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.
69 PP/01 Solaris build fix for Oracle's LDAP libraries.
70 Bugzilla 1109, patch from Stephen Usher.
72 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
74 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
77 TF/02 Fix a couple more cases where we did not log the error message
78 when unlink() failed. See also change 4.74-TF/03.
80 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
81 lock up or crash if it happened to be inside a call to libc when it
82 got a SIGUSR1 from exiwhat.
84 The SIGUSR1 handler appends the current process status to the process
85 log which is later printed by exiwhat. It used to use the general
86 purpose logging code to do this, but several functions it calls are
89 The new output code in the SIGUSR1 handler is specific to the process
90 log, and simple enough that it's easy to inspect for signal safety.
91 Removing some special cases also simplifies the general logging code.
92 Removing the spurious timestamps from the process log simplifies
95 TF/04 Improved ratelimit ACL condition.
97 The /noupdate option has been deprecated in favour of /readonly which
98 has clearer semantics. The /leaky, /strict, and /readonly update modes
99 are mutually exclusive. The update mode is no longer included in the
100 database key; it just determines when the database is updated. (This
101 means that when you upgrde Exim will forget old rate measurements.)
103 Exim now checks that the per_* options are used with an update mode that
104 makes sense for the current ACL. For example, when Exim is processing a
105 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
106 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
107 must specify per_mail/readonly. If you omit the update mode it defaults to
108 /leaky where that makes sense (as before) or /readonly where required.
110 The /noupdate option is now undocumented but still supported for
111 backwards compatibility. It is equivalent to /readonly except that in
112 ACLs where /readonly is required you may specify /leaky/noupdate or
113 /strict/noupdate which are treated the same as /readonly.
115 A useful new feature is the /count= option. This is a generalization
116 of the per_byte option, so that you can measure the throughput of other
117 aggregate values. For example, the per_byte option is now equivalent
118 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
120 The per_rcpt option has been generalized using the /count= mechanism
121 (though it's more complicated than the per_byte equivalence). When it is
122 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
123 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
124 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
125 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
126 Note that using per_rcpt with a non-readonly update mode in more than
127 one ACL will cause the recipients to be double-counted. (The per_mail
128 and per_byte options don't have this problem.)
130 The handling of very low rates has changed slightly. If the computed rate
131 is less than the event's count (usually one) then this event is the first
132 after a long gap. In this case the rate is set to the same as this event's
133 count, so that the first message of a spam run is counted properly.
135 The major new feature is a mechanism for counting the rate of unique
136 events. The new per_addr option counts the number of different
137 recipients that someone has sent messages to in the last time period. It
138 behaves like per_rcpt if all the recipient addresses are different, but
139 duplicate recipient addresses do not increase the measured rate. Like
140 the /count= option this is a general mechanism, so the per_addr option
141 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
142 example, measure the rate that a client uses different sender addresses
143 with the options per_mail/unique=$sender_address. There are further
144 details in the main documentation.
146 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
148 TF/06 Removed a few PCRE remnants.
150 TF/07 Automatically extract Exim's version number from tags in the git
151 repository when doing development or release builds.
153 PP/02 Raise smtp_cmd_buffer_size to 16kB.
154 Bugzilla 879. Patch from Paul Fisher.
156 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
157 Heavily based on revision 40f9a89a from Simon Arlott's tree.
160 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
162 PP/05 Variable $av_failed, true if the AV scanner deferred.
163 Bugzilla 1078. Patch from John Horne.
165 PP/06 Stop make process more reliably on build failure.
166 Bugzilla 1087. Patch from Heiko Schlittermann.
168 PP/07 Make maildir_use_size_file an _expandable_ boolean.
169 Bugzilla 1089. Patch from Heiko Schlittermann.
171 PP/08 Handle ${run} returning more data than OS pipe buffer size.
172 Bugzilla 1131. Patch from Holger Weiß.
174 PP/09 Handle IPv6 addresses with SPF.
175 Bugzilla 860. Patch from Wolfgang Breyha.
177 PP/10 GnuTLS: support TLS 1.2 & 1.1.
179 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
182 PP/11 match_* no longer expand right-hand-side by default.
183 New compile-time build option, EXPAND_LISTMATCH_RHS.
184 New expansion conditions, "inlist", "inlisti".
186 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
188 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
190 PP/14 fix log_write() format string regression from TF/03.
191 Bugzilla 1152. Patch from Dmitry Isaikin.
197 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
199 PP/02 Harmonised TLS library version reporting; only show if debugging.
200 Layout now matches that introduced for other libraries in 4.74 PP/03.
202 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
204 PP/04 New "dns_use_edns0" global option.
206 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
209 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
210 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
212 TK/01 Updated PolarSSL code to 0.14.2.
213 Bugzilla 1097. Patch from Andreas Metzler.
215 PP/07 Catch divide-by-zero in ${eval:...}.
218 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
221 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
222 format-string attack -- SECURITY: remote arbitrary code execution.
224 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
225 time unintentionally subject to list matching rules, letting the header
226 cause arbitrary Exim lookups (of items which can occur in lists, *not*
227 arbitrary string expansion). This allowed for information disclosure.
229 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
230 INT_MIN/-1 -- value coerced to INT_MAX.
236 NM/01 Workround for PCRE version dependency in version reporting
239 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
240 This fixes portability to compilers other than gcc, notably
241 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
243 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
244 makefiles for portability to HP-UX and POSIX correctness.
246 PP/01 Permit LOOKUP_foo enabling on the make command-line.
247 Also via indented variable definition in the Makefile.
248 (Debugging by Oliver Heesakkers).
250 PP/02 Restore caching of spamd results with expanded spamd_address.
251 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
253 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
254 Improves build reliability. Fix from: Frank Elsner
256 NM/02 Fix wide character breakage in the rfc2047 coding
257 Fixes bug 1064. Patch from Andrey N. Oktyabrski
259 NM/03 Allow underscore in dnslist lookups
260 Fixes bug 1026. Patch from Graeme Fowler
262 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
263 Code patches from Adam Ciarcinski of NetBSD.
265 NM/04 Fixed exiqgrep to cope with mailq missing size issue
268 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
269 is logged, to avoid truncation. Patch from John Horne.
271 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
272 Patch from Jakob Hirsch.
274 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
275 SQL string expansion failure details.
276 Patch from Andrey Oktyabrski.
278 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
279 Patch from Simon Arlott.
281 PP/09 New lookups functionality failed to compile on old gcc which rejects
282 extern declarations in function scope.
283 Patch from Oliver Fleischmann
285 PP/10 Use sig_atomic_t for flags set from signal handlers.
286 Check getgroups() return and improve debugging.
287 Fixed developed for diagnosis in bug 927 (which turned out to be
290 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
291 Patch from Mark Zealey.
293 PP/12 Bugzilla 1056: Improved spamd server selection.
294 Patch from Mark Zealey.
296 PP/13 Bugzilla 1086: Deal with maildir quota file races.
297 Based on patch from Heiko Schlittermann.
299 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
300 Patch from Uwe Doering, sign-off by Michael Haardt.
302 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
303 variable declaration deep within a block. Bug and patch from
306 PP/15 lookups-Makefile IRIX compatibilty coercion.
308 PP/16 Make DISABLE_DKIM build knob functional.
310 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
311 Patch by Simon Arlott
313 TF/03 Fix valgrind.h portability to C89 compilers that do not support
314 variable argument macros. Our copy now differs from upstream.
320 TF/01 Failure to get a lock on a hints database can have serious
321 consequences so log it to the panic log.
323 TF/02 Log LMTP confirmation messages in the same way as SMTP,
324 controlled using the smtp_confirmation log selector.
326 TF/03 Include the error message when we fail to unlink a spool file.
328 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
329 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
330 for maintaining out-of-tree patches for some time.
332 PP/01 Bugzilla 139: Documentation and portability issues.
333 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
334 Handle per-OS dynamic-module compilation flags.
336 PP/02 Let /dev/null have normal permissions.
337 The 4.73 fixes were a little too stringent and complained about the
338 permissions on /dev/null. Exempt it from some checks.
339 Reported by Andreas M. Kirchwitz.
341 PP/03 Report version information for many libraries, including
342 Exim version information for dynamically loaded libraries. Created
343 version.h, now support a version extension string for distributors
344 who patch heavily. Dynamic module ABI change.
346 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
347 privilege escalation vulnerability whereby the Exim run-time user
348 can cause root to append content of the attacker's choosing to
351 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
354 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
355 If dropping privileges for untrusted macros, we disabled normal logging
356 on the basis that it would fail; for the Exim run-time user, this is not
357 the case, and it resulted in successful deliveries going unlogged.
358 Fixed. Reported by Andreas Metzler.
364 PP/01 Date: & Message-Id: revert to normally being appended to a message,
365 only prepend for the Resent-* case. Fixes regression introduced in
366 Exim 4.70 by NM/22 for Bugzilla 607.
368 PP/02 Include check_rfc2047_length in configure.default because we're seeing
369 increasing numbers of administrators be bitten by this.
371 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
373 PP/03 Bugzilla 994: added openssl_options main configuration option.
375 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
377 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
379 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
381 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
382 without a peer certificate, leading to a segfault because of an
383 assumption that peers always have certificates. Be a little more
384 paranoid. Problem reported by Martin Tscholak.
386 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
387 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
388 NB: ClamAV planning to remove STREAM in "middle of 2010".
389 CL also introduces -bmalware, various -d+acl logging additions and
390 more caution in buffer sizes.
392 PP/09 Implemented reverse_ip expansion operator.
394 PP/10 Bugzilla 937: provide a "debug" ACL control.
396 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
398 PP/12 Bugzilla 973: Implement --version.
400 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
402 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
404 PP/15 Bugzilla 816: support multiple condition rules on Routers.
406 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
407 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
408 ignore trailing whitespace.
410 JJ/02 prevent non-panic DKIM error from being sent to paniclog
412 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
415 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
416 Notification from Dr Andrew Aitchison.
418 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
419 ExtendedDetectionInfo response format.
420 Notification from John Horne.
422 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
425 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
426 XSL and documented dependency on system catalogs, with examples of how
429 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
432 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
433 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
434 configuration file which is writeable by the Exim user or group.
436 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
437 of configuration files to cover files specified with the -C option if
438 they are going to be used with root privileges, not just the default
441 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
442 option (effectively making it always true).
444 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
445 files to be used while preserving root privileges.
447 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
448 that rogue child processes cannot use them.
450 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
451 run-time user, instead of root.
453 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
454 Exim run-time user without dropping privileges.
456 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
457 result string, instead of calling string_vformat() twice with the same
460 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
461 for other users. Others should always drop root privileges if they use
462 -C on the command line, even for a whitelisted configure file.
464 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
466 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
472 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
473 $data_path, and $header_path variables; fixed documentation bugs and
476 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
477 exipick to access non-standard spools, including the "frozen" queue
480 NM/01 Bugzilla 965: Support mysql stored procedures.
481 Patch from Alain Williams
483 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
485 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
486 Patch from Andreas Metzler
488 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
489 Patch from Kirill Miazine
491 NM/05 Bugzilla 671: Added umask to procmail example.
493 JJ/03 installed exipick 20100323.0, fixing doc bug
495 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
496 directory. Notification and patch from Dan Rosenberg.
498 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
500 TK/02 Improve log output when DKIM signing operation fails.
502 MH/01 Treat the transport option dkim_domain as a colon separated
503 list, not as a single string, and sign the message with each element,
504 omitting multiple occurences of the same signer.
506 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
507 Bugzilla 985, 986. Patch by Simon Arlott
509 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
510 Patch by Simon Arlott
512 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
513 MBX locking. Notification from Dan Rosenberg.
519 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
521 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
523 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
525 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
527 NM/04 Bugzilla 915: Flag broken perl installation during build.
533 TK/01 Added patch by Johannes Berg that expands the main option
534 "spamd_address" if it starts with a dollar sign.
536 TK/02 Write list of recipients to X-Envelope-Sender header when building
537 the mbox-format spool file for content scanning (suggested by Jakob
540 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
541 (http://www.dcc-servers.net/) support via dccifd. Activated by
542 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
544 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
545 by Mark Daniel Reidel <mr@df.eu>.
547 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
548 When building exim an external PCRE library is now needed -
549 PCRE is a system library on the majority of modern systems.
550 See entry on PCRE_LIBS in EDITME file.
552 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
553 conversation. Added nologin parameter to request.
554 Patch contributed by Kirill Miazine.
556 TF/01 Do not log submission mode rewrites if they do not change the address.
558 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
560 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
561 log files in place. Contributed by Roberto Lima.
563 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
565 TF/03 Bugzilla 615: When checking the local_parts router precondition
566 after a local_part_suffix or local_part_prefix option, Exim now
567 does not use the address's named list lookup cache, since this
568 contains cached lookups for the whole local part.
570 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
571 Robert Millan. Documentation is in experimental-spec.txt.
573 TF/04 Bugzilla 668: Fix parallel build (make -j).
575 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
577 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
578 Patch provided by Jan Srzednicki.
580 TF/05 Leading white space used to be stripped from $spam_report which
581 wrecked the formatting. Now it is preserved.
583 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
584 that they are available at delivery time.
586 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
588 TF/08 TLS error reporting now respects the incoming_interface and
589 incoming_port log selectors.
591 TF/09 Produce a more useful error message if an SMTP transport's hosts
592 setting expands to an empty string.
594 NM/06 Bugzilla 744: EXPN did not work under TLS.
595 Patch provided by Phil Pennock.
597 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
598 Patch provided by Richard Godbee.
600 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
601 acl_smtp_notquit, added index entry.
603 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
604 Patch provided by Eugene Bujak.
606 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
607 accept(). Patch provided by Maxim Dounin.
609 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
610 Patch provided by Phil Pennock.
612 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
614 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
615 Patch provided by Brad "anomie" Jorsch.
617 NM/14 Bugzilla 622: Added timeout setting to transport filter.
618 Patch provided by Dean Brooks.
620 TK/05 Add native DKIM support (does not depend on external libraries).
622 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
623 Patch provided by Graeme Fowler.
625 NM/16 Bugzilla 851: Documentation example syntax fix.
627 NM/17 Changed NOTICE file to remove references to embedded PCRE.
629 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
632 NM/19 Bugzilla 745: TLS version reporting.
633 Patch provided by Phil Pennock.
635 NM/20 Bugzilla 167: bool: condition support.
636 Patch provided by Phil Pennock.
638 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
639 clients. Patch provided by Phil Pennock.
641 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
642 Patch provided by Brad "anomie" Jorsch.
644 NM/23 Bugzilla 687: Fix misparses in eximstats.
645 Patch provided by Heiko Schlittermann.
647 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
648 Patch provided by Heiko Schlittermann.
650 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
651 plus update to original patch.
653 NM/26 Bugzilla 799: Documentation correction for ratelimit.
655 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
656 Patch provided by David Brownlee.
658 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
660 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
662 NM/30 Bugzilla 888: TLS documentation bugfixes.
664 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
666 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
667 Unlike the original bugzilla I have changed all shell scripts in src tree.
669 NM/33 Bugzilla 898: Transport filter timeout fix.
670 Patch by Todd Rinaldo.
672 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
673 Patch by Serge Demonchaux.
675 NM/35 Bugzilla 39: Base64 decode bug fixes.
676 Patch by Jakob Hirsch.
678 NM/36 Bugzilla 909: Correct connect() call in dcc code.
680 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
682 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
684 NM/39 Bugzilla 911: Fixed MakeLinks build script.
690 TK/01 Add preliminary DKIM support. Currently requires a forked version of
691 ALT-N's libdkim that I have put here:
692 http://duncanthrax.net/exim-experimental/
694 Note to Michael Haardt: I had to rename some vars in sieve.c. They
695 were called 'true' and it seems that C99 defines that as a reserved
696 keyword to be used with 'bool' variable types. That means you could
697 not include C99-style headers which use bools without triggering
698 build errors in sieve.c.
700 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
701 as mailq or other aliases. Changed the --help handling significantly
702 to do whats expected. exim_usage() emits usage/help information.
704 SC/01 Added the -bylocaldomain option to eximstats.
706 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
708 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
710 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
716 PH/01 Another patch from the Sieve maintainer.
718 PH/02 When an IPv6 address is converted to a string for single-key lookup
719 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
720 dots are used instead of colons so that keys in lsearch files need not
721 contain colons. This was done some time before quoting was made available
722 in lsearch files. However, iplsearch files do require colons in IPv6 keys
723 (notated using the quote facility) so as to distinguish them from IPv4
724 keys. This meant that lookups for IP addresses in host lists did not work
725 for iplsearch lookups.
727 This has been fixed by arranging for IPv6 addresses to be expressed with
728 colons if the lookup type is iplsearch. This is not incompatible, because
729 previously such lookups could never work.
731 The situation is now rather anomolous, since one *can* have colons in
732 ordinary lsearch keys. However, making the change in all cases is
733 incompatible and would probably break a number of configurations.
735 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
738 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
739 conversion specification without a maximum field width, thereby enabling
740 a rogue spamd server to cause a buffer overflow. While nobody in their
741 right mind would setup Exim to query an untrusted spamd server, an
742 attacker that gains access to a server running spamd could potentially
743 exploit this vulnerability to run arbitrary code as the Exim user.
745 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
746 $primary_hostname instead of what libspf2 thinks the hosts name is.
748 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
749 a directory entry by the name of the lookup key. Previously, if a
750 symlink pointed to a non-existing file or a file in a directory that
751 Exim lacked permissions to read, a lookup for a key matching that
752 symlink would fail. Now it is enough that a matching directory entry
753 exists, symlink or not. (Bugzilla 503.)
755 PH/03 The body_linecount and body_zerocount variables are now exported in the
758 PH/04 Added the $dnslist_matched variable.
760 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
761 This means they are set thereafter only if the connection becomes
764 PH/06 Added the client_condition to authenticators so that some can be skipped
765 by clients under certain conditions.
767 PH/07 The error message for a badly-placed control=no_multiline_responses left
768 "_responses" off the end of the name.
770 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
772 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
773 (without spaces) instead of just copying the configuration text.
775 PH/10 Added the /noupdate option to the ratelimit ACL condition.
777 PH/11 Added $max_received_linelength.
779 PH/12 Added +ignore_defer and +include_defer to host lists.
781 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
782 way in which PCRE > 7.0 is built.
784 PH/14 Implemented queue_only_load_latch.
786 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
787 MAIL command. The effect was to mangle the value on 64-bit systems.
789 PH/16 Another patch from the Sieve maintainer.
791 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
793 PH/18 If a system quota error occurred while trying to create the file for
794 a maildir delivery, the message "Mailbox is full" was not appended to the
795 bounce if the delivery eventually timed out. Change 4.67/27 below applied
796 only to a quota excession during the actual writing of the file.
798 PH/19 It seems that peer DN values may contain newlines (and other non-printing
799 characters?) which causes problems in log lines. The DN values are now
800 passed through string_printing() before being added to log lines.
802 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
803 and InterBase are left for another time.)
805 PH/21 Added message_body_newlines option.
807 PH/22 Guard against possible overflow in moan_check_errorcopy().
809 PH/23 POSIX allows open() to be a macro; guard against that.
811 PH/24 If the recipient of an error message contained an @ in the local part
812 (suitably quoted, of course), incorrect values were put in $domain and
813 $local_part during the evaluation of errors_copy.
819 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
820 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
823 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
824 issue a MAIL command.
826 PH/02 In an ACL statement such as
828 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
830 if a client was not listed at all, or was listed with a value other than
831 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
832 the condition was not true (as it should be), so access was not denied.
833 The bug was that the ! inversion was incorrectly passed on to the second
834 item. This has been fixed.
836 PH/03 Added additional dnslists conditions == and =& which are different from
837 = and & when the dns lookup returns more than one IP address.
839 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
840 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
842 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
843 FSYNC, which compiles an option called disable_fsync that allows for
844 bypassing fsync(). The documentation is heavily laced with warnings.
846 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
848 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
849 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
850 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
851 including adding "make clean"; (3) Added -fPIC when compiling the test
852 dynamically loaded module, to get rid of a warning.
854 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
855 message fails, move_frozen_messages = true and ignore_bounce_errors_after
856 = 0s. The bug is otherwise harmless.
858 PH/07 There was a bug in the dovecot authenticator such that the value of
859 $auth1 could be overwritten, and so not correctly preserved, after a
860 successful authentication. This usually meant that the value preserved by
861 the server_setid option was incorrect.
863 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
865 PH/09 Installed PCRE release 7.0.
867 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
868 run for batched SMTP input. It is now run at the start of every message
869 in the batch. While fixing this I discovered that the process information
870 (output by running exiwhat) was not always getting set for -bs and -bS
871 input. This is fixed, and it now also says "batched" for BSMTP.
873 PH/11 Added control=no_pipelining.
875 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
876 patch, slightly modified), and move the expansion of helo_data till after
877 the connection is made in the smtp transport (so it can use these
880 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
882 PH/14 Added log_selector = +pid.
884 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
886 PH/16 Add ${if forany and ${if forall.
888 PH/17 Added dsn_from option to vary the From: line in DSNs.
890 PH/18 Flush SMTP output before performing a callout, unless control =
891 no_callout_flush is set.
893 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
894 was true (the default) a successful delivery failed to delete the retry
895 item, thus causing premature timeout of the address. The bug is now
898 PH/20 Added hosts_avoid_pipelining to the smtp transport.
900 PH/21 Long custom messages for fakedefer and fakereject are now split up
901 into multiline reponses in the same way that messages for "deny" and
902 other ACL rejections are.
904 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
905 with slight modification.
907 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
908 draft, changing the syntax and factoring some duplicate code".
910 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
911 for deliveries of the second and subsequent messages over the same SMTP
914 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
915 ${reduce, with only minor "tidies".
917 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
919 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
920 expansion side effects.
922 PH/27 When a message times out after an over-quota error from an Exim-imposed
923 quota, the bounce message says "mailbox is full". This message was not
924 being given when it was a system quota that was exceeded. It now should
927 MH/03 Made $recipients available in local_scan(). local_scan() already has
928 better access to the recipient list through recipients_list[], but
929 $recipients can be useful in postmaster-provided expansion strings.
931 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
932 in the case of a MAIL command with additional options following the
933 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
934 were accidentally chopped off.
936 PH/29 SMTP synchronization checks are implemented when a command is read -
937 there is a check that no more input is waiting when there shouldn't be
938 any. However, for some commands, a delay in an ACL can mean that it is
939 some time before the response is written. In this time, more input might
940 arrive, invalidly. So now there are extra checks after an ACL has run for
941 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
942 pipelining has not been advertised.
944 PH/30 MH's patch to allow iscntrl() characters to be list separators.
946 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
947 returned in the SMTP response when smtp_return_error_details was false.
950 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
951 instead of the C I/O that was originally supplied, because problems were
954 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
955 Exim which did not show up earlier: it was assuming that a call to
956 SSL_CTX_set_info_callback() might give an error value. In fact, there is
957 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
958 was a macro that became an assignment, so it seemed to work. This has
959 changed to a proper function call with a void return, hence the compile
960 error. Exim's code has been fixed.
962 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
965 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
967 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
969 PH/37 If a message is not accepted after it has had an id assigned (e.g.
970 because it turns out to be too big or there is a timeout) there is no
971 "Completed" line in the log. When some messages of this type were
972 selected by exigrep, they were listed as "not completed". Others were
973 picked up by some special patterns. I have improved the selection
974 criteria to be more general.
976 PH/38 The host_find_failed option in the manualroute router can now be set
977 to "ignore", to completely ignore a host whose IP address cannot be
978 found. If all hosts are ignored, the behaviour is controlled by the new
979 host_all_ignored option.
981 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
982 homing or because of multiple MX records with /mx) generated more than
983 one IP address, and the following item turned out to be the local host,
984 all the secondary addresses of the first item were incorrectly removed
985 from the list, along with the local host and any following hosts (which
986 is what is supposed to happen).
988 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
989 whoever called Exim into the -H file. In the case of the daemon it was
990 behaving confusingly. When first started, it used values for whoever
991 started the daemon, but after a SIGHUP it used the Exim user (because it
992 calls itself on a restart). I have changed the code so that it now always
995 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
996 message are rejected with the same error (e.g. no authentication or bad
997 sender address), and a DATA command is nevertheless sent (as can happen
998 with PIPELINING or a stupid MUA), the error message that was given to the
999 RCPT commands is included in the rejection of the DATA command. This is
1000 intended to be helpful for MUAs that show only the final error to their
1003 PH/42 Another patch from the Sieve maintainer.
1005 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1006 Eximstats - Fixed some broken HTML links and added missing column headers
1008 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1009 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1011 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1013 PH/43 Yet another patch from the Sieve maintainer.
1015 PH/44 I found a way to check for a TCP/IP connection going away before sending
1016 the response to the final '.' that terminates a message, but only in the
1017 case where the client has not sent further data following the '.'
1018 (unfortunately, this is allowed). However, in many cases there won't be
1019 any further data because there won't be any more messages to send. A call
1020 to select() can be used: if it shows that the input is "ready", there is
1021 either input waiting, or the socket has been closed. An attempt to read
1022 the next input character can distinguish the two cases. Previously, Exim
1023 would have sent an OK response which the client would never have see.
1024 This could lead to message repetition. This fix should cure that, at
1025 least in a lot of common cases.
1027 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1028 advertised in response to EHLO.
1034 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1035 fixed by 4.65/MH/01 (is this a record?) are fixed:
1037 (i) An empty string was always treated as zero by the numeric comparison
1038 operators. This behaviour has been restored.
1040 (ii) It is documented that the numeric comparison operators always treat
1041 their arguments as decimal numbers. This was broken in that numbers
1042 starting with 0 were being interpreted as octal.
1044 While fixing these problems I realized that there was another issue that
1045 hadn't been noticed. Values of message_size_limit (both the global option
1046 and the transport option) were treated as octal if they started with 0.
1047 The documentation was vague. These values are now always treated as
1048 decimal, and I will make that clear in the documentation.
1054 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1055 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1058 MH/01 Don't check that the operands of numeric comparison operators are
1059 integers when their expansion is in "skipping" mode (fixes bug
1060 introduced by 4.64-PH/07).
1062 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1063 child addresses, Exim now panics and dies. Previously, because the count
1064 is held in a short int, deliveries were likely to be lost. As such a
1065 large number of recipients for a single message is ridiculous
1066 (performance will be very, very poor), I have chosen to impose a limit
1067 rather than extend the field.
1073 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1074 leftover -K file (the existence of which was triggered by #402).
1075 While we were at it, introduced process PID as part of the -K
1076 filename. This should rule out race conditions when creating
1079 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1080 processing considerably. Previous code took too long for large mails,
1081 triggering a timeout which in turn triggers #401.
1083 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1084 in the DK code in transports.c. sendfile() is not really portable,
1085 hence the _LINUX specificness.
1087 TF/01 In the add_headers option to the mail command in an Exim filter,
1088 there was a bug that Exim would claim a syntax error in any
1089 header after the first one which had an odd number of characters
1092 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1093 callout verification, Exim cached a "reject" for the entire domain. This
1094 is correct for most verifications, but it is not correct for a recipient
1095 verification with use_sender or use_postmaster set, because in that case
1096 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1097 case of MAIL FROM:<> rejection from other early rejections (e.g.
1098 rejection of HELO). When verifying a recipient using a non-null MAIL
1099 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1100 Whatever the result of the callout, the value of the domain cache is
1101 left unchanged (for any other kind of callout, getting as far as trying
1102 RCPT means that the domain itself is ok).
1104 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1107 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1108 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1109 session, the daemon ignores ECONNECT errors and logs others; it now
1110 ignores EPIPE as well.
1112 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1113 (quoted-printable decoding).
1115 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1116 later the small subsequent patch to fix an introduced bug.
1118 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1120 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1122 PH/08 An error is now given if message_size_limit is specified negative.
1124 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1125 to be given (somewhat) arbitrary names.
1127 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1130 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1131 miscellaneous code fixes
1133 PH/10 Added the log_reject_target ACL modifier to specify where to log
1136 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1137 hostname. This is wrong, because it relates to the incoming message (and
1138 probably the interface on which it is arriving) and not to the outgoing
1139 callout (which could be using a different interface). This has been
1140 changed to use the value of the helo_data option from the smtp transport
1141 instead - this is what is used when a message is actually being sent. If
1142 there is no remote transport (possible with a router that sets up host
1143 addresses), $smtp_active_hostname is used.
1145 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1146 tweaks were necessary in order to get it to work (see also 21 below):
1147 (a) The code assumed that strncpy() returns a negative number on buffer
1148 overflow, which isn't the case. Replaced with Exim's string_format()
1150 (b) There were several signed/unsigned issues. I just did the minimum
1151 hacking in of casts. There is scope for a larger refactoring.
1152 (c) The code used strcasecmp() which is not a standard C function.
1153 Replaced with Exim's strcmpic() function.
1154 (d) The code set only $1; it now sets $auth1 as well.
1155 (e) A simple test gave the error "authentication client didn't specify
1156 service in request". It would seem that Dovecot has changed its
1157 interface. Fortunately there's a specification; I followed it and
1158 changed what the client sends and it appears to be working now.
1160 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1163 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1164 address A is aliased to B and C, where B exists and C does not. Without
1165 -v the output is "A verified" because verification stops after a
1166 successful redirection if more than one address is generated. However,
1167 with -v the child addresses are also verified. Exim was outputting "A
1168 failed to verify" and then showing the successful verification for C,
1169 with its parentage. It now outputs "B failed to verify", showing B's
1170 parentage before showing the successful verification of C.
1172 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1173 look up a TXT record in a specific list after matching in a combined
1176 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1177 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1178 they consult the DNS. I had assumed they would set it the way they
1179 wanted; and indeed my experiments on Linux seem to show that in some
1180 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1181 To be on the safe side, however, I have now made the interface to
1182 host_find_byname() similar to host_find_bydns(), with an argument
1183 containing the DNS resolver options. The host_find_byname() function now
1184 sets these options at its start, just as host_find_bydns() does. The smtp
1185 transport options dns_qualify_single and dns_search_parents are passed to
1186 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1187 of host_find_byname() use the default settings of RES_DEFNAMES
1188 (qualify_single) but not RES_DNSRCH (search_parents).
1190 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1191 spool_read_header() do less string testing, by means of a preliminary
1192 switch on the second character of optional "-foo" lines. (This is
1193 overdue, caused by the large number of possibilities that now exist.
1194 Originally there were few.) While I was there, I also converted the
1195 str(n)cmp tests so they don't re-test the leading "-" and the first
1196 character, in the hope this might squeeze out yet more improvement.
1198 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1199 flag allowing group syntax was set by the header_syntax check but not
1200 turned off, possible causing trouble later; (2) The flag was not being
1201 set at all for the header_verify test, causing "group"-style headers to
1202 be rejected. I have now set it in this case, and also caused header_
1203 verify to ignore an empty address taken from a group. While doing this, I
1204 came across some other cases where the code for allowing group syntax
1205 while scanning a header line wasn't quite right (mostly, not resetting
1206 the flag correctly in the right place). These bugs could have caused
1207 trouble for malformed header lines. I hope it is now all correct.
1209 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1210 with the "reply" argument non-NULL. The code, however (which originally
1211 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1212 but it didn't always do it. This confused somebody who was copying the
1213 code for some other use. I have removed all the tests.
1215 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1216 feature that was used to support insecure browsers during the U.S. crypto
1217 embargo. It requires special client support, and Exim is probably the
1218 only MTA that supported it -- and would never use it because real RSA is
1219 always available. This code has been removed, because it had the bad
1220 effect of slowing Exim down by computing (never used) parameters for the
1221 RSA_EXPORT functionality.
1223 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1224 authenticator to fail if there's a tab character in the incoming data
1225 (there should never be unless someone is messing about, as it's supposed
1226 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1227 if the connection is using TLS or if the remote IP is the same as the
1228 local IP, and the "valid-client-cert option" if a client certificate has
1231 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1232 authenticators. This can be used for authorization after authentication
1233 succeeds. (In the case of plaintext, it servers for both authentication
1236 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1237 if any retry times were supplied.
1239 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1240 connection, where there is no client IP address to check. In this
1241 situation, the verify now always succeeds.
1243 PH/25 Applied John Jetmore's -Mset patch.
1245 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1247 PH/27 In a string expansion for a processed (not raw) header when multiple
1248 headers of the same name were present, leading whitespace was being
1249 removed from all of them, but trailing whitespace was being removed only
1250 from the last one. Now trailing whitespace is removed from each header
1251 before concatenation. Completely empty headers in a concatenation (as
1252 before) are ignored.
1254 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1255 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1257 PH/29 [Removed. This was a change that I later backed out, and forgot to
1258 correct the ChangeLog entry (that I had efficiently created) before
1259 committing the later change.]
1261 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1262 address errors (4xx response to RCPT) over the same connection as other
1263 messages routed to the same hosts. Such deliveries are always "forced",
1264 so retry times are not inspected. This resulted in far too many retries
1265 for the affected addresses. The effect occurred only when there were more
1266 hosts than the hosts_max_try setting in the smtp transport when it had
1267 the 4xx errors. Those hosts that it had tried were not added to the list
1268 of hosts for which the message was waiting, so if all were tried, there
1269 was no problem. Two fixes have been applied:
1271 (i) If there are any address or message errors in an SMTP delivery, none
1272 of the hosts (tried or untried) are now added to the list of hosts
1273 for which the message is waiting, so the message should not be a
1274 candidate for sending over the same connection that was used for a
1275 successful delivery of some other message. This seems entirely
1276 reasonable: after all the message is NOT "waiting for some host".
1277 This is so "obvious" that I'm not sure why it wasn't done
1278 previously. Hope I haven't missed anything, but it can't do any
1279 harm, as the worst effect is to miss an optimization.
1281 (ii) If, despite (i), such a delivery is accidentally attempted, the
1282 routing retry time is respected, so at least it doesn't keep
1283 hammering the server.
1285 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1286 in ${readsocket because some servers need this prod.
1288 PH/32 Added some extra debug output when updating a wait-xxx database.
1290 PH/33 The hint "could be header name not terminated by colon", which has been
1291 given for certain expansion errors for a long time, was not being given
1292 for the ${if def:h_colon_omitted{... case.
1294 PH/34 The spec says: "With one important exception, whenever a domain list is
1295 being scanned, $domain contains the subject domain." There was at least
1296 one case where this was not true.
1298 PH/35 The error "getsockname() failed: connection reset by peer" was being
1299 written to the panic log as well as the main log, but it isn't really
1300 panic-worthy as it just means the connection died rather early on. I have
1301 removed the panic log writing for the ECONNRESET error when getsockname()
1304 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1305 runs only) independently of the message's sender address. This meant
1306 that, if the 4xx error was in fact related to the sender, a different
1307 message to the same recipient with a different sender could confuse
1308 things. In particualar, this can happen when sending to a greylisting
1309 server, but other circumstances could also provoke similar problems.
1310 I have changed the default so that the retry time for these errors is now
1311 based a combination of the sender and recipient addresses. This change
1312 can be overridden by setting address_retry_include_sender=false in the
1315 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1316 remote server are returned as part of bounce messages. This was not
1317 happening for LMTP over a pipe (the lmtp transport), but now it is the
1318 same for both kinds of LMTP.
1320 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1321 in header lines that were in fact CNAMEs. This is no longer the case.
1323 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1324 and queue runs started by the daemon processed all messages. This has
1325 been fixed so that -R and -S can now usefully be given with -q<time>.
1327 PH/40 Import PCRE release 6.7 (fixes some bugs).
1329 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1331 PH/42 Give an error if -q is specified more than once.
1333 PH/43 Renamed the variables $interface_address and $interface_port as
1334 $received_ip_address and $received_port, to make it clear that these
1335 values apply to message reception, and not to the outgoing interface when
1336 a message is delivered. (The old names remain recognized, of course.)
1338 PH/44 There was no timeout on the connect() call when using a Unix domain
1339 socket in the ${readsocket expansion. There now is.
1341 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1342 be meaningful with "accept".
1344 SC/01 Eximstats V1.43
1345 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1347 SC/02 Eximstats V1.44
1348 Use a glob alias rather than an array ref in the generated
1349 parser. This improves both readability and performance.
1351 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1352 Collect SpamAssassin and rejection statistics.
1353 Don't display local sender or destination tables unless
1354 there is data to show.
1355 Added average volumes into the top table text output.
1357 SC/04 Eximstats V1.46
1358 Collect data on the number of addresses (recipients)
1359 as well as the number of messages.
1361 SC/05 Eximstats V1.47
1362 Added 'Message too big' to the list of mail rejection
1363 reasons (thanks to Marco Gaiarin).
1365 SC/06 Eximstats V1.48
1366 Mainlog lines which have GMT offsets and are too short to
1367 have a flag are now skipped.
1369 SC/07 Eximstats V1.49 (Alain Williams)
1370 Added the -emptyok flag.
1372 SC/08 Eximstats V1.50
1373 Fixes for obtaining the IP address from reject messages.
1375 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1376 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1377 whitesspace changes from 4.64-PH/27
1379 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1382 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1383 are found, allow negative numbers in numeric criteria)
1385 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1387 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1388 to match changes made in 4.64-PH/43
1390 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1392 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1393 "domain" is actually an IP address, and force a failure. This locks out
1394 those revolvers/nameservers that support "A-for-A" lookups, in
1395 contravention of the specifications.
1397 PH/48 When a host name was looked up from an IP address, and the subsequent
1398 forward lookup of the name timed out, the host name was left in
1399 $sender_host_name, contrary to the specification.
1401 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1402 restricted to single-key lookups, Exim was not diagnosing an error if
1403 * or *@ was used with a query-style lookup.
1405 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1407 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1408 long ago, but noone interested enough thought of it. Let's just say that
1409 the "1.1" means that there are some new functions that weren't there at
1410 some point in the past.
1412 PH/51 Error processing for expansion failure of helo_data from an smtp
1413 transport during callout processing was broken.
1415 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1416 tested/used via the -bh/-bhc/-bs options.
1418 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1419 bug, fixed in subsequent PCRE releases).
1421 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1422 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1424 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1430 SC/01 Use a glob alias rather than an array ref in eximstats generated
1431 parser. This improves both readability and performance.
1433 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1434 Don't display local sender or destination tables in eximstats unless
1435 there is data to show.
1436 Added average volumes into the eximstats top table text output.
1438 SC/03 Collect data on the number of addresses (recipients) as well
1439 as the number of messages in eximstats.
1441 TF/01 Correct an error in the documentation for the redirect router. Exim
1442 does (usually) call initgroups() when daemonizing.
1444 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1445 with consistent privilege compared to when running as a daemon.
1447 TF/03 Note in the spec that $authenticated_id is not set for local
1448 submissions from trusted users.
1450 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1451 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1453 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1454 by adding some example configuration directives to the default
1455 configuration file. A little bit of work is required to uncomment the
1456 directives and define how usernames and passwords are checked, but
1457 there is now a framework to start from.
1459 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1460 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1461 without this. I don't know how relevant this is to other LDAP libraries.
1463 PH/02 Add the verb name to the "unknown ACL verb" error.
1465 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1467 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1469 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1470 directory not expanded when it should be if an expanded home directory
1471 was set for the address (which is overridden by the transport).
1473 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1476 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1477 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1478 because it is too late at that time, and has no effect.
1480 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1481 security issue with \' (bugzilla #107). I could not use the
1482 PQescapeStringConn() function, because it needs a PGconn value as one of
1485 PH/08 When testing addresses using -bt, indicate those final addresses that
1486 are duplicates that would not cause an additional delivery. At least one
1487 person was confused, thinking that -bt output corresponded to deliveries.
1488 (Suppressing duplicates isn't a good idea as you lose the information
1489 about possibly different redirections that led to the duplicates.)
1491 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1492 systems where poll() doesn't work, in particular OS X.
1494 PH/10 Added more information to debugging output for retry time not reached.
1496 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1497 operations in malware.c.
1499 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1502 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1503 syslog_duplication set false, log lines that would normally be written
1504 both the the main log and to the reject log were not written to syslog at
1507 PH/14 In the default configuration, change the use of "message" in ACL warn
1508 statements to "add_header".
1510 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1511 not followed by a command (e.g. "seen endif").
1513 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1514 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1517 PH/17 Added extra conditions to the default value of delay_warning_condition
1521 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1522 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1523 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1526 The Auto-Submitted: and various List- headers are standardised, whereas I
1527 don't think Precedence: ever was.
1529 PH/18 Refactored debugging code in route_finduser() to show more information,
1530 in particular, the error code if getpwnam() issues one.
1532 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1533 This is apparently needed in addition to the PH/07 change above to avoid
1534 any possible encoding problems.
1536 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1537 but not after initializing Perl.
1539 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1540 output them only if debugging. By default they are written stderr,
1541 apparently, which is not desirable.
1543 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1546 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1549 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1551 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1552 authenticated or an ident call has been made. Suppress the default
1553 values for $authenticated_id and $authenticated_sender (but permit -oMai
1554 and -oMas) when testing with -bh.
1556 PH/24 Re-jigged the order of the tests in the default configuration so that the
1557 tests for valid domains and recipients precede the DNS black list and CSA
1558 tests, on the grounds that those ones are more expensive.
1560 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1561 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1562 This bug exists in every version of Exim that I still have, right back to
1565 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1566 However, an attempt to turn on case-sensitivity in a regex key by
1567 including (?-i) didn't work because the subject string was already
1568 lowercased, and the effects were non-intuitive. It turns out that a
1569 one-line patch can be used to allow (?-i) to work as expected.
1575 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1576 other effects) broke the use of negated acl sub-conditions.
1578 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1581 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1582 "Deny" causes Exim to reject the incoming connection with a 554 error.
1583 Unfortunately, if there is a major crisis, such as a disk failure,
1584 tcp-wrappers gives "deny", whereas what one would like would be some
1585 kind of temporary error. A kludge has been added to help with this.
1586 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1587 554 error is used if errno is still zero or contains ENOENT (which occurs
1588 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1591 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1593 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1594 errors. Otherwise a message that provokes a temporary error (when other
1595 messages do not) can cause a whole host to time out.
1597 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1598 addresses were routed directly to files or pipes from a redirect router.
1599 File deliveries just didn't batch; pipe deliveries might have suffered
1602 PH/06 A failure to get a lock for a hints database would erroneously always say
1603 "Failed to get write lock", even when it was really a read lock.
1605 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1606 of 0600. This has been changed to use the value of the lockfile_mode
1607 option (which defaults to 0600).
1609 PH/08 Applied small patch from the Sieve maintainer.
1611 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1612 folder from quota calculations, a direct delivery into this folder messed
1613 up the contents of the maildirsize file. This was because the regex was
1614 used only to exclude .Trash (or whatever) when the size of the mailbox
1615 was calculated. There was no check that a delivery was happening into an
1616 excluded directory. This bug has been fixed by ignoring all quota
1617 processing for deliveries into excluded directories.
1619 PH/10 Added the maildirfolder_create_regex option to appendfile.
1625 PH/01 The code for finding all the local interface addresses on a FreeBSD
1626 system running IPv6 was broken. This may well have applied to all BSD
1627 systems, as well as to others that have similar system calls. The broken
1628 code found IPv4 interfaces correctly, but gave incorrect values for the
1629 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1630 that it would not match correctly against @[] and not recognize the IPv6
1633 PH/02 The ipliteral router was not recognizing addresses of the form user@
1634 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1636 PH/03 Added disable_ipv6.
1638 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1639 decoded form, because it is most often used to construct To: headers
1640 lines in autoreplies, and the decoded form may well be syntactically
1641 invalid. However, $reply_address has leading white space removed, and all
1642 newlines turned into spaces so that the autoreply transport does not
1645 PH/05 If group was specified without a user on a router, and no group or user
1646 was specified on a transport, the group from the router was ignored.
1648 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1649 for visible compile-time settings that can be used to change these
1650 numbers, for those that want even more. Backwards compatibility with old
1651 spool files has been maintained. However, going back to a previous Exim
1652 release will lost any variables that are in spool files.
1654 PH/07 Two small changes when running in the test harness: increase delay when
1655 passing a TCP/IP connection to a new process, in case the original
1656 process has to generate a bounce, and remove special handling of
1657 127.0.0.2 (sic), which is no longer necessary.
1659 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1660 be the same on different OS.
1662 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1665 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1666 whether --show-vars was specified or not
1668 JJ/02 exipick: Added support for new ACL variable spool format introduced
1671 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1672 syntactically invalid From: or Reply-to: line, and a filter used this to
1673 generate an autoreply, and therefore failed to obtain an address for the
1674 autoreply, Exim could try to deliver to a non-existent relative file
1675 name, causing unrelated and misleading errors. What now happens is that
1676 it logs this as a hard delivery error, but does not attempt to create a
1679 PH/11 The exinext utility has a -C option for testing purposes, but although
1680 the given file was scanned by exinext itself; it wasn't being passed on
1681 when Exim was called.
1683 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1684 an end-of-file indication when reading a command response.
1686 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1687 compiled. In many other places in Exim, IPv6 addresses are always
1688 recognized, so I have changed this. It also means that IPv4 domain
1689 literals of the form [IPV4:n.n.n.n] are now always recognized.
1691 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1692 used if the router is not running as root, for example, when verifying at
1693 ACL time, or when using -bh. The debugging output from this situation was
1694 non-existent - all you got was a failure to exec. I have made two
1697 (a) Failures to set uid/gid, the current directory, or a process leader
1698 in a subprocess such as that created by queryprogram now generate
1699 suitable debugging ouput when -d is set.
1701 (b) The queryprogram router detects when it is not running as root,
1702 outputs suitable debugging information if -d is set, and then runs
1703 the subprocess without attempting to change uid/gid.
1705 PH/15 Minor change to Makefile for building test_host (undocumented testing
1708 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1709 additional section of a DNS packet that returns MX or SRV records.
1710 Instead, it always explicitly searches for A/AAAA records. This avoids
1711 major problems that occur when a DNS server includes only records of one
1712 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1713 fixed another bug: if SRV records were looked up and the corresponding
1714 address records were *not* found in the additional section, the port
1715 values from the SRV records were lost.
1717 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1718 using the correct key (the original address) when searching the retry
1719 rules in order to find which one to use for generating the retry hint.
1721 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1722 adding the default one. Similarly, if it contains a Reply-To: header, the
1723 errors_reply_to option, if set, is not used.
1725 PH/19 When calculating a retry time, Exim used to measure the "time since
1726 failure" by looking at the "first failed" field in the retry record. Now
1727 it does not use this if it is later than than the arrival time of the
1728 message. Instead it uses the arrival time. This makes for better
1729 behaviour in cases where some deliveries succeed, thus re-setting the
1730 "first failed" field. An example is a quota failure for a huge message
1731 when small messages continue to be delivered. Without this change, the
1732 "time since failure" will always be short, possible causing more frequent
1733 delivery attempts for the huge message than are intended.
1734 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1736 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1737 $1, $2, $3) because the numerical variables can be reset during some
1738 expansion items (e.g. "match"), thereby losing the authentication data.
1740 PH/21 Make -bV show the size of off_t variables so that the test suite can
1741 decide whether to run tests for quotas > 2G.
1743 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1744 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1745 filecount value is greater than 2G or if a quota value is greater than 2G
1746 on a system where the size of off_t is not greater than 4, a panic error
1749 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1750 never match. The debug and -bh output now contains an explicit error
1751 message indicating a malformed IPv4 address or mask.
1753 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1754 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1755 PH/23 above applies.
1757 PH/25 Do not write to syslog when running in the test harness. The only
1758 occasion when this arises is a failure to open the main or panic logs
1759 (for which there is an explicit test).
1761 PH/26 Added the /no_tell option to "control=freeze".
1763 PH/27 If a host name lookup failed very early in a connection, for example, if
1764 the IP address matched host_lookup and the reverse lookup yielded a name
1765 that did not have a forward lookup, an error message of the form "no IP
1766 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1767 could be logged. Now it outputs the IP address instead of "NULL".
1769 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1770 allows the sender and the authenticated sender to be set when
1771 submitting a message from within Exim. Since child_open_exim() is
1772 documented for local_scan(), the new function should be too.
1774 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1775 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1776 results in an empty string is now treated as unset.
1778 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1780 PH/31 Added sender_verify_fail logging option.
1782 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1783 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1784 tidied the source and removed it altogether.
1786 PH/33 When a queue run was abandoned because the load average was too high, a
1787 log line was always written; now it is written only if the queue_run log
1788 selector is set. In addition, the log line for abandonment now contains
1789 information about the queue run such as the pid. This is always present
1790 in "start" and "stop" lines but was omitted from the "abandon" line.
1792 PH/34 Omit spaces between a header name and the colon in the error message that
1793 is given when verify = headers_syntax fails (if there are lots of them,
1794 the message gets confusing).
1796 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1797 names, as there are now some PTR records that contain slashes. This check
1798 is only to protect against broken name servers that fall over on strange
1799 characters, so the fact that it applies to all lookups doesn't matter.
1801 PH/36 Now that the new test suite is complete, we can remove some of the
1802 special code in Exim that was needed for the old test suite. For example,
1803 sorting DNS records because real resolvers return them in an arbitrary
1804 order. The new test suite's fake resolver always returns records in the
1807 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1808 bounces) except when queue_only is set, to avoid logging races between
1809 the different processes.
1811 PH/38 Panic-die if .include specifies a non-absolute path.
1813 PH/39 A tweak to the "H" retry rule from its user.
1815 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1816 a label. They prevented compilation on older perls.
1818 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1819 a warning to be raised on newish perls.
1821 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1822 on queue. Changes to match documented behaviour of showing count of
1823 messages matching specified criteria.
1825 PH/40 Changed the default ident timeout from 30s to 5s.
1827 PH/41 Added support for the use of login_cap features, on those BSD systems
1828 that have them, for controlling the resources used by pipe deliveries.
1830 PH/42 The content-scanning code uses fopen() to create files in which to put
1831 message data. Previously it was not paying any attention to the mode of
1832 the files. Exim runs with umask(0) because the rest of the code creates
1833 files with open(), and sets the required mode explicitly. Thus, these
1834 files were ending up world-writeable. This was not a big issue, because,
1835 being within the spool directory, they were not world-accessible. I have
1836 created a function called modefopen, which takes an additional mode
1837 argument. It sets umask(777), creates the file, chmods it to the required
1838 mode, then resets the umask. All the relevant calls to fopen() in the
1839 content scanning code have been changed to use this function.
1841 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1842 to 24 hours. This avoids potential overflow problems when processing G
1843 and H retry rules. I suspect nobody ever tinkers with this value.
1845 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1847 PH/45 When the plaintext authenticator is running as a client, the server's
1848 challenges are checked to ensure they are valid base64 strings. By
1849 default, the authentication attempt is cancelled if an invalid string is
1850 received. Setting client_ignore_invalid_base64 true ignores these errors.
1851 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1852 they are received. Thus, the responses can be made to depend on the
1853 challenges. If an invalid string is ignored, an empty string is placed in
1856 PH/46 Messages that are created by the autoreply transport now contains a
1857 References: header, in accordance with RFCs 2822 and 3834.
1859 PH/47 Added authenticated_sender_force to the smtp transport.
1861 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1863 PH/49 Installed latest patch from the Sieve maintainer.
1865 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1866 was also set, the appendfile transport was unnecessarily scanning a
1867 directory of message files (e.g. for maildir delivery) to find the count
1868 of files (along with the size), even though it did not need this
1869 information. It now does the scan only if it needs to find either the
1870 size of the count of files.
1872 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1874 PH/52 Two bugs concerned with error handling when the smtp transport is
1877 (i) Exim was not creating retry information for temporary errors given
1878 for individual recipients after the DATA command when the smtp transport
1879 was used in LMTP mode. This meant that they could be retried too
1880 frequently, and not timed out correctly.
1882 (ii) Exim was setting the flag that allows error details to be returned
1883 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1884 recipients that were returned after the DATA command.
1886 PH/53 This is related to PH/52, but is more general: for any failing address,
1887 when detailed error information was permitted to be returned to the
1888 sender, but the error was temporary, then after the final timeout, only
1889 "retry timeout exceeded" was returned. Now it returns the full error as
1890 well as "retry timeout exceeded".
1892 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1893 do this, and (what is worse) MTAs that accept it.
1895 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1896 will now be deprecated.
1898 PH/56 New os.c-cygwin from the Cygwin maintainer.
1900 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1901 formats (previously only available in exim formats via -bpr, -bpru,
1902 and -bpra. Now also available in native and exiqgrep formats)
1904 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1905 with very large, slow to parse queues
1907 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1909 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1911 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1912 responses to authentication challenges, though it was showing the
1913 challenges; (ii) I've removed the CR characters from the debug output for
1916 PH/58 Allow for the insertion of a newline as well as a space when a string
1917 is turned into more than one encoded-word during RFC 2047 encoding. The
1918 Sieve code now uses this.
1920 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1921 data_4xx, lost_connection, tls_required.
1923 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1924 message was being sent as an SMTP response.
1926 PH/61 Add -l and -k options to exicyclog.
1928 PH/62 When verifying, if an address was redirected to one new address, so that
1929 verification continued, and the new address failed or deferred after
1930 having set something in $address_data, the value of $address_data was not
1931 passed back to the ACL. This was different to the case when no
1932 redirection occurred. The value is now passed back in both cases.
1934 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1935 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1936 use login_cap.h, so on its own it isn't the distinguishing feature. The
1937 new name refers directly to the setclassresources() function.
1939 PH/65 Added configuration files for NetBSD3.
1941 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1943 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1944 is preferred over IPv4.
1946 PH/68 The bounce_return_message and bounce_return_body options were not being
1947 honoured for bounces generated during the reception of non-SMTP messages.
1948 In particular, this applied to messages rejected by the ACL. This bug has
1949 been fixed. However, if bounce_return_message is true and bounce_return_
1950 body is false, the headers that are returned for a non-SMTP message
1951 include only those that have been read before the error was detected.
1952 (In the case of an ACL rejection, they have all been read.)
1954 PH/69 The HTML version of the specification is now built in a directory called
1955 spec_html instead of spec.html, because the latter looks like a path with
1956 a MIME-type, and this confuses some software.
1958 PH/70 Catch two compiler warnings in sieve.c.
1960 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1961 function verify_get_ident() calls ip_connect() to connect a socket, but
1962 if the "connect()" function timed out, ip_connect() used to close the
1963 socket. However, verify_get_ident() also closes the socket later, and in
1964 between Exim writes to the log, which may get opened at this point. When
1965 the socket was closed in ip_connect(), the log could get the same file
1966 descriptor number as the socket. This naturally causes chaos. The fix is
1967 not to close the socket in ip_connect(); the socket should be closed by
1968 the function that creates it. There was only one place in the code where
1969 this was missing, in the iplookup router, which I don't think anybody now
1970 uses, but I've fixed it anyway.
1972 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1973 well as to direct DNS lookups. Otherwise the handling of names in host
1974 lists is inconsistent and therefore confusing.
1980 PH/01 Two changes to the default runtime configuration:
1982 (1) Move the checks for relay_from_hosts and authenticated clients from
1983 after to before the (commented out) DNS black list checks.
1985 (2) Add control=submission to the relay_from_hosts and authenticated
1986 clients checks, on the grounds that messages accepted by these
1987 statements are most likely to be submissions.
1989 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1991 (1) Generate an error if the third argument for the ${prvs expansion is
1994 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1997 (3) Reset the variables that are obtained from the first argument of
1998 ${prvscheck and used in the second argument before leaving the code,
1999 because their memory is reclaimed, so using them afterwards may do
2002 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2003 one (it's much easier than Tom thought :-).
2005 (5) Because of (4), we can now allow for the use of $prvscheck_result
2006 inside the third argument.
2008 PH/03 For some reason, the default setting of PATH when running a command from
2009 a pipe transport was just "/usr/bin". I have changed it to
2012 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2013 anything to be listed in the output from -bV.
2015 PH/05 When a filter generated an autoreply, the entire To: header line was
2016 quoted in the delivery log line, like this:
2018 => >A.N.Other <ano@some.domain> <original@ddress> ...
2020 This has been changed so that it extracts the operative address. There
2021 may be more than one such address. If so, they are comma-separated, like
2024 => >ano@some.domain,ona@other.domain <original@ddress> ...
2026 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2027 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2028 not being looked up in the rDNS to get a host name, Exim was showing the
2029 IP address twice in Received: lines, even though the IP addresses were
2030 identical. For example:
2032 Received: from [1.2.3.4] (helo=[1.2.3.4])
2034 However, if the real host name was known, it was omitting the HELO data
2035 if it matched the actual IP address. This has been tidied up so that it
2036 doesn't show the same IP address twice.
2038 PH/07 When both +timestamp and +memory debugging was on, the value given by
2039 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2040 called by the debug printing, thereby overwriting the timestamp buffer.
2041 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2043 PH/08 When the original message was included in an autoreply transport, it
2044 always said "this is a copy of the message, including all the headers",
2045 even if body_only or headers_only was set. It now gives an appropriate
2048 PH/09 Applied a patch from the Sieve maintainer which:
2050 o fixes some comments
2051 o adds the (disabled) notify extension core
2052 o adds some debug output for the result of if/elsif tests
2053 o points to the current vacation draft in the documentation
2054 and documents the missing references header update
2058 o fixes a bug in processing the envelope test (when testing
2059 multiple envelope elements, the last element determinted the
2062 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2063 Electronic Mail") by including:
2065 Auto-submitted: auto-generated
2067 in the messages that it generates (bounce messages and others, such as
2068 warnings). In the case of bounce messages for non-SMTP mesages, there was
2069 also a typo: it was using "Auto_submitted" (underscore instead of
2070 hyphen). Since every message generated by Exim is necessarily in response
2071 to another message, thes have all been changed to:
2073 Auto-Submitted: auto-replied
2075 in accordance with these statements in the RFC:
2077 The auto-replied keyword:
2079 - SHOULD be used on messages sent in direct response to another
2080 message by an automatic process,
2082 - MUST NOT be used on manually-generated messages,
2084 - MAY be used on Delivery Status Notifications (DSNs) and Message
2085 Disposition Notifications (MDNs),
2087 - MUST NOT be used on messages generated by automatic or periodic
2088 processes, except for messages which are automatic responses to
2091 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2092 to the default Received: header definition.
2094 PH/12 Added log selector acl_warn_skipped (default on).
2096 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2097 variables because (a) they are in the wrong storage pool and (b) even if
2098 they were copied, it wouldn't work properly because of the caching.
2100 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2101 checking when decoding. Apparently there are clients that generate
2102 overlong encoded strings. Why am I not surprised?
2104 PH/15 If the first argument of "${if match_address" was not empty, but did not
2105 contain an "@" character, Exim crashed. Now it writes a panic log message
2106 and treats the condition as false.
2108 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2110 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2111 extension "envelope-auth". The code is finished and in agreement with
2112 other implementations, but there is no documentation so far and in fact,
2113 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2114 not changing the active code.
2116 Print executed "if" and "elsif" statements when debugging is used. This
2117 helps a great deal to understand what a filter does.
2119 Document more things not specified clearly in RFC3028. I had all this
2120 sorted out, when out of a sudden new issues came to my mind. Oops."
2122 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2125 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2126 canonical form (as documented). However, after a host name lookup from
2127 the IP address, check_host() was doing a simple string comparison with
2128 addresses acquired from the DNS when checking that the found name did
2129 have the original IP as one of its addresses. Since any found IPv6
2130 addresses are likely to be in abbreviated form, the comparison could
2131 fail. Luckily, there already exists a function for doing the comparison
2132 by converting both addresses to binary, so now that is used instead of
2133 the text comparison.
2135 PH/20 There was another similar case to PH/19, when a complete host name was
2136 given in a host list; looking up its IP address could give an abbreviated
2137 form, whereas the current host's name might or might not be abbreviated.
2138 The same fix has been applied.
2144 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2145 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2148 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2149 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2151 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2153 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2154 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2155 Also, for "Auto-Submitted", treat anything other than "no" as
2156 non-personal, in accordance with RFC 3834. (Previously it treated
2157 anything starting "auto-" as non-personal.)
2159 TF/01 The control=submission/name=... option had a problem with syntax
2160 errors if the name included a slash character. The /name= option
2161 now slurps the rest of the string, so it can include any characters
2162 but it must come last in the list of options (after /sender_retain
2165 PH/05 Some modifications to the interface to the fake nameserver for the new
2173 TK/01 Added the "success_on_redirect" address verification option. See
2174 NewStuff for rationale and an example.
2176 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2178 PH/02 Patch to exigrep to allow it to work on syslog lines.
2180 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2181 fread() to skip over the body file's header line, because in Cygwin the
2182 header line is locked and is inaccessible.
2184 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2185 co-exist for some time) to make it clear that it is the Exim ID that is
2186 referenced, not the Message-ID: header line.
2188 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2189 string_format() function, because snprintf() does not exist on all
2192 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2193 ${stat: expansion item.
2195 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2196 protocol synchronization error", to keep the pedants happy.
2198 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2199 well as for IRIX systems, when gcc is being used. See the host.c source
2202 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2204 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2207 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2208 transport and to the smtp transport in LMTP mode.
2210 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2212 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2213 run a filter in a subprocess. This could lead to confusion in subsequent
2214 lookups in the parent process. There should also be a search_tidyup() at
2215 the end of the subprocess.
2217 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2218 only if the host matched helo_try_verify_hosts, which caused the
2219 verification to occur when the EHLO/HELO command was issued. The ACL just
2220 tested the remembered result. Now, if a previous verification attempt has
2221 not happened, "verify = helo" does it there and then.
2223 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2225 TK/03 Fix log output including CR from clamd.
2227 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2228 exist provoked a memory error which could cause a segfault.
2230 PH/15 Installed PCRE 6.2
2232 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2234 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2235 of the problem. Specifically, suggested +O2 rather than +O1 for the
2238 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2240 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2241 record, the port was not correctly logged when the outgoing_port log
2242 selector was set (it logged the transort's default port).
2244 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2245 fallback_hosts, and "hosts" in the smtp transport.
2247 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2248 host errors such as "Connection refused".
2250 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2251 authentication with radiusclient 0.4.9:
2253 - Error returned from rc_read_config was caught wrongly
2254 - Username/password not passed on to radius server due to wrong length.
2256 The presumption is that some radiusclient API changes for 4.51/PH/17
2257 were not taken care of correctly. The code is still untested by me (my
2258 Linux distribution still has 0.3.2 of radiusclient), but it was
2259 contributed by a Radius user.
2261 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2262 expanding the "port" option of the smtp transport.
2264 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2265 while reading a MIME header. Thanks to Tom Hughes for a patch.
2267 PH/24 Include config.h inside local_scan.h so that configuration settings are
2270 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2271 that in an ACL for RCPT (for example), you can examine exactly what was
2274 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2275 commands, but it was not correctly comparing the address with the actual
2276 client host address. Thus, it would show the EHLO address in Received:
2277 header lines when this was not necessary.
2279 PH/27 Added the % operator to ${eval:}.
2281 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2282 it should be ignoring failures (because with -C, for example, it has lost
2283 privilege). It wasn't ignoring creation failures other than "already
2286 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2289 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2290 because an input line was too long, either on its own, or by virtue of
2291 too many continuations, the temporary file was not being removed, and the
2292 return code was incorrect.
2294 PH/31 Missing "BOOL" in function definition in filtertest.c.
2296 PH/32 Applied Sieve patches from the maintainer.
2298 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2300 PH/33 Added "verify = not_blind".
2302 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2303 Local/Makefile (with some defaults set). These are used in built scripts
2304 such as exicyclog, but they have never been used in the exim_install
2305 script (though there are many overriding facilities there). I have
2306 arranged that the exim_install script now takes note of these two
2309 PH/35 Installed configuration files for Dragonfly.
2311 PH/36 When a locally submitted message by a trusted user did not contain a
2312 From: header, and the sender address was obtained from -f or from an SMTP
2313 MAIL command, and the trusted user did not use -F to supply a sender
2314 name, $originator_name was incorrectly used when constructing a From:
2315 header. Furthermore, $originator_name was used for submission mode
2316 messages from external hosts without From: headers in a similar way,
2317 which is clearly wrong.
2319 PH/37 Added control=suppress_local_fixups.
2321 PH/38 When log_selector = +received_sender was set, and the addition of the
2322 sender made the log line's construction buffer exactly full, or one byte
2323 less than full, an overflow happened when the terminating "\n" was
2326 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2327 when the result of a list match is failure because a DNS lookup failed.
2329 PH/40 RM_COMMAND is now used in the building process.
2331 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2332 the "build-* directories that it finds.
2334 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2335 address in a domain literal was a prefix of an interface address.
2337 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2338 when verifying a sender address, unless rewrite_headers is false.
2340 PH/44 Wrote a long comment about why errors_to addresses are verified as
2341 recipients, not senders.
2343 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2344 the ratelimit ACL was added.
2346 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2348 PH/46 Added extra information about PostgreSQL errors to the error string.
2350 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2351 avoiding the need to install special zones in a real server. This is
2352 backwards compatible; if it can't find the fake resolver, it drops back.
2353 Thus, both old and new test suites can be run.
2355 TF/02 Added util/ratelimit.pl
2357 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2358 clock is set back in time.
2360 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2361 Candler <B.Candler@pobox.com>.
2363 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2364 OK for addresses that are the result of redirections.
2366 PH/48 A number of further additions for the benefit of the new test suite,
2367 including a fake gethostbyname() that interfaces to the fake DNS resolver
2370 TF/06 The fix for widen_domains has also been applied to qualify_single and
2371 search_parents which are the other dnslookup options that can cause
2374 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2377 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2379 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2380 service" if it encountered an MX record with an empty target hostname.
2381 The message is now "an MX or SRV record indicated no SMTP service".
2383 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2384 if the DNS of the sending site is misconfigured. This is quite a
2385 common situation. This change restores the behaviour of treating a
2386 helo verification defer as a failure.
2388 PH/51 If self=fail was set on a router, the bounce message did not include the
2389 actual error message.
2395 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2397 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2398 command itself ended in error, the underlying message about the transport
2399 filter timeout was being overwritten with the pipe command error. Now the
2400 underlying error message should be appended to the second error message.
2402 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2404 PH/02 Reduce the amount of output that "make" produces by default. Full output
2405 can still be requested.
2407 PH/03 The warning log line about a condition test deferring for a "warn" verb
2408 was being output only once per connection, rather than after each
2409 occurrence (because it was using the same function as for successful
2410 "warn" verbs). This seems wrong, so I have changed it.
2412 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2413 it should not have, which might have caused a crash in the right
2414 circumstances, but probably never did.
2416 PH/04 Installed a modified version of Tony Finch's patch to make submission
2417 mode fix the return path as well as the Sender: header line, and to
2418 add a /name= option so that you can make the user's friendly name appear
2421 TF/03 Added the control = fakedefer ACL modifier.
2423 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2424 Mark Lowes for thorough testing.
2426 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2428 TK/03 Merged latest SRS patch from Miles Wilton.
2430 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2431 with the definition in sysexits.h (which is #included earlier).
2432 Fortunately, Exim does not actually use EX_OK. The code used to try to
2433 preserve the sysexits.h value, by assumimg that macro definitions were
2434 scanned for macro replacements. I have been disabused of this notion,
2435 so now the code just undefines EX_OK before #including unistd.h.
2437 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2438 in the smtp transport. When a block could not be written in a single
2439 write() function, the timeout was being re-applied to each part-write.
2440 This seems wrong - if the receiver was accepting one byte at a time it
2441 would take for ever. The timeout is now adjusted when this happens. It
2442 doesn't have to be particularly precise.
2444 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2445 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2447 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2448 without a domain if the check to <postmaster@domain> fails.
2450 SC/01 Eximstats: added -xls and the ability to specify output files
2451 (patch written by Frank Heydlauf).
2453 SC/02 Eximstats: use FileHandles for outputing results.
2455 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2457 SC/04 Eximstats: fixed display of large numbers with -nvr option
2459 SC/05 Eximstats: fixed merging of reports with empty tables.
2461 SC/06 Eximstats: added the -include_original_destination flag
2463 SC/07 Eximstats: removed tabs and trailing whitespace.
2465 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2467 TK/06 MBOX spool code: Add real "From " MBOX separator line
2468 so the .eml file is really in mbox format (even though
2469 most programs do not really care). Patch from Alex Miller.
2471 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2472 The latter is generated from $received_to and is only set if the
2473 message has one envelope recipient. SA can use these headers,
2474 obviously out-of-the-box. Patch from Alex Miller.
2476 PH/08 The ${def test on a variable was returning false if the variable's
2477 value was "0", contrary to what the specification has always said!
2478 The result should be true unless the variable is empty.
2480 PH/09 The syntax error of a character other than { following "${if
2481 def:variable_name" (after optional whitespace) was not being diagnosed.
2482 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2483 accidental colon was present, for example, could give incorrect results.
2485 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2486 result is used (not including appendfile, where other changes are about
2489 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2490 This involved changing a lot of size variables from int to off_t. It
2491 should work with maildirs and everything.
2493 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2494 spamd dying while we are connected to it.
2496 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2499 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2500 function for BDB 4.3.
2502 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2504 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2505 the sources for PCRE have been re-arranged and more files are now
2508 PH/15 The code I had for printing potentially long long variables in PH/11
2509 above was not the best (it lost precision). The length of off_t variables
2510 is now inspected at build time, and an appropriate printing format (%ld
2511 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2512 to be "long long int" or "long int". This is needed for the internal
2513 formatting function string_vformat().
2515 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2516 the configuration file to be ":syslog", then the script "guesses" where
2517 the logs files are, rather than using the compiled in default. In our
2518 case the guess is not the same as the compiled default, so the script
2519 suddenly stopped working when I started to use syslog. The patch checks
2520 to see if log_file_path is "". If so, it attempts to read it from exim
2521 with no configuration file to get the compiled in version, before it
2522 falls back to the previous guessing code."
2524 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2525 implementing BATV in an Exim configuration. See NewStuff for the gory
2528 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2529 Makefile that are specific to HP-UX.
2531 PH/18 If the "use_postmaster" option was set for a recipient callout together
2532 with the "random" option, the postmaster address was used as the MAIL
2533 FROM address for the random test, but not for the subsequent recipient
2534 test. It is now used for both.
2536 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2537 patch removes a few documentation additions to RFC 3028, because the
2538 latest draft now contains them. It adds the new en;ascii-case comparator
2539 and a new error check for 8bit text in MIME parts. Comparator and
2540 require names are now matched exactly. I enabled the subaddress
2541 extension, but it is not well tested yet (read: it works for me)."
2543 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2544 rework some of the code of TK/09 above to avoid the hardwired use of
2545 "%lld" and "long long". Replaced the call to snprintf() with a call to
2548 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2549 records point to non-existent hosts", "retry timeout exceeded", and
2550 "retry time not reached for any host after a long failure period".
2552 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2553 experimental DomainKeys support:
2555 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2556 (2) On an error such as an illegally used "control", the wrong name for
2557 the control was given.
2559 These problems did NOT occur unless DomainKeys support was compiled.
2561 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2563 PH/24 Added ${if match_ip condition.
2565 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2566 databases so that it will be absolutely obvious if a crash occurs in the
2567 DB library. This is a regular occurrence (often caused by mis-matched
2570 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2571 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2572 picked up on a user's system that detects such things. There doesn't seem
2573 to be a gcc warning option for this - only an attribute that has to be
2574 put on the function's prototype. It seems that in Fedora Core 4 they have
2575 set this on a number of new functions. No doubt there will be more in due
2578 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2579 specify a transport. However, if an address that was verified by such a
2580 router was the subject of a callout, Exim crashed because it tried to
2581 read the rcpt_include_affixes from the non-existent transport. Now it
2582 just assumes that the setting of that option is false. This bug was
2583 introduced by 4.51/PH/31.
2585 PH/28 Changed -d+all to exclude +memory, because that information is very
2586 rarely of interest, but it makes the output a lot bigger. People tend to
2587 do -d+all out of habit.
2589 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2590 code in os-type was giving problems when libc.so lives in lib64, like on
2593 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2594 aren't the modern standard, and it seems that some systems' include files
2595 don't always have them. Exim was already checking for some of the newer
2596 ones like T_AAAA, and defining it itself. I've added checks for all the
2597 record types that Exim uses.
2599 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2600 not automatically generating a new one, as it is supposed to. This
2601 prevented TLS from working. If the file did exist, but contained invalid
2602 data, a new version was generated, as expected. It was only the case of a
2603 non-existent file that was broken.
2605 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2606 with a change in libdomainkeys > 0.64.
2608 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2609 from DNS. If the selector record carries the flag, it now has
2610 precedence over the domain-wide flag.
2612 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2614 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2615 the use of a port name that isn't defined in /etc/services) occurred, the
2616 message was deferred as in a normal delivery, and thus remained on the
2617 spool, instead of being failed because of the mua_wrapper setting. This
2618 is now fixed, and I tidied up some of the mua_wrapper messages at the
2621 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2622 the messages in a hash of arrays rather than using individual hashes.
2623 This is a bit cleaner and results in dramatic memory savings, albeit
2624 at a slight CPU cost.
2626 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2627 as requested by Marc Sherman.
2629 SC/10 Eximstats: added histograms for user specified patterns as requested
2632 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2634 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2635 fopen() in the content-scanning modules that did not already have it.
2641 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2642 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2644 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2646 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2648 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2649 to test Sieve filters that use "vacation".
2651 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2652 that changes the way the GnuTLS parameters are stored in the cache file.
2653 The new format can be generated externally. For backward compatibility,
2654 if the data in the cache doesn't make sense, Exim assumes it has read an
2655 old-format file, and it generates new data and writes a new file. This
2656 means that you can't go back to an older release without removing the
2659 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2660 work if there are any delivery delays because "one_time" forces the
2661 parent to be marked "delivered", so its unseen clone is never tried
2662 again. For this reason, Exim now forbids the simultaneous setting of
2665 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2666 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2667 entry if you want to know the details. The fix, however, neglected to
2668 consider the case when local delivery batching is involved. The test for
2669 "previously delivered" was not happening when checking to see if an
2670 address could be batched with a previous (undelivered) one; under
2671 certain circumstances this could lead to multiple deliveries to the same
2674 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2675 in its include files, and this causes problems building Exim.
2677 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2678 header_syntax) but Exim was just ignoring anything given after a slash.
2679 In particular, this caused confusion with an attempt to use "verify =
2680 reverse_host_lookup/defer_ok". An error is now given when options are
2681 supplied for verify items that do not have them. (Maybe reverse_host_
2682 lookup should have a defer_ok option, but that's a different point.)
2684 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2685 defined by RFC 821) to 2048, because there were problems with some AUTH
2686 commands, and RFC 1869 says the size should be increased for extended
2687 SMTP commands that take arguments.
2689 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2692 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2693 "unknown" error; now it says that the functionality isn't in the binary.
2695 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2696 an address' error message when a string expansion fails (syntax or
2697 whatever). Otherwise the password may appear in the log. Following change
2698 PH/42 below, there is no longer a chance of it appearing in a bounce
2701 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2703 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2704 of its addresses was ever tried. (Bugzilla bug #2.)
2706 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2707 the result incorrectly in the debug output. (It correctly added a newline
2708 to what was transported.)
2710 TF/01 Added $received_time.
2712 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2713 commented out examples of how to interface to a virus scanner and to
2714 SpamAssassin. Also added commented examples of av_scanner and
2715 spamd_address settings.
2717 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2718 and controls are allowed in which ACLs. There were a couple of minor
2719 errors. Some of the entries in the conditions table (which is a table of
2720 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2721 negation of where the condition IS allowed.
2723 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2725 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2726 header file does not have a version number, so I've had to invent a new
2727 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2728 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2729 radiusclient), but it was contributed by a Radius user.
2731 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2732 files or directories, for OpenSSL.
2734 PH/19 When an Exim process that is running as root has to create an Exim log
2735 file, it does so in a subprocess that runs as exim:exim so as to get the
2736 ownership right at creation (otherwise, other Exim processes might see
2737 the file with the wrong ownership). There was no test for failure of this
2738 fork() call, which would lead to the process getting stuck as it waited
2739 for a non-existent subprocess. Forks do occasionally fail when resources
2740 run out. I reviewed all the other calls to fork(); they all seem to check
2743 PH/20 When checking for unexpected SMTP input at connect time (before writing
2744 the banner), Exim was not dealing correctly with a non-positive return
2745 from the read() function. If the client had disconnected by this time,
2746 the result was a log entry for a synchronization error with an empty
2747 string after "input=" when read() returned zero. If read() returned -1
2748 (an event I could not check), uninitialized data bytes were printed.
2749 There were reports of junk text (parts of files, etc) appearing after
2752 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2754 PH/22 Added support for macro redefinition, and (re)definition in between
2755 driver and ACL definitions.
2757 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2758 forgetting to use the resulting value; it was using the unexpanded value.
2760 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2761 hadn't been configured. The fix is from Juergen Kreileder, who
2762 understands it better than I do:
2764 "Here's what I see happening with three configured cyrus_sasl
2765 authenticators configured (plain, login, cram-md5):
2767 On startup auth_cyrus_sasl_init() gets called for each of these.
2768 This means three calls to sasl_listmech() without a specified mech_list.
2769 => SASL tests which mechs of all available mechs actually work
2770 => three warnings about OTP not working
2771 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2773 With the patch, sasl_listmech() also gets called three times. But now
2774 SASL's mech_list option is set to the server_mech specified in the the
2775 authenticator. Or in other words, the answer from sasl_listmech()
2776 gets limited to just the mech you're testing for (which is different
2778 => the return list contains just 'plain' or 'login', 'cram-md5' or
2779 nothing depending on the value of ob->server_mech.
2781 I've just tested the patch: Authentication still works fine,
2782 unavailable mechs specified in the exim configuration are still
2783 caught, and the auth.log warnings about OTP are gone."
2785 PH/25 When debugging is enabled, the contents of the command line are added
2786 to the debugging output, even when log_selector=+arguments is not
2789 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2790 answer is "GNU", and only if the return is "GNU/something" is the answer
2793 PH/27 $acl_verify_message is now set immediately after the failure of a
2794 verification in an ACL, and so is available in subsequent modifiers. In
2795 particular, the message can be preserved by coding like this:
2797 warn !verify = sender
2798 set acl_m0 = $acl_verify_message
2800 Previously, $acl_verify_message was set only while expanding "message"
2801 and "log_message" when a very denied access.
2803 PH/28 Modified OS/os.c-Linux with
2805 -#ifndef OS_LOAD_AVERAGE
2806 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2808 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2809 nomenclature these days.)
2811 PH/29 Installed patch from the Sieve maintainer that adds the options
2812 sieve_useraddress and sieve_subaddress to the redirect router.
2814 PH/30 In these circumstances:
2815 . Two addresses routed to the same list of hosts;
2816 . First host does not offer TLS;
2817 . First host accepts first address;
2818 . First host gives temporary error to second address;
2819 . Second host offers TLS and a TLS session is established;
2820 . Second host accepts second address.
2821 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2822 and peerdn, if requested) that were in fact used only for the second
2825 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2826 attention to any local part prefix or suffix that was matched by the
2827 router that accepted the address. It now behaves in the same way as it
2828 does for delivery: the affixes are removed from the local part unless
2829 rcpt_include_affixes is set on the transport.
2831 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2832 timeout during the DATA phase of an incoming message.
2834 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2835 applied a patch sanctioned by the Sieve maintainer.
2837 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2838 the uid or gid is negative. A case of a negative gid caused this to be
2839 noticed. The fix allows for either to be negative.
2841 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2842 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2845 PH/36 The change PH/12 above was broken. Fixed it.
2847 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2848 the grounds that routing the same address twice would always produce the
2849 same answer. This might have been true once, but it is certainly no
2850 longer true now. Routing a child address may depend on the previous
2851 routing that produced that child. Some complicated redirection strategies
2852 went wrong when messages had multiple recipients, and made Exim's
2853 behaviour dependent on the order in which the addresses were given.
2855 I have moved the duplicate checking until after the routing is complete.
2856 Exim scans the addresses that are assigned to local and remote
2857 transports, and removes any duplicates. This means that more work will be
2858 done, as duplicates will always all be routed, but duplicates are
2859 presumably rare, so I don't expect this is of any significance.
2861 For deliveries to pipes, files, and autoreplies, the duplicate checking
2862 still happens during the routing process, since they are not going to be
2865 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2866 It corrects a timeout issue with spamd. This is Ian's comment: "The
2867 background is that sometimes spamd either never reads data from a
2868 connection it has accepted, or it never writes response data. The exiscan
2869 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2870 blindly assumes that writes won't block so it may never time out."
2872 PH/39 Allow G after quota size as well as K and M.
2874 PH/40 The value set for $authenticated_id in an authenticator may not contain
2875 binary zeroes or newlines because the value is written to log lines and
2876 to spool files. There was no check on this. Now the value is run through
2877 the string_printing() function so that such characters are converted to
2878 printable escape sequences.
2880 PH/41 $message_linecount is a new variable that contains the total number of
2881 lines in the message. Compare $body_linecount, which is the count for the
2884 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2885 bounce and delay warning messages, except in certain special cases, which
2888 (a) An SMTP error message from a remote host;
2889 (b) A message specified in a :fail: redirection;
2890 (c) A message specified in a "fail" command in a system filter;
2891 (d) A message specified in a FAIL return from the queryprogram router;
2892 (e) A message specified by the cannot_route_message router option.
2894 In these cases only, Exim does include the error details in bounce and
2895 warning messages. There are also a few cases where bland messages such
2896 as "unrouteable address" or "local delivery error" are given.
2898 PH/43 $value is now also set for the "else" part of a ${run expansion.
2900 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2901 being worked on, but at least Exim now implements the latest version to
2904 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2905 process to complete was treated as a delivery failure, a timeout while
2906 writing the message to the pipe was logged, but erroneously treated as a
2907 successful delivery. Such timeouts include transport filter timeouts. For
2908 consistency with the overall process timeout, these timeouts are now
2909 treated as errors, giving rise to delivery failures by default. However,
2910 there is now a new Boolean option for the pipe transport called
2911 timeout_defer, which, if set TRUE, converts the failures into defers for
2912 both kinds of timeout. A transport filter timeout is now identified in
2915 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2916 systems where "make" and "gmake" are different, calling "gmake" at top
2917 level broke things. I've arranged for the value of $(MAKE) to be passed
2918 from the Makefile to this script so that it can call the same version of
2922 A note about Exim versions 4.44 and 4.50
2923 ----------------------------------------
2925 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2926 changes of various kinds. As a consequence, a big documentation update was
2927 needed. This delayed the release for rather longer than seemed good, especially
2928 in the light of a couple of (minor) security issues. Therefore, the changes
2929 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2930 release. So 4.44 and 4.50 are in effect two different branches that both start
2933 I have left the 4.50 change log unchanged; it contains all the changes since
2934 4.43. The change log for 4.44 is below; many of its items are identical to
2935 those for 4.50. This seems to be the most sensible way to preserve the
2936 historical information.
2942 1. Minor wording change to the doc/README.SIEVE file.
2944 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2945 computation of the current number of files was incorrect.
2947 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2948 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2951 4. Give more explanation in the error message when the command for a transport
2952 filter fails to execute.
2954 5. There are several places where Exim runs a non-Exim command in a
2955 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2956 was being done only for the command run by the queryprogram router. It is
2957 now done for all such subprocesses. The other cases are: ${run, transport
2958 filters, and the commands run by the lmtp and pipe transports.
2960 6. Added CONFIGURE_GROUP build-time option.
2962 7. Some older OS have a limit of 256 on the maximum number of file
2963 descriptors. Exim was using setrlimit() to set 1000 as a large value
2964 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2965 systems. I've change it so that if it can't get 1000, it tries for 256.
2967 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2968 was an oversight, and furthermore, ever since the addition of extra
2969 controls (e.g. 4.43/32), the checks on when to allow different forms of
2970 "control" were broken. There should now be diagnostics for all cases when a
2971 control that does not make sense is encountered.
2973 9. Added the /retain_sender option to "control=submission".
2975 10. $recipients is now available in the predata ACL (oversight).
2977 11. Tidy the search cache before the fork to do a delivery from a message
2978 received from the command line. Otherwise the child will trigger a lookup
2979 failure and thereby defer the delivery if it tries to use (for example) a
2980 cached ldap connection that the parent has called unbind on.
2982 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2983 of $address_data from the recipient verification was clobbered by the
2984 sender verification.
2986 13. The value of address_data from a sender verification is now available in
2987 $sender_address_data in subsequent conditions in the ACL statement.
2989 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2991 15. Added a new option "connect=<time>" to callout options, to set a different
2994 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2995 was its contents. (It was OK if the option was not defined at all.)
2997 17. A "Completed" log line is now written for messages that are removed from
2998 the spool by the -Mrm option.
3000 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3001 information about exactly what failed.
3003 19. Added -dd to debug only the daemon process.
3005 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3006 handles timeouts, both on the server side and network timeouts. Renamed the
3007 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3009 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3010 It is now set to "smtps".
3012 22. $host_address is now set to the target address during the checking of
3013 ignore_target_hosts.
3015 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3016 being passed; this would have caused $sender_host_name to have been used if
3017 matching the list had actually called for a host name (not very likely,
3018 since this list is usually IP addresses). A host name is now passed as
3021 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3022 code that specifies a non-restarting handler (typically sigaction() in
3023 modern systems) in an attempt to fix a rare and obscure crash bug.
3025 25. Narrowed the window for a race in the daemon that could cause it to ignore
3026 SIGCHLD signals. This is not a major problem, because they are used only to
3027 wake it up if nothing else does.
3029 26. A malformed maildirsize file could cause Exim to calculate negative values
3030 for the mailbox size or file count. Odd effects could occur as a result.
3031 The maildirsize information is now recalculated if the size or filecount
3034 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3035 support for a long time. Removed HAVE_SYS_VFS_H.
3037 28. Installed the latest version of exipick from John Jetmore.
3039 29. In an address list, if the pattern was not a regular expression, an empty
3040 subject address (from a bounce message) matched only if the pattern was an
3041 empty string. Non-empty patterns were not even tested. This was the wrong
3042 because it is perfectly reasonable to use an empty address as part of a
3043 database query. An empty address is now tested by patterns that are
3044 lookups. However, all the other forms of pattern expect the subject to
3045 contain a local part and a domain, and therefore, for them, an empty
3046 address still always fails if the pattern is not itself empty.
3048 30. Exim went into a mad DNS loop when attempting to do a callout where the
3049 host was specified on an smtp transport, and looking it up yielded more
3050 than one IP address.
3052 31. Re-factored the code for checking spool and log partition space into a
3053 function that finds that data and another that does the check. The former
3054 is then used to implement four new variables: $spool_space, $log_space,
3055 $spool_inodes, and $log_inodes.
3057 32. The RFC2047 encoding function was originally intended for short strings
3058 such as real names; it was not keeping to the 75-character limit for
3059 encoded words that the RFC imposes. It now respects the limit, and
3060 generates multiple encoded words if necessary. To be on the safe side, I
3061 have increased the buffer size for the ${rfc2047: expansion operator from
3064 33. It is now permitted to omit both strings after an "if" condition; if the
3065 condition is true, the result is "true". As before, when the second string
3066 is omitted, a false condition yields an empty string. This makes it less
3067 cumbersome to write custom ACL and router conditions.
3069 34. Failure to deliver a bounce message always caused it to be frozen, even if
3070 there was an errors_to setting on the router. The errors_to setting is now
3073 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3074 canonical form (fully expanded) before being placed in
3075 $sender_host_address.
3077 36. The table in the code that translates DNS record types into text (T_A to
3078 "A" for instance) was missing entries for NS and CNAME. It is just possible
3079 that this could have caused confusion if both these types were looked up
3080 for the same domain, because the text type is used as part of Exim's
3081 per-process caching. But the chance of anyone hitting this buglet seems
3084 37. The dnsdb lookup has been extended in a number of ways.
3086 (1) There is a new type, "zns", which walks up the domain tree until it
3087 finds some nameserver records. It should be used with care.
3089 (2) There is a new type, "mxh", which is like "mx" except that it returns
3090 just the host names, not the priorities.
3092 (3) It is now possible to give a list of domains (or IP addresses) to be
3093 looked up. The behaviour when one of the lookups defers can be
3094 controlled by a keyword.
3096 (4) It is now possible to specify the separator character for use when
3097 multiple records are returned.
3099 38. The dnslists ACL condition has been extended: it is now possible to supply
3100 a list of IP addresses and/or domains to be looked up in a particular DNS
3103 39. Added log_selector=+queue_time_overall.
3105 40. When running the queue in the test harness, wait just a tad after forking a
3106 delivery process, to get repeatability of debugging output.
3108 41. Include certificate and key file names in error message when GnuTLS fails
3109 to set them up, because the GnuTLS error message doesn't include the name
3110 of the failing file when there is a problem reading it.
3112 42. Allow both -bf and -bF in the same test run.
3114 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3116 44. The "Exiscan patch" is now merged into the mainline Exim source.
3118 45. Sometimes the final signoff response after QUIT could fail to get
3119 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3120 < 0 before doing a fflush(). This bug looks as though it goes back to the
3121 introduction of TLS in release 3.20, but "sometimes" must have been rare
3122 because the tests only now provoked it.
3124 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3125 (this can affect the format of dates).
3127 47. exim_tidydb, when checking for the continued existence of a message for
3128 which it has found a message-specific retry record, was not finding
3129 messages that were in split spool directories. Consequently, it was
3130 deleting retry records that should have stayed in existence.
3132 48. Steve fixed some bugs in eximstats.
3134 49. The SPA authentication driver was not abandoning authentication and moving
3135 on to the next authenticator when an expansion was forced to fail,
3136 contradicting the general specification for all authenticators. Instead it
3137 was generating a temporary error. It now behaves as specified.
3139 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3140 (the order specifies the preference for clients). The order is now AES256,
3141 AES128, 3DES, ARCFOUR128.
3143 51. Small patch to Sieve code - explicitly set From: when generating an
3146 52. Exim crashed if a remote delivery caused a very long error message to be
3147 recorded - for instance if somebody sent an entire SpamAssassin report back
3148 as a large number of 550 error lines. This bug was coincidentally fixed by
3149 increasing the size of one of Exim's internal buffers (big_buffer) that
3150 happened as part of the Exiscan merge. However, to be on the safe side, I
3151 have made the code more robust (and fixed the comments that describe what
3154 53. Now that there can be additional text after "Completed" in log lines (if
3155 the queue_time_overall log selector is set), a one-byte patch to exigrep
3156 was needed to allow it to recognize "Completed" as not the last thing in
3159 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3160 patch that reportedly fixes this has been added. I am not expert enough to
3161 create a test for it. This is what the patch creator wrote:
3163 "I found a little strange behaviour of ldap code when working with
3164 Windows 2003 AD Domain, where users was placed in more than one
3165 Organization Units. When I tried to give exim partial DN, the exit code
3166 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3167 But simultaneously result of request was absolutely normal ldap result,
3168 so I produce this patch..."
3170 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3171 REFERENCE, so I have modified the code to exclude the patch when that macro
3174 55. Some experimental protocols are using DNS PTR records for new purposes. The
3175 keys for these records are domain names, not reversed IP addresses. The
3176 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3177 leaves it alone. Component reversal etc. now happens only for IP addresses.
3180 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3182 57. Double the size of the debug message buffer (to 2048) so that more of very
3183 long debug lines gets shown.
3185 58. The exicyclog utility now does better if the number of log files to keep
3186 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3188 59. Two changes related to the smtp_active_hostname option:
3190 (1) $smtp_active_hostname is now available as a variable.
3191 (2) The default for smtp_banner uses $smtp_active_hostname instead
3192 of $primary_hostname.
3194 60. The host_aton() function is supposed to be passed a string that is known
3195 to be a valid IP address. However, in the case of IPv6 addresses, it was
3196 not checking this. This is a hostage to fortune. Exim now panics and dies
3197 if the condition is not met. A case was found where this could be provoked
3198 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3199 components; fortuitously, this particular loophole had already been fixed
3200 by change 4.50/55 above.
3202 If there are any other similar loopholes, the new check in host_aton()
3203 itself should stop them being exploited. The report I received stated that
3204 data on the command line could provoke the exploit when Exim was running as
3205 exim, but did not say which command line option was involved. All I could
3206 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3207 running as the user.
3210 61. There was a buffer overflow vulnerability in the SPA authentication code
3211 (which came originally from the Samba project). I have added a test to the
3212 spa_base64_to_bits() function which I hope fixes it.
3215 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3216 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3218 63. The daemon start-up calls getloadavg() while still root for those OS that
3219 need the first call to be done as root, but it missed one case: when
3220 deliver_queue_load_max is set with deliver_drop_privilege. This is
3221 necessary for the benefit of the queue runner, because there is no re-exec
3222 when deliver_drop_privilege is set.
3224 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3225 This has been fixed.
3227 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3228 was in use, was not putting the data itself into the right store pool;
3229 consequently, it could be overwritten for a subsequent message in the same
3230 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3233 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3235 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3236 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3237 the function were treating the return as a boolean value, which happened to
3238 work because 0=false and not-0=true, but is not correct code.
3240 68. The host_aton() function was not handling scoped IPv6 addresses (those
3241 with, for example, "%eth0" on the end) correctly.
3243 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3244 negated items (that is, ~something) in unsigned ints. Some compilers
3245 apparently mutter when there is no cast.
3247 70. If an address verification called from an ACL failed, and did not produce a
3248 user-specific message (i.e. there was only a "system" message), nothing was
3249 put in $acl_verify_message. In this situation, it now puts the system
3252 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3253 session; change 4.31/43 added the unwanted input to the log line - except
3254 that it did not do this in the start of session case. It now does.
3256 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3257 This is wrong and can cause the other end to generate a synchronization
3258 error if it is another Exim or anything else that does the synchronization
3259 check. A QUIT command is no longer sent after a timeout.
3261 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3262 during host lookups.
3264 74. The defer_ok option of callout verification was not working if it was used
3265 when verifying addresses in header lines, that is, for this case:
3267 verify = header_sender/callout=defer_ok
3269 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3270 those file descriptors could be used for SMTP connections. If anything
3271 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3272 could be sent to the SMTP client, causing chaos. The daemon now opens
3273 stdin, stdout, and stderr to /dev/null when it puts itself into the
3276 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3277 log by default. The user can override this with suitable Perl magic.
3279 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3280 the log message when discard triggers, was not working for the DATA ACL or
3281 for the non-SMTP ACL.
3283 78. Error message wording change in sieve.c.
3285 79. If smtp_accept_max_per_host was set, the number of connections could be
3286 restricted to fewer than expected, because the daemon was trying to set up
3287 a new connection before checking whether the processes handling previous
3288 connections had finished. The check for completed processes is now done
3289 earlier. On busy systems, this bug wouldn't be noticed because something
3290 else would have woken the daemon, and it would have reaped the completed
3293 80. If a message was submitted locally by a user whose login name contained one
3294 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3295 It caused a spool format error. I have fixed the spool reading code. A
3296 related problem was that the "from" clause in the Received: line became
3297 illegal because of the space(s). It is now covered by ${quote_local_part.
3299 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3302 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3304 83. Patch from Sieve maintainer to fix unterminated string problem in
3305 "vacation" handling.
3307 84. Some minor changes to the Linux configuration files to help with other
3308 OS variants using glibc.
3310 85. One more patch for Sieve to update vacation handling to latest spec.
3313 ----------------------------------------------------
3314 See the note above about the 4.44 and 4.50 releases.
3315 ----------------------------------------------------
3321 1. Change 4.43/35 introduced a bug that caused file counts to be
3322 incorrectly computed when quota_filecount was set in an appendfile
3325 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3326 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3329 3. Give more explanation in the error message when the command for a transport
3330 filter fails to execute.
3332 4. There are several places where Exim runs a non-Exim command in a
3333 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3334 was being done only for the command run by the queryprogram router. It is
3335 now done for all such subprocesses. The other cases are: ${run, transport
3336 filters, and the commands run by the lmtp and pipe transports.
3338 5. Some older OS have a limit of 256 on the maximum number of file
3339 descriptors. Exim was using setrlimit() to set 1000 as a large value
3340 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3341 systems. I've change it so that if it can't get 1000, it tries for 256.
3343 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3344 was an oversight, and furthermore, ever since the addition of extra
3345 controls (e.g. 4.43/32), the checks on when to allow different forms of
3346 "control" were broken. There should now be diagnostics for all cases when a
3347 control that does not make sense is encountered.
3349 7. $recipients is now available in the predata ACL (oversight).
3351 8. Tidy the search cache before the fork to do a delivery from a message
3352 received from the command line. Otherwise the child will trigger a lookup
3353 failure and thereby defer the delivery if it tries to use (for example) a
3354 cached ldap connection that the parent has called unbind on.
3356 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3357 of $address_data from the recipient verification was clobbered by the
3358 sender verification.
3360 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3361 was its contents. (It was OK if the option was not defined at all.)
3363 11. A "Completed" log line is now written for messages that are removed from
3364 the spool by the -Mrm option.
3366 12. $host_address is now set to the target address during the checking of
3367 ignore_target_hosts.
3369 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3370 being passed; this would have caused $sender_host_name to have been used if
3371 matching the list had actually called for a host name (not very likely,
3372 since this list is usually IP addresses). A host name is now passed as
3375 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3376 code that specifies a non-restarting handler (typically sigaction() in
3377 modern systems) in an attempt to fix a rare and obscure crash bug.
3379 15. Narrowed the window for a race in the daemon that could cause it to ignore
3380 SIGCHLD signals. This is not a major problem, because they are used only to
3381 wake it up if nothing else does.
3383 16. A malformed maildirsize file could cause Exim to calculate negative values
3384 for the mailbox size or file count. Odd effects could occur as a result.
3385 The maildirsize information is now recalculated if the size or filecount
3388 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3389 support for a long time. Removed HAVE_SYS_VFS_H.
3391 18. Updated exipick to current release from John Jetmore.
3393 19. Allow an empty sender to be matched against a lookup in an address list.
3394 Previously the only cases considered were a regular expression, or an
3397 20. Exim went into a mad DNS lookup loop when doing a callout where the
3398 host was specified on the transport, if the DNS lookup yielded more than
3401 21. The RFC2047 encoding function was originally intended for short strings
3402 such as real names; it was not keeping to the 75-character limit for
3403 encoded words that the RFC imposes. It now respects the limit, and
3404 generates multiple encoded words if necessary. To be on the safe side, I
3405 have increased the buffer size for the ${rfc2047: expansion operator from
3408 22. Failure to deliver a bounce message always caused it to be frozen, even if
3409 there was an errors_to setting on the router. The errors_to setting is now
3412 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3413 canonical form (fully expanded) before being placed in
3414 $sender_host_address.
3416 24. Updated eximstats to version 1.33
3418 25. Include certificate and key file names in error message when GnuTLS fails
3419 to set them up, because the GnuTLS error message doesn't include the name
3420 of the failing file when there is a problem reading it.
3422 26. Expand error message when OpenSSL has problems setting up cert/key files.
3425 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3426 (this can affect the format of dates).
3428 28. exim_tidydb, when checking for the continued existence of a message for
3429 which it has found a message-specific retry record, was not finding
3430 messages that were in split spool directories. Consequently, it was
3431 deleting retry records that should have stayed in existence.
3433 29. eximstats updated to version 1.35
3434 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3435 1.35 - bugfix such that pie charts by volume are generated correctly
3437 30. The SPA authentication driver was not abandoning authentication and moving
3438 on to the next authenticator when an expansion was forced to fail,
3439 contradicting the general specification for all authenticators. Instead it
3440 was generating a temporary error. It now behaves as specified.
3442 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3443 (the order specifies the preference for clients). The order is now AES256,
3444 AES128, 3DES, ARCFOUR128.
3446 31. Small patch to Sieve code - explicitly set From: when generating an
3449 32. Exim crashed if a remote delivery caused a very long error message to be
3450 recorded - for instance if somebody sent an entire SpamAssassin report back
3451 as a large number of 550 error lines. This bug was coincidentally fixed by
3452 increasing the size of one of Exim's internal buffers (big_buffer) that
3453 happened as part of the Exiscan merge. However, to be on the safe side, I
3454 have made the code more robust (and fixed the comments that describe what
3457 33. Some experimental protocols are using DNS PTR records for new purposes. The
3458 keys for these records are domain names, not reversed IP addresses. The
3459 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3460 leaves it alone. Component reversal etc. now happens only for IP addresses.
3463 34. The host_aton() function is supposed to be passed a string that is known
3464 to be a valid IP address. However, in the case of IPv6 addresses, it was
3465 not checking this. This is a hostage to fortune. Exim now panics and dies
3466 if the condition is not met. A case was found where this could be provoked
3467 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3468 components; fortuitously, this particular loophole had already been fixed
3469 by change 4.50/55 or 4.44/33 above.
3471 If there are any other similar loopholes, the new check in host_aton()
3472 itself should stop them being exploited. The report I received stated that
3473 data on the command line could provoke the exploit when Exim was running as
3474 exim, but did not say which command line option was involved. All I could
3475 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3476 running as the user.
3479 35. There was a buffer overflow vulnerability in the SPA authentication code
3480 (which came originally from the Samba project). I have added a test to the
3481 spa_base64_to_bits() function which I hope fixes it.
3484 36. The daemon start-up calls getloadavg() while still root for those OS that
3485 need the first call to be done as root, but it missed one case: when
3486 deliver_queue_load_max is set with deliver_drop_privilege. This is
3487 necessary for the benefit of the queue runner, because there is no re-exec
3488 when deliver_drop_privilege is set.
3490 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3491 was in use, was not putting the data itself into the right store pool;
3492 consequently, it could be overwritten for a subsequent message in the same
3493 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3496 38. Sometimes the final signoff response after QUIT could fail to get
3497 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3498 < 0 before doing a fflush(). This bug looks as though it goes back to the
3499 introduction of TLS in release 3.20, but "sometimes" must have been rare
3500 because the tests only now provoked it.
3506 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3507 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3508 Now it returns an integer. A number of calls were still expecting a T/F
3509 return. Fortuitously, in all cases, the tests worked in OK situations,
3510 which is the norm. However, things would have gone wrong on any write
3511 failures on the smtp file descriptor. This function is used when sending
3512 messages over SMTP and also when doing verify callouts.
3514 2. When Exim is called to do synchronous delivery of a locally submitted
3515 message (the -odf or -odi options), it no longer closes stderr before doing
3518 3. Implemented the mua_wrapper option.
3520 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3522 5. Implemented the functions header_remove(), header_testname(),
3523 header_add_at_position(), and receive_remove_recipient(), and exported them
3526 6. If an ACL "warn" statement specified the addition of headers, Exim already
3527 inserted X-ACL-Warn: at the start if there was no header name. However, it
3528 was not making this test for the second and subsequent header lines if
3529 there were newlines in the string. This meant that an invalid header could
3530 be inserted if Exim was badly configured.
3532 7. Allow an ACL "warn" statement to add header lines at the start or after all
3533 the Received: headers, as well as at the end.
3535 8. Added the rcpt_4xx retry error code.
3537 9. Added postmaster_mailfrom=xxx to callout verification option.
3539 10. Added mailfrom=xxxx to the callout verification option, for verify=
3542 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3543 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3545 12. Inserted some casts to stop certain compilers warning when using pointer
3546 differences as field lengths or precisions in printf-type calls (mostly
3547 affecting debugging statements).
3549 13. Added optional readline() support for -be (dynamically loaded).
3551 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3552 same clock tick as a message's arrival, so that its received time was the
3553 same as the "first fail" time on the retry record, and that message
3554 remained on the queue past the ultimate address timeout, every queue runner
3555 would try a delivery (because it was past the ultimate address timeout) but
3556 after another failure, the ultimate address timeout, which should have then
3557 bounced the address, did not kick in. This was a "< instead of <=" error;
3558 in most cases the first failure would have been in the next clock tick
3559 after the received time, and all would be well.
3561 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3562 being recognized when the domain list was tested by the match_domain
3563 condition in an expansion string.
3565 16. Added the ${str2b64: operator.
3567 17. Exim was always calling setrlimit() to set a large limit for the number of
3568 processes, without checking whether the existing limit was already
3569 adequate. (It did check for the limit on file descriptors.) Furthermore,
3570 errors from getrlimit() and setrlimit() were being ignored. Now they are
3571 logged to the main and panic logs, but Exim does carry on, to try to do its
3572 job under whatever limits there are.
3574 18. Imported PCRE 5.0.
3576 19. Trivial typo in log message " temporarily refused connection" (the leading
3579 20. If the log selector return_path_on_delivery was set and an address was
3580 redirected to /dev/null, the delivery process crashed because it assumed
3581 that a return path would always be set for a "successful" delivery. In this
3582 case, the whole delivery is bypassed as an optimization, and therefore no
3585 21. Internal re-arrangement: the function for sending a challenge and reading
3586 a response while authentication was assuming a zero-terminated challenge
3587 string. It's now changed to take a pointer and a length, to allow for
3588 binary data in such strings.
3590 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3592 23. Exim was not respecting finduser_retries when seeking the login of the
3593 uid under which it was called; it was always trying 10 times. (The default
3594 setting of finduser_retries is zero.) Also, it was sleeping after the final
3595 failure, which is pointless.
3597 24. Implemented tls_on_connect_ports.
3599 25. Implemented acl_smtp_predata.
3601 26. If the domain in control=submission is set empty, Exim assumes that the
3602 authenticated id is a complete email address when it generates From: or
3603 Sender: header lines.
3605 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3606 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3607 chown and chgrp in /bin and hostname in /usr/bin.
3609 28. Exim was keeping the "process log" file open after each use, just as it
3610 does for the main log. This opens the possibility of it remaining open for
3611 long periods when the USR1 signal hits a daemon. Occasional processlog
3612 errors were reported, that could have been caused by this. Anyway, it seems
3613 much more sensible not to leave this file open at all, so that is what now
3616 29. The long-running daemon process does not normally write to the log once it
3617 has entered its main loop, and it closes the log before doing so. This is
3618 so that log files can straightforwardly be renamed and moved. However,
3619 there are a couple of unusual error situations where the daemon does write
3620 log entries, and I had neglected to close the log afterwards.
3622 30. The text of an SMTP error response that was received during a remote
3623 delivery was being truncated at 512 bytes. This is too short for some of
3624 the long messages that one sometimes sees. I've increased the limit to
3627 31. It is now possible to make retry rules that apply only when a message has a
3628 specific sender, in particular, an empty sender.
3630 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3631 it possible to be selective about when SMTP synchronization is enforced.
3633 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3635 32. Implemented hosts_connection_nolog.
3637 33. Added an ACL for QUIT.
3639 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3642 35. Added mailbox_size and mailbox_filecount to appendfile.
3644 36. Added control = no_multiline_responses to ACLs.
3646 37. There was a bug in the logic of the code that waits for the clock to tick
3647 in the case where the clock went backwards by a substantial amount such
3648 that the microsecond fraction of "now" was more than the microsecond
3649 fraction of "then" (but the whole seconds number was less).
3651 38. Added support for the libradius Radius client library this is found on
3652 FreeBSD (previously only the radiusclient library was supported).
3658 1. When certain lookups returned multiple values in the form name=value, the
3659 quoting of the values was not always being done properly. Specifically:
3660 (a) If the value started with a double quote, but contained no whitespace,
3662 (b) If the value contained whitespace other than a space character (i.e.
3663 tabs or newlines or carriage returns) it was not quoted.
3664 This fix has been applied to the mysql and pgsql lookups by writing a
3665 separate quoting function and calling it from the lookup code. The fix
3666 should probably also be applied to nisplus, ibase and oracle lookups, but
3667 since I cannot test any of those, I have not disturbed their existing code.
3669 2. A hit in the callout cache for a specific address caused a log line with no
3670 reason for rejecting RCPT. Now it says "Previous (cached) callout
3671 verification failure".
3673 3. There was an off-by-one bug in the queryprogram router. An over-long
3674 return line was truncated at 256 instead of 255 characters, thereby
3675 overflowing its buffer with the terminating zero. As well as fixing this, I
3676 have increased the buffer size to 1024 (and made a note to document this).
3678 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3679 when Exim is waiting for an SMTP response from a remote server, Exim
3680 restarts its select() call on the socket, thereby resetting its timeout.
3681 This is not a problem when such interrupts are rare. Somebody set up a cron
3682 job to run exiwhat every 2 minutes, which is less than the normal select()
3683 timeout (5 or 10 minutes). This meant that the select() timeout never
3684 kicked in because it was always reset. I have fixed this by comparing the
3685 time when an interrupt arrives with the time at the start of the first call
3686 to select(). If more time than the timeout has elapsed, the interrupt is
3687 treated as a timeout.
3689 5. Some internal re-factoring in preparation for the addition of Sieve
3690 extensions (by MH). In particular, the "personal" test is moved to a
3691 separate function, and given an option for scanning Cc: and Bcc: (which is
3692 not set for Exim filters).
3694 6. When Exim created an email address using the login of the caller as the
3695 local part (e.g. when creating a From: or Sender: header line), it was not
3696 quoting the local part when it contained special characters such as @.
3698 7. Installed new OpenBSD configuration files.
3700 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3701 try to make them clearer.
3703 9. Callout options, other than the timeout value, were being ignored when
3704 verifying sender addresses in header lines. For example, when using
3706 verify = header_sender/callout=no_cache
3708 the cache was (incorrectly) being used.
3710 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3711 only the Cygwin environment.
3713 11. When return_path_on_delivery was set as a log selector, if different remote
3714 addresses in the same message used different return paths and parallel
3715 remote delivery occurred, the wrong values would sometimes be logged.
3716 (Whenever a remote delivery process finished, the return path value from
3717 the most recently started remote delivery process was logged.)
3719 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3720 lines when AUTH and/or TLS are in use. This is the "received protocol"
3721 field. Exim used to use "asmtp" for authenticated SMTP, without any
3722 indication (in the protocol name) for TLS use. Now it follows the RFC and
3723 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3724 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3725 names appear in log lines as well as in Received: header lines.
3727 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3728 extensions, and comparison tests, and to fix some bugs.
3730 14. Changes to the "personal" filter test:
3732 (1) The test was buggy in that it was just doing the equivalent of
3733 "contains" tests on header lines. For example, if a user's address was
3734 anne@some.where, the "personal" test would incorrectly be true for
3736 To: susanne@some.where
3738 This test is now done by extracting each address from the header in turn,
3739 and checking the entire address. Other tests that are part of "personal"
3740 are now done using regular expressions (for example, to check local parts
3741 of addresses in From: header lines).
3743 (2) The list of non-personal local parts in From: addresses has been
3744 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3745 taken from the Sieve specification recommendations.
3747 (3) If the message contains any header line starting with "List-" it is
3748 treated as non-personal.
3750 (4) The test for "circular" in the Subject: header line has been removed
3751 because it now seems ill-conceived.
3753 15. Minor typos in src/EDITME comments corrected.
3755 16. Installed latest exipick from John Jetmore.
3757 17. If headers_add on a router specified a text string that was too long for
3758 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3759 of string_sprintf() is now avoided.
3761 18. $message_body_size was not set (it was always zero) when running the DATA
3762 ACL and the local_scan() function.
3764 19. For the "mail" command in an Exim filter, no default was being set for
3765 the once_repeat time, causing a random time value to be used if "once" was
3766 specified. (If the value happened to be <= 0, no repeat happened.) The
3767 default is now 0s, meaning "never repeat". The "vacation" command was OK
3768 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3769 (I found it when inspecting the code).
3771 20. There is now an overall timeout for performing a callout verification. It
3772 defaults to 4 times the callout timeout, which applies to individual SMTP
3773 commands during the callout. The overall timeout applies when there is more
3774 than one host that can be tried. The timeout is checked before trying the
3775 next host. This prevents very long delays if there are a large number of
3776 hosts and all are timing out (e.g. when the network connections are timing
3777 out). The value of the overall timeout can be changed by specifying an
3778 additional sub-option for "callout", called "maxwait". For example:
3780 verify = sender/callout=5s,maxwait=20s
3782 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3783 to the end before writing, but this should make it even safer).
3785 22. Exim was forgetting that it had advertised PIPELINING for the second and
3786 subsequent messages on an SMTP connection. It was also not resetting its
3787 memory on STARTTLS and an internal HELO.
3789 23. When Exim logs an SMTP synchronization error within a session, it now
3790 records whether PIPELINING has been advertised or not.
3792 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3793 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3794 rather than long int.
3796 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3798 26. Added the never_mail option to autoreply.
3804 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3805 crash if the getsockname() call failed; this can happen if a connection is
3806 closed very soon after it is established. The problem was simply in the
3807 order in which certain operations were done, causing Exim to try to write
3808 to the SMTP stream before it had set up the file descriptor. The bug has
3809 been fixed by making things happen in the correct order.
3815 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3816 before the connection was closed, thus losing the rejection response.
3818 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3819 some early Solaris releases, but causes trouble in current releases where
3820 socklen_t is defined.
3822 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3825 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3828 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3829 that releasing the top of it at the end releases what was used for sub-
3830 expansions (unless the block got too big). However, discard this block if
3831 the first thing is a variable or header, so that we can use its block when
3832 it is dynamic (useful for very large $message_headers, for example).
3834 6. Lookups now cache *every* query, not just the most recent. A new, separate
3835 store pool is used for this. It can be recovered when all lookup caches are
3836 flushed. Lookups now release memory at the end of their result strings.
3837 This has involved some general refactoring of the lookup sources.
3839 7. Some code has been added to the store_xxx() functions to reduce the amount
3840 of flapping under certain conditions.
3842 8. log_incoming_interface used to affect only the <= reception log lines. Now
3843 it causes the local interface and port to be added to several more SMTP log
3844 lines, for example "SMTP connection from", and rejection lines.
3846 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3848 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3850 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3851 could be overwritten at the end of the current message (or the start of a
3852 new message if it was set in a HELO ACL). The value is now preserved for
3853 the duration of the SMTP connection.
3855 12. If a transport had a headers_rewrite setting, and a matching header line
3856 contained an unqualified address, that address was qualified, even if it
3857 did not match any rewriting rules. The underlying bug was that the values
3858 of the flags that permit the existence of unqualified sender and recipient
3859 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3860 non-local messages, and by -bnq for local messages) were not being
3861 preserved with the message after it was received.
3863 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3864 "next input=" as part of the text comprising the host identity instead of
3865 the correct text. The code was using the same buffer for two different
3866 strings. However, depending on which order the printing function evaluated
3867 its arguments, the bug did not always show up. Under Linux, for example, my
3868 test suite worked just fine.
3870 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3871 doesn't work with some older versions of Perl. It has been changed to "my",
3872 which in any case is probably the better facility to use.
3874 15. A really picky compiler found some instances of statements for creating
3875 error messages that either had too many or two few arguments for the format
3878 16. The size of the buffer for calls to the DNS resolver has been increased
3879 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3880 for addresses that have a lot of PTR records. This alleviates a problem; it
3881 does not fully solve it.
3883 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3884 buffer now truncates the list and logs the incident, which is the same
3885 action as happens when Exim is looking up a host name and its aliases.
3886 Previously in this situation something unpredictable would happen;
3887 sometimes it was "internal error: store_reset failed".
3889 18. If a server dropped the connection unexpectedly when an Exim client was
3890 using GnuTLS and trying to read a response, the client delivery process
3891 crashed while trying to generate an error log message.
3893 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3894 string, for example:
3896 warn message = H1: something\nH2: something
3898 the text was added as a single header line from Exim's point of view
3899 though it ended up OK in the delivered message. However, searching for the
3900 second and subsequent header lines using $h_h2: did not work. This has been
3901 fixed. Similarly, if a system filter added multiple headers in this way,
3902 the routers could not see them.
3904 20. Expanded the error message when iplsearch is called with an invalid key to
3905 suggest using net-iplsearch in a host list.
3907 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3908 ACLs are no longer actually imposed (and a message to that effect is
3911 22. If a "gecos" field in a passwd entry contained escaped characters, in
3912 particular, if it contained a \" sequence, Exim got it wrong when building
3913 a From: or a Sender: header from that name. A second bug also caused
3914 incorrect handling when an unquoted " was present following a character
3915 that needed quoting.
3917 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3918 was not being matched caselessly.
3920 24. Arranged for all hyphens in the exim.8 source to be escaped with
3923 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3924 properly. Recipient callout cache records were still being keyed to include
3925 the sender, even when use_sender was set false. This led to far more
3926 callouts that were necessary. The sender is no longer included in the key
3927 when use_sender is false.
3929 26. Added "control = submission" modifier to ACLs.
3931 27. Added the ${base62d: operator to decode base 62 numbers.
3933 28. dnsdb lookups can now access SRV records.
3935 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3936 the configuration file.
3938 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3939 (-v) mode. This makes the output for a verbose queue run more intelligible.
3941 31. Added a use_postmaster feature to recipient callouts.
3943 32. Added the $body_zerocount variable, containing the number of binary zero
3944 bytes in the message body.
3946 33. The time of last modification of the "new" subdirectory is now used as the
3947 "mailbox time last read" when there is a quota error for a maildir
3950 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3952 35. Added +ignore_unknown as a special item in host lists.
3954 36. Code for decoding IPv6 addresses in host lists is now included, even if
3955 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3956 address was recognized as an IP address, but was then not correctly decoded
3957 into binary, causing unexpected and incorrect effects when compared with
3964 1. Very minor rewording of debugging text in manualroute to say "list of
3965 hosts" instead of "hostlist".
3967 2. If verify=header_syntax was set, and a header line with an unqualified
3968 address (no domain) and a large number of spaces between the end of the
3969 name and the colon was received, the reception process suffered a buffer
3970 overflow, and (when I tested it) crashed. This was caused by some obsolete
3971 code that should have been removed. The fix is to remove it!
3973 3. When running in the test harness, delay a bit after writing a bounce
3974 message to get a bit more predictability in the log output.
3976 4. Added a call to search_tidyup() just before forking a reception process. In
3977 theory, someone could use a lookup in the expansion of smtp_accept_max_
3978 per_host which, without the tidyup, could leave open a database connection.
3980 5. Added the variables $recipient_data and $sender_data which get set from a
3981 lookup success in an ACL "recipients" or "senders" condition, or a router
3982 "senders" option, similar to $domain_data and $local_part_data.
3984 6. Moved the writing of debug_print from before to after the "senders" test
3987 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3988 problems for message scanning, either using a data ACL, or using
3989 local_scan() because the Received: header was not generated till after they
3990 were called (in order to set the time as the time of reception completion).
3991 I have revised the way this works. The header is now generated after the
3992 body is received, but before the ACL or local_scan() are called. After they
3993 are run, the timestamp in the header is updated.
3999 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4000 before starting a queue runner without re-exec. This happened only when
4001 deliver_drop_privilege was set or when the Exim user was set to root. The
4002 effect of the bug was that timeouts during subsequent deliveries caused
4003 crashes instead of being properly handled. The handler is now left at its
4004 default (and expected) setting.
4006 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4007 message, again when deliver_drop_privilege is set or Exim is run as root.
4008 The bug described in (1) was not present in this case, but the tidying up
4009 of the other signals was missing. I have made the two cases consistent.
4011 3. The ignore_target_hosts setting on a manualroute router was being ignored
4012 for hosts that were looked up using the /MX notation.
4014 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4017 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4018 operated on the sender address. After changing the $sender_address to <>
4019 for the sender address verify, Exim was re-instated it as the original
4020 (before rewriting) address, but remembering that it had rewritten it, so it
4021 wasn't rewriting it again. This bug also had the effect of breaking the
4022 sender address verification caching when the sender address was rewritten.
4024 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4025 This has been changed so that if the ip literal address matches
4026 ignore_target_hosts, the router declines.
4028 7. Added expansion conditions match_domain, match_address, and match_local_
4029 part (NOT match_host).
4031 8. The placeholder for the Received: header didn't have a length field set.
4033 9. Added code to Exim itself and to exim_lock to test for a specific race
4034 condition that could lead to file corruption when using MBX delivery. The
4035 issue is with the lockfile that is created in /tmp. If this file is removed
4036 after a process has opened it but before that process has acquired a lock,
4037 there is the potential for a second process to recreate the file and also
4038 acquire a lock. This could lead to two Exim processes writing to the file
4039 at the same time. The added code performs the same test as UW imapd; it
4040 checks after acquiring the lock that its file descriptor still refers to
4041 the same named file.
4043 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4044 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4047 11. Added the smtp_active_hostname option. If used, this will typically be made
4048 to depend on the incoming interface address. Because $interface_address is
4049 not set up until the daemon has forked a reception process, error responses
4050 that can happen earlier (such as "too many connections") no longer contain
4053 12. If an expansion in a condition on a "warn" statement fails because a lookup
4054 defers, the "warn" statement is abandoned, and the next ACL statement is
4055 processed. Previously this caused the whole ACL to be aborted.
4057 13. Added the iplsearch lookup type.
4059 14. Added ident_timeout as a log selector.
4061 15. Added tls_certificate_verified as a log selector.
4063 16. Added a global option tls_require_ciphers (compare the smtp transport
4064 option of the same name). This controls incoming TLS connections.
4066 17. I finally figured out how to make tls_require_ciphers do a similar thing
4067 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4068 before starting the TLS session.
4070 18. Tabs are now shown as \t in -bP output.
4072 19. If the log selector return_path_on_delivery was set, Exim crashed when
4073 bouncing a message because it had too many Received: header lines.
4075 20. If two routers both had headers_remove settings, and the first one included
4076 a superfluous trailing colon, the final name in the first list and the
4077 first name in the second list were incorrectly joined into one item (with a
4078 colon in the middle).
4084 1. Added -C and -D options to the exinext utility, mainly to make it easier
4085 to include in the automated testing, but these could be helpful when
4086 multiple configurations are in use.
4088 2. The exinext utility was not formatting the output nicely when there was
4089 an alternate port involved in the retry record key, nor when there was a
4090 message id as well (for retries that were specific to a specific message
4091 and a specific host). It was also confused by IPv6 addresses, because of
4092 the additional colons they contain. I have fixed the IPv4 problem, and
4093 patched it up to do a reasonable job for IPv6.
4095 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4096 delivery, the log line now contains "pipelined" if PIPELINING was used.
4098 4. An SMTP transport process used to panic and die if the bind() call to set
4099 an explicit outgoing interface failed. This has been changed; it is now
4100 treated in the same way as a connect() failure.
4102 5. A reference to $sender_host_name in the part of a conditional expansion
4103 that was being skipped was still causing a DNS lookup. This no longer
4106 6. The def: expansion condition was not recognizing references to header lines
4107 that used bh_ and bheader_.
4109 7. Added the _cache feature to named lists.
4111 8. The code for checking quota_filecount in the appendfile transport was
4112 allowing one more file than it should have been.
4114 9. For compatibility with Sendmail, the command line option
4122 and sets the incoming protocol and host name (for trusted callers). The
4123 host name and its colon can be omitted when only the protocol is to be set.
4124 Note the Exim already has two private options, -pd and -ps, that refer to
4125 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4126 "s", but I don't think that's a major issue.
4128 10. A number of refactoring changes to the code, none of which should affect
4131 (a) The number of logging options was getting close to filling up the
4132 32-bit word that was used as a bit map. I have split them into two classes:
4133 those that are passed in the argument to log_write(), and those that are
4134 only ever tested independently outside of that function. These are now in
4135 separate 32-bit words, so there is plenty of room for expansion again.
4136 There is no change in the user interface or the logging behaviour.
4138 (b) When building, for example, log lines, the code previously used a
4139 macro that called string_cat() twice, in order to add two strings. This is
4140 not really sufficiently general. Furthermore, there was one instance where
4141 it was actually wrong because one of the argument was used twice, and in
4142 one call a function was used. (As it happened, calling the function twice
4143 did not affect the overall behaviour.) The macro has been replaced by a
4144 function that can join an arbitrary number of extra strings onto a growing
4147 (c) The code for expansion conditions now uses a table and a binary chop
4148 instead of a serial search (which was left over from when there were very
4149 few conditions). Also, it now recognizes conditions like "pam" even when
4150 the relevant support is not compiled in: a suitably worded error message is
4151 given if an attempt is made to use such a condition.
4153 11. Added ${time_interval:xxxxx}.
4155 12. A bug was causing one of the ddress fields not to be passed back correctly
4156 from remote delivery subprocesses. The field in question was not being
4157 subsequently used, so this caused to problems in practice.
4159 13. Added new log selectors queue_time and deliver_time.
4161 14. Might have fixed a bug in maildirsizefile handling that threw up
4162 "unexpected character" debug warnings, and recalculated the data
4163 unnecessarily. In any case, I expanded the warning message to give more
4166 15. Added the message "Restricted characters in address" to the statements in
4167 the default ACL that block characters like @ and % in local parts.
4169 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4170 Three changes have been made:
4172 (a) There was a serious bug; a negative response to MAIL caused the whole
4173 recipient domain to be cached as invalid, thereby blocking all messages
4174 to all local parts at the same domain, from all senders. This bug has
4175 been fixed. The domain is no longer cached after a negative response to
4176 MAIL if the sender used is not empty.
4178 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4181 (c) A new callout option, "use_sender" has been added for people who want
4182 the modified behaviour.
4188 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4191 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4192 indeed breaks things for older releases.
4194 3. Added additional logging to the case where there is a problem reading data
4195 from a filter that is running in a subprocess using a pipe, in order to
4196 try to track down a specific problem.
4198 4. Testing facility fudge: when running in the test harness and attempting
4199 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4200 getting "No route to host". Convert this to a timeout.
4202 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4205 6. Some OS don't have socklen_t but use size_t instead. This affects the
4206 fifth argument of getsockopt() amongst other things. This is now
4207 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4208 can be set for individual OS. I have set it for SunOS5, OSF1, and
4209 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4210 some earlier ones do not.
4212 7. Change 4.30/15 was not doing the test caselessly.
4214 8. The standard form for an IPv6 address literal was being rejected by address
4215 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4216 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4217 this, as well as the form without the "IPv6" on the front (but only when
4218 address literals are enabled, of course).
4220 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4222 10. Exim crashed if a message with an empty sender address specified by -f
4223 encountered a router with an errors_to setting. This could be provoked only
4224 by a command such as
4228 where an empty string was supplied; "<>" did not hit this bug.
4230 11. Installed PCRE release 4.5.
4232 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4233 remained set. It is now erased.
4235 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4236 times from message ids (which are base 36 rather than the normal 62).
4238 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4239 were being counted as actual protocol errors, and logged if the log
4240 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4241 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4242 and DATA following a set of rejected RCPTs do not count as protocol errors.
4243 In other words, Exim assumes they were pipelined, though this may not
4244 actually be the case. Of course, in all cases the client gets an
4245 appropriate error code.
4247 15. If a lookup fails in an ACL condition, a message about the failure may
4248 be available; it is used if testing the ACL cannot continue, because most
4249 such messages specify what the cause of the deferral is. However, some
4250 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4251 that caused an old message to be retained and used if a later statement
4252 caused a defer, replacing the real cause of the deferral.
4254 16. If an IP address had so many PTR records that the DNS lookup buffer
4255 was not large enough to hold them, Exim could crash while trying to process
4256 the truncated data. It now detects and logs this case.
4258 17. Further to 4.21/58, another change has been made: if (and only if) the
4259 first line of a message (the first header line) ends with CRLF, a bare LF
4260 in a subsequent header line has a space inserted after it, so as not to
4261 terminate the header.
4263 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4264 unnecessarily, used atoi() instead of strtol(), and didn't check the
4265 termination when getting file sizes from file names by regex.
4267 19. Completely re-implemented the support for maildirsize files, in the light
4268 of a number of problems with the previous contributed implementation
4269 (4.30/29). In particular:
4271 . If the quota is zero, the maildirsize file is maintained, but no quota is
4274 . If the maildir directory does not exist, it is created before any attempt
4275 to write a maildirsize file.
4277 . The quota value in the file is just a cache; if the quota is changed in
4278 the transport, the new value overrides.
4280 . A regular expression is available for excluding directories from the
4283 20. The autoreply transport checks the characters in options that define the
4284 message's headers; it allows continued headers, but it was checking with
4285 isspace() after an embedded newline instead of explicitly looking for a
4288 21. If all the "regular" hosts to which an address was routed had passed their
4289 expiry times, and had not reached their retry times, the address was
4290 bounced, even if fallback hosts were defined. Now Exim should go on to try
4293 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4294 equivalent code in the SMTP transport. Some hosts send humungous responses
4295 to HELO/EHLO, more than 1024 it seems.
4297 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4298 gives compiler warnings in some environments. I've now done it "properly",
4301 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4302 (because of problems with the built-in one) was declared to return uschar *
4303 instead of char *, causing compiler failure.
4305 25. Fixed a file descriptor leak when processing alias/forward files.
4307 26. Fixed a minor format string issue in dbfn.c.
4309 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4311 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4312 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4315 29. When a sender address is verified, it is cached, to save repeating the test
4316 when there is more than one recipient in a message. However, when the
4317 verification involves a callout, it is possible for different callout
4318 options to be set for different recipients. It is too complicated to keep
4319 track of this in the cache, so now Exim always runs a verification when a
4320 callout is required, relying on the callout cache for the optimization.
4321 The overhead is duplication of the address routing, but this should not be
4324 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4325 to be verified with callout=postmaster, and the main callout worked but the
4326 postmaster check failed, the verification correctly failed. However, if a
4327 subsequent RCPT command asked for sender verification *without* the
4328 postmaster check, incorrect caching caused this verification also to fail,
4331 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4332 it was not caching the DNS options (qualify_single, search_parents) that
4333 were used when the lookup failed. A subsequent lookup with different
4334 options therefore always gave the same answer, though there were cases
4335 where it should not have. (Example: a "domains = !$mx_any" option on a
4336 dnslookup router: the "domains" option is always processed without any
4337 widening, but the router might have qualify_single set.) Now Exim uses the
4338 cached value only when the same options are set.
4340 32. Added John Jetmore's "exipick" utility to the distribution.
4342 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4343 than a timeout (e.g. a certificate is required, and is not provided), an
4344 Exim server now closes the connection immediately. Previously it waited for
4345 the client to close - but if the client is SSL, it seems that they each
4346 wait for each other, leading to a delay before one of them times out.
4348 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4349 maintained 0.8.x compatibility because I don't think many are using it, and
4350 it is clearly obsolete.
4352 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4355 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4356 client certificate was expired. A simple patch fixes this, though I don't
4357 understand the full logic of why the verify callback is called multiple
4360 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4361 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4362 which causes problems with some clients (such as the Certicom SSL Plus
4363 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4364 disables the coutermeasure allowing Eudora to connect."
4366 38. Exim was not checking that a write() to a log file succeeded. This could
4367 lead to Bad Things if a log got too big, in particular if it hit a file
4368 size limit. Exim now panics and dies if it cannot write to a log file, just
4369 as it does if it cannot open a log file.
4371 39. Modified OS/Makefile-Linux so that it now contains
4373 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4375 The two -D definitions ensure that Exim is compiled with large file
4376 support, which makes it possible to handle log files that are bigger than
4379 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4380 instance) a domain was checked against a named list that involved a lookup,
4381 causing $domain_data to be set, then another domain was checked against the
4382 same list, then the first domain was re-checked, the value of $domain_data
4383 after the final check could be wrong. In particular, if the second check
4384 failed, it could be set empty. This bug probably also applied to
4387 41. The strip_trailing_dot option was not being applied to the address given
4388 with the -f command-line option.
4390 42. The code for reading a message's header from the spool was incrementing
4391 $received_count, but never initializing it. This meant that the value was
4392 incorrect (doubled) while delivering a message in the same process in which
4393 it was received. In the most common configuration of Exim, this never
4394 happens - a fresh exec is done - but it can happen when
4395 deliver_drop_privilege is set.
4397 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4398 it now includes up to 150 characters of the unexpected data in the log
4401 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4402 and building data strings. The size of both of these buffers was 10 000
4403 bytes - far larger than anybody would *ever* want, thought I. Needless to
4404 say, somebody hit the limit. I have increased the maximum line length to
4405 20 000 and the maximum data length of concatenated lines to 100 000. I have
4406 also fixed two bugs, because there was no checking on these buffers. Tsk,
4407 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4408 buffer is too small.
4410 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4411 lsearch lookups. Now it does.
4413 46. When parsing a route_list item in a manualroute router, a fixed-length
4414 buffer was used for the list of hosts. I made this 1024 bytes long,
4415 thinking that nobody would ever have a list of hosts that long. Wrong.
4416 Somebody had a whole pile of complicated expansion conditions, and the
4417 string was silently truncated, leading to an expansion error. It turns out
4418 that it is easier to change to an unlimited length (owing to other changes
4419 that have happened since this code was originally written) than to build
4420 structure for giving a limitation error. The length of the item that
4421 expands into the list of hosts is now unlimited.
4423 47. The lsearch lookup could not handle data where the length of text line was
4424 more than 4095 characters. Such lines were truncated, leading to shortened
4425 data being returned. It should now handle lines of any length.
4427 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4428 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4431 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4432 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4433 no longer happens in comment lines. A list of replacements is now placed
4434 at the head of all of the source files, except those whose only change is
4435 to replace PERL_COMMAND in the very first #! line.
4437 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4438 messages on the queue, with a bottom-up merge sort, using code contributed
4439 by Michael Haardt. This should make operations like -bp somewhat faster on
4440 large queues. It won't affect queue runners, except when queue_run_in_order
4443 51. Installed eximstats 1.31 in the distribution.
4445 52. Added support for SRV lookups to the dnslookup router.
4447 53. If an ACL referred to $message_body or $message_body_end, the value was not
4448 reset for any messages that followed in the same SMTP session.
4450 54. The store-handling optimization for building very long strings was not
4451 differentiating between the different store pools. I don't think this
4452 actually made any difference in practice, but I've tidied it.
4454 55. While running the routers to verify a sender address, $sender_address
4455 was still set to the sender address. This is wrong, because when routing to
4456 send a bounce to the sender, it would be empty. Therefore, I have changed
4457 it so that, while verifying a sender address, $sender_address is set to <>.
4458 (There is no change to what happens when verifying a recipient address.)
4460 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4461 target A or AAAA records (if not already returned) without resetting the
4462 qualify_single or search_parents options of the DNS resolver. These are
4463 inappropriate in this case because the targets of MX and SRV records must
4464 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4465 target that, when qualified, matched something in the local domain. These
4466 two options are now turned off when doing these lookups.
4468 57. It seems that at least some releases of Reiserfs (which does not have the
4469 concept of a fixed number of inodes) returns zero and not -1 for the
4470 number of available inodes. This interacted badly with check_spool_inodes,
4471 which assumed that -1 was the "no such thing" setting. What I have done is
4472 to check that the total number of inodes is greater than zero before doing
4473 the test of how many are available.
4475 58. When a "warn" ACL statement has a log_message modifier, the message is
4476 remembered, and not repeated. This is to avoid a lot of repetition when a
4477 message has many recipients that cause the same warning to be written.
4478 Howewer, Exim was preserving the list of already written lines for an
4479 entire SMTP session, which doesn't seem right. The memory is now reset if a
4480 new message is started.
4482 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4483 debugging output unless log_rewrite was also set.
4485 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4486 of (int)(handle) when we know that handle contains (void *)(-1).
4488 61. The Exim daemon panic-logs an error return when it closes the incoming
4489 connection. However "connection reset by peer" seems to be common, and
4490 isn't really an error worthy of noting specially, so that particular error
4493 62. When Exim is trying to find all the local interfaces, it used to panic and
4494 die if the ioctl to get the interface flags failed. However, it seems that
4495 on at least one OS (Solaris 9) it is possible to have an interface that is
4496 included in the list of interfaces, but for which you get a failure error
4497 for this call. This happens when the interface is not "plumbed" into a
4498 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4499 failure of the "get flags" call assumes that the interface is down.
4501 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4502 makes life easier for people who are doing arithmetic on fields extracted
4503 from dates, where you often get leading zeros that should not be
4504 interpreted as octal.
4506 64. Added qualify_domain to the redirect router, to override the global
4509 65. If a pathologically long header line contained very many addresses (the
4510 report of this problem mentioned 10 000) and each of them was rewritten,
4511 Exim could use up a very large amount of memory. (It kept on making new
4512 copies of the header line as it rewrote, and never released the old ones.)
4513 At the expense of a bit more processing, the header rewriting function has
4514 been changed so that it no longer eats memory in this way.
4516 66. The generation of the Received: header has been moved from the time that a
4517 message starts to be received, to the time that it finishes. The timestamp
4518 in the Received: header should now be very close to that of the <= log
4519 line. There are two side-effects of this change:
4521 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4522 logged header lines no longer include the local Received: line, because
4523 it has not yet been created. The same applies to a copy of the message
4524 that is returned to a non-SMTP sender when a message is rejected.
4526 (b) When a filter file is tested using -bf, no additional Received: header
4527 is added to the test message. After some thought, I decided that this
4530 This change does not affect the value of $received_for. It is still set
4531 after address rewriting, but before local_scan() is called.
4533 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4535 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4536 gave an unhelpful panic error message, and a defer error. I have managed to
4537 change this behaviour so that it now rejects any supplied certificate,
4538 which seems right, as the list of acceptable certificates is empty.
4540 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4541 gave an unhelpful defer error. I have not managed to make this reject any
4542 supplied certificates, but the error message it gives is "no certificate
4543 supplied", which is not helpful.
4545 70. exigrep's output now also includes lines that are not associated with any
4546 message, but which match the given pattern. Implemented by a patch from
4547 Martin Sluka, which also tidied up the Perl a bit.
4549 71. Recipient callout verification, like sender verification, was using <> in
4550 the MAIL FROM command. This isn't really the right thing, since the actual
4551 sender may affect whether the remote host accepts the recipient or not. I
4552 have changed it to use the actual sender in the callout; this means that
4553 the cache record is now keyed on a recipient/sender pair, not just the
4554 recipient address. There doesn't seem to be a real danger of callout loops,
4555 since a callout by the remote host to check the sender would use <>.
4556 [SEE ABOVE: changed after hitting problems.]
4558 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4559 temporary errors. However, in the case of such a code being given after
4560 the end of a data transmission (i.e. after ".") Exim was failing to write
4561 a retry record for the message. (Yes, there was some broken host that was
4562 actually sending 8xx at this point.)
4564 73. An unknown lookup type in a host list could cause Exim to panic-die when
4565 the list was checked. (An example that provoked this was putting <; in the
4566 middle of a list instead of at the start.) If this happened during a DATA
4567 ACL check, a -D file could be left lying around. This kind of configuration
4568 error no longer causes Exim to die; instead it causes a defer errror. The
4569 incident is still logged to the main and panic logs.
4571 74. Buglet left over from Exim 3 conversion. The message "too many messages
4572 in one connection" was written to the rejectlog but not the mainlog, except
4573 when address rewriting (yes!) was being logged.
4575 75. Added write_rejectlog option.
4577 76. When a system filter was run not as root (that is, when system_filter_user
4578 was set), the values of the $n variables were not being returned to the
4579 main process; thus, they were not subsequently available in the $sn
4582 77. Added +return_path_on_delivery log selector.
4584 78. A connection timeout was being treated differently from recipients deferred
4585 when testing hosts_max_try with a message that was older than the host's
4586 retry timeout. (The host should not be counted, thus allowing all hosts to
4587 be tried at least once before bouncing.) This may have been the cause of an
4588 occasionally reported bug whereby a message would remain on the queue
4589 longer than the retry timeout, but would be bounced if a delivery was
4590 forced. I say "may" because I never totally pinned down the problem;
4591 setting up timeout/retry tests is difficult. See also the next item.
4593 79. The ultimate address timeout was not being applied to errors that involved
4594 a combination of host plus message (for example, a timeout on a MAIL
4595 command). When an address resolved to a number of possible hosts, and they
4596 were not all tried for each delivery (e.g. because of hosts_max_try), a
4597 message could remain on the queue longer than the retry timeout.
4599 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4602 81. Fixed an obscure SMTP outgoing bug which required at least the following
4603 conditions: (a) there was another message waiting for the same server;
4604 (b) the server returned 5xx to all RCPT commands in the first message so
4605 that the message was not completed; (c) the server dropped the connection
4606 or gave a negative response to the RSET that Exim sends to abort the
4607 transaction. The observed case was a dropped connection after DATA that had
4608 been sent in pipelining mode. That is, the server had advertised PIPELINING
4609 but was not implementing it correctly. The effect of the bug was incorrect
4610 behaviour, such as trying another host, and this could lead to a crash.
4616 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4617 and daemon.c were passed as pointers to ints; they should have been
4618 pointers to socklen_t variables (which are typically unsigned ints).
4620 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4623 3. Fixed a really odd bug that affected only the testing scheme; patching a
4624 certain fixed string in the binary changed the value of another string that
4625 happened to be identical to the end of the original first string.
4627 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4628 name", it returns that address as the IP address. On some operating
4629 systems (e.g. Solaris), it also passes back the IP address string as the
4630 "host name". However, on others (e.g. Linux), it passes back an empty
4631 string. Exim wasn't checking for this, and was changing the host name to an
4632 empty string, assuming it had been canonicized.
4634 5. Although rare, it is permitted to have more than one PTR record for a given
4635 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4636 all the names associated with an address, because they do in Solaris.
4637 However, it seems that they do not in Linux for data that comes from the
4638 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4639 I found this out when I moved to a new Linux workstation and tried to run
4640 the Exim test suite.
4642 To get round this problem I have changed the code so that it now does its
4643 own call to the DNS to look up PTR records when searching for a host name.
4644 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4645 addresses that are only in /etc/hosts are still found.
4647 This behaviour is, however, controlled by an option called host_lookup_
4648 order, which defaults to "bydns:byaddr". If people want to use the other
4649 order, or indeed, just use one or the other means of lookup, they can
4650 specify it in this variable.
4652 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4653 some operating systems, this comes back from gethostbyaddr() as an empty
4654 string, and this is what Exim used to test for. However, it seems that in
4655 other systems, "." is yielded. Exim now tests for this case too.
4657 7. The values of check_spool_space and check_log_space are now held internally
4658 as a number of kilobytes instead of an absolute number of bytes. If a
4659 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4660 kilobyte. This means that much larger values can be stored.
4662 8. Exim monitor: an attempt to get the action menu when not actually pointing
4663 at a message produces an empty menu entitled "No message selected". This
4664 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4665 no entries in it ("Shell widget menu has zero width and/or height"). So I
4666 have added a single, blank menu entry in this case.
4668 9. Added ${quote_local_part.
4670 10. MIME decoding is now applied to the contents of Subject: header lines when
4673 11. Now that a reference to $sender_host_address automatically causes a reverse
4674 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4675 host lookup before query-style lookups in lists that might use this
4676 variable. This has therefore been abolished, and the "net-" prefix is no
4677 longer necessary for query-style lookups.
4679 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4680 have been a typo for LFLAGS, so it has been changed.
4682 13. The install script calls Exim with "-C /dev/null" in order to find the
4683 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4684 to be output. Howeve, since Exim outputs its version number before the
4685 error, it didn't break the script. It just looked ugly. I fixed this by
4686 always allowing "-C /dev/null" if the caller is root.
4688 14. Ignore overlarge ACL variable number when reading spool file - insurance
4689 against a later release with more variables having written the file.
4691 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4692 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4693 this, as well as the form without the "IPv6" on the front.
4695 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4696 OS/Makefile-Darwin file.
4698 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4699 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4701 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4702 variable. However, some OS use ints for inodes. I've added cast to long int
4703 to get rid of the compiler warning.
4705 19. I had forgotten to lock out "/../" in configuration file names when
4706 ALT_CONFIG_PREFIX was set.
4708 20. Routers used for verification do not need to specify transports. However,
4709 if such a router generated a host list, and callout was configured, Exim
4710 crashed, because it could not find a port number from the (non-existent)
4711 transport. It now assumes port 25 in this circumstance.
4713 21. Added the -t option to exigrep.
4715 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4716 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4717 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4718 them is set, LOOKUP_LSEARCH is forced.
4720 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4722 24. Added sender and host information to the "rejected by local_scan()" log
4723 line; previously there was no indication of these.
4725 25. Added .include_if_exists.
4727 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4728 Linux. It turns out that not all file systems support this. Apparently some
4729 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4730 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4731 not supported on the file descriptor, is now ignored when Exim is trying to
4732 sync a directory. This applies only to Linux.
4734 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4736 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4737 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4740 29. Added support for maildirsize files from supplied patch (modified a bit).
4742 30. The use of :fail: followed by an empty string could lead Exim to respond to
4743 sender verification failures with (e.g.):
4745 550 Verification failed for <xxx>
4746 550 Sender verify failed
4748 where the first response line was missing the '-' that indicates it is not
4749 the final line of the response.
4751 31. The loop for finding the name of the user that called Exim had a hardwired
4752 limit of 10; it now uses the value of finduser_retries, which is used for
4753 all other user lookups.
4755 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4758 33. Exim was neglecting to zero errno before one call of strtol() when
4759 expanding a string and expecting an integer value. On some systems this
4760 resulted in spurious "integer overflow" errors. Also, it was casting the
4761 result into an int without checking.
4763 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4764 did not work. The code looks as if it has *never* worked, though it appears
4765 to have been documented since at least releast 1.62. I have made it work.
4767 35. The "timeout_DNS" error in retry rules, also documented since at least
4768 1.62, also never worked. As it isn't clear exactly what this means, and
4769 clearly it isn't a major issue, I have abolished the feature by treating it
4770 as "timeout", and writing a warning to the main and panic logs.
4772 36. The display of retry rules for -brt wasn't always showing the error code
4775 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4776 timeout_connect_A, timeout_connect_MX.
4778 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4779 to the empty sender.
4781 39. The daemon was not analysing the content of -oX till after it had closed
4782 stderr and disconnected from the controlling terminal. This meant that any
4783 syntax errors were only noted on the panic log, and the return code from
4784 the command was 0. By re-arranging the code a little, I've made the
4785 decoding happen first, so such errors now appear on stderr, and the return
4786 code is 1. However, the actual setting up of the sockets still happens in
4787 the disconnected process, so errors there are still only recorded on the
4790 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4791 connections (as happens on some IP stacks) was logged at start up time as
4792 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4793 it from "IPv6 and IPv4", which means that two separate sockets are being
4796 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4797 says whether AF_INET or AF_INET6 was passed as an argument.
4799 42. Exiwhat output was messed up when time zones were included in log
4802 43. Exiwhat now gives more information about the daemon's listening ports,
4803 and whether -tls-on-connect was used.
4805 44. The "port" option of the smtp transport is now expanded.
4807 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4808 silently ignored. Now an error message is written to the main and panic
4811 46. There's a new ACL modifier called "logwrite" which writes to a log file
4812 as soon as it is encountered.
4814 47. Added $local_user_uid and $local_user_gid at routing time.
4816 48. Exim crashed when trying to verify a sender address that was being
4819 49. Exim was recognizing only a space character after ".include". It now also
4820 recognizes a tab character.
4822 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4823 extracting the relevant information from the specification. The man page no
4824 longer contains scrambled data for the -d option, and I've added a section
4825 at the front about calling Exim under different names.
4827 51. Added "extra_headers" argument to the "mail" command in filter files.
4829 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4832 53. Installed eximstats 1.29.
4834 54. Added transport_filter_timeout as a generic transport option.
4836 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4837 Cc: header lines. This was required by RFC 822, but it not required by RFC
4840 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4841 incoming messages that did not have them. Now it does so only if the
4842 message originates locally, that is, if there is no associated remote host
4843 address. When Resent- header lines are present, this applies to the Resent-
4844 lines rather than the non-Resent- lines.
4846 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4847 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4849 58. Messages for configuration errors now include the name of the main
4850 configuration file - useful now that there may be more than one file in a
4851 list (.included file names were always shown).
4853 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4854 for those rare installations that do not start the daemon as root or run it
4855 setuid root. I've cut out the call to initgroups() if the daemon is not
4858 60. The Exim user and group can now be bound into the binary as text strings
4859 that are looked up at the start of Exim's processing.
4861 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4863 62. Added $mailstore_basename variable.
4865 63. Installed patch to sieve.c from Michael Haardt.
4867 64. When Exim failed to open the panic log after failing to open the main log,
4868 the original message it was trying to log was written to stderr and debug
4869 output, but if they were not available (the usual case in production), it
4870 was lost. Now it is written to syslog before the two lines that record the
4871 failures to open the logs.
4873 65. Users' Exim filters run in subprocesses under the user's uid. It is
4874 possible for a "deliver" command or an alias in a "personal" command to
4875 provoke an address rewrite. If logging of address rewriting is configured,
4876 this fails because the process is not running as root or exim. There may be
4877 a better way of dealing with this, but for the moment (because 4.30 needs
4878 to be released), I have disabled address rewrite logging when running a
4879 filter in a non-root, non-exim process.
4885 1. The buildconfig auxiliary program wasn't quoting the value set for
4886 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4887 not defined. This bug was masked in 4.22 by the effect that was fixed in
4890 2. Some messages that were rejected after a message id was allocated were
4891 shown as "incomplete" by exigrep. It no longer does this for messages that
4892 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4894 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4895 have allow_domain_literals set, the ID did not get logged in the <= line.
4896 Domain literals are now always recognized in Message-ID: header lines.
4898 4. The first argument for a ${extract expansion item is the key name or field
4899 number. Leading and trailing spaces in this item were not being ignored,
4900 causing some misleading effects.
4902 5. When deliver_drop_privilege was set, single queue runner processes started
4903 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4904 same command in the process it spins off) were not dropping privilege.
4906 6. When the daemon running as "exim" started a queue runner, it always
4907 re-executed Exim in the spun-off process. This is a waste of effort when
4908 deliver_drop_privilege is set. The new process now just calls the
4909 queue-runner function directly.
4915 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4918 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4919 ignored. Though the use of -oP was forcing the writing of a pid file, it
4920 was always written to the default place.
4922 3. If the message "no IP address found for host xxxx" is generated during
4923 incoming verification, it is now followed by identification of the incoming
4924 connection (so you can more easily find what provoked it).
4926 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4928 5. Added some features to "harden" Exim a bit more against certain attacks:
4930 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4931 be put in Local/Makefile. This is like the never_users runtime option,
4932 but it cannot be overridden. The default setting is "root".
4934 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4935 prefix string with which any file named in a -C command line option
4938 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4939 is retained for -C and -D only if the caller of Exim is root. Without
4940 it, the exim user may also use -C and -D and retain privilege.
4942 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4943 command line option is disabled.
4945 6. Macro names set by the -D option must start with an upper case letter, just
4946 like macro names defined in the configuration file.
4948 7. Added "dereference=" facility to LDAP.
4950 8. Two instances of the typo "uknown" in the source files are fixed.
4952 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4953 the Configure-Makefile script screwed up while processing it.
4955 10. Incorporated PCRE 4.4.
4957 11. The SMTP synchronization check was not operating right at the start of an
4958 SMTP session. For example, it could not catch a HELO sent before the client
4959 waited for the greeting. There is now a check for outstanding input at the
4960 point when the greeting is written. Because of the duplex, asynchronous
4961 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4962 way, but not yet received, when the check is performed.
4964 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4965 on TCP/IP sockets, because this apparently causes some broken clients to
4968 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4969 unchanged) from the Cygwin maintainer.
4971 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4972 was supported instead of testing for mbx. Effectively a typo.
4974 15. The spa authenticator server code was not checking that the input it
4975 received was valid base64.
4977 16. The debug output line for the "set" modifier in ACLs was not showing the
4978 name of the variable that was being set.
4980 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4982 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4983 lookup on its own. Something else was needed to trigger the lookup. For
4984 example, a match in host_lookup or the need for a host name in a host list.
4985 Now, if $sender_host_name is referenced and the host name has not yet been
4986 looked up, a lookup is performed. If the lookup fails, the variable remains
4987 empty, and $host_lookup_failed is set to "1".
4989 19. Added "eqi" as a case-independent comparison operator.
4991 20. The saslauthd authentication condition could segfault if neither service
4992 nor realm was specified.
4994 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4995 error message that was logged was misleading, and incoming SMTP
4996 connections were dropped. The message is now more accurate, and temporary
4997 errors are given to SMTP connections.
4999 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5000 (except RSET) with a 421 error, until QUIT is received. However, it was
5001 failing to send a response to QUIT.
5003 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5004 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5005 ACL is now run after the helo_try_verify_hosts code.
5007 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5008 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5009 case-independent, so other case variants are also recognized). Apparently
5010 some systems use these upper case variants.
5012 25. If more than two messages were waiting for the same host, and a transport
5013 filter was specified for the transport, Exim sent two messages over the
5014 same TCP/IP connection, and then failed with "socket operation on non-
5015 socket" when it tried to send the third.
5017 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5019 27. The extern definition of crypt16() in expand.c was not being excluded when
5020 the OS had its own crypt16() function.
5022 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5023 as a preferred synonym for return_size_limit, both as an option and as an
5026 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5028 30. Changed the default configuration ACL to relax the local part checking rule
5029 for addresses that are not in any local domains. For these addresses,
5030 slashes and pipe symbols are allowed within local parts, but the sequence
5031 /../ is explicitly forbidden.
5033 31. SPA server authentication was not clearing the challenge buffer before
5036 32. log_message in a "warn" ACL statement was writing to the reject log as
5037 well as to the main log, which contradicts the documentation and doesn't
5038 seem right (because no rejection is happening). So I have stopped it.
5040 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5041 However, I am unable to do any testing of this.
5043 34. Fixed an infelicity in the appendfile transport. When checking directories
5044 for a mailbox, to see if any needed to be created, it was accidentally
5045 using path names with one or more superfluous leading slashes; tracing
5046 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5048 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5049 contents are added to the log line that is written for every discarded
5050 recipient. (Previously a log_message setting was ignored.)
5052 36. The ${quote: operator now quotes the string if it is empty.
5054 37. The install script runs exim in order to find its version number. If for
5055 some reason other than non-existence or emptiness, which it checks, it
5056 could not run './exim', it was installing it with an empty version number,
5057 i.e. as "exim-". This error state is now caught, and the installation is
5060 38. An argument was missing from the function that creates an error message
5061 when Exim fails to connect to the socket for saslauthd authentication.
5062 This could cause Exim to crash, or give a corrupted message.
5064 39. Added isip, isip4, and isip6 to ${if conditions.
5066 40. The ACL variables $acl_xx are now saved with the message, and can be
5067 accessed later in routers, transports, and filters.
5069 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5070 strings in the file are not string-expanded.
5072 42. If a MAIL command specified a SIZE value that was too large to fit into an
5073 int variable, the check against message_size_limit failed. Such values are
5074 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5075 day this will have to be increased, but I don't think I want to be around
5076 when emails are that large.
5083 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5084 iconv() is not standard in FreeBSD.
5086 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5087 IPv6 enabled. The observed symptom was a segmentation fault on return from
5088 the function os_common_find_running_interfaces() in src/os.c.
5090 3. In the check_special_case() function in daemon.c I had used "errno" as an
5091 argument name, which causes warnings on some systems. This was basically a
5092 typo, since it was named "eno" in the comments!
5094 4. The code that waits for the clock to tick (at a resolution of some fraction
5095 of a second) so as to ensure message-id uniqueness was always waiting for
5096 at least one whole tick, when it could have waited for less. [This is
5097 almost certainly not relevant at current processor speeds, where it is
5098 unlikely to ever wait at all. But we try to future-proof.]
5100 5. The function that sleeps for a time interval that includes fractions of a
5101 second contained a race. It did not block SIGALRM between setting the
5102 timer, and suspending (a couple of lines later). If the interval was short
5103 and the sigsuspend() was delayed until after it had expired, the suspension
5104 never ended. On busy systems this could lead to processes getting stuck for
5107 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5108 process, before it forks any delivery processes. The open lookup caching
5109 mechanism meant that the open file or database connection was passed into
5110 the delivery process. The problem was that delivery processes always tidy
5111 up cached lookup data. This could cause a problem for the next delivery
5112 process started by the queue runner, because the external queue runner
5113 process does not know about the closure. So the next delivery process
5114 still has data in the lookup cache. In the case of a file lookup, there was
5115 no problem because closing a file descriptor in a subprocess doesn't affect
5116 the parent. However, if the lookup was caching a connection to a database,
5117 the connection was closed, and the second delivery process was likely to
5118 see errors such as "PGSQL: query failed: server closed the connection
5119 unexpectedly". The problem has been fixed by closing all cached lookups
5120 in a queue runner before running a delivery process.
5122 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5123 seem to have the "const" qualifier which it has on other OS. I've
5126 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5127 *of the same type* (client or server) with the same public name that an
5128 error should be diagnosed.
5130 9. When Exim looked up a host name for an IP address, but failed to find the
5131 original IP address when looking up the host name (a safety check), it
5132 output the message "<ip address> does not match any IP for NULL", which was
5133 confusing, to say the least. The bug was that the host name should have
5134 appeared instead of "NULL".
5136 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5137 user that is built into the binary, and the -C or -D options is used, root
5138 privilege is dropped before the configuration file is read. In addition,
5139 logging is switched to stderr instead of the normal log files. If the
5140 configuration then re-defines the Exim user, the unprivileged environment
5141 is probably not what is expected, so Exim logs a panic warning message (but
5144 However, if deliver_drop_privilege is set, the unprivileged state may well
5145 be exactly what is intended, so the warning has been cut out in that case,
5146 and Exim is allowed to try to write to its normal log files.
5152 1. smtp_return_error_details was not giving details for temporary sender
5153 or receiver verification errors.
5155 2. Diagnose a configuration error if two authenticators have the same public
5158 3. Exim used not to create the message log file for a message until the first
5159 delivery attempt. This could be confusing when incoming messages were held
5160 for policy or load reasons. The message log file is now created at the time
5161 the message is received, and an initial "Received" line is written to it.
5163 4. The automatically generated man page for command line options had a minor
5164 bug that caused no ill effects; however, a more serious problem was that
5165 the procedure for building the man page automatically didn't always
5166 operate. Consequently, release 4.20 contains an out-of-date version. This
5167 shouldn't happen again.
5169 5. When building Exim with embedded Perl support, the script that builds the
5170 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5171 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5173 6. The freeze_tell option was not being used for messages that were frozen on
5174 arrival, either by an ACL or by local_scan().
5176 7. Added the smtp_incomplete_transaction log selector.
5178 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5179 was accepting AUTH without a new EHLO.
5181 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5182 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5185 10. Logging of TCP/IP connections (when configured) now happens in the main
5186 daemon process instead of the child process, so that the TCP/IP connection
5187 count is more accurate (but it can never be perfect).
5189 11. The use of "drop" in a nested ACL was not being handled correctly in the
5190 outer ACL. Now, if condition failure induced by the nested "drop" causes
5191 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5192 or "require"), the connection is dropped.
5194 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5195 that yield "discard" can now be used with an "accept" or a "discard" verb,
5196 but an error is generated for any others (because I can't see a useful way
5197 to define what should happen).
5199 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5200 are now processed in the same way as lines in the Exim configuration file.
5201 In particular, continuation lines are supported.
5203 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5205 15. Added -ti meaning -t -i.
5207 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5208 domains, and warn by logging if others are found.
5210 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5211 returned from GIFCONF when Exim is trying to find the list of interfaces on
5212 a host. The code in os.c has been modified to copy each ifreq to an aligned
5213 structure in all cases.
5215 Also, in some cases, the returned ifreq's were being copied to a 'struct
5216 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5217 means the last couple of bytes of an IPv6 address could be chopped if the
5218 ifreq contained only a normal sockaddr (14 bytes storage).
5220 18. Named domain lists were not supported in the hosts_treat_as_local option.
5221 An entry such as +xxxx was not recognized, and was treated as a literal
5224 19. Ensure that header lines added by a DATA ACL are included in the reject log
5225 if the ACL subsequently rejects the message.
5227 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5228 MD5 (which is deprecated).
5230 21. When testing a filter file using -bf, Exim was writing a message when it
5231 took the sender from a "From " line in the message, but it was not doing so
5232 when it took $return_path from a Return-Path: header line. It now does.
5234 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5235 with a valid header line field name (a series of printing characters
5236 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5238 23. Changed "disc" in the source to "disk" to conform to the documentation and
5239 the book and for uniformity.
5241 24. Ignore Sendmail's -Ooption=value command line item.
5243 25. When execve() failed while trying to run a command in a pipe transport,
5244 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5245 could be confused with a return value of 69 from the command itself. This
5246 has been changed to 127, the value the shell returns if it is asked to run
5247 a non-existent command. The wording for the related log line suggests a
5248 non-existent command as the problem.
5250 26. If received_header_text expands to an empty string, do not add a Received:
5251 header line to the message. (Well, it adds a token one on the spool, but
5252 marks it "old" so that it doesn't get used or transmitted.)
5254 27. Installed eximstats 1.28 (addition of -nt option).
5256 28. There was no check for failure on the call to getsockname() in the daemon
5257 code. This can fail if there is a shortage of resources on the system, with
5258 ENOMEM, for example. A temporary error is now given on failure.
5260 29. Contrary to the C standard, it seems that in some environments, the
5261 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5262 program. Exim now does this explicitly; it affects the formatting of
5263 timestamps using strftime().
5265 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5266 complaints. I've now initialized all the variables, to avoid this.
5268 32. Header lines added by a system filter were not being "seen" during
5269 transport-time rewrites.
5271 33. The info_callback() function passed to OpenSSL is set up with type void
5272 (*)(SSL *, int, int), as described somewhere. However, when calling the
5273 function (actually a macro) that sets it up, the type void(*)() is
5274 expected. I've put in a cast to prevent warnings from picky compilers.
5276 34. If a DNS black list lookup found a CNAME record, but there were no A
5277 records associated with the domain it pointed at, Exim crashed.
5279 35. If a DNS black list lookup returned more than one A record, Exim ignored
5280 all but the first. It now scans all returned addresses if a particular IP
5281 value is being sought. In this situation, the contents of the
5282 $dnslist_value variable are a list of all the addresses, separated by a
5285 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5286 to accept a host name and all its aliases if the forward lookup for any of
5287 them yielded the IP address of the incoming connection. Now it accepts only
5288 those names whose forward lookup yields the correct IP address. Any other
5289 names are discarded. This closes a loophole whereby a rogue DNS
5290 administrator could create reverse DNS records to break through a
5291 wildcarded host restriction in an ACL.
5293 37. If a user filter or a system filter that ran in a subprocess used any of
5294 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5295 the wrong values were passed to the pipe command ($thisaddress had the
5296 value of $0, $0 had the value of $1, etc). This bug was introduced by
5297 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5299 38. Improved the line breaking for long SMTP error messages from ACLs.
5300 Previously, if there was no break point between 40 and 75 characters, Exim
5301 left the rest of the message alone. Two changes have been made: (a) I've
5302 reduced the minimum length to 35 characters; (b) if it can't find a break
5303 point between 35 and 75 characters, it looks ahead and uses the first one
5304 that it finds. This may give the occasional overlong line, but at least the
5305 remaining text gets split now.
5307 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5308 file descriptors might be low, and that setting 1000 would always raise it.
5309 It turns out that in some environments, the limit is already over 1000 and
5310 that lowering it causes trouble. So now Exim takes care not to decrease it.
5312 40. When delivering a message, the value of $return_path is set to $sender_
5313 address at the start of routing (routers may change the value). By an
5314 oversight, this default was not being set up when an address was tested by
5315 -bt or -bv, which affected the outcome if any router or filter referred to
5318 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5319 log only when "message" or "log_message" are set. However, if one of the
5320 conditions was an address verification, or a call to a nested ACL, the
5321 messages generated by the underlying test were being passed through. This
5322 no longer happens. The underlying message is available in $acl_verify_
5323 message for both "message" and "log_message" expansions, so it can be
5324 passed through if needed.
5326 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5327 new expansion $bh_ to give the encoded byte string without charset
5328 translation. Translation happens only if iconv() is available; HAVE_ICONV
5329 indicates this at build time. HEADERS_CHARSET gives the charset to
5330 translate to; headers_charset can change it in the configuration, and
5331 "headers charset" can change it in an individual filter file.
5333 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5334 that creates RFC 2047 encoded "words" labels them as that charset instead
5335 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5336 expansion operator; (ii) when Exim creates a From: line for a local
5337 message; (iii) when a header line is rewritten to include a "phrase" part.
5339 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5340 buggy, causing it to skip the first lines of messages whose message ID
5341 ended in 'D'. This would not have bitten before Exim release 4.14, because
5342 message IDs were unlikely to end in 'D' before then. The effect was to have
5343 incorrect size information for certain domains.
5345 45. #include "config.h" was missing at the start of the crypt16.c module. This
5346 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5349 46. If there was a timeout during a "random" callout check, Exim treated it as
5350 a failure of the random address, and carried on sending RSET and the real
5351 address. If the delay was just some slowness somewhere, the response to the
5352 original RCPT would be taken as a response to RSET and so on, causing
5353 mayhem of various kinds.
5355 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5356 when I implemented it. It didn't allow for the fact that some option values
5357 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5358 the right test for positive values.
5360 48. Domain names in DNS records are case-independent. Exim always looks them up
5361 in lower case. Some resolvers return domain names in exactly the case they
5362 appear in the zone file, that is, they may contain uppercase letters. Not
5363 all resolvers do this - some return always lower case. Exim was treating a
5364 change of case by a resolver as a change of domain, similar to a widening
5365 of a domain abbreviation. This triggered its re-routing code and so it was
5366 trying to route what was effectively the same domain again. This normally
5367 caused routing to fail (because the router wouldn't handle the domain
5368 twice). Now Exim checks for this case specially, and just changes the
5369 casing of the domain that it ultimately uses when it transmits the message
5372 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5375 50. If a filter generated a file delivery with a non-absolute name (possible if
5376 no home directory exists for the router), the forbid_file option was not
5379 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5380 the existing equality matching.
5382 52. Exim was using ints instead of ino_t variables in some places where it was
5383 dealing with inode numbers.
5385 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5386 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5387 called TMPDIR, and if it finds it is different, it changes its value.
5389 54. The smtp_printf() function is now made available to local_scan() so
5390 additional output lines can be written before returning. There is also an
5391 smtp_fflush() function to enable the detection of a dropped connection.
5392 The variables smtp_input and smtp_batched_input are exported to
5395 55. Changed the default runtime configuration: the message "Unknown user"
5396 has been removed from the ACL, and instead placed on the localuser router,
5397 using the cannot_route_message feature. This means that any verification
5398 failures that generate their own messages won't get overridden. Similarly,
5399 the "Unrouteable address" message that was in the ACL for unverifiable
5400 relay addresses has also been removed.
5402 56. Added hosts_avoid_esmtp to the smtp transport.
5404 57. The exicyclog script was not checking for the esoteric option
5405 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5406 will work only if exicyclog is run under the appropriate euid.
5408 58. Following a discussion on the list, the rules by which Exim recognises line
5409 endings on incoming messages have been changed. The -dropcr and drop_cr
5410 options are now no-ops, retained only for backwards compatibility. The
5411 following line terminators are recognized: LF CRLF CR. However, special
5412 processing applies to CR:
5414 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5415 nor a local message in the state where . is a terminator.
5417 (ii) If a bare CR is encountered in a header line, an extra space is added
5418 after the line terminator so as not to end the header. The reasoning
5419 behind this is that bare CRs in header lines are most likely either
5420 to be mistakes, or people trying to play silly games.
5422 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5423 was being incorrectly given as 18 bytes larger than it should have been.
5424 This is a VOB (very old bug).
5426 60. This may never have affected anything current, but just in case it has:
5427 When the local host is found other than at the start of a list of hosts,
5428 the local host, those with the same MX, and any that follow, are discarded.
5429 When the list in question was part of a longer list of hosts, the following
5430 hosts (not currently being processed) were also being discarded. This no
5431 longer happens. I'm not sure if this situation could ever has previously
5434 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5437 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5438 header. This is something that is recommended in a new Internet Draft, and
5439 is something that is documented as being done by Sendmail. There are two
5440 possible values. For messages generated by the autoreply transport, Exim
5443 Auto-Submitted: auto-replied
5445 whereas for all other generated messages (e.g. bounces) it adds
5447 Auto-Submitted: auto-generated
5449 63. The "personal" condition in filters now includes a test for the
5450 Auto-Submitted: header. If it contains the string "auto-" the message it
5451 not considered personal.
5453 64. Added rcpt_include_affixes as a generic transport option.
5455 65. Added queue_only_override (default true).
5457 66. Added the syslog_duplication option.
5459 67. If what should have been the first header line of a message consisted of
5460 a space followed by a colon, Exim was mis-interpreting it as a header line.
5461 It isn't of course - it is syntactically invalid and should therefore be
5462 treated as the start of the message body. The misbehaviour could have
5463 caused a number of strange effects, including loss of data in subsequent
5464 header lines, and spool format errors.
5466 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5467 client host had authenticated. This control can now be exercised by an ACL
5468 for more flexibility.
5470 69. By default, callouts do not happen when testing with -bh. There is now a
5471 variant, -bhc, which does actually run the callout code, including
5472 consulting and updating the callout cache.
5474 70. Added support for saslauthd authentication, courtesy of Alexander
5477 71. If statvfs() failed on the spool or log directories while checking their
5478 size for availability, Exim confusingly gave the error "space shortage".
5479 Furthermore, in debugging mode it crashed with a floating point exception.
5480 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5481 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5482 serious problem, Exim now writes to the main and panic logs when this
5483 happens, with details of the failure. It then refuses to accept the
5484 incoming message, giving the message "spool directory problem" or "log
5485 directory problem" with a 421 code for SMTP messages.
5487 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5488 0, 1, and 2 exist, because some OS complain for execs without them (see
5489 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5490 descriptors. However, the code omitted to check that the open succeeded,
5491 causing mysterious errors if for some reason the permissions on /dev/null
5492 got screwed. Now Exim writes a message to the main and panic logs, and
5493 bombs out if it can't open /dev/null.
5495 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5496 interact so that it is all more flexible. It is supposed to remain
5497 backwards compatible. Also added extra_local_interfaces.
5499 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5500 to bomb out with an assertion failure - to the client this appears as a
5501 connection drop. This problem occurs in the part of the code that was taken
5502 from the Samba project. Fortunately, the assertion is in a very simple
5503 function, so I have fixed this by reproducing the function inline in the
5504 one place where it is called, and arranging for authentication to fail
5505 instead of killing the process with assert().
5507 75. The SPA client code was not working when the server requested OEM rather
5508 than Unicode encoding.
5510 76. Added code to make require_files with a specific uid setting more usable in
5511 the case where statting the file as root fails - usually a non-root-mounted
5512 NFS file system. When this happens and the failure is EACCES, Exim now
5513 forks a subprocess and does the per-uid checking as the relevant uid.
5515 77. Added process_log_path.
5517 78. If log_file_path was not explicitly set, a setting of check_log_space or
5518 check_log_inodes was ignored.
5520 79. If a space check for the spool or log partitions fails, the incident is now
5521 logged. Of course, in the latter case the data may get lost...
5523 80. Added the %p formatting code to string_format() so that it can be used to
5524 print addresses in debug_print(). Adjusted all the address printing in the
5525 debugging in store.c to use %p rather than %d.
5527 81. There was a concern that a line of code in smtp_in.c could overflow a
5528 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5529 initially expressed, the concern was not well-founded, because trailing
5530 spaces are removed early. However, if the trailing spaces were followed by
5531 a NULL, they did not get removed, so the overflow was possible. Two fixes
5534 (a) I re-wrote the offending code in a cleaner fashion.
5535 (b) If an incoming SMTP command contains a NULL character, it is rejected
5538 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5539 runs initgroups(), so that if the Exim user is in any additional groups,
5540 they will be used during message reception.
5546 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob