the daemon is not
4882 60. The Exim user and group can now be bound into the binary as text strings
4883 that are looked up at the start of Exim's processing.
4885 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4887 62. Added $mailstore_basename variable.
4889 63. Installed patch to sieve.c from Michael Haardt.
4891 64. When Exim failed to open the panic log after failing to open the main log,
4892 the original message it was trying to log was written to stderr and debug
4893 output, but if they were not available (the usual case in production), it
4894 was lost. Now it is written to syslog before the two lines that record the
4895 failures to open the logs.
4897 65. Users' Exim filters run in subprocesses under the user's uid. It is
4898 possible for a "deliver" command or an alias in a "personal" command to
4899 provoke an address rewrite. If logging of address rewriting is configured,
4900 this fails because the process is not running as root or exim. There may be
4901 a better way of dealing with this, but for the moment (because 4.30 needs
4902 to be released), I have disabled address rewrite logging when running a
4903 filter in a non-root, non-exim process.
4909 1. The buildconfig auxiliary program wasn't quoting the value set for
4910 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4911 not defined. This bug was masked in 4.22 by the effect that was fixed in
4914 2. Some messages that were rejected after a message id was allocated were
4915 shown as "incomplete" by exigrep. It no longer does this for messages that
4916 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4918 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4919 have allow_domain_literals set, the ID did not get logged in the <= line.
4920 Domain literals are now always recognized in Message-ID: header lines.
4922 4. The first argument for a ${extract expansion item is the key name or field
4923 number. Leading and trailing spaces in this item were not being ignored,
4924 causing some misleading effects.
4926 5. When deliver_drop_privilege was set, single queue runner processes started
4927 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4928 same command in the process it spins off) were not dropping privilege.
4930 6. When the daemon running as "exim" started a queue runner, it always
4931 re-executed Exim in the spun-off process. This is a waste of effort when
4932 deliver_drop_privilege is set. The new process now just calls the
4933 queue-runner function directly.
4939 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4942 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4943 ignored. Though the use of -oP was forcing the writing of a pid file, it
4944 was always written to the default place.
4946 3. If the message "no IP address found for host xxxx" is generated during
4947 incoming verification, it is now followed by identification of the incoming
4948 connection (so you can more easily find what provoked it).
4950 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4952 5. Added some features to "harden" Exim a bit more against certain attacks:
4954 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4955 be put in Local/Makefile. This is like the never_users runtime option,
4956 but it cannot be overridden. The default setting is "root".
4958 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4959 prefix string with which any file named in a -C command line option
4962 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4963 is retained for -C and -D only if the caller of Exim is root. Without
4964 it, the exim user may also use -C and -D and retain privilege.
4966 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4967 command line option is disabled.
4969 6. Macro names set by the -D option must start with an upper case letter, just
4970 like macro names defined in the configuration file.
4972 7. Added "dereference=" facility to LDAP.
4974 8. Two instances of the typo "uknown" in the source files are fixed.
4976 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4977 the Configure-Makefile script screwed up while processing it.
4979 10. Incorporated PCRE 4.4.
4981 11. The SMTP synchronization check was not operating right at the start of an
4982 SMTP session. For example, it could not catch a HELO sent before the client
4983 waited for the greeting. There is now a check for outstanding input at the
4984 point when the greeting is written. Because of the duplex, asynchronous
4985 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4986 way, but not yet received, when the check is performed.
4988 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4989 on TCP/IP sockets, because this apparently causes some broken clients to
4992 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4993 unchanged) from the Cygwin maintainer.
4995 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4996 was supported instead of testing for mbx. Effectively a typo.
4998 15. The spa authenticator server code was not checking that the input it
4999 received was valid base64.
5001 16. The debug output line for the "set" modifier in ACLs was not showing the
5002 name of the variable that was being set.
5004 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5006 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5007 lookup on its own. Something else was needed to trigger the lookup. For
5008 example, a match in host_lookup or the need for a host name in a host list.
5009 Now, if $sender_host_name is referenced and the host name has not yet been
5010 looked up, a lookup is performed. If the lookup fails, the variable remains
5011 empty, and $host_lookup_failed is set to "1".
5013 19. Added "eqi" as a case-independent comparison operator.
5015 20. The saslauthd authentication condition could segfault if neither service
5016 nor realm was specified.
5018 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5019 error message that was logged was misleading, and incoming SMTP
5020 connections were dropped. The message is now more accurate, and temporary
5021 errors are given to SMTP connections.
5023 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5024 (except RSET) with a 421 error, until QUIT is received. However, it was
5025 failing to send a response to QUIT.
5027 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5028 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5029 ACL is now run after the helo_try_verify_hosts code.
5031 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5032 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5033 case-independent, so other case variants are also recognized). Apparently
5034 some systems use these upper case variants.
5036 25. If more than two messages were waiting for the same host, and a transport
5037 filter was specified for the transport, Exim sent two messages over the
5038 same TCP/IP connection, and then failed with "socket operation on non-
5039 socket" when it tried to send the third.
5041 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5043 27. The extern definition of crypt16() in expand.c was not being excluded when
5044 the OS had its own crypt16() function.
5046 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5047 as a preferred synonym for return_size_limit, both as an option and as an
5050 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5052 30. Changed the default configuration ACL to relax the local part checking rule
5053 for addresses that are not in any local domains. For these addresses,
5054 slashes and pipe symbols are allowed within local parts, but the sequence
5055 /../ is explicitly forbidden.
5057 31. SPA server authentication was not clearing the challenge buffer before
5060 32. log_message in a "warn" ACL statement was writing to the reject log as
5061 well as to the main log, which contradicts the documentation and doesn't
5062 seem right (because no rejection is happening). So I have stopped it.
5064 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5065 However, I am unable to do any testing of this.
5067 34. Fixed an infelicity in the appendfile transport. When checking directories
5068 for a mailbox, to see if any needed to be created, it was accidentally
5069 using path names with one or more superfluous leading slashes; tracing
5070 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5072 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5073 contents are added to the log line that is written for every discarded
5074 recipient. (Previously a log_message setting was ignored.)
5076 36. The ${quote: operator now quotes the string if it is empty.
5078 37. The install script runs exim in order to find its version number. If for
5079 some reason other than non-existence or emptiness, which it checks, it
5080 could not run './exim', it was installing it with an empty version number,
5081 i.e. as "exim-". This error state is now caught, and the installation is
5084 38. An argument was missing from the function that creates an error message
5085 when Exim fails to connect to the socket for saslauthd authentication.
5086 This could cause Exim to crash, or give a corrupted message.
5088 39. Added isip, isip4, and isip6 to ${if conditions.
5090 40. The ACL variables $acl_xx are now saved with the message, and can be
5091 accessed later in routers, transports, and filters.
5093 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5094 strings in the file are not string-expanded.
5096 42. If a MAIL command specified a SIZE value that was too large to fit into an
5097 int variable, the check against message_size_limit failed. Such values are
5098 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5099 day this will have to be increased, but I don't think I want to be around
5100 when emails are that large.
5107 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5108 iconv() is not standard in FreeBSD.
5110 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5111 IPv6 enabled. The observed symptom was a segmentation fault on return from
5112 the function os_common_find_running_interfaces() in src/os.c.
5114 3. In the check_special_case() function in daemon.c I had used "errno" as an
5115 argument name, which causes warnings on some systems. This was basically a
5116 typo, since it was named "eno" in the comments!
5118 4. The code that waits for the clock to tick (at a resolution of some fraction
5119 of a second) so as to ensure message-id uniqueness was always waiting for
5120 at least one whole tick, when it could have waited for less. [This is
5121 almost certainly not relevant at current processor speeds, where it is
5122 unlikely to ever wait at all. But we try to future-proof.]
5124 5. The function that sleeps for a time interval that includes fractions of a
5125 second contained a race. It did not block SIGALRM between setting the
5126 timer, and suspending (a couple of lines later). If the interval was short
5127 and the sigsuspend() was delayed until after it had expired, the suspension
5128 never ended. On busy systems this could lead to processes getting stuck for
5131 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5132 process, before it forks any delivery processes. The open lookup caching
5133 mechanism meant that the open file or database connection was passed into
5134 the delivery process. The problem was that delivery processes always tidy
5135 up cached lookup data. This could cause a problem for the next delivery
5136 process started by the queue runner, because the external queue runner
5137 process does not know about the closure. So the next delivery process
5138 still has data in the lookup cache. In the case of a file lookup, there was
5139 no problem because closing a file descriptor in a subprocess doesn't affect
5140 the parent. However, if the lookup was caching a connection to a database,
5141 the connection was closed, and the second delivery process was likely to
5142 see errors such as "PGSQL: query failed: server closed the connection
5143 unexpectedly". The problem has been fixed by closing all cached lookups
5144 in a queue runner before running a delivery process.
5146 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5147 seem to have the "const" qualifier which it has on other OS. I've
5150 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5151 *of the same type* (client or server) with the same public name that an
5152 error should be diagnosed.
5154 9. When Exim looked up a host name for an IP address, but failed to find the
5155 original IP address when looking up the host name (a safety check), it
5156 output the message "<ip address> does not match any IP for NULL", which was
5157 confusing, to say the least. The bug was that the host name should have
5158 appeared instead of "NULL".
5160 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5161 user that is built into the binary, and the -C or -D options is used, root
5162 privilege is dropped before the configuration file is read. In addition,
5163 logging is switched to stderr instead of the normal log files. If the
5164 configuration then re-defines the Exim user, the unprivileged environment
5165 is probably not what is expected, so Exim logs a panic warning message (but
5168 However, if deliver_drop_privilege is set, the unprivileged state may well
5169 be exactly what is intended, so the warning has been cut out in that case,
5170 and Exim is allowed to try to write to its normal log files.
5176 1. smtp_return_error_details was not giving details for temporary sender
5177 or receiver verification errors.
5179 2. Diagnose a configuration error if two authenticators have the same public
5182 3. Exim used not to create the message log file for a message until the first
5183 delivery attempt. This could be confusing when incoming messages were held
5184 for policy or load reasons. The message log file is now created at the time
5185 the message is received, and an initial "Received" line is written to it.
5187 4. The automatically generated man page for command line options had a minor
5188 bug that caused no ill effects; however, a more serious problem was that
5189 the procedure for building the man page automatically didn't always
5190 operate. Consequently, release 4.20 contains an out-of-date version. This
5191 shouldn't happen again.
5193 5. When building Exim with embedded Perl support, the script that builds the
5194 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5195 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5197 6. The freeze_tell option was not being used for messages that were frozen on
5198 arrival, either by an ACL or by local_scan().
5200 7. Added the smtp_incomplete_transaction log selector.
5202 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5203 was accepting AUTH without a new EHLO.
5205 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5206 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5209 10. Logging of TCP/IP connections (when configured) now happens in the main
5210 daemon process instead of the child process, so that the TCP/IP connection
5211 count is more accurate (but it can never be perfect).
5213 11. The use of "drop" in a nested ACL was not being handled correctly in the
5214 outer ACL. Now, if condition failure induced by the nested "drop" causes
5215 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5216 or "require"), the connection is dropped.
5218 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5219 that yield "discard" can now be used with an "accept" or a "discard" verb,
5220 but an error is generated for any others (because I can't see a useful way
5221 to define what should happen).
5223 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5224 are now processed in the same way as lines in the Exim configuration file.
5225 In particular, continuation lines are supported.
5227 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5229 15. Added -ti meaning -t -i.
5231 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5232 domains, and warn by logging if others are found.
5234 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5235 returned from GIFCONF when Exim is trying to find the list of interfaces on
5236 a host. The code in os.c has been modified to copy each ifreq to an aligned
5237 structure in all cases.
5239 Also, in some cases, the returned ifreq's were being copied to a 'struct
5240 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5241 means the last couple of bytes of an IPv6 address could be chopped if the
5242 ifreq contained only a normal sockaddr (14 bytes storage).
5244 18. Named domain lists were not supported in the hosts_treat_as_local option.
5245 An entry such as +xxxx was not recognized, and was treated as a literal
5248 19. Ensure that header lines added by a DATA ACL are included in the reject log
5249 if the ACL subsequently rejects the message.
5251 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5252 MD5 (which is deprecated).
5254 21. When testing a filter file using -bf, Exim was writing a message when it
5255 took the sender from a "From " line in the message, but it was not doing so
5256 when it took $return_path from a Return-Path: header line. It now does.
5258 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5259 with a valid header line field name (a series of printing characters
5260 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5262 23. Changed "disc" in the source to "disk" to conform to the documentation and
5263 the book and for uniformity.
5265 24. Ignore Sendmail's -Ooption=value command line item.
5267 25. When execve() failed while trying to run a command in a pipe transport,
5268 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5269 could be confused with a return value of 69 from the command itself. This
5270 has been changed to 127, the value the shell returns if it is asked to run
5271 a non-existent command. The wording for the related log line suggests a
5272 non-existent command as the problem.
5274 26. If received_header_text expands to an empty string, do not add a Received:
5275 header line to the message. (Well, it adds a token one on the spool, but
5276 marks it "old" so that it doesn't get used or transmitted.)
5278 27. Installed eximstats 1.28 (addition of -nt option).
5280 28. There was no check for failure on the call to getsockname() in the daemon
5281 code. This can fail if there is a shortage of resources on the system, with
5282 ENOMEM, for example. A temporary error is now given on failure.
5284 29. Contrary to the C standard, it seems that in some environments, the
5285 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5286 program. Exim now does this explicitly; it affects the formatting of
5287 timestamps using strftime().
5289 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5290 complaints. I've now initialized all the variables, to avoid this.
5292 32. Header lines added by a system filter were not being "seen" during
5293 transport-time rewrites.
5295 33. The info_callback() function passed to OpenSSL is set up with type void
5296 (*)(SSL *, int, int), as described somewhere. However, when calling the
5297 function (actually a macro) that sets it up, the type void(*)() is
5298 expected. I've put in a cast to prevent warnings from picky compilers.
5300 34. If a DNS black list lookup found a CNAME record, but there were no A
5301 records associated with the domain it pointed at, Exim crashed.
5303 35. If a DNS black list lookup returned more than one A record, Exim ignored
5304 all but the first. It now scans all returned addresses if a particular IP
5305 value is being sought. In this situation, the contents of the
5306 $dnslist_value variable are a list of all the addresses, separated by a
5309 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5310 to accept a host name and all its aliases if the forward lookup for any of
5311 them yielded the IP address of the incoming connection. Now it accepts only
5312 those names whose forward lookup yields the correct IP address. Any other
5313 names are discarded. This closes a loophole whereby a rogue DNS
5314 administrator could create reverse DNS records to break through a
5315 wildcarded host restriction in an ACL.
5317 37. If a user filter or a system filter that ran in a subprocess used any of
5318 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5319 the wrong values were passed to the pipe command ($thisaddress had the
5320 value of $0, $0 had the value of $1, etc). This bug was introduced by
5321 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5323 38. Improved the line breaking for long SMTP error messages from ACLs.
5324 Previously, if there was no break point between 40 and 75 characters, Exim
5325 left the rest of the message alone. Two changes have been made: (a) I've
5326 reduced the minimum length to 35 characters; (b) if it can't find a break
5327 point between 35 and 75 characters, it looks ahead and uses the first one
5328 that it finds. This may give the occasional overlong line, but at least the
5329 remaining text gets split now.
5331 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5332 file descriptors might be low, and that setting 1000 would always raise it.
5333 It turns out that in some environments, the limit is already over 1000 and
5334 that lowering it causes trouble. So now Exim takes care not to decrease it.
5336 40. When delivering a message, the value of $return_path is set to $sender_
5337 address at the start of routing (routers may change the value). By an
5338 oversight, this default was not being set up when an address was tested by
5339 -bt or -bv, which affected the outcome if any router or filter referred to
5342 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5343 log only when "message" or "log_message" are set. However, if one of the
5344 conditions was an address verification, or a call to a nested ACL, the
5345 messages generated by the underlying test were being passed through. This
5346 no longer happens. The underlying message is available in $acl_verify_
5347 message for both "message" and "log_message" expansions, so it can be
5348 passed through if needed.
5350 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5351 new expansion $bh_ to give the encoded byte string without charset
5352 translation. Translation happens only if iconv() is available; HAVE_ICONV
5353 indicates this at build time. HEADERS_CHARSET gives the charset to
5354 translate to; headers_charset can change it in the configuration, and
5355 "headers charset" can change it in an individual filter file.
5357 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5358 that creates RFC 2047 encoded "words" labels them as that charset instead
5359 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5360 expansion operator; (ii) when Exim creates a From: line for a local
5361 message; (iii) when a header line is rewritten to include a "phrase" part.
5363 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5364 buggy, causing it to skip the first lines of messages whose message ID
5365 ended in 'D'. This would not have bitten before Exim release 4.14, because
5366 message IDs were unlikely to end in 'D' before then. The effect was to have
5367 incorrect size information for certain domains.
5369 45. #include "config.h" was missing at the start of the crypt16.c module. This
5370 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5373 46. If there was a timeout during a "random" callout check, Exim treated it as
5374 a failure of the random address, and carried on sending RSET and the real
5375 address. If the delay was just some slowness somewhere, the response to the
5376 original RCPT would be taken as a response to RSET and so on, causing
5377 mayhem of various kinds.
5379 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5380 when I implemented it. It didn't allow for the fact that some option values
5381 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5382 the right test for positive values.
5384 48. Domain names in DNS records are case-independent. Exim always looks them up
5385 in lower case. Some resolvers return domain names in exactly the case they
5386 appear in the zone file, that is, they may contain uppercase letters. Not
5387 all resolvers do this - some return always lower case. Exim was treating a
5388 change of case by a resolver as a change of domain, similar to a widening
5389 of a domain abbreviation. This triggered its re-routing code and so it was
5390 trying to route what was effectively the same domain again. This normally
5391 caused routing to fail (because the router wouldn't handle the domain
5392 twice). Now Exim checks for this case specially, and just changes the
5393 casing of the domain that it ultimately uses when it transmits the message
5396 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5399 50. If a filter generated a file delivery with a non-absolute name (possible if
5400 no home directory exists for the router), the forbid_file option was not
5403 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5404 the existing equality matching.
5406 52. Exim was using ints instead of ino_t variables in some places where it was
5407 dealing with inode numbers.
5409 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5410 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5411 called TMPDIR, and if it finds it is different, it changes its value.
5413 54. The smtp_printf() function is now made available to local_scan() so
5414 additional output lines can be written before returning. There is also an
5415 smtp_fflush() function to enable the detection of a dropped connection.
5416 The variables smtp_input and smtp_batched_input are exported to
5419 55. Changed the default runtime configuration: the message "Unknown user"
5420 has been removed from the ACL, and instead placed on the localuser router,
5421 using the cannot_route_message feature. This means that any verification
5422 failures that generate their own messages won't get overridden. Similarly,
5423 the "Unrouteable address" message that was in the ACL for unverifiable
5424 relay addresses has also been removed.
5426 56. Added hosts_avoid_esmtp to the smtp transport.
5428 57. The exicyclog script was not checking for the esoteric option
5429 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5430 will work only if exicyclog is run under the appropriate euid.
5432 58. Following a discussion on the list, the rules by which Exim recognises line
5433 endings on incoming messages have been changed. The -dropcr and drop_cr
5434 options are now no-ops, retained only for backwards compatibility. The
5435 following line terminators are recognized: LF CRLF CR. However, special
5436 processing applies to CR:
5438 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5439 nor a local message in the state where . is a terminator.
5441 (ii) If a bare CR is encountered in a header line, an extra space is added
5442 after the line terminator so as not to end the header. The reasoning
5443 behind this is that bare CRs in header lines are most likely either
5444 to be mistakes, or people trying to play silly games.
5446 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5447 was being incorrectly given as 18 bytes larger than it should have been.
5448 This is a VOB (very old bug).
5450 60. This may never have affected anything current, but just in case it has:
5451 When the local host is found other than at the start of a list of hosts,
5452 the local host, those with the same MX, and any that follow, are discarded.
5453 When the list in question was part of a longer list of hosts, the following
5454 hosts (not currently being processed) were also being discarded. This no
5455 longer happens. I'm not sure if this situation could ever has previously
5458 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5461 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5462 header. This is something that is recommended in a new Internet Draft, and
5463 is something that is documented as being done by Sendmail. There are two
5464 possible values. For messages generated by the autoreply transport, Exim
5467 Auto-Submitted: auto-replied
5469 whereas for all other generated messages (e.g. bounces) it adds
5471 Auto-Submitted: auto-generated
5473 63. The "personal" condition in filters now includes a test for the
5474 Auto-Submitted: header. If it contains the string "auto-" the message it
5475 not considered personal.
5477 64. Added rcpt_include_affixes as a generic transport option.
5479 65. Added queue_only_override (default true).
5481 66. Added the syslog_duplication option.
5483 67. If what should have been the first header line of a message consisted of
5484 a space followed by a colon, Exim was mis-interpreting it as a header line.
5485 It isn't of course - it is syntactically invalid and should therefore be
5486 treated as the start of the message body. The misbehaviour could have
5487 caused a number of strange effects, including loss of data in subsequent
5488 header lines, and spool format errors.
5490 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5491 client host had authenticated. This control can now be exercised by an ACL
5492 for more flexibility.
5494 69. By default, callouts do not happen when testing with -bh. There is now a
5495 variant, -bhc, which does actually run the callout code, including
5496 consulting and updating the callout cache.
5498 70. Added support for saslauthd authentication, courtesy of Alexander
5501 71. If statvfs() failed on the spool or log directories while checking their
5502 size for availability, Exim confusingly gave the error "space shortage".
5503 Furthermore, in debugging mode it crashed with a floating point exception.
5504 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5505 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5506 serious problem, Exim now writes to the main and panic logs when this
5507 happens, with details of the failure. It then refuses to accept the
5508 incoming message, giving the message "spool directory problem" or "log
5509 directory problem" with a 421 code for SMTP messages.
5511 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5512 0, 1, and 2 exist, because some OS complain for execs without them (see
5513 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5514 descriptors. However, the code omitted to check that the open succeeded,
5515 causing mysterious errors if for some reason the permissions on /dev/null
5516 got screwed. Now Exim writes a message to the main and panic logs, and
5517 bombs out if it can't open /dev/null.
5519 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5520 interact so that it is all more flexible. It is supposed to remain
5521 backwards compatible. Also added extra_local_interfaces.
5523 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5524 to bomb out with an assertion failure - to the client this appears as a
5525 connection drop. This problem occurs in the part of the code that was taken
5526 from the Samba project. Fortunately, the assertion is in a very simple
5527 function, so I have fixed this by reproducing the function inline in the
5528 one place where it is called, and arranging for authentication to fail
5529 instead of killing the process with assert().
5531 75. The SPA client code was not working when the server requested OEM rather
5532 than Unicode encoding.
5534 76. Added code to make require_files with a specific uid setting more usable in
5535 the case where statting the file as root fails - usually a non-root-mounted
5536 NFS file system. When this happens and the failure is EACCES, Exim now
5537 forks a subprocess and does the per-uid checking as the relevant uid.
5539 77. Added process_log_path.
5541 78. If log_file_path was not explicitly set, a setting of check_log_space or
5542 check_log_inodes was ignored.
5544 79. If a space check for the spool or log partitions fails, the incident is now
5545 logged. Of course, in the latter case the data may get lost...
5547 80. Added the %p formatting code to string_format() so that it can be used to
5548 print addresses in debug_print(). Adjusted all the address printing in the
5549 debugging in store.c to use %p rather than %d.
5551 81. There was a concern that a line of code in smtp_in.c could overflow a
5552 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5553 initially expressed, the concern was not well-founded, because trailing
5554 spaces are removed early. However, if the trailing spaces were followed by
5555 a NULL, they did not get removed, so the overflow was possible. Two fixes
5558 (a) I re-wrote the offending code in a cleaner fashion.
5559 (b) If an incoming SMTP command contains a NULL character, it is rejected
5562 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5563 runs initgroups(), so that if the Exim user is in any additional groups,
5564 they will be used during message reception.
5570 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob