1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Solaris build fix for Oracle's LDAP libraries.
8 Bugzilla 1109, patch from Stephen Usher.
10 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
12 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
15 TF/02 Fix a couple more cases where we did not log the error message
16 when unlink() failed. See also change 4.74-TF/03.
18 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
19 lock up or crash if it happened to be inside a call to libc when it
20 got a SIGUSR1 from exiwhat.
22 The SIGUSR1 handler appends the current process status to the process
23 log which is later printed by exiwhat. It used to use the general
24 purpose logging code to do this, but several functions it calls are
27 The new output code in the SIGUSR1 handler is specific to the process
28 log, and simple enough that it's easy to inspect for signal safety.
29 Removing some special cases also simplifies the general logging code.
30 Removing the spurious timestamps from the process log simplifies
33 TF/04 Improved ratelimit ACL condition.
35 The /noupdate option has been deprecated in favour of /readonly which
36 has clearer semantics. The /leaky, /strict, and /readonly update modes
37 are mutually exclusive. The update mode is no longer included in the
38 database key; it just determines when the database is updated. (This
39 means that when you upgrde Exim will forget old rate measurements.)
41 Exim now checks that the per_* options are used with an update mode that
42 makes sense for the current ACL. For example, when Exim is processing a
43 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
44 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
45 must specify per_mail/readonly. If you omit the update mode it defaults to
46 /leaky where that makes sense (as before) or /readonly where required.
48 The /noupdate option is now undocumented but still supported for
49 backwards compatibility. It is equivalent to /readonly except that in
50 ACLs where /readonly is required you may specify /leaky/noupdate or
51 /strict/noupdate which are treated the same as /readonly.
53 A useful new feature is the /count= option. This is a generalization
54 of the per_byte option, so that you can measure the throughput of other
55 aggregate values. For example, the per_byte option is now equivalent
56 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
58 The per_rcpt option has been generalized using the /count= mechanism
59 (though it's more complicated than the per_byte equivalence). When it is
60 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
61 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
62 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
63 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
64 Note that using per_rcpt with a non-readonly update mode in more than
65 one ACL will cause the recipients to be double-counted. (The per_mail
66 and per_byte options don't have this problem.)
68 The handling of very low rates has changed slightly. If the computed rate
69 is less than the event's count (usually one) then this event is the first
70 after a long gap. In this case the rate is set to the same as this event's
71 count, so that the first message of a spam run is counted properly.
73 The major new feature is a mechanism for counting the rate of unique
74 events. The new per_addr option counts the number of different
75 recipients that someone has sent messages to in the last time period. It
76 behaves like per_rcpt if all the recipient addresses are different, but
77 duplicate recipient addresses do not increase the measured rate. Like
78 the /count= option this is a general mechanism, so the per_addr option
79 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
80 example, measure the rate that a client uses different sender addresses
81 with the options per_mail/unique=$sender_address. There are further
82 details in the main documentation.
84 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
86 TF/06 Removed a few PCRE remnants.
88 TF/07 Automatically extract Exim's version number from tags in the git
89 repository when doing development or release builds.
91 PP/02 Raise smtp_cmd_buffer_size to 16kB. Patch from Paul Fisher.
94 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
95 Heavily based on revision 40f9a89a from Simon Arlott's tree.
98 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
100 PP/05 Variable $av_failed, true if the AV scanner deferred.
101 Patch from John Horne.
108 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
110 PP/02 Harmonised TLS library version reporting; only show if debugging.
111 Layout now matches that introduced for other libraries in 4.74 PP/03.
113 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
115 PP/04 New "dns_use_edns0" global option.
117 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
120 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
121 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
123 TK/01 Updated PolarSSL code to 0.14.2.
124 Bugzilla 1097. Patch from Andreas Metzler.
126 PP/07 Catch divide-by-zero in ${eval:...}.
129 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
132 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
133 format-string attack -- SECURITY: remote arbitrary code execution.
135 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
136 time unintentionally subject to list matching rules, letting the header
137 cause arbitrary Exim lookups (of items which can occur in lists, *not*
138 arbitrary string expansion). This allowed for information disclosure.
140 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
141 INT_MIN/-1 -- value coerced to INT_MAX.
147 NM/01 Workround for PCRE version dependency in version reporting
150 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
151 This fixes portability to compilers other than gcc, notably
152 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
154 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
155 makefiles for portability to HP-UX and POSIX correctness.
157 PP/01 Permit LOOKUP_foo enabling on the make command-line.
158 Also via indented variable definition in the Makefile.
159 (Debugging by Oliver Heesakkers).
161 PP/02 Restore caching of spamd results with expanded spamd_address.
162 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
164 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
165 Improves build reliability. Fix from: Frank Elsner
167 NM/02 Fix wide character breakage in the rfc2047 coding
168 Fixes bug 1064. Patch from Andrey N. Oktyabrski
170 NM/03 Allow underscore in dnslist lookups
171 Fixes bug 1026. Patch from Graeme Fowler
173 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
174 Code patches from Adam Ciarcinski of NetBSD.
176 NM/04 Fixed exiqgrep to cope with mailq missing size issue
179 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
180 is logged, to avoid truncation. Patch from John Horne.
182 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
183 Patch from Jakob Hirsch.
185 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
186 SQL string expansion failure details.
187 Patch from Andrey Oktyabrski.
189 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
190 Patch from Simon Arlott.
192 PP/09 New lookups functionality failed to compile on old gcc which rejects
193 extern declarations in function scope.
194 Patch from Oliver Fleischmann
196 PP/10 Use sig_atomic_t for flags set from signal handlers.
197 Check getgroups() return and improve debugging.
198 Fixed developed for diagnosis in bug 927 (which turned out to be
201 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
202 Patch from Mark Zealey.
204 PP/12 Bugzilla 1056: Improved spamd server selection.
205 Patch from Mark Zealey.
207 PP/13 Bugzilla 1086: Deal with maildir quota file races.
208 Based on patch from Heiko Schlittermann.
210 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
211 Patch from Uwe Doering, sign-off by Michael Haardt.
213 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
214 variable declaration deep within a block. Bug and patch from
217 PP/15 lookups-Makefile IRIX compatibilty coercion.
219 PP/16 Make DISABLE_DKIM build knob functional.
221 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
222 Patch by Simon Arlott
224 TF/03 Fix valgrind.h portability to C89 compilers that do not support
225 variable argument macros. Our copy now differs from upstream.
231 TF/01 Failure to get a lock on a hints database can have serious
232 consequences so log it to the panic log.
234 TF/02 Log LMTP confirmation messages in the same way as SMTP,
235 controlled using the smtp_confirmation log selector.
237 TF/03 Include the error message when we fail to unlink a spool file.
239 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
240 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
241 for maintaining out-of-tree patches for some time.
243 PP/01 Bugzilla 139: Documentation and portability issues.
244 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
245 Handle per-OS dynamic-module compilation flags.
247 PP/02 Let /dev/null have normal permissions.
248 The 4.73 fixes were a little too stringent and complained about the
249 permissions on /dev/null. Exempt it from some checks.
250 Reported by Andreas M. Kirchwitz.
252 PP/03 Report version information for many libraries, including
253 Exim version information for dynamically loaded libraries. Created
254 version.h, now support a version extension string for distributors
255 who patch heavily. Dynamic module ABI change.
257 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
258 privilege escalation vulnerability whereby the Exim run-time user
259 can cause root to append content of the attacker's choosing to
262 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
265 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
266 If dropping privileges for untrusted macros, we disabled normal logging
267 on the basis that it would fail; for the Exim run-time user, this is not
268 the case, and it resulted in successful deliveries going unlogged.
269 Fixed. Reported by Andreas Metzler.
275 PP/01 Date: & Message-Id: revert to normally being appended to a message,
276 only prepend for the Resent-* case. Fixes regression introduced in
277 Exim 4.70 by NM/22 for Bugzilla 607.
279 PP/02 Include check_rfc2047_length in configure.default because we're seeing
280 increasing numbers of administrators be bitten by this.
282 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
284 PP/03 Bugzilla 994: added openssl_options main configuration option.
286 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
288 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
290 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
292 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
293 without a peer certificate, leading to a segfault because of an
294 assumption that peers always have certificates. Be a little more
295 paranoid. Problem reported by Martin Tscholak.
297 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
298 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
299 NB: ClamAV planning to remove STREAM in "middle of 2010".
300 CL also introduces -bmalware, various -d+acl logging additions and
301 more caution in buffer sizes.
303 PP/09 Implemented reverse_ip expansion operator.
305 PP/10 Bugzilla 937: provide a "debug" ACL control.
307 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
309 PP/12 Bugzilla 973: Implement --version.
311 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
313 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
315 PP/15 Bugzilla 816: support multiple condition rules on Routers.
317 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
318 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
319 ignore trailing whitespace.
321 JJ/02 prevent non-panic DKIM error from being sent to paniclog
323 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
326 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
327 Notification from Dr Andrew Aitchison.
329 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
330 ExtendedDetectionInfo response format.
331 Notification from John Horne.
333 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
336 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
337 XSL and documented dependency on system catalogs, with examples of how
340 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
343 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
344 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
345 configuration file which is writeable by the Exim user or group.
347 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
348 of configuration files to cover files specified with the -C option if
349 they are going to be used with root privileges, not just the default
352 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
353 option (effectively making it always true).
355 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
356 files to be used while preserving root privileges.
358 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
359 that rogue child processes cannot use them.
361 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
362 run-time user, instead of root.
364 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
365 Exim run-time user without dropping privileges.
367 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
368 result string, instead of calling string_vformat() twice with the same
371 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
372 for other users. Others should always drop root privileges if they use
373 -C on the command line, even for a whitelisted configure file.
375 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
377 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
383 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
384 $data_path, and $header_path variables; fixed documentation bugs and
387 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
388 exipick to access non-standard spools, including the "frozen" queue
391 NM/01 Bugzilla 965: Support mysql stored procedures.
392 Patch from Alain Williams
394 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
396 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
397 Patch from Andreas Metzler
399 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
400 Patch from Kirill Miazine
402 NM/05 Bugzilla 671: Added umask to procmail example.
404 JJ/03 installed exipick 20100323.0, fixing doc bug
406 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
407 directory. Notification and patch from Dan Rosenberg.
409 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
411 TK/02 Improve log output when DKIM signing operation fails.
413 MH/01 Treat the transport option dkim_domain as a colon separated
414 list, not as a single string, and sign the message with each element,
415 omitting multiple occurences of the same signer.
417 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
418 Bugzilla 985, 986. Patch by Simon Arlott
420 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
421 Patch by Simon Arlott
423 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
424 MBX locking. Notification from Dan Rosenberg.
430 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
432 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
434 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
436 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
438 NM/04 Bugzilla 915: Flag broken perl installation during build.
444 TK/01 Added patch by Johannes Berg that expands the main option
445 "spamd_address" if it starts with a dollar sign.
447 TK/02 Write list of recipients to X-Envelope-Sender header when building
448 the mbox-format spool file for content scanning (suggested by Jakob
451 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
452 (http://www.dcc-servers.net/) support via dccifd. Activated by
453 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
455 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
456 by Mark Daniel Reidel <mr@df.eu>.
458 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
459 When building exim an external PCRE library is now needed -
460 PCRE is a system library on the majority of modern systems.
461 See entry on PCRE_LIBS in EDITME file.
463 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
464 conversation. Added nologin parameter to request.
465 Patch contributed by Kirill Miazine.
467 TF/01 Do not log submission mode rewrites if they do not change the address.
469 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
471 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
472 log files in place. Contributed by Roberto Lima.
474 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
476 TF/03 Bugzilla 615: When checking the local_parts router precondition
477 after a local_part_suffix or local_part_prefix option, Exim now
478 does not use the address's named list lookup cache, since this
479 contains cached lookups for the whole local part.
481 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
482 Robert Millan. Documentation is in experimental-spec.txt.
484 TF/04 Bugzilla 668: Fix parallel build (make -j).
486 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
488 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
489 Patch provided by Jan Srzednicki.
491 TF/05 Leading white space used to be stripped from $spam_report which
492 wrecked the formatting. Now it is preserved.
494 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
495 that they are available at delivery time.
497 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
499 TF/08 TLS error reporting now respects the incoming_interface and
500 incoming_port log selectors.
502 TF/09 Produce a more useful error message if an SMTP transport's hosts
503 setting expands to an empty string.
505 NM/06 Bugzilla 744: EXPN did not work under TLS.
506 Patch provided by Phil Pennock.
508 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
509 Patch provided by Richard Godbee.
511 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
512 acl_smtp_notquit, added index entry.
514 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
515 Patch provided by Eugene Bujak.
517 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
518 accept(). Patch provided by Maxim Dounin.
520 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
521 Patch provided by Phil Pennock.
523 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
525 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
526 Patch provided by Brad "anomie" Jorsch.
528 NM/14 Bugzilla 622: Added timeout setting to transport filter.
529 Patch provided by Dean Brooks.
531 TK/05 Add native DKIM support (does not depend on external libraries).
533 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
534 Patch provided by Graeme Fowler.
536 NM/16 Bugzilla 851: Documentation example syntax fix.
538 NM/17 Changed NOTICE file to remove references to embedded PCRE.
540 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
543 NM/19 Bugzilla 745: TLS version reporting.
544 Patch provided by Phil Pennock.
546 NM/20 Bugzilla 167: bool: condition support.
547 Patch provided by Phil Pennock.
549 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
550 clients. Patch provided by Phil Pennock.
552 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
553 Patch provided by Brad "anomie" Jorsch.
555 NM/23 Bugzilla 687: Fix misparses in eximstats.
556 Patch provided by Heiko Schlittermann.
558 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
559 Patch provided by Heiko Schlittermann.
561 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
562 plus update to original patch.
564 NM/26 Bugzilla 799: Documentation correction for ratelimit.
566 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
567 Patch provided by David Brownlee.
569 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
571 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
573 NM/30 Bugzilla 888: TLS documentation bugfixes.
575 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
577 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
578 Unlike the original bugzilla I have changed all shell scripts in src tree.
580 NM/33 Bugzilla 898: Transport filter timeout fix.
581 Patch by Todd Rinaldo.
583 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
584 Patch by Serge Demonchaux.
586 NM/35 Bugzilla 39: Base64 decode bug fixes.
587 Patch by Jakob Hirsch.
589 NM/36 Bugzilla 909: Correct connect() call in dcc code.
591 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
593 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
595 NM/39 Bugzilla 911: Fixed MakeLinks build script.
601 TK/01 Add preliminary DKIM support. Currently requires a forked version of
602 ALT-N's libdkim that I have put here:
603 http://duncanthrax.net/exim-experimental/
605 Note to Michael Haardt: I had to rename some vars in sieve.c. They
606 were called 'true' and it seems that C99 defines that as a reserved
607 keyword to be used with 'bool' variable types. That means you could
608 not include C99-style headers which use bools without triggering
609 build errors in sieve.c.
611 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
612 as mailq or other aliases. Changed the --help handling significantly
613 to do whats expected. exim_usage() emits usage/help information.
615 SC/01 Added the -bylocaldomain option to eximstats.
617 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
619 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
621 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
627 PH/01 Another patch from the Sieve maintainer.
629 PH/02 When an IPv6 address is converted to a string for single-key lookup
630 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
631 dots are used instead of colons so that keys in lsearch files need not
632 contain colons. This was done some time before quoting was made available
633 in lsearch files. However, iplsearch files do require colons in IPv6 keys
634 (notated using the quote facility) so as to distinguish them from IPv4
635 keys. This meant that lookups for IP addresses in host lists did not work
636 for iplsearch lookups.
638 This has been fixed by arranging for IPv6 addresses to be expressed with
639 colons if the lookup type is iplsearch. This is not incompatible, because
640 previously such lookups could never work.
642 The situation is now rather anomolous, since one *can* have colons in
643 ordinary lsearch keys. However, making the change in all cases is
644 incompatible and would probably break a number of configurations.
646 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
649 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
650 conversion specification without a maximum field width, thereby enabling
651 a rogue spamd server to cause a buffer overflow. While nobody in their
652 right mind would setup Exim to query an untrusted spamd server, an
653 attacker that gains access to a server running spamd could potentially
654 exploit this vulnerability to run arbitrary code as the Exim user.
656 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
657 $primary_hostname instead of what libspf2 thinks the hosts name is.
659 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
660 a directory entry by the name of the lookup key. Previously, if a
661 symlink pointed to a non-existing file or a file in a directory that
662 Exim lacked permissions to read, a lookup for a key matching that
663 symlink would fail. Now it is enough that a matching directory entry
664 exists, symlink or not. (Bugzilla 503.)
666 PH/03 The body_linecount and body_zerocount variables are now exported in the
669 PH/04 Added the $dnslist_matched variable.
671 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
672 This means they are set thereafter only if the connection becomes
675 PH/06 Added the client_condition to authenticators so that some can be skipped
676 by clients under certain conditions.
678 PH/07 The error message for a badly-placed control=no_multiline_responses left
679 "_responses" off the end of the name.
681 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
683 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
684 (without spaces) instead of just copying the configuration text.
686 PH/10 Added the /noupdate option to the ratelimit ACL condition.
688 PH/11 Added $max_received_linelength.
690 PH/12 Added +ignore_defer and +include_defer to host lists.
692 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
693 way in which PCRE > 7.0 is built.
695 PH/14 Implemented queue_only_load_latch.
697 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
698 MAIL command. The effect was to mangle the value on 64-bit systems.
700 PH/16 Another patch from the Sieve maintainer.
702 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
704 PH/18 If a system quota error occurred while trying to create the file for
705 a maildir delivery, the message "Mailbox is full" was not appended to the
706 bounce if the delivery eventually timed out. Change 4.67/27 below applied
707 only to a quota excession during the actual writing of the file.
709 PH/19 It seems that peer DN values may contain newlines (and other non-printing
710 characters?) which causes problems in log lines. The DN values are now
711 passed through string_printing() before being added to log lines.
713 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
714 and InterBase are left for another time.)
716 PH/21 Added message_body_newlines option.
718 PH/22 Guard against possible overflow in moan_check_errorcopy().
720 PH/23 POSIX allows open() to be a macro; guard against that.
722 PH/24 If the recipient of an error message contained an @ in the local part
723 (suitably quoted, of course), incorrect values were put in $domain and
724 $local_part during the evaluation of errors_copy.
730 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
731 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
734 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
735 issue a MAIL command.
737 PH/02 In an ACL statement such as
739 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
741 if a client was not listed at all, or was listed with a value other than
742 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
743 the condition was not true (as it should be), so access was not denied.
744 The bug was that the ! inversion was incorrectly passed on to the second
745 item. This has been fixed.
747 PH/03 Added additional dnslists conditions == and =& which are different from
748 = and & when the dns lookup returns more than one IP address.
750 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
751 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
753 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
754 FSYNC, which compiles an option called disable_fsync that allows for
755 bypassing fsync(). The documentation is heavily laced with warnings.
757 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
759 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
760 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
761 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
762 including adding "make clean"; (3) Added -fPIC when compiling the test
763 dynamically loaded module, to get rid of a warning.
765 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
766 message fails, move_frozen_messages = true and ignore_bounce_errors_after
767 = 0s. The bug is otherwise harmless.
769 PH/07 There was a bug in the dovecot authenticator such that the value of
770 $auth1 could be overwritten, and so not correctly preserved, after a
771 successful authentication. This usually meant that the value preserved by
772 the server_setid option was incorrect.
774 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
776 PH/09 Installed PCRE release 7.0.
778 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
779 run for batched SMTP input. It is now run at the start of every message
780 in the batch. While fixing this I discovered that the process information
781 (output by running exiwhat) was not always getting set for -bs and -bS
782 input. This is fixed, and it now also says "batched" for BSMTP.
784 PH/11 Added control=no_pipelining.
786 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
787 patch, slightly modified), and move the expansion of helo_data till after
788 the connection is made in the smtp transport (so it can use these
791 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
793 PH/14 Added log_selector = +pid.
795 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
797 PH/16 Add ${if forany and ${if forall.
799 PH/17 Added dsn_from option to vary the From: line in DSNs.
801 PH/18 Flush SMTP output before performing a callout, unless control =
802 no_callout_flush is set.
804 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
805 was true (the default) a successful delivery failed to delete the retry
806 item, thus causing premature timeout of the address. The bug is now
809 PH/20 Added hosts_avoid_pipelining to the smtp transport.
811 PH/21 Long custom messages for fakedefer and fakereject are now split up
812 into multiline reponses in the same way that messages for "deny" and
813 other ACL rejections are.
815 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
816 with slight modification.
818 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
819 draft, changing the syntax and factoring some duplicate code".
821 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
822 for deliveries of the second and subsequent messages over the same SMTP
825 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
826 ${reduce, with only minor "tidies".
828 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
830 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
831 expansion side effects.
833 PH/27 When a message times out after an over-quota error from an Exim-imposed
834 quota, the bounce message says "mailbox is full". This message was not
835 being given when it was a system quota that was exceeded. It now should
838 MH/03 Made $recipients available in local_scan(). local_scan() already has
839 better access to the recipient list through recipients_list[], but
840 $recipients can be useful in postmaster-provided expansion strings.
842 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
843 in the case of a MAIL command with additional options following the
844 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
845 were accidentally chopped off.
847 PH/29 SMTP synchronization checks are implemented when a command is read -
848 there is a check that no more input is waiting when there shouldn't be
849 any. However, for some commands, a delay in an ACL can mean that it is
850 some time before the response is written. In this time, more input might
851 arrive, invalidly. So now there are extra checks after an ACL has run for
852 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
853 pipelining has not been advertised.
855 PH/30 MH's patch to allow iscntrl() characters to be list separators.
857 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
858 returned in the SMTP response when smtp_return_error_details was false.
861 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
862 instead of the C I/O that was originally supplied, because problems were
865 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
866 Exim which did not show up earlier: it was assuming that a call to
867 SSL_CTX_set_info_callback() might give an error value. In fact, there is
868 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
869 was a macro that became an assignment, so it seemed to work. This has
870 changed to a proper function call with a void return, hence the compile
871 error. Exim's code has been fixed.
873 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
876 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
878 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
880 PH/37 If a message is not accepted after it has had an id assigned (e.g.
881 because it turns out to be too big or there is a timeout) there is no
882 "Completed" line in the log. When some messages of this type were
883 selected by exigrep, they were listed as "not completed". Others were
884 picked up by some special patterns. I have improved the selection
885 criteria to be more general.
887 PH/38 The host_find_failed option in the manualroute router can now be set
888 to "ignore", to completely ignore a host whose IP address cannot be
889 found. If all hosts are ignored, the behaviour is controlled by the new
890 host_all_ignored option.
892 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
893 homing or because of multiple MX records with /mx) generated more than
894 one IP address, and the following item turned out to be the local host,
895 all the secondary addresses of the first item were incorrectly removed
896 from the list, along with the local host and any following hosts (which
897 is what is supposed to happen).
899 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
900 whoever called Exim into the -H file. In the case of the daemon it was
901 behaving confusingly. When first started, it used values for whoever
902 started the daemon, but after a SIGHUP it used the Exim user (because it
903 calls itself on a restart). I have changed the code so that it now always
906 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
907 message are rejected with the same error (e.g. no authentication or bad
908 sender address), and a DATA command is nevertheless sent (as can happen
909 with PIPELINING or a stupid MUA), the error message that was given to the
910 RCPT commands is included in the rejection of the DATA command. This is
911 intended to be helpful for MUAs that show only the final error to their
914 PH/42 Another patch from the Sieve maintainer.
916 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
917 Eximstats - Fixed some broken HTML links and added missing column headers
919 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
920 columns for Rejects, Temp Rejects, Ham, and Spam rows.
922 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
924 PH/43 Yet another patch from the Sieve maintainer.
926 PH/44 I found a way to check for a TCP/IP connection going away before sending
927 the response to the final '.' that terminates a message, but only in the
928 case where the client has not sent further data following the '.'
929 (unfortunately, this is allowed). However, in many cases there won't be
930 any further data because there won't be any more messages to send. A call
931 to select() can be used: if it shows that the input is "ready", there is
932 either input waiting, or the socket has been closed. An attempt to read
933 the next input character can distinguish the two cases. Previously, Exim
934 would have sent an OK response which the client would never have see.
935 This could lead to message repetition. This fix should cure that, at
936 least in a lot of common cases.
938 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
939 advertised in response to EHLO.
945 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
946 fixed by 4.65/MH/01 (is this a record?) are fixed:
948 (i) An empty string was always treated as zero by the numeric comparison
949 operators. This behaviour has been restored.
951 (ii) It is documented that the numeric comparison operators always treat
952 their arguments as decimal numbers. This was broken in that numbers
953 starting with 0 were being interpreted as octal.
955 While fixing these problems I realized that there was another issue that
956 hadn't been noticed. Values of message_size_limit (both the global option
957 and the transport option) were treated as octal if they started with 0.
958 The documentation was vague. These values are now always treated as
959 decimal, and I will make that clear in the documentation.
965 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
966 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
969 MH/01 Don't check that the operands of numeric comparison operators are
970 integers when their expansion is in "skipping" mode (fixes bug
971 introduced by 4.64-PH/07).
973 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
974 child addresses, Exim now panics and dies. Previously, because the count
975 is held in a short int, deliveries were likely to be lost. As such a
976 large number of recipients for a single message is ridiculous
977 (performance will be very, very poor), I have chosen to impose a limit
978 rather than extend the field.
984 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
985 leftover -K file (the existence of which was triggered by #402).
986 While we were at it, introduced process PID as part of the -K
987 filename. This should rule out race conditions when creating
990 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
991 processing considerably. Previous code took too long for large mails,
992 triggering a timeout which in turn triggers #401.
994 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
995 in the DK code in transports.c. sendfile() is not really portable,
996 hence the _LINUX specificness.
998 TF/01 In the add_headers option to the mail command in an Exim filter,
999 there was a bug that Exim would claim a syntax error in any
1000 header after the first one which had an odd number of characters
1003 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1004 callout verification, Exim cached a "reject" for the entire domain. This
1005 is correct for most verifications, but it is not correct for a recipient
1006 verification with use_sender or use_postmaster set, because in that case
1007 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1008 case of MAIL FROM:<> rejection from other early rejections (e.g.
1009 rejection of HELO). When verifying a recipient using a non-null MAIL
1010 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1011 Whatever the result of the callout, the value of the domain cache is
1012 left unchanged (for any other kind of callout, getting as far as trying
1013 RCPT means that the domain itself is ok).
1015 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1018 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1019 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1020 session, the daemon ignores ECONNECT errors and logs others; it now
1021 ignores EPIPE as well.
1023 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1024 (quoted-printable decoding).
1026 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1027 later the small subsequent patch to fix an introduced bug.
1029 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1031 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1033 PH/08 An error is now given if message_size_limit is specified negative.
1035 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1036 to be given (somewhat) arbitrary names.
1038 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1041 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1042 miscellaneous code fixes
1044 PH/10 Added the log_reject_target ACL modifier to specify where to log
1047 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1048 hostname. This is wrong, because it relates to the incoming message (and
1049 probably the interface on which it is arriving) and not to the outgoing
1050 callout (which could be using a different interface). This has been
1051 changed to use the value of the helo_data option from the smtp transport
1052 instead - this is what is used when a message is actually being sent. If
1053 there is no remote transport (possible with a router that sets up host
1054 addresses), $smtp_active_hostname is used.
1056 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1057 tweaks were necessary in order to get it to work (see also 21 below):
1058 (a) The code assumed that strncpy() returns a negative number on buffer
1059 overflow, which isn't the case. Replaced with Exim's string_format()
1061 (b) There were several signed/unsigned issues. I just did the minimum
1062 hacking in of casts. There is scope for a larger refactoring.
1063 (c) The code used strcasecmp() which is not a standard C function.
1064 Replaced with Exim's strcmpic() function.
1065 (d) The code set only $1; it now sets $auth1 as well.
1066 (e) A simple test gave the error "authentication client didn't specify
1067 service in request". It would seem that Dovecot has changed its
1068 interface. Fortunately there's a specification; I followed it and
1069 changed what the client sends and it appears to be working now.
1071 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1074 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1075 address A is aliased to B and C, where B exists and C does not. Without
1076 -v the output is "A verified" because verification stops after a
1077 successful redirection if more than one address is generated. However,
1078 with -v the child addresses are also verified. Exim was outputting "A
1079 failed to verify" and then showing the successful verification for C,
1080 with its parentage. It now outputs "B failed to verify", showing B's
1081 parentage before showing the successful verification of C.
1083 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1084 look up a TXT record in a specific list after matching in a combined
1087 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1088 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1089 they consult the DNS. I had assumed they would set it the way they
1090 wanted; and indeed my experiments on Linux seem to show that in some
1091 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1092 To be on the safe side, however, I have now made the interface to
1093 host_find_byname() similar to host_find_bydns(), with an argument
1094 containing the DNS resolver options. The host_find_byname() function now
1095 sets these options at its start, just as host_find_bydns() does. The smtp
1096 transport options dns_qualify_single and dns_search_parents are passed to
1097 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1098 of host_find_byname() use the default settings of RES_DEFNAMES
1099 (qualify_single) but not RES_DNSRCH (search_parents).
1101 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1102 spool_read_header() do less string testing, by means of a preliminary
1103 switch on the second character of optional "-foo" lines. (This is
1104 overdue, caused by the large number of possibilities that now exist.
1105 Originally there were few.) While I was there, I also converted the
1106 str(n)cmp tests so they don't re-test the leading "-" and the first
1107 character, in the hope this might squeeze out yet more improvement.
1109 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1110 flag allowing group syntax was set by the header_syntax check but not
1111 turned off, possible causing trouble later; (2) The flag was not being
1112 set at all for the header_verify test, causing "group"-style headers to
1113 be rejected. I have now set it in this case, and also caused header_
1114 verify to ignore an empty address taken from a group. While doing this, I
1115 came across some other cases where the code for allowing group syntax
1116 while scanning a header line wasn't quite right (mostly, not resetting
1117 the flag correctly in the right place). These bugs could have caused
1118 trouble for malformed header lines. I hope it is now all correct.
1120 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1121 with the "reply" argument non-NULL. The code, however (which originally
1122 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1123 but it didn't always do it. This confused somebody who was copying the
1124 code for some other use. I have removed all the tests.
1126 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1127 feature that was used to support insecure browsers during the U.S. crypto
1128 embargo. It requires special client support, and Exim is probably the
1129 only MTA that supported it -- and would never use it because real RSA is
1130 always available. This code has been removed, because it had the bad
1131 effect of slowing Exim down by computing (never used) parameters for the
1132 RSA_EXPORT functionality.
1134 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1135 authenticator to fail if there's a tab character in the incoming data
1136 (there should never be unless someone is messing about, as it's supposed
1137 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1138 if the connection is using TLS or if the remote IP is the same as the
1139 local IP, and the "valid-client-cert option" if a client certificate has
1142 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1143 authenticators. This can be used for authorization after authentication
1144 succeeds. (In the case of plaintext, it servers for both authentication
1147 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1148 if any retry times were supplied.
1150 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1151 connection, where there is no client IP address to check. In this
1152 situation, the verify now always succeeds.
1154 PH/25 Applied John Jetmore's -Mset patch.
1156 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1158 PH/27 In a string expansion for a processed (not raw) header when multiple
1159 headers of the same name were present, leading whitespace was being
1160 removed from all of them, but trailing whitespace was being removed only
1161 from the last one. Now trailing whitespace is removed from each header
1162 before concatenation. Completely empty headers in a concatenation (as
1163 before) are ignored.
1165 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1166 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1168 PH/29 [Removed. This was a change that I later backed out, and forgot to
1169 correct the ChangeLog entry (that I had efficiently created) before
1170 committing the later change.]
1172 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1173 address errors (4xx response to RCPT) over the same connection as other
1174 messages routed to the same hosts. Such deliveries are always "forced",
1175 so retry times are not inspected. This resulted in far too many retries
1176 for the affected addresses. The effect occurred only when there were more
1177 hosts than the hosts_max_try setting in the smtp transport when it had
1178 the 4xx errors. Those hosts that it had tried were not added to the list
1179 of hosts for which the message was waiting, so if all were tried, there
1180 was no problem. Two fixes have been applied:
1182 (i) If there are any address or message errors in an SMTP delivery, none
1183 of the hosts (tried or untried) are now added to the list of hosts
1184 for which the message is waiting, so the message should not be a
1185 candidate for sending over the same connection that was used for a
1186 successful delivery of some other message. This seems entirely
1187 reasonable: after all the message is NOT "waiting for some host".
1188 This is so "obvious" that I'm not sure why it wasn't done
1189 previously. Hope I haven't missed anything, but it can't do any
1190 harm, as the worst effect is to miss an optimization.
1192 (ii) If, despite (i), such a delivery is accidentally attempted, the
1193 routing retry time is respected, so at least it doesn't keep
1194 hammering the server.
1196 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1197 in ${readsocket because some servers need this prod.
1199 PH/32 Added some extra debug output when updating a wait-xxx database.
1201 PH/33 The hint "could be header name not terminated by colon", which has been
1202 given for certain expansion errors for a long time, was not being given
1203 for the ${if def:h_colon_omitted{... case.
1205 PH/34 The spec says: "With one important exception, whenever a domain list is
1206 being scanned, $domain contains the subject domain." There was at least
1207 one case where this was not true.
1209 PH/35 The error "getsockname() failed: connection reset by peer" was being
1210 written to the panic log as well as the main log, but it isn't really
1211 panic-worthy as it just means the connection died rather early on. I have
1212 removed the panic log writing for the ECONNRESET error when getsockname()
1215 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1216 runs only) independently of the message's sender address. This meant
1217 that, if the 4xx error was in fact related to the sender, a different
1218 message to the same recipient with a different sender could confuse
1219 things. In particualar, this can happen when sending to a greylisting
1220 server, but other circumstances could also provoke similar problems.
1221 I have changed the default so that the retry time for these errors is now
1222 based a combination of the sender and recipient addresses. This change
1223 can be overridden by setting address_retry_include_sender=false in the
1226 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1227 remote server are returned as part of bounce messages. This was not
1228 happening for LMTP over a pipe (the lmtp transport), but now it is the
1229 same for both kinds of LMTP.
1231 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1232 in header lines that were in fact CNAMEs. This is no longer the case.
1234 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1235 and queue runs started by the daemon processed all messages. This has
1236 been fixed so that -R and -S can now usefully be given with -q<time>.
1238 PH/40 Import PCRE release 6.7 (fixes some bugs).
1240 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1242 PH/42 Give an error if -q is specified more than once.
1244 PH/43 Renamed the variables $interface_address and $interface_port as
1245 $received_ip_address and $received_port, to make it clear that these
1246 values apply to message reception, and not to the outgoing interface when
1247 a message is delivered. (The old names remain recognized, of course.)
1249 PH/44 There was no timeout on the connect() call when using a Unix domain
1250 socket in the ${readsocket expansion. There now is.
1252 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1253 be meaningful with "accept".
1255 SC/01 Eximstats V1.43
1256 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1258 SC/02 Eximstats V1.44
1259 Use a glob alias rather than an array ref in the generated
1260 parser. This improves both readability and performance.
1262 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1263 Collect SpamAssassin and rejection statistics.
1264 Don't display local sender or destination tables unless
1265 there is data to show.
1266 Added average volumes into the top table text output.
1268 SC/04 Eximstats V1.46
1269 Collect data on the number of addresses (recipients)
1270 as well as the number of messages.
1272 SC/05 Eximstats V1.47
1273 Added 'Message too big' to the list of mail rejection
1274 reasons (thanks to Marco Gaiarin).
1276 SC/06 Eximstats V1.48
1277 Mainlog lines which have GMT offsets and are too short to
1278 have a flag are now skipped.
1280 SC/07 Eximstats V1.49 (Alain Williams)
1281 Added the -emptyok flag.
1283 SC/08 Eximstats V1.50
1284 Fixes for obtaining the IP address from reject messages.
1286 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1287 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1288 whitesspace changes from 4.64-PH/27
1290 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1293 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1294 are found, allow negative numbers in numeric criteria)
1296 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1298 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1299 to match changes made in 4.64-PH/43
1301 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1303 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1304 "domain" is actually an IP address, and force a failure. This locks out
1305 those revolvers/nameservers that support "A-for-A" lookups, in
1306 contravention of the specifications.
1308 PH/48 When a host name was looked up from an IP address, and the subsequent
1309 forward lookup of the name timed out, the host name was left in
1310 $sender_host_name, contrary to the specification.
1312 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1313 restricted to single-key lookups, Exim was not diagnosing an error if
1314 * or *@ was used with a query-style lookup.
1316 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1318 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1319 long ago, but noone interested enough thought of it. Let's just say that
1320 the "1.1" means that there are some new functions that weren't there at
1321 some point in the past.
1323 PH/51 Error processing for expansion failure of helo_data from an smtp
1324 transport during callout processing was broken.
1326 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1327 tested/used via the -bh/-bhc/-bs options.
1329 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1330 bug, fixed in subsequent PCRE releases).
1332 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1333 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1335 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1341 SC/01 Use a glob alias rather than an array ref in eximstats generated
1342 parser. This improves both readability and performance.
1344 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1345 Don't display local sender or destination tables in eximstats unless
1346 there is data to show.
1347 Added average volumes into the eximstats top table text output.
1349 SC/03 Collect data on the number of addresses (recipients) as well
1350 as the number of messages in eximstats.
1352 TF/01 Correct an error in the documentation for the redirect router. Exim
1353 does (usually) call initgroups() when daemonizing.
1355 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1356 with consistent privilege compared to when running as a daemon.
1358 TF/03 Note in the spec that $authenticated_id is not set for local
1359 submissions from trusted users.
1361 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1362 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1364 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1365 by adding some example configuration directives to the default
1366 configuration file. A little bit of work is required to uncomment the
1367 directives and define how usernames and passwords are checked, but
1368 there is now a framework to start from.
1370 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1371 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1372 without this. I don't know how relevant this is to other LDAP libraries.
1374 PH/02 Add the verb name to the "unknown ACL verb" error.
1376 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1378 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1380 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1381 directory not expanded when it should be if an expanded home directory
1382 was set for the address (which is overridden by the transport).
1384 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1387 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1388 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1389 because it is too late at that time, and has no effect.
1391 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1392 security issue with \' (bugzilla #107). I could not use the
1393 PQescapeStringConn() function, because it needs a PGconn value as one of
1396 PH/08 When testing addresses using -bt, indicate those final addresses that
1397 are duplicates that would not cause an additional delivery. At least one
1398 person was confused, thinking that -bt output corresponded to deliveries.
1399 (Suppressing duplicates isn't a good idea as you lose the information
1400 about possibly different redirections that led to the duplicates.)
1402 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1403 systems where poll() doesn't work, in particular OS X.
1405 PH/10 Added more information to debugging output for retry time not reached.
1407 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1408 operations in malware.c.
1410 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1413 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1414 syslog_duplication set false, log lines that would normally be written
1415 both the the main log and to the reject log were not written to syslog at
1418 PH/14 In the default configuration, change the use of "message" in ACL warn
1419 statements to "add_header".
1421 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1422 not followed by a command (e.g. "seen endif").
1424 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1425 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1428 PH/17 Added extra conditions to the default value of delay_warning_condition
1432 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1433 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1434 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1437 The Auto-Submitted: and various List- headers are standardised, whereas I
1438 don't think Precedence: ever was.
1440 PH/18 Refactored debugging code in route_finduser() to show more information,
1441 in particular, the error code if getpwnam() issues one.
1443 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1444 This is apparently needed in addition to the PH/07 change above to avoid
1445 any possible encoding problems.
1447 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1448 but not after initializing Perl.
1450 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1451 output them only if debugging. By default they are written stderr,
1452 apparently, which is not desirable.
1454 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1457 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1460 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1462 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1463 authenticated or an ident call has been made. Suppress the default
1464 values for $authenticated_id and $authenticated_sender (but permit -oMai
1465 and -oMas) when testing with -bh.
1467 PH/24 Re-jigged the order of the tests in the default configuration so that the
1468 tests for valid domains and recipients precede the DNS black list and CSA
1469 tests, on the grounds that those ones are more expensive.
1471 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1472 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1473 This bug exists in every version of Exim that I still have, right back to
1476 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1477 However, an attempt to turn on case-sensitivity in a regex key by
1478 including (?-i) didn't work because the subject string was already
1479 lowercased, and the effects were non-intuitive. It turns out that a
1480 one-line patch can be used to allow (?-i) to work as expected.
1486 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1487 other effects) broke the use of negated acl sub-conditions.
1489 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1492 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1493 "Deny" causes Exim to reject the incoming connection with a 554 error.
1494 Unfortunately, if there is a major crisis, such as a disk failure,
1495 tcp-wrappers gives "deny", whereas what one would like would be some
1496 kind of temporary error. A kludge has been added to help with this.
1497 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1498 554 error is used if errno is still zero or contains ENOENT (which occurs
1499 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1502 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1504 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1505 errors. Otherwise a message that provokes a temporary error (when other
1506 messages do not) can cause a whole host to time out.
1508 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1509 addresses were routed directly to files or pipes from a redirect router.
1510 File deliveries just didn't batch; pipe deliveries might have suffered
1513 PH/06 A failure to get a lock for a hints database would erroneously always say
1514 "Failed to get write lock", even when it was really a read lock.
1516 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1517 of 0600. This has been changed to use the value of the lockfile_mode
1518 option (which defaults to 0600).
1520 PH/08 Applied small patch from the Sieve maintainer.
1522 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1523 folder from quota calculations, a direct delivery into this folder messed
1524 up the contents of the maildirsize file. This was because the regex was
1525 used only to exclude .Trash (or whatever) when the size of the mailbox
1526 was calculated. There was no check that a delivery was happening into an
1527 excluded directory. This bug has been fixed by ignoring all quota
1528 processing for deliveries into excluded directories.
1530 PH/10 Added the maildirfolder_create_regex option to appendfile.
1536 PH/01 The code for finding all the local interface addresses on a FreeBSD
1537 system running IPv6 was broken. This may well have applied to all BSD
1538 systems, as well as to others that have similar system calls. The broken
1539 code found IPv4 interfaces correctly, but gave incorrect values for the
1540 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1541 that it would not match correctly against @[] and not recognize the IPv6
1544 PH/02 The ipliteral router was not recognizing addresses of the form user@
1545 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1547 PH/03 Added disable_ipv6.
1549 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1550 decoded form, because it is most often used to construct To: headers
1551 lines in autoreplies, and the decoded form may well be syntactically
1552 invalid. However, $reply_address has leading white space removed, and all
1553 newlines turned into spaces so that the autoreply transport does not
1556 PH/05 If group was specified without a user on a router, and no group or user
1557 was specified on a transport, the group from the router was ignored.
1559 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1560 for visible compile-time settings that can be used to change these
1561 numbers, for those that want even more. Backwards compatibility with old
1562 spool files has been maintained. However, going back to a previous Exim
1563 release will lost any variables that are in spool files.
1565 PH/07 Two small changes when running in the test harness: increase delay when
1566 passing a TCP/IP connection to a new process, in case the original
1567 process has to generate a bounce, and remove special handling of
1568 127.0.0.2 (sic), which is no longer necessary.
1570 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1571 be the same on different OS.
1573 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1576 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1577 whether --show-vars was specified or not
1579 JJ/02 exipick: Added support for new ACL variable spool format introduced
1582 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1583 syntactically invalid From: or Reply-to: line, and a filter used this to
1584 generate an autoreply, and therefore failed to obtain an address for the
1585 autoreply, Exim could try to deliver to a non-existent relative file
1586 name, causing unrelated and misleading errors. What now happens is that
1587 it logs this as a hard delivery error, but does not attempt to create a
1590 PH/11 The exinext utility has a -C option for testing purposes, but although
1591 the given file was scanned by exinext itself; it wasn't being passed on
1592 when Exim was called.
1594 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1595 an end-of-file indication when reading a command response.
1597 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1598 compiled. In many other places in Exim, IPv6 addresses are always
1599 recognized, so I have changed this. It also means that IPv4 domain
1600 literals of the form [IPV4:n.n.n.n] are now always recognized.
1602 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1603 used if the router is not running as root, for example, when verifying at
1604 ACL time, or when using -bh. The debugging output from this situation was
1605 non-existent - all you got was a failure to exec. I have made two
1608 (a) Failures to set uid/gid, the current directory, or a process leader
1609 in a subprocess such as that created by queryprogram now generate
1610 suitable debugging ouput when -d is set.
1612 (b) The queryprogram router detects when it is not running as root,
1613 outputs suitable debugging information if -d is set, and then runs
1614 the subprocess without attempting to change uid/gid.
1616 PH/15 Minor change to Makefile for building test_host (undocumented testing
1619 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1620 additional section of a DNS packet that returns MX or SRV records.
1621 Instead, it always explicitly searches for A/AAAA records. This avoids
1622 major problems that occur when a DNS server includes only records of one
1623 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1624 fixed another bug: if SRV records were looked up and the corresponding
1625 address records were *not* found in the additional section, the port
1626 values from the SRV records were lost.
1628 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1629 using the correct key (the original address) when searching the retry
1630 rules in order to find which one to use for generating the retry hint.
1632 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1633 adding the default one. Similarly, if it contains a Reply-To: header, the
1634 errors_reply_to option, if set, is not used.
1636 PH/19 When calculating a retry time, Exim used to measure the "time since
1637 failure" by looking at the "first failed" field in the retry record. Now
1638 it does not use this if it is later than than the arrival time of the
1639 message. Instead it uses the arrival time. This makes for better
1640 behaviour in cases where some deliveries succeed, thus re-setting the
1641 "first failed" field. An example is a quota failure for a huge message
1642 when small messages continue to be delivered. Without this change, the
1643 "time since failure" will always be short, possible causing more frequent
1644 delivery attempts for the huge message than are intended.
1645 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1647 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1648 $1, $2, $3) because the numerical variables can be reset during some
1649 expansion items (e.g. "match"), thereby losing the authentication data.
1651 PH/21 Make -bV show the size of off_t variables so that the test suite can
1652 decide whether to run tests for quotas > 2G.
1654 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1655 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1656 filecount value is greater than 2G or if a quota value is greater than 2G
1657 on a system where the size of off_t is not greater than 4, a panic error
1660 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1661 never match. The debug and -bh output now contains an explicit error
1662 message indicating a malformed IPv4 address or mask.
1664 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1665 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1666 PH/23 above applies.
1668 PH/25 Do not write to syslog when running in the test harness. The only
1669 occasion when this arises is a failure to open the main or panic logs
1670 (for which there is an explicit test).
1672 PH/26 Added the /no_tell option to "control=freeze".
1674 PH/27 If a host name lookup failed very early in a connection, for example, if
1675 the IP address matched host_lookup and the reverse lookup yielded a name
1676 that did not have a forward lookup, an error message of the form "no IP
1677 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1678 could be logged. Now it outputs the IP address instead of "NULL".
1680 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1681 allows the sender and the authenticated sender to be set when
1682 submitting a message from within Exim. Since child_open_exim() is
1683 documented for local_scan(), the new function should be too.
1685 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1686 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1687 results in an empty string is now treated as unset.
1689 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1691 PH/31 Added sender_verify_fail logging option.
1693 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1694 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1695 tidied the source and removed it altogether.
1697 PH/33 When a queue run was abandoned because the load average was too high, a
1698 log line was always written; now it is written only if the queue_run log
1699 selector is set. In addition, the log line for abandonment now contains
1700 information about the queue run such as the pid. This is always present
1701 in "start" and "stop" lines but was omitted from the "abandon" line.
1703 PH/34 Omit spaces between a header name and the colon in the error message that
1704 is given when verify = headers_syntax fails (if there are lots of them,
1705 the message gets confusing).
1707 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1708 names, as there are now some PTR records that contain slashes. This check
1709 is only to protect against broken name servers that fall over on strange
1710 characters, so the fact that it applies to all lookups doesn't matter.
1712 PH/36 Now that the new test suite is complete, we can remove some of the
1713 special code in Exim that was needed for the old test suite. For example,
1714 sorting DNS records because real resolvers return them in an arbitrary
1715 order. The new test suite's fake resolver always returns records in the
1718 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1719 bounces) except when queue_only is set, to avoid logging races between
1720 the different processes.
1722 PH/38 Panic-die if .include specifies a non-absolute path.
1724 PH/39 A tweak to the "H" retry rule from its user.
1726 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1727 a label. They prevented compilation on older perls.
1729 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1730 a warning to be raised on newish perls.
1732 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1733 on queue. Changes to match documented behaviour of showing count of
1734 messages matching specified criteria.
1736 PH/40 Changed the default ident timeout from 30s to 5s.
1738 PH/41 Added support for the use of login_cap features, on those BSD systems
1739 that have them, for controlling the resources used by pipe deliveries.
1741 PH/42 The content-scanning code uses fopen() to create files in which to put
1742 message data. Previously it was not paying any attention to the mode of
1743 the files. Exim runs with umask(0) because the rest of the code creates
1744 files with open(), and sets the required mode explicitly. Thus, these
1745 files were ending up world-writeable. This was not a big issue, because,
1746 being within the spool directory, they were not world-accessible. I have
1747 created a function called modefopen, which takes an additional mode
1748 argument. It sets umask(777), creates the file, chmods it to the required
1749 mode, then resets the umask. All the relevant calls to fopen() in the
1750 content scanning code have been changed to use this function.
1752 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1753 to 24 hours. This avoids potential overflow problems when processing G
1754 and H retry rules. I suspect nobody ever tinkers with this value.
1756 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1758 PH/45 When the plaintext authenticator is running as a client, the server's
1759 challenges are checked to ensure they are valid base64 strings. By
1760 default, the authentication attempt is cancelled if an invalid string is
1761 received. Setting client_ignore_invalid_base64 true ignores these errors.
1762 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1763 they are received. Thus, the responses can be made to depend on the
1764 challenges. If an invalid string is ignored, an empty string is placed in
1767 PH/46 Messages that are created by the autoreply transport now contains a
1768 References: header, in accordance with RFCs 2822 and 3834.
1770 PH/47 Added authenticated_sender_force to the smtp transport.
1772 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1774 PH/49 Installed latest patch from the Sieve maintainer.
1776 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1777 was also set, the appendfile transport was unnecessarily scanning a
1778 directory of message files (e.g. for maildir delivery) to find the count
1779 of files (along with the size), even though it did not need this
1780 information. It now does the scan only if it needs to find either the
1781 size of the count of files.
1783 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1785 PH/52 Two bugs concerned with error handling when the smtp transport is
1788 (i) Exim was not creating retry information for temporary errors given
1789 for individual recipients after the DATA command when the smtp transport
1790 was used in LMTP mode. This meant that they could be retried too
1791 frequently, and not timed out correctly.
1793 (ii) Exim was setting the flag that allows error details to be returned
1794 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1795 recipients that were returned after the DATA command.
1797 PH/53 This is related to PH/52, but is more general: for any failing address,
1798 when detailed error information was permitted to be returned to the
1799 sender, but the error was temporary, then after the final timeout, only
1800 "retry timeout exceeded" was returned. Now it returns the full error as
1801 well as "retry timeout exceeded".
1803 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1804 do this, and (what is worse) MTAs that accept it.
1806 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1807 will now be deprecated.
1809 PH/56 New os.c-cygwin from the Cygwin maintainer.
1811 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1812 formats (previously only available in exim formats via -bpr, -bpru,
1813 and -bpra. Now also available in native and exiqgrep formats)
1815 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1816 with very large, slow to parse queues
1818 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1820 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1822 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1823 responses to authentication challenges, though it was showing the
1824 challenges; (ii) I've removed the CR characters from the debug output for
1827 PH/58 Allow for the insertion of a newline as well as a space when a string
1828 is turned into more than one encoded-word during RFC 2047 encoding. The
1829 Sieve code now uses this.
1831 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1832 data_4xx, lost_connection, tls_required.
1834 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1835 message was being sent as an SMTP response.
1837 PH/61 Add -l and -k options to exicyclog.
1839 PH/62 When verifying, if an address was redirected to one new address, so that
1840 verification continued, and the new address failed or deferred after
1841 having set something in $address_data, the value of $address_data was not
1842 passed back to the ACL. This was different to the case when no
1843 redirection occurred. The value is now passed back in both cases.
1845 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1846 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1847 use login_cap.h, so on its own it isn't the distinguishing feature. The
1848 new name refers directly to the setclassresources() function.
1850 PH/65 Added configuration files for NetBSD3.
1852 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1854 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1855 is preferred over IPv4.
1857 PH/68 The bounce_return_message and bounce_return_body options were not being
1858 honoured for bounces generated during the reception of non-SMTP messages.
1859 In particular, this applied to messages rejected by the ACL. This bug has
1860 been fixed. However, if bounce_return_message is true and bounce_return_
1861 body is false, the headers that are returned for a non-SMTP message
1862 include only those that have been read before the error was detected.
1863 (In the case of an ACL rejection, they have all been read.)
1865 PH/69 The HTML version of the specification is now built in a directory called
1866 spec_html instead of spec.html, because the latter looks like a path with
1867 a MIME-type, and this confuses some software.
1869 PH/70 Catch two compiler warnings in sieve.c.
1871 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1872 function verify_get_ident() calls ip_connect() to connect a socket, but
1873 if the "connect()" function timed out, ip_connect() used to close the
1874 socket. However, verify_get_ident() also closes the socket later, and in
1875 between Exim writes to the log, which may get opened at this point. When
1876 the socket was closed in ip_connect(), the log could get the same file
1877 descriptor number as the socket. This naturally causes chaos. The fix is
1878 not to close the socket in ip_connect(); the socket should be closed by
1879 the function that creates it. There was only one place in the code where
1880 this was missing, in the iplookup router, which I don't think anybody now
1881 uses, but I've fixed it anyway.
1883 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1884 well as to direct DNS lookups. Otherwise the handling of names in host
1885 lists is inconsistent and therefore confusing.
1891 PH/01 Two changes to the default runtime configuration:
1893 (1) Move the checks for relay_from_hosts and authenticated clients from
1894 after to before the (commented out) DNS black list checks.
1896 (2) Add control=submission to the relay_from_hosts and authenticated
1897 clients checks, on the grounds that messages accepted by these
1898 statements are most likely to be submissions.
1900 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1902 (1) Generate an error if the third argument for the ${prvs expansion is
1905 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1908 (3) Reset the variables that are obtained from the first argument of
1909 ${prvscheck and used in the second argument before leaving the code,
1910 because their memory is reclaimed, so using them afterwards may do
1913 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1914 one (it's much easier than Tom thought :-).
1916 (5) Because of (4), we can now allow for the use of $prvscheck_result
1917 inside the third argument.
1919 PH/03 For some reason, the default setting of PATH when running a command from
1920 a pipe transport was just "/usr/bin". I have changed it to
1923 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1924 anything to be listed in the output from -bV.
1926 PH/05 When a filter generated an autoreply, the entire To: header line was
1927 quoted in the delivery log line, like this:
1929 => >A.N.Other <ano@some.domain> <original@ddress> ...
1931 This has been changed so that it extracts the operative address. There
1932 may be more than one such address. If so, they are comma-separated, like
1935 => >ano@some.domain,ona@other.domain <original@ddress> ...
1937 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1938 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1939 not being looked up in the rDNS to get a host name, Exim was showing the
1940 IP address twice in Received: lines, even though the IP addresses were
1941 identical. For example:
1943 Received: from [1.2.3.4] (helo=[1.2.3.4])
1945 However, if the real host name was known, it was omitting the HELO data
1946 if it matched the actual IP address. This has been tidied up so that it
1947 doesn't show the same IP address twice.
1949 PH/07 When both +timestamp and +memory debugging was on, the value given by
1950 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1951 called by the debug printing, thereby overwriting the timestamp buffer.
1952 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1954 PH/08 When the original message was included in an autoreply transport, it
1955 always said "this is a copy of the message, including all the headers",
1956 even if body_only or headers_only was set. It now gives an appropriate
1959 PH/09 Applied a patch from the Sieve maintainer which:
1961 o fixes some comments
1962 o adds the (disabled) notify extension core
1963 o adds some debug output for the result of if/elsif tests
1964 o points to the current vacation draft in the documentation
1965 and documents the missing references header update
1969 o fixes a bug in processing the envelope test (when testing
1970 multiple envelope elements, the last element determinted the
1973 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1974 Electronic Mail") by including:
1976 Auto-submitted: auto-generated
1978 in the messages that it generates (bounce messages and others, such as
1979 warnings). In the case of bounce messages for non-SMTP mesages, there was
1980 also a typo: it was using "Auto_submitted" (underscore instead of
1981 hyphen). Since every message generated by Exim is necessarily in response
1982 to another message, thes have all been changed to:
1984 Auto-Submitted: auto-replied
1986 in accordance with these statements in the RFC:
1988 The auto-replied keyword:
1990 - SHOULD be used on messages sent in direct response to another
1991 message by an automatic process,
1993 - MUST NOT be used on manually-generated messages,
1995 - MAY be used on Delivery Status Notifications (DSNs) and Message
1996 Disposition Notifications (MDNs),
1998 - MUST NOT be used on messages generated by automatic or periodic
1999 processes, except for messages which are automatic responses to
2002 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2003 to the default Received: header definition.
2005 PH/12 Added log selector acl_warn_skipped (default on).
2007 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2008 variables because (a) they are in the wrong storage pool and (b) even if
2009 they were copied, it wouldn't work properly because of the caching.
2011 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2012 checking when decoding. Apparently there are clients that generate
2013 overlong encoded strings. Why am I not surprised?
2015 PH/15 If the first argument of "${if match_address" was not empty, but did not
2016 contain an "@" character, Exim crashed. Now it writes a panic log message
2017 and treats the condition as false.
2019 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2021 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2022 extension "envelope-auth". The code is finished and in agreement with
2023 other implementations, but there is no documentation so far and in fact,
2024 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2025 not changing the active code.
2027 Print executed "if" and "elsif" statements when debugging is used. This
2028 helps a great deal to understand what a filter does.
2030 Document more things not specified clearly in RFC3028. I had all this
2031 sorted out, when out of a sudden new issues came to my mind. Oops."
2033 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2036 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2037 canonical form (as documented). However, after a host name lookup from
2038 the IP address, check_host() was doing a simple string comparison with
2039 addresses acquired from the DNS when checking that the found name did
2040 have the original IP as one of its addresses. Since any found IPv6
2041 addresses are likely to be in abbreviated form, the comparison could
2042 fail. Luckily, there already exists a function for doing the comparison
2043 by converting both addresses to binary, so now that is used instead of
2044 the text comparison.
2046 PH/20 There was another similar case to PH/19, when a complete host name was
2047 given in a host list; looking up its IP address could give an abbreviated
2048 form, whereas the current host's name might or might not be abbreviated.
2049 The same fix has been applied.
2055 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2056 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2059 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2060 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2062 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2064 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2065 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2066 Also, for "Auto-Submitted", treat anything other than "no" as
2067 non-personal, in accordance with RFC 3834. (Previously it treated
2068 anything starting "auto-" as non-personal.)
2070 TF/01 The control=submission/name=... option had a problem with syntax
2071 errors if the name included a slash character. The /name= option
2072 now slurps the rest of the string, so it can include any characters
2073 but it must come last in the list of options (after /sender_retain
2076 PH/05 Some modifications to the interface to the fake nameserver for the new
2084 TK/01 Added the "success_on_redirect" address verification option. See
2085 NewStuff for rationale and an example.
2087 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2089 PH/02 Patch to exigrep to allow it to work on syslog lines.
2091 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2092 fread() to skip over the body file's header line, because in Cygwin the
2093 header line is locked and is inaccessible.
2095 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2096 co-exist for some time) to make it clear that it is the Exim ID that is
2097 referenced, not the Message-ID: header line.
2099 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2100 string_format() function, because snprintf() does not exist on all
2103 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2104 ${stat: expansion item.
2106 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2107 protocol synchronization error", to keep the pedants happy.
2109 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2110 well as for IRIX systems, when gcc is being used. See the host.c source
2113 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2115 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2118 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2119 transport and to the smtp transport in LMTP mode.
2121 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2123 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2124 run a filter in a subprocess. This could lead to confusion in subsequent
2125 lookups in the parent process. There should also be a search_tidyup() at
2126 the end of the subprocess.
2128 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2129 only if the host matched helo_try_verify_hosts, which caused the
2130 verification to occur when the EHLO/HELO command was issued. The ACL just
2131 tested the remembered result. Now, if a previous verification attempt has
2132 not happened, "verify = helo" does it there and then.
2134 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2136 TK/03 Fix log output including CR from clamd.
2138 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2139 exist provoked a memory error which could cause a segfault.
2141 PH/15 Installed PCRE 6.2
2143 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2145 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2146 of the problem. Specifically, suggested +O2 rather than +O1 for the
2149 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2151 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2152 record, the port was not correctly logged when the outgoing_port log
2153 selector was set (it logged the transort's default port).
2155 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2156 fallback_hosts, and "hosts" in the smtp transport.
2158 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2159 host errors such as "Connection refused".
2161 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2162 authentication with radiusclient 0.4.9:
2164 - Error returned from rc_read_config was caught wrongly
2165 - Username/password not passed on to radius server due to wrong length.
2167 The presumption is that some radiusclient API changes for 4.51/PH/17
2168 were not taken care of correctly. The code is still untested by me (my
2169 Linux distribution still has 0.3.2 of radiusclient), but it was
2170 contributed by a Radius user.
2172 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2173 expanding the "port" option of the smtp transport.
2175 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2176 while reading a MIME header. Thanks to Tom Hughes for a patch.
2178 PH/24 Include config.h inside local_scan.h so that configuration settings are
2181 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2182 that in an ACL for RCPT (for example), you can examine exactly what was
2185 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2186 commands, but it was not correctly comparing the address with the actual
2187 client host address. Thus, it would show the EHLO address in Received:
2188 header lines when this was not necessary.
2190 PH/27 Added the % operator to ${eval:}.
2192 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2193 it should be ignoring failures (because with -C, for example, it has lost
2194 privilege). It wasn't ignoring creation failures other than "already
2197 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2200 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2201 because an input line was too long, either on its own, or by virtue of
2202 too many continuations, the temporary file was not being removed, and the
2203 return code was incorrect.
2205 PH/31 Missing "BOOL" in function definition in filtertest.c.
2207 PH/32 Applied Sieve patches from the maintainer.
2209 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2211 PH/33 Added "verify = not_blind".
2213 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2214 Local/Makefile (with some defaults set). These are used in built scripts
2215 such as exicyclog, but they have never been used in the exim_install
2216 script (though there are many overriding facilities there). I have
2217 arranged that the exim_install script now takes note of these two
2220 PH/35 Installed configuration files for Dragonfly.
2222 PH/36 When a locally submitted message by a trusted user did not contain a
2223 From: header, and the sender address was obtained from -f or from an SMTP
2224 MAIL command, and the trusted user did not use -F to supply a sender
2225 name, $originator_name was incorrectly used when constructing a From:
2226 header. Furthermore, $originator_name was used for submission mode
2227 messages from external hosts without From: headers in a similar way,
2228 which is clearly wrong.
2230 PH/37 Added control=suppress_local_fixups.
2232 PH/38 When log_selector = +received_sender was set, and the addition of the
2233 sender made the log line's construction buffer exactly full, or one byte
2234 less than full, an overflow happened when the terminating "\n" was
2237 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2238 when the result of a list match is failure because a DNS lookup failed.
2240 PH/40 RM_COMMAND is now used in the building process.
2242 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2243 the "build-* directories that it finds.
2245 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2246 address in a domain literal was a prefix of an interface address.
2248 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2249 when verifying a sender address, unless rewrite_headers is false.
2251 PH/44 Wrote a long comment about why errors_to addresses are verified as
2252 recipients, not senders.
2254 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2255 the ratelimit ACL was added.
2257 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2259 PH/46 Added extra information about PostgreSQL errors to the error string.
2261 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2262 avoiding the need to install special zones in a real server. This is
2263 backwards compatible; if it can't find the fake resolver, it drops back.
2264 Thus, both old and new test suites can be run.
2266 TF/02 Added util/ratelimit.pl
2268 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2269 clock is set back in time.
2271 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2272 Candler <B.Candler@pobox.com>.
2274 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2275 OK for addresses that are the result of redirections.
2277 PH/48 A number of further additions for the benefit of the new test suite,
2278 including a fake gethostbyname() that interfaces to the fake DNS resolver
2281 TF/06 The fix for widen_domains has also been applied to qualify_single and
2282 search_parents which are the other dnslookup options that can cause
2285 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2288 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2290 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2291 service" if it encountered an MX record with an empty target hostname.
2292 The message is now "an MX or SRV record indicated no SMTP service".
2294 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2295 if the DNS of the sending site is misconfigured. This is quite a
2296 common situation. This change restores the behaviour of treating a
2297 helo verification defer as a failure.
2299 PH/51 If self=fail was set on a router, the bounce message did not include the
2300 actual error message.
2306 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2308 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2309 command itself ended in error, the underlying message about the transport
2310 filter timeout was being overwritten with the pipe command error. Now the
2311 underlying error message should be appended to the second error message.
2313 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2315 PH/02 Reduce the amount of output that "make" produces by default. Full output
2316 can still be requested.
2318 PH/03 The warning log line about a condition test deferring for a "warn" verb
2319 was being output only once per connection, rather than after each
2320 occurrence (because it was using the same function as for successful
2321 "warn" verbs). This seems wrong, so I have changed it.
2323 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2324 it should not have, which might have caused a crash in the right
2325 circumstances, but probably never did.
2327 PH/04 Installed a modified version of Tony Finch's patch to make submission
2328 mode fix the return path as well as the Sender: header line, and to
2329 add a /name= option so that you can make the user's friendly name appear
2332 TF/03 Added the control = fakedefer ACL modifier.
2334 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2335 Mark Lowes for thorough testing.
2337 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2339 TK/03 Merged latest SRS patch from Miles Wilton.
2341 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2342 with the definition in sysexits.h (which is #included earlier).
2343 Fortunately, Exim does not actually use EX_OK. The code used to try to
2344 preserve the sysexits.h value, by assumimg that macro definitions were
2345 scanned for macro replacements. I have been disabused of this notion,
2346 so now the code just undefines EX_OK before #including unistd.h.
2348 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2349 in the smtp transport. When a block could not be written in a single
2350 write() function, the timeout was being re-applied to each part-write.
2351 This seems wrong - if the receiver was accepting one byte at a time it
2352 would take for ever. The timeout is now adjusted when this happens. It
2353 doesn't have to be particularly precise.
2355 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2356 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2358 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2359 without a domain if the check to <postmaster@domain> fails.
2361 SC/01 Eximstats: added -xls and the ability to specify output files
2362 (patch written by Frank Heydlauf).
2364 SC/02 Eximstats: use FileHandles for outputing results.
2366 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2368 SC/04 Eximstats: fixed display of large numbers with -nvr option
2370 SC/05 Eximstats: fixed merging of reports with empty tables.
2372 SC/06 Eximstats: added the -include_original_destination flag
2374 SC/07 Eximstats: removed tabs and trailing whitespace.
2376 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2378 TK/06 MBOX spool code: Add real "From " MBOX separator line
2379 so the .eml file is really in mbox format (even though
2380 most programs do not really care). Patch from Alex Miller.
2382 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2383 The latter is generated from $received_to and is only set if the
2384 message has one envelope recipient. SA can use these headers,
2385 obviously out-of-the-box. Patch from Alex Miller.
2387 PH/08 The ${def test on a variable was returning false if the variable's
2388 value was "0", contrary to what the specification has always said!
2389 The result should be true unless the variable is empty.
2391 PH/09 The syntax error of a character other than { following "${if
2392 def:variable_name" (after optional whitespace) was not being diagnosed.
2393 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2394 accidental colon was present, for example, could give incorrect results.
2396 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2397 result is used (not including appendfile, where other changes are about
2400 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2401 This involved changing a lot of size variables from int to off_t. It
2402 should work with maildirs and everything.
2404 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2405 spamd dying while we are connected to it.
2407 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2410 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2411 function for BDB 4.3.
2413 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2415 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2416 the sources for PCRE have been re-arranged and more files are now
2419 PH/15 The code I had for printing potentially long long variables in PH/11
2420 above was not the best (it lost precision). The length of off_t variables
2421 is now inspected at build time, and an appropriate printing format (%ld
2422 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2423 to be "long long int" or "long int". This is needed for the internal
2424 formatting function string_vformat().
2426 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2427 the configuration file to be ":syslog", then the script "guesses" where
2428 the logs files are, rather than using the compiled in default. In our
2429 case the guess is not the same as the compiled default, so the script
2430 suddenly stopped working when I started to use syslog. The patch checks
2431 to see if log_file_path is "". If so, it attempts to read it from exim
2432 with no configuration file to get the compiled in version, before it
2433 falls back to the previous guessing code."
2435 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2436 implementing BATV in an Exim configuration. See NewStuff for the gory
2439 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2440 Makefile that are specific to HP-UX.
2442 PH/18 If the "use_postmaster" option was set for a recipient callout together
2443 with the "random" option, the postmaster address was used as the MAIL
2444 FROM address for the random test, but not for the subsequent recipient
2445 test. It is now used for both.
2447 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2448 patch removes a few documentation additions to RFC 3028, because the
2449 latest draft now contains them. It adds the new en;ascii-case comparator
2450 and a new error check for 8bit text in MIME parts. Comparator and
2451 require names are now matched exactly. I enabled the subaddress
2452 extension, but it is not well tested yet (read: it works for me)."
2454 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2455 rework some of the code of TK/09 above to avoid the hardwired use of
2456 "%lld" and "long long". Replaced the call to snprintf() with a call to
2459 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2460 records point to non-existent hosts", "retry timeout exceeded", and
2461 "retry time not reached for any host after a long failure period".
2463 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2464 experimental DomainKeys support:
2466 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2467 (2) On an error such as an illegally used "control", the wrong name for
2468 the control was given.
2470 These problems did NOT occur unless DomainKeys support was compiled.
2472 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2474 PH/24 Added ${if match_ip condition.
2476 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2477 databases so that it will be absolutely obvious if a crash occurs in the
2478 DB library. This is a regular occurrence (often caused by mis-matched
2481 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2482 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2483 picked up on a user's system that detects such things. There doesn't seem
2484 to be a gcc warning option for this - only an attribute that has to be
2485 put on the function's prototype. It seems that in Fedora Core 4 they have
2486 set this on a number of new functions. No doubt there will be more in due
2489 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2490 specify a transport. However, if an address that was verified by such a
2491 router was the subject of a callout, Exim crashed because it tried to
2492 read the rcpt_include_affixes from the non-existent transport. Now it
2493 just assumes that the setting of that option is false. This bug was
2494 introduced by 4.51/PH/31.
2496 PH/28 Changed -d+all to exclude +memory, because that information is very
2497 rarely of interest, but it makes the output a lot bigger. People tend to
2498 do -d+all out of habit.
2500 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2501 code in os-type was giving problems when libc.so lives in lib64, like on
2504 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2505 aren't the modern standard, and it seems that some systems' include files
2506 don't always have them. Exim was already checking for some of the newer
2507 ones like T_AAAA, and defining it itself. I've added checks for all the
2508 record types that Exim uses.
2510 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2511 not automatically generating a new one, as it is supposed to. This
2512 prevented TLS from working. If the file did exist, but contained invalid
2513 data, a new version was generated, as expected. It was only the case of a
2514 non-existent file that was broken.
2516 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2517 with a change in libdomainkeys > 0.64.
2519 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2520 from DNS. If the selector record carries the flag, it now has
2521 precedence over the domain-wide flag.
2523 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2525 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2526 the use of a port name that isn't defined in /etc/services) occurred, the
2527 message was deferred as in a normal delivery, and thus remained on the
2528 spool, instead of being failed because of the mua_wrapper setting. This
2529 is now fixed, and I tidied up some of the mua_wrapper messages at the
2532 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2533 the messages in a hash of arrays rather than using individual hashes.
2534 This is a bit cleaner and results in dramatic memory savings, albeit
2535 at a slight CPU cost.
2537 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2538 as requested by Marc Sherman.
2540 SC/10 Eximstats: added histograms for user specified patterns as requested
2543 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2545 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2546 fopen() in the content-scanning modules that did not already have it.
2552 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2553 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2555 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2557 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2559 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2560 to test Sieve filters that use "vacation".
2562 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2563 that changes the way the GnuTLS parameters are stored in the cache file.
2564 The new format can be generated externally. For backward compatibility,
2565 if the data in the cache doesn't make sense, Exim assumes it has read an
2566 old-format file, and it generates new data and writes a new file. This
2567 means that you can't go back to an older release without removing the
2570 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2571 work if there are any delivery delays because "one_time" forces the
2572 parent to be marked "delivered", so its unseen clone is never tried
2573 again. For this reason, Exim now forbids the simultaneous setting of
2576 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2577 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2578 entry if you want to know the details. The fix, however, neglected to
2579 consider the case when local delivery batching is involved. The test for
2580 "previously delivered" was not happening when checking to see if an
2581 address could be batched with a previous (undelivered) one; under
2582 certain circumstances this could lead to multiple deliveries to the same
2585 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2586 in its include files, and this causes problems building Exim.
2588 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2589 header_syntax) but Exim was just ignoring anything given after a slash.
2590 In particular, this caused confusion with an attempt to use "verify =
2591 reverse_host_lookup/defer_ok". An error is now given when options are
2592 supplied for verify items that do not have them. (Maybe reverse_host_
2593 lookup should have a defer_ok option, but that's a different point.)
2595 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2596 defined by RFC 821) to 2048, because there were problems with some AUTH
2597 commands, and RFC 1869 says the size should be increased for extended
2598 SMTP commands that take arguments.
2600 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2603 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2604 "unknown" error; now it says that the functionality isn't in the binary.
2606 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2607 an address' error message when a string expansion fails (syntax or
2608 whatever). Otherwise the password may appear in the log. Following change
2609 PH/42 below, there is no longer a chance of it appearing in a bounce
2612 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2614 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2615 of its addresses was ever tried. (Bugzilla bug #2.)
2617 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2618 the result incorrectly in the debug output. (It correctly added a newline
2619 to what was transported.)
2621 TF/01 Added $received_time.
2623 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2624 commented out examples of how to interface to a virus scanner and to
2625 SpamAssassin. Also added commented examples of av_scanner and
2626 spamd_address settings.
2628 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2629 and controls are allowed in which ACLs. There were a couple of minor
2630 errors. Some of the entries in the conditions table (which is a table of
2631 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2632 negation of where the condition IS allowed.
2634 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2636 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2637 header file does not have a version number, so I've had to invent a new
2638 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2639 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2640 radiusclient), but it was contributed by a Radius user.
2642 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2643 files or directories, for OpenSSL.
2645 PH/19 When an Exim process that is running as root has to create an Exim log
2646 file, it does so in a subprocess that runs as exim:exim so as to get the
2647 ownership right at creation (otherwise, other Exim processes might see
2648 the file with the wrong ownership). There was no test for failure of this
2649 fork() call, which would lead to the process getting stuck as it waited
2650 for a non-existent subprocess. Forks do occasionally fail when resources
2651 run out. I reviewed all the other calls to fork(); they all seem to check
2654 PH/20 When checking for unexpected SMTP input at connect time (before writing
2655 the banner), Exim was not dealing correctly with a non-positive return
2656 from the read() function. If the client had disconnected by this time,
2657 the result was a log entry for a synchronization error with an empty
2658 string after "input=" when read() returned zero. If read() returned -1
2659 (an event I could not check), uninitialized data bytes were printed.
2660 There were reports of junk text (parts of files, etc) appearing after
2663 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2665 PH/22 Added support for macro redefinition, and (re)definition in between
2666 driver and ACL definitions.
2668 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2669 forgetting to use the resulting value; it was using the unexpanded value.
2671 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2672 hadn't been configured. The fix is from Juergen Kreileder, who
2673 understands it better than I do:
2675 "Here's what I see happening with three configured cyrus_sasl
2676 authenticators configured (plain, login, cram-md5):
2678 On startup auth_cyrus_sasl_init() gets called for each of these.
2679 This means three calls to sasl_listmech() without a specified mech_list.
2680 => SASL tests which mechs of all available mechs actually work
2681 => three warnings about OTP not working
2682 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2684 With the patch, sasl_listmech() also gets called three times. But now
2685 SASL's mech_list option is set to the server_mech specified in the the
2686 authenticator. Or in other words, the answer from sasl_listmech()
2687 gets limited to just the mech you're testing for (which is different
2689 => the return list contains just 'plain' or 'login', 'cram-md5' or
2690 nothing depending on the value of ob->server_mech.
2692 I've just tested the patch: Authentication still works fine,
2693 unavailable mechs specified in the exim configuration are still
2694 caught, and the auth.log warnings about OTP are gone."
2696 PH/25 When debugging is enabled, the contents of the command line are added
2697 to the debugging output, even when log_selector=+arguments is not
2700 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2701 answer is "GNU", and only if the return is "GNU/something" is the answer
2704 PH/27 $acl_verify_message is now set immediately after the failure of a
2705 verification in an ACL, and so is available in subsequent modifiers. In
2706 particular, the message can be preserved by coding like this:
2708 warn !verify = sender
2709 set acl_m0 = $acl_verify_message
2711 Previously, $acl_verify_message was set only while expanding "message"
2712 and "log_message" when a very denied access.
2714 PH/28 Modified OS/os.c-Linux with
2716 -#ifndef OS_LOAD_AVERAGE
2717 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2719 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2720 nomenclature these days.)
2722 PH/29 Installed patch from the Sieve maintainer that adds the options
2723 sieve_useraddress and sieve_subaddress to the redirect router.
2725 PH/30 In these circumstances:
2726 . Two addresses routed to the same list of hosts;
2727 . First host does not offer TLS;
2728 . First host accepts first address;
2729 . First host gives temporary error to second address;
2730 . Second host offers TLS and a TLS session is established;
2731 . Second host accepts second address.
2732 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2733 and peerdn, if requested) that were in fact used only for the second
2736 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2737 attention to any local part prefix or suffix that was matched by the
2738 router that accepted the address. It now behaves in the same way as it
2739 does for delivery: the affixes are removed from the local part unless
2740 rcpt_include_affixes is set on the transport.
2742 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2743 timeout during the DATA phase of an incoming message.
2745 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2746 applied a patch sanctioned by the Sieve maintainer.
2748 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2749 the uid or gid is negative. A case of a negative gid caused this to be
2750 noticed. The fix allows for either to be negative.
2752 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2753 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2756 PH/36 The change PH/12 above was broken. Fixed it.
2758 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2759 the grounds that routing the same address twice would always produce the
2760 same answer. This might have been true once, but it is certainly no
2761 longer true now. Routing a child address may depend on the previous
2762 routing that produced that child. Some complicated redirection strategies
2763 went wrong when messages had multiple recipients, and made Exim's
2764 behaviour dependent on the order in which the addresses were given.
2766 I have moved the duplicate checking until after the routing is complete.
2767 Exim scans the addresses that are assigned to local and remote
2768 transports, and removes any duplicates. This means that more work will be
2769 done, as duplicates will always all be routed, but duplicates are
2770 presumably rare, so I don't expect this is of any significance.
2772 For deliveries to pipes, files, and autoreplies, the duplicate checking
2773 still happens during the routing process, since they are not going to be
2776 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2777 It corrects a timeout issue with spamd. This is Ian's comment: "The
2778 background is that sometimes spamd either never reads data from a
2779 connection it has accepted, or it never writes response data. The exiscan
2780 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2781 blindly assumes that writes won't block so it may never time out."
2783 PH/39 Allow G after quota size as well as K and M.
2785 PH/40 The value set for $authenticated_id in an authenticator may not contain
2786 binary zeroes or newlines because the value is written to log lines and
2787 to spool files. There was no check on this. Now the value is run through
2788 the string_printing() function so that such characters are converted to
2789 printable escape sequences.
2791 PH/41 $message_linecount is a new variable that contains the total number of
2792 lines in the message. Compare $body_linecount, which is the count for the
2795 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2796 bounce and delay warning messages, except in certain special cases, which
2799 (a) An SMTP error message from a remote host;
2800 (b) A message specified in a :fail: redirection;
2801 (c) A message specified in a "fail" command in a system filter;
2802 (d) A message specified in a FAIL return from the queryprogram router;
2803 (e) A message specified by the cannot_route_message router option.
2805 In these cases only, Exim does include the error details in bounce and
2806 warning messages. There are also a few cases where bland messages such
2807 as "unrouteable address" or "local delivery error" are given.
2809 PH/43 $value is now also set for the "else" part of a ${run expansion.
2811 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2812 being worked on, but at least Exim now implements the latest version to
2815 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2816 process to complete was treated as a delivery failure, a timeout while
2817 writing the message to the pipe was logged, but erroneously treated as a
2818 successful delivery. Such timeouts include transport filter timeouts. For
2819 consistency with the overall process timeout, these timeouts are now
2820 treated as errors, giving rise to delivery failures by default. However,
2821 there is now a new Boolean option for the pipe transport called
2822 timeout_defer, which, if set TRUE, converts the failures into defers for
2823 both kinds of timeout. A transport filter timeout is now identified in
2826 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2827 systems where "make" and "gmake" are different, calling "gmake" at top
2828 level broke things. I've arranged for the value of $(MAKE) to be passed
2829 from the Makefile to this script so that it can call the same version of
2833 A note about Exim versions 4.44 and 4.50
2834 ----------------------------------------
2836 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2837 changes of various kinds. As a consequence, a big documentation update was
2838 needed. This delayed the release for rather longer than seemed good, especially
2839 in the light of a couple of (minor) security issues. Therefore, the changes
2840 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2841 release. So 4.44 and 4.50 are in effect two different branches that both start
2844 I have left the 4.50 change log unchanged; it contains all the changes since
2845 4.43. The change log for 4.44 is below; many of its items are identical to
2846 those for 4.50. This seems to be the most sensible way to preserve the
2847 historical information.
2853 1. Minor wording change to the doc/README.SIEVE file.
2855 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2856 computation of the current number of files was incorrect.
2858 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2859 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2862 4. Give more explanation in the error message when the command for a transport
2863 filter fails to execute.
2865 5. There are several places where Exim runs a non-Exim command in a
2866 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2867 was being done only for the command run by the queryprogram router. It is
2868 now done for all such subprocesses. The other cases are: ${run, transport
2869 filters, and the commands run by the lmtp and pipe transports.
2871 6. Added CONFIGURE_GROUP build-time option.
2873 7. Some older OS have a limit of 256 on the maximum number of file
2874 descriptors. Exim was using setrlimit() to set 1000 as a large value
2875 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2876 systems. I've change it so that if it can't get 1000, it tries for 256.
2878 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2879 was an oversight, and furthermore, ever since the addition of extra
2880 controls (e.g. 4.43/32), the checks on when to allow different forms of
2881 "control" were broken. There should now be diagnostics for all cases when a
2882 control that does not make sense is encountered.
2884 9. Added the /retain_sender option to "control=submission".
2886 10. $recipients is now available in the predata ACL (oversight).
2888 11. Tidy the search cache before the fork to do a delivery from a message
2889 received from the command line. Otherwise the child will trigger a lookup
2890 failure and thereby defer the delivery if it tries to use (for example) a
2891 cached ldap connection that the parent has called unbind on.
2893 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2894 of $address_data from the recipient verification was clobbered by the
2895 sender verification.
2897 13. The value of address_data from a sender verification is now available in
2898 $sender_address_data in subsequent conditions in the ACL statement.
2900 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2902 15. Added a new option "connect=<time>" to callout options, to set a different
2905 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2906 was its contents. (It was OK if the option was not defined at all.)
2908 17. A "Completed" log line is now written for messages that are removed from
2909 the spool by the -Mrm option.
2911 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2912 information about exactly what failed.
2914 19. Added -dd to debug only the daemon process.
2916 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2917 handles timeouts, both on the server side and network timeouts. Renamed the
2918 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2920 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2921 It is now set to "smtps".
2923 22. $host_address is now set to the target address during the checking of
2924 ignore_target_hosts.
2926 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2927 being passed; this would have caused $sender_host_name to have been used if
2928 matching the list had actually called for a host name (not very likely,
2929 since this list is usually IP addresses). A host name is now passed as
2932 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2933 code that specifies a non-restarting handler (typically sigaction() in
2934 modern systems) in an attempt to fix a rare and obscure crash bug.
2936 25. Narrowed the window for a race in the daemon that could cause it to ignore
2937 SIGCHLD signals. This is not a major problem, because they are used only to
2938 wake it up if nothing else does.
2940 26. A malformed maildirsize file could cause Exim to calculate negative values
2941 for the mailbox size or file count. Odd effects could occur as a result.
2942 The maildirsize information is now recalculated if the size or filecount
2945 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2946 support for a long time. Removed HAVE_SYS_VFS_H.
2948 28. Installed the latest version of exipick from John Jetmore.
2950 29. In an address list, if the pattern was not a regular expression, an empty
2951 subject address (from a bounce message) matched only if the pattern was an
2952 empty string. Non-empty patterns were not even tested. This was the wrong
2953 because it is perfectly reasonable to use an empty address as part of a
2954 database query. An empty address is now tested by patterns that are
2955 lookups. However, all the other forms of pattern expect the subject to
2956 contain a local part and a domain, and therefore, for them, an empty
2957 address still always fails if the pattern is not itself empty.
2959 30. Exim went into a mad DNS loop when attempting to do a callout where the
2960 host was specified on an smtp transport, and looking it up yielded more
2961 than one IP address.
2963 31. Re-factored the code for checking spool and log partition space into a
2964 function that finds that data and another that does the check. The former
2965 is then used to implement four new variables: $spool_space, $log_space,
2966 $spool_inodes, and $log_inodes.
2968 32. The RFC2047 encoding function was originally intended for short strings
2969 such as real names; it was not keeping to the 75-character limit for
2970 encoded words that the RFC imposes. It now respects the limit, and
2971 generates multiple encoded words if necessary. To be on the safe side, I
2972 have increased the buffer size for the ${rfc2047: expansion operator from
2975 33. It is now permitted to omit both strings after an "if" condition; if the
2976 condition is true, the result is "true". As before, when the second string
2977 is omitted, a false condition yields an empty string. This makes it less
2978 cumbersome to write custom ACL and router conditions.
2980 34. Failure to deliver a bounce message always caused it to be frozen, even if
2981 there was an errors_to setting on the router. The errors_to setting is now
2984 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2985 canonical form (fully expanded) before being placed in
2986 $sender_host_address.
2988 36. The table in the code that translates DNS record types into text (T_A to
2989 "A" for instance) was missing entries for NS and CNAME. It is just possible
2990 that this could have caused confusion if both these types were looked up
2991 for the same domain, because the text type is used as part of Exim's
2992 per-process caching. But the chance of anyone hitting this buglet seems
2995 37. The dnsdb lookup has been extended in a number of ways.
2997 (1) There is a new type, "zns", which walks up the domain tree until it
2998 finds some nameserver records. It should be used with care.
3000 (2) There is a new type, "mxh", which is like "mx" except that it returns
3001 just the host names, not the priorities.
3003 (3) It is now possible to give a list of domains (or IP addresses) to be
3004 looked up. The behaviour when one of the lookups defers can be
3005 controlled by a keyword.
3007 (4) It is now possible to specify the separator character for use when
3008 multiple records are returned.
3010 38. The dnslists ACL condition has been extended: it is now possible to supply
3011 a list of IP addresses and/or domains to be looked up in a particular DNS
3014 39. Added log_selector=+queue_time_overall.
3016 40. When running the queue in the test harness, wait just a tad after forking a
3017 delivery process, to get repeatability of debugging output.
3019 41. Include certificate and key file names in error message when GnuTLS fails
3020 to set them up, because the GnuTLS error message doesn't include the name
3021 of the failing file when there is a problem reading it.
3023 42. Allow both -bf and -bF in the same test run.
3025 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3027 44. The "Exiscan patch" is now merged into the mainline Exim source.
3029 45. Sometimes the final signoff response after QUIT could fail to get
3030 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3031 < 0 before doing a fflush(). This bug looks as though it goes back to the
3032 introduction of TLS in release 3.20, but "sometimes" must have been rare
3033 because the tests only now provoked it.
3035 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3036 (this can affect the format of dates).
3038 47. exim_tidydb, when checking for the continued existence of a message for
3039 which it has found a message-specific retry record, was not finding
3040 messages that were in split spool directories. Consequently, it was
3041 deleting retry records that should have stayed in existence.
3043 48. Steve fixed some bugs in eximstats.
3045 49. The SPA authentication driver was not abandoning authentication and moving
3046 on to the next authenticator when an expansion was forced to fail,
3047 contradicting the general specification for all authenticators. Instead it
3048 was generating a temporary error. It now behaves as specified.
3050 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3051 (the order specifies the preference for clients). The order is now AES256,
3052 AES128, 3DES, ARCFOUR128.
3054 51. Small patch to Sieve code - explicitly set From: when generating an
3057 52. Exim crashed if a remote delivery caused a very long error message to be
3058 recorded - for instance if somebody sent an entire SpamAssassin report back
3059 as a large number of 550 error lines. This bug was coincidentally fixed by
3060 increasing the size of one of Exim's internal buffers (big_buffer) that
3061 happened as part of the Exiscan merge. However, to be on the safe side, I
3062 have made the code more robust (and fixed the comments that describe what
3065 53. Now that there can be additional text after "Completed" in log lines (if
3066 the queue_time_overall log selector is set), a one-byte patch to exigrep
3067 was needed to allow it to recognize "Completed" as not the last thing in
3070 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3071 patch that reportedly fixes this has been added. I am not expert enough to
3072 create a test for it. This is what the patch creator wrote:
3074 "I found a little strange behaviour of ldap code when working with
3075 Windows 2003 AD Domain, where users was placed in more than one
3076 Organization Units. When I tried to give exim partial DN, the exit code
3077 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3078 But simultaneously result of request was absolutely normal ldap result,
3079 so I produce this patch..."
3081 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3082 REFERENCE, so I have modified the code to exclude the patch when that macro
3085 55. Some experimental protocols are using DNS PTR records for new purposes. The
3086 keys for these records are domain names, not reversed IP addresses. The
3087 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3088 leaves it alone. Component reversal etc. now happens only for IP addresses.
3091 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3093 57. Double the size of the debug message buffer (to 2048) so that more of very
3094 long debug lines gets shown.
3096 58. The exicyclog utility now does better if the number of log files to keep
3097 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3099 59. Two changes related to the smtp_active_hostname option:
3101 (1) $smtp_active_hostname is now available as a variable.
3102 (2) The default for smtp_banner uses $smtp_active_hostname instead
3103 of $primary_hostname.
3105 60. The host_aton() function is supposed to be passed a string that is known
3106 to be a valid IP address. However, in the case of IPv6 addresses, it was
3107 not checking this. This is a hostage to fortune. Exim now panics and dies
3108 if the condition is not met. A case was found where this could be provoked
3109 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3110 components; fortuitously, this particular loophole had already been fixed
3111 by change 4.50/55 above.
3113 If there are any other similar loopholes, the new check in host_aton()
3114 itself should stop them being exploited. The report I received stated that
3115 data on the command line could provoke the exploit when Exim was running as
3116 exim, but did not say which command line option was involved. All I could
3117 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3118 running as the user.
3121 61. There was a buffer overflow vulnerability in the SPA authentication code
3122 (which came originally from the Samba project). I have added a test to the
3123 spa_base64_to_bits() function which I hope fixes it.
3126 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3127 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3129 63. The daemon start-up calls getloadavg() while still root for those OS that
3130 need the first call to be done as root, but it missed one case: when
3131 deliver_queue_load_max is set with deliver_drop_privilege. This is
3132 necessary for the benefit of the queue runner, because there is no re-exec
3133 when deliver_drop_privilege is set.
3135 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3136 This has been fixed.
3138 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3139 was in use, was not putting the data itself into the right store pool;
3140 consequently, it could be overwritten for a subsequent message in the same
3141 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3144 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3146 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3147 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3148 the function were treating the return as a boolean value, which happened to
3149 work because 0=false and not-0=true, but is not correct code.
3151 68. The host_aton() function was not handling scoped IPv6 addresses (those
3152 with, for example, "%eth0" on the end) correctly.
3154 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3155 negated items (that is, ~something) in unsigned ints. Some compilers
3156 apparently mutter when there is no cast.
3158 70. If an address verification called from an ACL failed, and did not produce a
3159 user-specific message (i.e. there was only a "system" message), nothing was
3160 put in $acl_verify_message. In this situation, it now puts the system
3163 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3164 session; change 4.31/43 added the unwanted input to the log line - except
3165 that it did not do this in the start of session case. It now does.
3167 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3168 This is wrong and can cause the other end to generate a synchronization
3169 error if it is another Exim or anything else that does the synchronization
3170 check. A QUIT command is no longer sent after a timeout.
3172 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3173 during host lookups.
3175 74. The defer_ok option of callout verification was not working if it was used
3176 when verifying addresses in header lines, that is, for this case:
3178 verify = header_sender/callout=defer_ok
3180 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3181 those file descriptors could be used for SMTP connections. If anything
3182 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3183 could be sent to the SMTP client, causing chaos. The daemon now opens
3184 stdin, stdout, and stderr to /dev/null when it puts itself into the
3187 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3188 log by default. The user can override this with suitable Perl magic.
3190 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3191 the log message when discard triggers, was not working for the DATA ACL or
3192 for the non-SMTP ACL.
3194 78. Error message wording change in sieve.c.
3196 79. If smtp_accept_max_per_host was set, the number of connections could be
3197 restricted to fewer than expected, because the daemon was trying to set up
3198 a new connection before checking whether the processes handling previous
3199 connections had finished. The check for completed processes is now done
3200 earlier. On busy systems, this bug wouldn't be noticed because something
3201 else would have woken the daemon, and it would have reaped the completed
3204 80. If a message was submitted locally by a user whose login name contained one
3205 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3206 It caused a spool format error. I have fixed the spool reading code. A
3207 related problem was that the "from" clause in the Received: line became
3208 illegal because of the space(s). It is now covered by ${quote_local_part.
3210 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3213 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3215 83. Patch from Sieve maintainer to fix unterminated string problem in
3216 "vacation" handling.
3218 84. Some minor changes to the Linux configuration files to help with other
3219 OS variants using glibc.
3221 85. One more patch for Sieve to update vacation handling to latest spec.
3224 ----------------------------------------------------
3225 See the note above about the 4.44 and 4.50 releases.
3226 ----------------------------------------------------
3232 1. Change 4.43/35 introduced a bug that caused file counts to be
3233 incorrectly computed when quota_filecount was set in an appendfile
3236 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3237 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3240 3. Give more explanation in the error message when the command for a transport
3241 filter fails to execute.
3243 4. There are several places where Exim runs a non-Exim command in a
3244 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3245 was being done only for the command run by the queryprogram router. It is
3246 now done for all such subprocesses. The other cases are: ${run, transport
3247 filters, and the commands run by the lmtp and pipe transports.
3249 5. Some older OS have a limit of 256 on the maximum number of file
3250 descriptors. Exim was using setrlimit() to set 1000 as a large value
3251 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3252 systems. I've change it so that if it can't get 1000, it tries for 256.
3254 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3255 was an oversight, and furthermore, ever since the addition of extra
3256 controls (e.g. 4.43/32), the checks on when to allow different forms of
3257 "control" were broken. There should now be diagnostics for all cases when a
3258 control that does not make sense is encountered.
3260 7. $recipients is now available in the predata ACL (oversight).
3262 8. Tidy the search cache before the fork to do a delivery from a message
3263 received from the command line. Otherwise the child will trigger a lookup
3264 failure and thereby defer the delivery if it tries to use (for example) a
3265 cached ldap connection that the parent has called unbind on.
3267 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3268 of $address_data from the recipient verification was clobbered by the
3269 sender verification.
3271 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3272 was its contents. (It was OK if the option was not defined at all.)
3274 11. A "Completed" log line is now written for messages that are removed from
3275 the spool by the -Mrm option.
3277 12. $host_address is now set to the target address during the checking of
3278 ignore_target_hosts.
3280 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3281 being passed; this would have caused $sender_host_name to have been used if
3282 matching the list had actually called for a host name (not very likely,
3283 since this list is usually IP addresses). A host name is now passed as
3286 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3287 code that specifies a non-restarting handler (typically sigaction() in
3288 modern systems) in an attempt to fix a rare and obscure crash bug.
3290 15. Narrowed the window for a race in the daemon that could cause it to ignore
3291 SIGCHLD signals. This is not a major problem, because they are used only to
3292 wake it up if nothing else does.
3294 16. A malformed maildirsize file could cause Exim to calculate negative values
3295 for the mailbox size or file count. Odd effects could occur as a result.
3296 The maildirsize information is now recalculated if the size or filecount
3299 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3300 support for a long time. Removed HAVE_SYS_VFS_H.
3302 18. Updated exipick to current release from John Jetmore.
3304 19. Allow an empty sender to be matched against a lookup in an address list.
3305 Previously the only cases considered were a regular expression, or an
3308 20. Exim went into a mad DNS lookup loop when doing a callout where the
3309 host was specified on the transport, if the DNS lookup yielded more than
3312 21. The RFC2047 encoding function was originally intended for short strings
3313 such as real names; it was not keeping to the 75-character limit for
3314 encoded words that the RFC imposes. It now respects the limit, and
3315 generates multiple encoded words if necessary. To be on the safe side, I
3316 have increased the buffer size for the ${rfc2047: expansion operator from
3319 22. Failure to deliver a bounce message always caused it to be frozen, even if
3320 there was an errors_to setting on the router. The errors_to setting is now
3323 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3324 canonical form (fully expanded) before being placed in
3325 $sender_host_address.
3327 24. Updated eximstats to version 1.33
3329 25. Include certificate and key file names in error message when GnuTLS fails
3330 to set them up, because the GnuTLS error message doesn't include the name
3331 of the failing file when there is a problem reading it.
3333 26. Expand error message when OpenSSL has problems setting up cert/key files.
3336 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3337 (this can affect the format of dates).
3339 28. exim_tidydb, when checking for the continued existence of a message for
3340 which it has found a message-specific retry record, was not finding
3341 messages that were in split spool directories. Consequently, it was
3342 deleting retry records that should have stayed in existence.
3344 29. eximstats updated to version 1.35
3345 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3346 1.35 - bugfix such that pie charts by volume are generated correctly
3348 30. The SPA authentication driver was not abandoning authentication and moving
3349 on to the next authenticator when an expansion was forced to fail,
3350 contradicting the general specification for all authenticators. Instead it
3351 was generating a temporary error. It now behaves as specified.
3353 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3354 (the order specifies the preference for clients). The order is now AES256,
3355 AES128, 3DES, ARCFOUR128.
3357 31. Small patch to Sieve code - explicitly set From: when generating an
3360 32. Exim crashed if a remote delivery caused a very long error message to be
3361 recorded - for instance if somebody sent an entire SpamAssassin report back
3362 as a large number of 550 error lines. This bug was coincidentally fixed by
3363 increasing the size of one of Exim's internal buffers (big_buffer) that
3364 happened as part of the Exiscan merge. However, to be on the safe side, I
3365 have made the code more robust (and fixed the comments that describe what
3368 33. Some experimental protocols are using DNS PTR records for new purposes. The
3369 keys for these records are domain names, not reversed IP addresses. The
3370 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3371 leaves it alone. Component reversal etc. now happens only for IP addresses.
3374 34. The host_aton() function is supposed to be passed a string that is known
3375 to be a valid IP address. However, in the case of IPv6 addresses, it was
3376 not checking this. This is a hostage to fortune. Exim now panics and dies
3377 if the condition is not met. A case was found where this could be provoked
3378 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3379 components; fortuitously, this particular loophole had already been fixed
3380 by change 4.50/55 or 4.44/33 above.
3382 If there are any other similar loopholes, the new check in host_aton()
3383 itself should stop them being exploited. The report I received stated that
3384 data on the command line could provoke the exploit when Exim was running as
3385 exim, but did not say which command line option was involved. All I could
3386 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3387 running as the user.
3390 35. There was a buffer overflow vulnerability in the SPA authentication code
3391 (which came originally from the Samba project). I have added a test to the
3392 spa_base64_to_bits() function which I hope fixes it.
3395 36. The daemon start-up calls getloadavg() while still root for those OS that
3396 need the first call to be done as root, but it missed one case: when
3397 deliver_queue_load_max is set with deliver_drop_privilege. This is
3398 necessary for the benefit of the queue runner, because there is no re-exec
3399 when deliver_drop_privilege is set.
3401 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3402 was in use, was not putting the data itself into the right store pool;
3403 consequently, it could be overwritten for a subsequent message in the same
3404 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3407 38. Sometimes the final signoff response after QUIT could fail to get
3408 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3409 < 0 before doing a fflush(). This bug looks as though it goes back to the
3410 introduction of TLS in release 3.20, but "sometimes" must have been rare
3411 because the tests only now provoked it.
3417 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3418 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3419 Now it returns an integer. A number of calls were still expecting a T/F
3420 return. Fortuitously, in all cases, the tests worked in OK situations,
3421 which is the norm. However, things would have gone wrong on any write
3422 failures on the smtp file descriptor. This function is used when sending
3423 messages over SMTP and also when doing verify callouts.
3425 2. When Exim is called to do synchronous delivery of a locally submitted
3426 message (the -odf or -odi options), it no longer closes stderr before doing
3429 3. Implemented the mua_wrapper option.
3431 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3433 5. Implemented the functions header_remove(), header_testname(),
3434 header_add_at_position(), and receive_remove_recipient(), and exported them
3437 6. If an ACL "warn" statement specified the addition of headers, Exim already
3438 inserted X-ACL-Warn: at the start if there was no header name. However, it
3439 was not making this test for the second and subsequent header lines if
3440 there were newlines in the string. This meant that an invalid header could
3441 be inserted if Exim was badly configured.
3443 7. Allow an ACL "warn" statement to add header lines at the start or after all
3444 the Received: headers, as well as at the end.
3446 8. Added the rcpt_4xx retry error code.
3448 9. Added postmaster_mailfrom=xxx to callout verification option.
3450 10. Added mailfrom=xxxx to the callout verification option, for verify=
3453 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3454 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3456 12. Inserted some casts to stop certain compilers warning when using pointer
3457 differences as field lengths or precisions in printf-type calls (mostly
3458 affecting debugging statements).
3460 13. Added optional readline() support for -be (dynamically loaded).
3462 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3463 same clock tick as a message's arrival, so that its received time was the
3464 same as the "first fail" time on the retry record, and that message
3465 remained on the queue past the ultimate address timeout, every queue runner
3466 would try a delivery (because it was past the ultimate address timeout) but
3467 after another failure, the ultimate address timeout, which should have then
3468 bounced the address, did not kick in. This was a "< instead of <=" error;
3469 in most cases the first failure would have been in the next clock tick
3470 after the received time, and all would be well.
3472 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3473 being recognized when the domain list was tested by the match_domain
3474 condition in an expansion string.
3476 16. Added the ${str2b64: operator.
3478 17. Exim was always calling setrlimit() to set a large limit for the number of
3479 processes, without checking whether the existing limit was already
3480 adequate. (It did check for the limit on file descriptors.) Furthermore,
3481 errors from getrlimit() and setrlimit() were being ignored. Now they are
3482 logged to the main and panic logs, but Exim does carry on, to try to do its
3483 job under whatever limits there are.
3485 18. Imported PCRE 5.0.
3487 19. Trivial typo in log message " temporarily refused connection" (the leading
3490 20. If the log selector return_path_on_delivery was set and an address was
3491 redirected to /dev/null, the delivery process crashed because it assumed
3492 that a return path would always be set for a "successful" delivery. In this
3493 case, the whole delivery is bypassed as an optimization, and therefore no
3496 21. Internal re-arrangement: the function for sending a challenge and reading
3497 a response while authentication was assuming a zero-terminated challenge
3498 string. It's now changed to take a pointer and a length, to allow for
3499 binary data in such strings.
3501 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3503 23. Exim was not respecting finduser_retries when seeking the login of the
3504 uid under which it was called; it was always trying 10 times. (The default
3505 setting of finduser_retries is zero.) Also, it was sleeping after the final
3506 failure, which is pointless.
3508 24. Implemented tls_on_connect_ports.
3510 25. Implemented acl_smtp_predata.
3512 26. If the domain in control=submission is set empty, Exim assumes that the
3513 authenticated id is a complete email address when it generates From: or
3514 Sender: header lines.
3516 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3517 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3518 chown and chgrp in /bin and hostname in /usr/bin.
3520 28. Exim was keeping the "process log" file open after each use, just as it
3521 does for the main log. This opens the possibility of it remaining open for
3522 long periods when the USR1 signal hits a daemon. Occasional processlog
3523 errors were reported, that could have been caused by this. Anyway, it seems
3524 much more sensible not to leave this file open at all, so that is what now
3527 29. The long-running daemon process does not normally write to the log once it
3528 has entered its main loop, and it closes the log before doing so. This is
3529 so that log files can straightforwardly be renamed and moved. However,
3530 there are a couple of unusual error situations where the daemon does write
3531 log entries, and I had neglected to close the log afterwards.
3533 30. The text of an SMTP error response that was received during a remote
3534 delivery was being truncated at 512 bytes. This is too short for some of
3535 the long messages that one sometimes sees. I've increased the limit to
3538 31. It is now possible to make retry rules that apply only when a message has a
3539 specific sender, in particular, an empty sender.
3541 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3542 it possible to be selective about when SMTP synchronization is enforced.
3544 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3546 32. Implemented hosts_connection_nolog.
3548 33. Added an ACL for QUIT.
3550 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3553 35. Added mailbox_size and mailbox_filecount to appendfile.
3555 36. Added control = no_multiline_responses to ACLs.
3557 37. There was a bug in the logic of the code that waits for the clock to tick
3558 in the case where the clock went backwards by a substantial amount such
3559 that the microsecond fraction of "now" was more than the microsecond
3560 fraction of "then" (but the whole seconds number was less).
3562 38. Added support for the libradius Radius client library this is found on
3563 FreeBSD (previously only the radiusclient library was supported).
3569 1. When certain lookups returned multiple values in the form name=value, the
3570 quoting of the values was not always being done properly. Specifically:
3571 (a) If the value started with a double quote, but contained no whitespace,
3573 (b) If the value contained whitespace other than a space character (i.e.
3574 tabs or newlines or carriage returns) it was not quoted.
3575 This fix has been applied to the mysql and pgsql lookups by writing a
3576 separate quoting function and calling it from the lookup code. The fix
3577 should probably also be applied to nisplus, ibase and oracle lookups, but
3578 since I cannot test any of those, I have not disturbed their existing code.
3580 2. A hit in the callout cache for a specific address caused a log line with no
3581 reason for rejecting RCPT. Now it says "Previous (cached) callout
3582 verification failure".
3584 3. There was an off-by-one bug in the queryprogram router. An over-long
3585 return line was truncated at 256 instead of 255 characters, thereby
3586 overflowing its buffer with the terminating zero. As well as fixing this, I
3587 have increased the buffer size to 1024 (and made a note to document this).
3589 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3590 when Exim is waiting for an SMTP response from a remote server, Exim
3591 restarts its select() call on the socket, thereby resetting its timeout.
3592 This is not a problem when such interrupts are rare. Somebody set up a cron
3593 job to run exiwhat every 2 minutes, which is less than the normal select()
3594 timeout (5 or 10 minutes). This meant that the select() timeout never
3595 kicked in because it was always reset. I have fixed this by comparing the
3596 time when an interrupt arrives with the time at the start of the first call
3597 to select(). If more time than the timeout has elapsed, the interrupt is
3598 treated as a timeout.
3600 5. Some internal re-factoring in preparation for the addition of Sieve
3601 extensions (by MH). In particular, the "personal" test is moved to a
3602 separate function, and given an option for scanning Cc: and Bcc: (which is
3603 not set for Exim filters).
3605 6. When Exim created an email address using the login of the caller as the
3606 local part (e.g. when creating a From: or Sender: header line), it was not
3607 quoting the local part when it contained special characters such as @.
3609 7. Installed new OpenBSD configuration files.
3611 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3612 try to make them clearer.
3614 9. Callout options, other than the timeout value, were being ignored when
3615 verifying sender addresses in header lines. For example, when using
3617 verify = header_sender/callout=no_cache
3619 the cache was (incorrectly) being used.
3621 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3622 only the Cygwin environment.
3624 11. When return_path_on_delivery was set as a log selector, if different remote
3625 addresses in the same message used different return paths and parallel
3626 remote delivery occurred, the wrong values would sometimes be logged.
3627 (Whenever a remote delivery process finished, the return path value from
3628 the most recently started remote delivery process was logged.)
3630 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3631 lines when AUTH and/or TLS are in use. This is the "received protocol"
3632 field. Exim used to use "asmtp" for authenticated SMTP, without any
3633 indication (in the protocol name) for TLS use. Now it follows the RFC and
3634 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3635 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3636 names appear in log lines as well as in Received: header lines.
3638 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3639 extensions, and comparison tests, and to fix some bugs.
3641 14. Changes to the "personal" filter test:
3643 (1) The test was buggy in that it was just doing the equivalent of
3644 "contains" tests on header lines. For example, if a user's address was
3645 anne@some.where, the "personal" test would incorrectly be true for
3647 To: susanne@some.where
3649 This test is now done by extracting each address from the header in turn,
3650 and checking the entire address. Other tests that are part of "personal"
3651 are now done using regular expressions (for example, to check local parts
3652 of addresses in From: header lines).
3654 (2) The list of non-personal local parts in From: addresses has been
3655 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3656 taken from the Sieve specification recommendations.
3658 (3) If the message contains any header line starting with "List-" it is
3659 treated as non-personal.
3661 (4) The test for "circular" in the Subject: header line has been removed
3662 because it now seems ill-conceived.
3664 15. Minor typos in src/EDITME comments corrected.
3666 16. Installed latest exipick from John Jetmore.
3668 17. If headers_add on a router specified a text string that was too long for
3669 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3670 of string_sprintf() is now avoided.
3672 18. $message_body_size was not set (it was always zero) when running the DATA
3673 ACL and the local_scan() function.
3675 19. For the "mail" command in an Exim filter, no default was being set for
3676 the once_repeat time, causing a random time value to be used if "once" was
3677 specified. (If the value happened to be <= 0, no repeat happened.) The
3678 default is now 0s, meaning "never repeat". The "vacation" command was OK
3679 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3680 (I found it when inspecting the code).
3682 20. There is now an overall timeout for performing a callout verification. It
3683 defaults to 4 times the callout timeout, which applies to individual SMTP
3684 commands during the callout. The overall timeout applies when there is more
3685 than one host that can be tried. The timeout is checked before trying the
3686 next host. This prevents very long delays if there are a large number of
3687 hosts and all are timing out (e.g. when the network connections are timing
3688 out). The value of the overall timeout can be changed by specifying an
3689 additional sub-option for "callout", called "maxwait". For example:
3691 verify = sender/callout=5s,maxwait=20s
3693 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3694 to the end before writing, but this should make it even safer).
3696 22. Exim was forgetting that it had advertised PIPELINING for the second and
3697 subsequent messages on an SMTP connection. It was also not resetting its
3698 memory on STARTTLS and an internal HELO.
3700 23. When Exim logs an SMTP synchronization error within a session, it now
3701 records whether PIPELINING has been advertised or not.
3703 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3704 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3705 rather than long int.
3707 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3709 26. Added the never_mail option to autoreply.
3715 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3716 crash if the getsockname() call failed; this can happen if a connection is
3717 closed very soon after it is established. The problem was simply in the
3718 order in which certain operations were done, causing Exim to try to write
3719 to the SMTP stream before it had set up the file descriptor. The bug has
3720 been fixed by making things happen in the correct order.
3726 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3727 before the connection was closed, thus losing the rejection response.
3729 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3730 some early Solaris releases, but causes trouble in current releases where
3731 socklen_t is defined.
3733 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3736 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3739 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3740 that releasing the top of it at the end releases what was used for sub-
3741 expansions (unless the block got too big). However, discard this block if
3742 the first thing is a variable or header, so that we can use its block when
3743 it is dynamic (useful for very large $message_headers, for example).
3745 6. Lookups now cache *every* query, not just the most recent. A new, separate
3746 store pool is used for this. It can be recovered when all lookup caches are
3747 flushed. Lookups now release memory at the end of their result strings.
3748 This has involved some general refactoring of the lookup sources.
3750 7. Some code has been added to the store_xxx() functions to reduce the amount
3751 of flapping under certain conditions.
3753 8. log_incoming_interface used to affect only the <= reception log lines. Now
3754 it causes the local interface and port to be added to several more SMTP log
3755 lines, for example "SMTP connection from", and rejection lines.
3757 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3759 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3761 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3762 could be overwritten at the end of the current message (or the start of a
3763 new message if it was set in a HELO ACL). The value is now preserved for
3764 the duration of the SMTP connection.
3766 12. If a transport had a headers_rewrite setting, and a matching header line
3767 contained an unqualified address, that address was qualified, even if it
3768 did not match any rewriting rules. The underlying bug was that the values
3769 of the flags that permit the existence of unqualified sender and recipient
3770 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3771 non-local messages, and by -bnq for local messages) were not being
3772 preserved with the message after it was received.
3774 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3775 "next input=" as part of the text comprising the host identity instead of
3776 the correct text. The code was using the same buffer for two different
3777 strings. However, depending on which order the printing function evaluated
3778 its arguments, the bug did not always show up. Under Linux, for example, my
3779 test suite worked just fine.
3781 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3782 doesn't work with some older versions of Perl. It has been changed to "my",
3783 which in any case is probably the better facility to use.
3785 15. A really picky compiler found some instances of statements for creating
3786 error messages that either had too many or two few arguments for the format
3789 16. The size of the buffer for calls to the DNS resolver has been increased
3790 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3791 for addresses that have a lot of PTR records. This alleviates a problem; it
3792 does not fully solve it.
3794 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3795 buffer now truncates the list and logs the incident, which is the same
3796 action as happens when Exim is looking up a host name and its aliases.
3797 Previously in this situation something unpredictable would happen;
3798 sometimes it was "internal error: store_reset failed".
3800 18. If a server dropped the connection unexpectedly when an Exim client was
3801 using GnuTLS and trying to read a response, the client delivery process
3802 crashed while trying to generate an error log message.
3804 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3805 string, for example:
3807 warn message = H1: something\nH2: something
3809 the text was added as a single header line from Exim's point of view
3810 though it ended up OK in the delivered message. However, searching for the
3811 second and subsequent header lines using $h_h2: did not work. This has been
3812 fixed. Similarly, if a system filter added multiple headers in this way,
3813 the routers could not see them.
3815 20. Expanded the error message when iplsearch is called with an invalid key to
3816 suggest using net-iplsearch in a host list.
3818 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3819 ACLs are no longer actually imposed (and a message to that effect is
3822 22. If a "gecos" field in a passwd entry contained escaped characters, in
3823 particular, if it contained a \" sequence, Exim got it wrong when building
3824 a From: or a Sender: header from that name. A second bug also caused
3825 incorrect handling when an unquoted " was present following a character
3826 that needed quoting.
3828 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3829 was not being matched caselessly.
3831 24. Arranged for all hyphens in the exim.8 source to be escaped with
3834 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3835 properly. Recipient callout cache records were still being keyed to include
3836 the sender, even when use_sender was set false. This led to far more
3837 callouts that were necessary. The sender is no longer included in the key
3838 when use_sender is false.
3840 26. Added "control = submission" modifier to ACLs.
3842 27. Added the ${base62d: operator to decode base 62 numbers.
3844 28. dnsdb lookups can now access SRV records.
3846 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3847 the configuration file.
3849 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3850 (-v) mode. This makes the output for a verbose queue run more intelligible.
3852 31. Added a use_postmaster feature to recipient callouts.
3854 32. Added the $body_zerocount variable, containing the number of binary zero
3855 bytes in the message body.
3857 33. The time of last modification of the "new" subdirectory is now used as the
3858 "mailbox time last read" when there is a quota error for a maildir
3861 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3863 35. Added +ignore_unknown as a special item in host lists.
3865 36. Code for decoding IPv6 addresses in host lists is now included, even if
3866 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3867 address was recognized as an IP address, but was then not correctly decoded
3868 into binary, causing unexpected and incorrect effects when compared with
3875 1. Very minor rewording of debugging text in manualroute to say "list of
3876 hosts" instead of "hostlist".
3878 2. If verify=header_syntax was set, and a header line with an unqualified
3879 address (no domain) and a large number of spaces between the end of the
3880 name and the colon was received, the reception process suffered a buffer
3881 overflow, and (when I tested it) crashed. This was caused by some obsolete
3882 code that should have been removed. The fix is to remove it!
3884 3. When running in the test harness, delay a bit after writing a bounce
3885 message to get a bit more predictability in the log output.
3887 4. Added a call to search_tidyup() just before forking a reception process. In
3888 theory, someone could use a lookup in the expansion of smtp_accept_max_
3889 per_host which, without the tidyup, could leave open a database connection.
3891 5. Added the variables $recipient_data and $sender_data which get set from a
3892 lookup success in an ACL "recipients" or "senders" condition, or a router
3893 "senders" option, similar to $domain_data and $local_part_data.
3895 6. Moved the writing of debug_print from before to after the "senders" test
3898 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3899 problems for message scanning, either using a data ACL, or using
3900 local_scan() because the Received: header was not generated till after they
3901 were called (in order to set the time as the time of reception completion).
3902 I have revised the way this works. The header is now generated after the
3903 body is received, but before the ACL or local_scan() are called. After they
3904 are run, the timestamp in the header is updated.
3910 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3911 before starting a queue runner without re-exec. This happened only when
3912 deliver_drop_privilege was set or when the Exim user was set to root. The
3913 effect of the bug was that timeouts during subsequent deliveries caused
3914 crashes instead of being properly handled. The handler is now left at its
3915 default (and expected) setting.
3917 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3918 message, again when deliver_drop_privilege is set or Exim is run as root.
3919 The bug described in (1) was not present in this case, but the tidying up
3920 of the other signals was missing. I have made the two cases consistent.
3922 3. The ignore_target_hosts setting on a manualroute router was being ignored
3923 for hosts that were looked up using the /MX notation.
3925 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3928 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3929 operated on the sender address. After changing the $sender_address to <>
3930 for the sender address verify, Exim was re-instated it as the original
3931 (before rewriting) address, but remembering that it had rewritten it, so it
3932 wasn't rewriting it again. This bug also had the effect of breaking the
3933 sender address verification caching when the sender address was rewritten.
3935 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3936 This has been changed so that if the ip literal address matches
3937 ignore_target_hosts, the router declines.
3939 7. Added expansion conditions match_domain, match_address, and match_local_
3940 part (NOT match_host).
3942 8. The placeholder for the Received: header didn't have a length field set.
3944 9. Added code to Exim itself and to exim_lock to test for a specific race
3945 condition that could lead to file corruption when using MBX delivery. The
3946 issue is with the lockfile that is created in /tmp. If this file is removed
3947 after a process has opened it but before that process has acquired a lock,
3948 there is the potential for a second process to recreate the file and also
3949 acquire a lock. This could lead to two Exim processes writing to the file
3950 at the same time. The added code performs the same test as UW imapd; it
3951 checks after acquiring the lock that its file descriptor still refers to
3952 the same named file.
3954 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3955 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3958 11. Added the smtp_active_hostname option. If used, this will typically be made
3959 to depend on the incoming interface address. Because $interface_address is
3960 not set up until the daemon has forked a reception process, error responses
3961 that can happen earlier (such as "too many connections") no longer contain
3964 12. If an expansion in a condition on a "warn" statement fails because a lookup
3965 defers, the "warn" statement is abandoned, and the next ACL statement is
3966 processed. Previously this caused the whole ACL to be aborted.
3968 13. Added the iplsearch lookup type.
3970 14. Added ident_timeout as a log selector.
3972 15. Added tls_certificate_verified as a log selector.
3974 16. Added a global option tls_require_ciphers (compare the smtp transport
3975 option of the same name). This controls incoming TLS connections.
3977 17. I finally figured out how to make tls_require_ciphers do a similar thing
3978 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3979 before starting the TLS session.
3981 18. Tabs are now shown as \t in -bP output.
3983 19. If the log selector return_path_on_delivery was set, Exim crashed when
3984 bouncing a message because it had too many Received: header lines.
3986 20. If two routers both had headers_remove settings, and the first one included
3987 a superfluous trailing colon, the final name in the first list and the
3988 first name in the second list were incorrectly joined into one item (with a
3989 colon in the middle).
3995 1. Added -C and -D options to the exinext utility, mainly to make it easier
3996 to include in the automated testing, but these could be helpful when
3997 multiple configurations are in use.
3999 2. The exinext utility was not formatting the output nicely when there was
4000 an alternate port involved in the retry record key, nor when there was a
4001 message id as well (for retries that were specific to a specific message
4002 and a specific host). It was also confused by IPv6 addresses, because of
4003 the additional colons they contain. I have fixed the IPv4 problem, and
4004 patched it up to do a reasonable job for IPv6.
4006 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4007 delivery, the log line now contains "pipelined" if PIPELINING was used.
4009 4. An SMTP transport process used to panic and die if the bind() call to set
4010 an explicit outgoing interface failed. This has been changed; it is now
4011 treated in the same way as a connect() failure.
4013 5. A reference to $sender_host_name in the part of a conditional expansion
4014 that was being skipped was still causing a DNS lookup. This no longer
4017 6. The def: expansion condition was not recognizing references to header lines
4018 that used bh_ and bheader_.
4020 7. Added the _cache feature to named lists.
4022 8. The code for checking quota_filecount in the appendfile transport was
4023 allowing one more file than it should have been.
4025 9. For compatibility with Sendmail, the command line option
4033 and sets the incoming protocol and host name (for trusted callers). The
4034 host name and its colon can be omitted when only the protocol is to be set.
4035 Note the Exim already has two private options, -pd and -ps, that refer to
4036 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4037 "s", but I don't think that's a major issue.
4039 10. A number of refactoring changes to the code, none of which should affect
4042 (a) The number of logging options was getting close to filling up the
4043 32-bit word that was used as a bit map. I have split them into two classes:
4044 those that are passed in the argument to log_write(), and those that are
4045 only ever tested independently outside of that function. These are now in
4046 separate 32-bit words, so there is plenty of room for expansion again.
4047 There is no change in the user interface or the logging behaviour.
4049 (b) When building, for example, log lines, the code previously used a
4050 macro that called string_cat() twice, in order to add two strings. This is
4051 not really sufficiently general. Furthermore, there was one instance where
4052 it was actually wrong because one of the argument was used twice, and in
4053 one call a function was used. (As it happened, calling the function twice
4054 did not affect the overall behaviour.) The macro has been replaced by a
4055 function that can join an arbitrary number of extra strings onto a growing
4058 (c) The code for expansion conditions now uses a table and a binary chop
4059 instead of a serial search (which was left over from when there were very
4060 few conditions). Also, it now recognizes conditions like "pam" even when
4061 the relevant support is not compiled in: a suitably worded error message is
4062 given if an attempt is made to use such a condition.
4064 11. Added ${time_interval:xxxxx}.
4066 12. A bug was causing one of the ddress fields not to be passed back correctly
4067 from remote delivery subprocesses. The field in question was not being
4068 subsequently used, so this caused to problems in practice.
4070 13. Added new log selectors queue_time and deliver_time.
4072 14. Might have fixed a bug in maildirsizefile handling that threw up
4073 "unexpected character" debug warnings, and recalculated the data
4074 unnecessarily. In any case, I expanded the warning message to give more
4077 15. Added the message "Restricted characters in address" to the statements in
4078 the default ACL that block characters like @ and % in local parts.
4080 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4081 Three changes have been made:
4083 (a) There was a serious bug; a negative response to MAIL caused the whole
4084 recipient domain to be cached as invalid, thereby blocking all messages
4085 to all local parts at the same domain, from all senders. This bug has
4086 been fixed. The domain is no longer cached after a negative response to
4087 MAIL if the sender used is not empty.
4089 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4092 (c) A new callout option, "use_sender" has been added for people who want
4093 the modified behaviour.
4099 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4102 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4103 indeed breaks things for older releases.
4105 3. Added additional logging to the case where there is a problem reading data
4106 from a filter that is running in a subprocess using a pipe, in order to
4107 try to track down a specific problem.
4109 4. Testing facility fudge: when running in the test harness and attempting
4110 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4111 getting "No route to host". Convert this to a timeout.
4113 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4116 6. Some OS don't have socklen_t but use size_t instead. This affects the
4117 fifth argument of getsockopt() amongst other things. This is now
4118 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4119 can be set for individual OS. I have set it for SunOS5, OSF1, and
4120 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4121 some earlier ones do not.
4123 7. Change 4.30/15 was not doing the test caselessly.
4125 8. The standard form for an IPv6 address literal was being rejected by address
4126 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4127 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4128 this, as well as the form without the "IPv6" on the front (but only when
4129 address literals are enabled, of course).
4131 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4133 10. Exim crashed if a message with an empty sender address specified by -f
4134 encountered a router with an errors_to setting. This could be provoked only
4135 by a command such as
4139 where an empty string was supplied; "<>" did not hit this bug.
4141 11. Installed PCRE release 4.5.
4143 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4144 remained set. It is now erased.
4146 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4147 times from message ids (which are base 36 rather than the normal 62).
4149 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4150 were being counted as actual protocol errors, and logged if the log
4151 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4152 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4153 and DATA following a set of rejected RCPTs do not count as protocol errors.
4154 In other words, Exim assumes they were pipelined, though this may not
4155 actually be the case. Of course, in all cases the client gets an
4156 appropriate error code.
4158 15. If a lookup fails in an ACL condition, a message about the failure may
4159 be available; it is used if testing the ACL cannot continue, because most
4160 such messages specify what the cause of the deferral is. However, some
4161 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4162 that caused an old message to be retained and used if a later statement
4163 caused a defer, replacing the real cause of the deferral.
4165 16. If an IP address had so many PTR records that the DNS lookup buffer
4166 was not large enough to hold them, Exim could crash while trying to process
4167 the truncated data. It now detects and logs this case.
4169 17. Further to 4.21/58, another change has been made: if (and only if) the
4170 first line of a message (the first header line) ends with CRLF, a bare LF
4171 in a subsequent header line has a space inserted after it, so as not to
4172 terminate the header.
4174 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4175 unnecessarily, used atoi() instead of strtol(), and didn't check the
4176 termination when getting file sizes from file names by regex.
4178 19. Completely re-implemented the support for maildirsize files, in the light
4179 of a number of problems with the previous contributed implementation
4180 (4.30/29). In particular:
4182 . If the quota is zero, the maildirsize file is maintained, but no quota is
4185 . If the maildir directory does not exist, it is created before any attempt
4186 to write a maildirsize file.
4188 . The quota value in the file is just a cache; if the quota is changed in
4189 the transport, the new value overrides.
4191 . A regular expression is available for excluding directories from the
4194 20. The autoreply transport checks the characters in options that define the
4195 message's headers; it allows continued headers, but it was checking with
4196 isspace() after an embedded newline instead of explicitly looking for a
4199 21. If all the "regular" hosts to which an address was routed had passed their
4200 expiry times, and had not reached their retry times, the address was
4201 bounced, even if fallback hosts were defined. Now Exim should go on to try
4204 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4205 equivalent code in the SMTP transport. Some hosts send humungous responses
4206 to HELO/EHLO, more than 1024 it seems.
4208 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4209 gives compiler warnings in some environments. I've now done it "properly",
4212 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4213 (because of problems with the built-in one) was declared to return uschar *
4214 instead of char *, causing compiler failure.
4216 25. Fixed a file descriptor leak when processing alias/forward files.
4218 26. Fixed a minor format string issue in dbfn.c.
4220 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4222 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4223 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4226 29. When a sender address is verified, it is cached, to save repeating the test
4227 when there is more than one recipient in a message. However, when the
4228 verification involves a callout, it is possible for different callout
4229 options to be set for different recipients. It is too complicated to keep
4230 track of this in the cache, so now Exim always runs a verification when a
4231 callout is required, relying on the callout cache for the optimization.
4232 The overhead is duplication of the address routing, but this should not be
4235 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4236 to be verified with callout=postmaster, and the main callout worked but the
4237 postmaster check failed, the verification correctly failed. However, if a
4238 subsequent RCPT command asked for sender verification *without* the
4239 postmaster check, incorrect caching caused this verification also to fail,
4242 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4243 it was not caching the DNS options (qualify_single, search_parents) that
4244 were used when the lookup failed. A subsequent lookup with different
4245 options therefore always gave the same answer, though there were cases
4246 where it should not have. (Example: a "domains = !$mx_any" option on a
4247 dnslookup router: the "domains" option is always processed without any
4248 widening, but the router might have qualify_single set.) Now Exim uses the
4249 cached value only when the same options are set.
4251 32. Added John Jetmore's "exipick" utility to the distribution.
4253 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4254 than a timeout (e.g. a certificate is required, and is not provided), an
4255 Exim server now closes the connection immediately. Previously it waited for
4256 the client to close - but if the client is SSL, it seems that they each
4257 wait for each other, leading to a delay before one of them times out.
4259 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4260 maintained 0.8.x compatibility because I don't think many are using it, and
4261 it is clearly obsolete.
4263 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4266 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4267 client certificate was expired. A simple patch fixes this, though I don't
4268 understand the full logic of why the verify callback is called multiple
4271 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4272 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4273 which causes problems with some clients (such as the Certicom SSL Plus
4274 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4275 disables the coutermeasure allowing Eudora to connect."
4277 38. Exim was not checking that a write() to a log file succeeded. This could
4278 lead to Bad Things if a log got too big, in particular if it hit a file
4279 size limit. Exim now panics and dies if it cannot write to a log file, just
4280 as it does if it cannot open a log file.
4282 39. Modified OS/Makefile-Linux so that it now contains
4284 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4286 The two -D definitions ensure that Exim is compiled with large file
4287 support, which makes it possible to handle log files that are bigger than
4290 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4291 instance) a domain was checked against a named list that involved a lookup,
4292 causing $domain_data to be set, then another domain was checked against the
4293 same list, then the first domain was re-checked, the value of $domain_data
4294 after the final check could be wrong. In particular, if the second check
4295 failed, it could be set empty. This bug probably also applied to
4298 41. The strip_trailing_dot option was not being applied to the address given
4299 with the -f command-line option.
4301 42. The code for reading a message's header from the spool was incrementing
4302 $received_count, but never initializing it. This meant that the value was
4303 incorrect (doubled) while delivering a message in the same process in which
4304 it was received. In the most common configuration of Exim, this never
4305 happens - a fresh exec is done - but it can happen when
4306 deliver_drop_privilege is set.
4308 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4309 it now includes up to 150 characters of the unexpected data in the log
4312 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4313 and building data strings. The size of both of these buffers was 10 000
4314 bytes - far larger than anybody would *ever* want, thought I. Needless to
4315 say, somebody hit the limit. I have increased the maximum line length to
4316 20 000 and the maximum data length of concatenated lines to 100 000. I have
4317 also fixed two bugs, because there was no checking on these buffers. Tsk,
4318 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4319 buffer is too small.
4321 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4322 lsearch lookups. Now it does.
4324 46. When parsing a route_list item in a manualroute router, a fixed-length
4325 buffer was used for the list of hosts. I made this 1024 bytes long,
4326 thinking that nobody would ever have a list of hosts that long. Wrong.
4327 Somebody had a whole pile of complicated expansion conditions, and the
4328 string was silently truncated, leading to an expansion error. It turns out
4329 that it is easier to change to an unlimited length (owing to other changes
4330 that have happened since this code was originally written) than to build
4331 structure for giving a limitation error. The length of the item that
4332 expands into the list of hosts is now unlimited.
4334 47. The lsearch lookup could not handle data where the length of text line was
4335 more than 4095 characters. Such lines were truncated, leading to shortened
4336 data being returned. It should now handle lines of any length.
4338 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4339 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4342 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4343 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4344 no longer happens in comment lines. A list of replacements is now placed
4345 at the head of all of the source files, except those whose only change is
4346 to replace PERL_COMMAND in the very first #! line.
4348 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4349 messages on the queue, with a bottom-up merge sort, using code contributed
4350 by Michael Haardt. This should make operations like -bp somewhat faster on
4351 large queues. It won't affect queue runners, except when queue_run_in_order
4354 51. Installed eximstats 1.31 in the distribution.
4356 52. Added support for SRV lookups to the dnslookup router.
4358 53. If an ACL referred to $message_body or $message_body_end, the value was not
4359 reset for any messages that followed in the same SMTP session.
4361 54. The store-handling optimization for building very long strings was not
4362 differentiating between the different store pools. I don't think this
4363 actually made any difference in practice, but I've tidied it.
4365 55. While running the routers to verify a sender address, $sender_address
4366 was still set to the sender address. This is wrong, because when routing to
4367 send a bounce to the sender, it would be empty. Therefore, I have changed
4368 it so that, while verifying a sender address, $sender_address is set to <>.
4369 (There is no change to what happens when verifying a recipient address.)
4371 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4372 target A or AAAA records (if not already returned) without resetting the
4373 qualify_single or search_parents options of the DNS resolver. These are
4374 inappropriate in this case because the targets of MX and SRV records must
4375 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4376 target that, when qualified, matched something in the local domain. These
4377 two options are now turned off when doing these lookups.
4379 57. It seems that at least some releases of Reiserfs (which does not have the
4380 concept of a fixed number of inodes) returns zero and not -1 for the
4381 number of available inodes. This interacted badly with check_spool_inodes,
4382 which assumed that -1 was the "no such thing" setting. What I have done is
4383 to check that the total number of inodes is greater than zero before doing
4384 the test of how many are available.
4386 58. When a "warn" ACL statement has a log_message modifier, the message is
4387 remembered, and not repeated. This is to avoid a lot of repetition when a
4388 message has many recipients that cause the same warning to be written.
4389 Howewer, Exim was preserving the list of already written lines for an
4390 entire SMTP session, which doesn't seem right. The memory is now reset if a
4391 new message is started.
4393 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4394 debugging output unless log_rewrite was also set.
4396 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4397 of (int)(handle) when we know that handle contains (void *)(-1).
4399 61. The Exim daemon panic-logs an error return when it closes the incoming
4400 connection. However "connection reset by peer" seems to be common, and
4401 isn't really an error worthy of noting specially, so that particular error
4404 62. When Exim is trying to find all the local interfaces, it used to panic and
4405 die if the ioctl to get the interface flags failed. However, it seems that
4406 on at least one OS (Solaris 9) it is possible to have an interface that is
4407 included in the list of interfaces, but for which you get a failure error
4408 for this call. This happens when the interface is not "plumbed" into a
4409 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4410 failure of the "get flags" call assumes that the interface is down.
4412 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4413 makes life easier for people who are doing arithmetic on fields extracted
4414 from dates, where you often get leading zeros that should not be
4415 interpreted as octal.
4417 64. Added qualify_domain to the redirect router, to override the global
4420 65. If a pathologically long header line contained very many addresses (the
4421 report of this problem mentioned 10 000) and each of them was rewritten,
4422 Exim could use up a very large amount of memory. (It kept on making new
4423 copies of the header line as it rewrote, and never released the old ones.)
4424 At the expense of a bit more processing, the header rewriting function has
4425 been changed so that it no longer eats memory in this way.
4427 66. The generation of the Received: header has been moved from the time that a
4428 message starts to be received, to the time that it finishes. The timestamp
4429 in the Received: header should now be very close to that of the <= log
4430 line. There are two side-effects of this change:
4432 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4433 logged header lines no longer include the local Received: line, because
4434 it has not yet been created. The same applies to a copy of the message
4435 that is returned to a non-SMTP sender when a message is rejected.
4437 (b) When a filter file is tested using -bf, no additional Received: header
4438 is added to the test message. After some thought, I decided that this
4441 This change does not affect the value of $received_for. It is still set
4442 after address rewriting, but before local_scan() is called.
4444 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4446 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4447 gave an unhelpful panic error message, and a defer error. I have managed to
4448 change this behaviour so that it now rejects any supplied certificate,
4449 which seems right, as the list of acceptable certificates is empty.
4451 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4452 gave an unhelpful defer error. I have not managed to make this reject any
4453 supplied certificates, but the error message it gives is "no certificate
4454 supplied", which is not helpful.
4456 70. exigrep's output now also includes lines that are not associated with any
4457 message, but which match the given pattern. Implemented by a patch from
4458 Martin Sluka, which also tidied up the Perl a bit.
4460 71. Recipient callout verification, like sender verification, was using <> in
4461 the MAIL FROM command. This isn't really the right thing, since the actual
4462 sender may affect whether the remote host accepts the recipient or not. I
4463 have changed it to use the actual sender in the callout; this means that
4464 the cache record is now keyed on a recipient/sender pair, not just the
4465 recipient address. There doesn't seem to be a real danger of callout loops,
4466 since a callout by the remote host to check the sender would use <>.
4467 [SEE ABOVE: changed after hitting problems.]
4469 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4470 temporary errors. However, in the case of such a code being given after
4471 the end of a data transmission (i.e. after ".") Exim was failing to write
4472 a retry record for the message. (Yes, there was some broken host that was
4473 actually sending 8xx at this point.)
4475 73. An unknown lookup type in a host list could cause Exim to panic-die when
4476 the list was checked. (An example that provoked this was putting <; in the
4477 middle of a list instead of at the start.) If this happened during a DATA
4478 ACL check, a -D file could be left lying around. This kind of configuration
4479 error no longer causes Exim to die; instead it causes a defer errror. The
4480 incident is still logged to the main and panic logs.
4482 74. Buglet left over from Exim 3 conversion. The message "too many messages
4483 in one connection" was written to the rejectlog but not the mainlog, except
4484 when address rewriting (yes!) was being logged.
4486 75. Added write_rejectlog option.
4488 76. When a system filter was run not as root (that is, when system_filter_user
4489 was set), the values of the $n variables were not being returned to the
4490 main process; thus, they were not subsequently available in the $sn
4493 77. Added +return_path_on_delivery log selector.
4495 78. A connection timeout was being treated differently from recipients deferred
4496 when testing hosts_max_try with a message that was older than the host's
4497 retry timeout. (The host should not be counted, thus allowing all hosts to
4498 be tried at least once before bouncing.) This may have been the cause of an
4499 occasionally reported bug whereby a message would remain on the queue
4500 longer than the retry timeout, but would be bounced if a delivery was
4501 forced. I say "may" because I never totally pinned down the problem;
4502 setting up timeout/retry tests is difficult. See also the next item.
4504 79. The ultimate address timeout was not being applied to errors that involved
4505 a combination of host plus message (for example, a timeout on a MAIL
4506 command). When an address resolved to a number of possible hosts, and they
4507 were not all tried for each delivery (e.g. because of hosts_max_try), a
4508 message could remain on the queue longer than the retry timeout.
4510 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4513 81. Fixed an obscure SMTP outgoing bug which required at least the following
4514 conditions: (a) there was another message waiting for the same server;
4515 (b) the server returned 5xx to all RCPT commands in the first message so
4516 that the message was not completed; (c) the server dropped the connection
4517 or gave a negative response to the RSET that Exim sends to abort the
4518 transaction. The observed case was a dropped connection after DATA that had
4519 been sent in pipelining mode. That is, the server had advertised PIPELINING
4520 but was not implementing it correctly. The effect of the bug was incorrect
4521 behaviour, such as trying another host, and this could lead to a crash.
4527 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4528 and daemon.c were passed as pointers to ints; they should have been
4529 pointers to socklen_t variables (which are typically unsigned ints).
4531 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4534 3. Fixed a really odd bug that affected only the testing scheme; patching a
4535 certain fixed string in the binary changed the value of another string that
4536 happened to be identical to the end of the original first string.
4538 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4539 name", it returns that address as the IP address. On some operating
4540 systems (e.g. Solaris), it also passes back the IP address string as the
4541 "host name". However, on others (e.g. Linux), it passes back an empty
4542 string. Exim wasn't checking for this, and was changing the host name to an
4543 empty string, assuming it had been canonicized.
4545 5. Although rare, it is permitted to have more than one PTR record for a given
4546 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4547 all the names associated with an address, because they do in Solaris.
4548 However, it seems that they do not in Linux for data that comes from the
4549 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4550 I found this out when I moved to a new Linux workstation and tried to run
4551 the Exim test suite.
4553 To get round this problem I have changed the code so that it now does its
4554 own call to the DNS to look up PTR records when searching for a host name.
4555 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4556 addresses that are only in /etc/hosts are still found.
4558 This behaviour is, however, controlled by an option called host_lookup_
4559 order, which defaults to "bydns:byaddr". If people want to use the other
4560 order, or indeed, just use one or the other means of lookup, they can
4561 specify it in this variable.
4563 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4564 some operating systems, this comes back from gethostbyaddr() as an empty
4565 string, and this is what Exim used to test for. However, it seems that in
4566 other systems, "." is yielded. Exim now tests for this case too.
4568 7. The values of check_spool_space and check_log_space are now held internally
4569 as a number of kilobytes instead of an absolute number of bytes. If a
4570 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4571 kilobyte. This means that much larger values can be stored.
4573 8. Exim monitor: an attempt to get the action menu when not actually pointing
4574 at a message produces an empty menu entitled "No message selected". This
4575 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4576 no entries in it ("Shell widget menu has zero width and/or height"). So I
4577 have added a single, blank menu entry in this case.
4579 9. Added ${quote_local_part.
4581 10. MIME decoding is now applied to the contents of Subject: header lines when
4584 11. Now that a reference to $sender_host_address automatically causes a reverse
4585 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4586 host lookup before query-style lookups in lists that might use this
4587 variable. This has therefore been abolished, and the "net-" prefix is no
4588 longer necessary for query-style lookups.
4590 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4591 have been a typo for LFLAGS, so it has been changed.
4593 13. The install script calls Exim with "-C /dev/null" in order to find the
4594 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4595 to be output. Howeve, since Exim outputs its version number before the
4596 error, it didn't break the script. It just looked ugly. I fixed this by
4597 always allowing "-C /dev/null" if the caller is root.
4599 14. Ignore overlarge ACL variable number when reading spool file - insurance
4600 against a later release with more variables having written the file.
4602 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4603 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4604 this, as well as the form without the "IPv6" on the front.
4606 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4607 OS/Makefile-Darwin file.
4609 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4610 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4612 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4613 variable. However, some OS use ints for inodes. I've added cast to long int
4614 to get rid of the compiler warning.
4616 19. I had forgotten to lock out "/../" in configuration file names when
4617 ALT_CONFIG_PREFIX was set.
4619 20. Routers used for verification do not need to specify transports. However,
4620 if such a router generated a host list, and callout was configured, Exim
4621 crashed, because it could not find a port number from the (non-existent)
4622 transport. It now assumes port 25 in this circumstance.
4624 21. Added the -t option to exigrep.
4626 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4627 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4628 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4629 them is set, LOOKUP_LSEARCH is forced.
4631 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4633 24. Added sender and host information to the "rejected by local_scan()" log
4634 line; previously there was no indication of these.
4636 25. Added .include_if_exists.
4638 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4639 Linux. It turns out that not all file systems support this. Apparently some
4640 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4641 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4642 not supported on the file descriptor, is now ignored when Exim is trying to
4643 sync a directory. This applies only to Linux.
4645 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4647 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4648 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4651 29. Added support for maildirsize files from supplied patch (modified a bit).
4653 30. The use of :fail: followed by an empty string could lead Exim to respond to
4654 sender verification failures with (e.g.):
4656 550 Verification failed for <xxx>
4657 550 Sender verify failed
4659 where the first response line was missing the '-' that indicates it is not
4660 the final line of the response.
4662 31. The loop for finding the name of the user that called Exim had a hardwired
4663 limit of 10; it now uses the value of finduser_retries, which is used for
4664 all other user lookups.
4666 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4669 33. Exim was neglecting to zero errno before one call of strtol() when
4670 expanding a string and expecting an integer value. On some systems this
4671 resulted in spurious "integer overflow" errors. Also, it was casting the
4672 result into an int without checking.
4674 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4675 did not work. The code looks as if it has *never* worked, though it appears
4676 to have been documented since at least releast 1.62. I have made it work.
4678 35. The "timeout_DNS" error in retry rules, also documented since at least
4679 1.62, also never worked. As it isn't clear exactly what this means, and
4680 clearly it isn't a major issue, I have abolished the feature by treating it
4681 as "timeout", and writing a warning to the main and panic logs.
4683 36. The display of retry rules for -brt wasn't always showing the error code
4686 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4687 timeout_connect_A, timeout_connect_MX.
4689 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4690 to the empty sender.
4692 39. The daemon was not analysing the content of -oX till after it had closed
4693 stderr and disconnected from the controlling terminal. This meant that any
4694 syntax errors were only noted on the panic log, and the return code from
4695 the command was 0. By re-arranging the code a little, I've made the
4696 decoding happen first, so such errors now appear on stderr, and the return
4697 code is 1. However, the actual setting up of the sockets still happens in
4698 the disconnected process, so errors there are still only recorded on the
4701 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4702 connections (as happens on some IP stacks) was logged at start up time as
4703 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4704 it from "IPv6 and IPv4", which means that two separate sockets are being
4707 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4708 says whether AF_INET or AF_INET6 was passed as an argument.
4710 42. Exiwhat output was messed up when time zones were included in log
4713 43. Exiwhat now gives more information about the daemon's listening ports,
4714 and whether -tls-on-connect was used.
4716 44. The "port" option of the smtp transport is now expanded.
4718 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4719 silently ignored. Now an error message is written to the main and panic
4722 46. There's a new ACL modifier called "logwrite" which writes to a log file
4723 as soon as it is encountered.
4725 47. Added $local_user_uid and $local_user_gid at routing time.
4727 48. Exim crashed when trying to verify a sender address that was being
4730 49. Exim was recognizing only a space character after ".include". It now also
4731 recognizes a tab character.
4733 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4734 extracting the relevant information from the specification. The man page no
4735 longer contains scrambled data for the -d option, and I've added a section
4736 at the front about calling Exim under different names.
4738 51. Added "extra_headers" argument to the "mail" command in filter files.
4740 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4743 53. Installed eximstats 1.29.
4745 54. Added transport_filter_timeout as a generic transport option.
4747 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4748 Cc: header lines. This was required by RFC 822, but it not required by RFC
4751 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4752 incoming messages that did not have them. Now it does so only if the
4753 message originates locally, that is, if there is no associated remote host
4754 address. When Resent- header lines are present, this applies to the Resent-
4755 lines rather than the non-Resent- lines.
4757 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4758 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4760 58. Messages for configuration errors now include the name of the main
4761 configuration file - useful now that there may be more than one file in a
4762 list (.included file names were always shown).
4764 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4765 for those rare installations that do not start the daemon as root or run it
4766 setuid root. I've cut out the call to initgroups() if the daemon is not
4769 60. The Exim user and group can now be bound into the binary as text strings
4770 that are looked up at the start of Exim's processing.
4772 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4774 62. Added $mailstore_basename variable.
4776 63. Installed patch to sieve.c from Michael Haardt.
4778 64. When Exim failed to open the panic log after failing to open the main log,
4779 the original message it was trying to log was written to stderr and debug
4780 output, but if they were not available (the usual case in production), it
4781 was lost. Now it is written to syslog before the two lines that record the
4782 failures to open the logs.
4784 65. Users' Exim filters run in subprocesses under the user's uid. It is
4785 possible for a "deliver" command or an alias in a "personal" command to
4786 provoke an address rewrite. If logging of address rewriting is configured,
4787 this fails because the process is not running as root or exim. There may be
4788 a better way of dealing with this, but for the moment (because 4.30 needs
4789 to be released), I have disabled address rewrite logging when running a
4790 filter in a non-root, non-exim process.
4796 1. The buildconfig auxiliary program wasn't quoting the value set for
4797 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4798 not defined. This bug was masked in 4.22 by the effect that was fixed in
4801 2. Some messages that were rejected after a message id was allocated were
4802 shown as "incomplete" by exigrep. It no longer does this for messages that
4803 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4805 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4806 have allow_domain_literals set, the ID did not get logged in the <= line.
4807 Domain literals are now always recognized in Message-ID: header lines.
4809 4. The first argument for a ${extract expansion item is the key name or field
4810 number. Leading and trailing spaces in this item were not being ignored,
4811 causing some misleading effects.
4813 5. When deliver_drop_privilege was set, single queue runner processes started
4814 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4815 same command in the process it spins off) were not dropping privilege.
4817 6. When the daemon running as "exim" started a queue runner, it always
4818 re-executed Exim in the spun-off process. This is a waste of effort when
4819 deliver_drop_privilege is set. The new process now just calls the
4820 queue-runner function directly.
4826 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4829 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4830 ignored. Though the use of -oP was forcing the writing of a pid file, it
4831 was always written to the default place.
4833 3. If the message "no IP address found for host xxxx" is generated during
4834 incoming verification, it is now followed by identification of the incoming
4835 connection (so you can more easily find what provoked it).
4837 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4839 5. Added some features to "harden" Exim a bit more against certain attacks:
4841 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4842 be put in Local/Makefile. This is like the never_users runtime option,
4843 but it cannot be overridden. The default setting is "root".
4845 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4846 prefix string with which any file named in a -C command line option
4849 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4850 is retained for -C and -D only if the caller of Exim is root. Without
4851 it, the exim user may also use -C and -D and retain privilege.
4853 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4854 command line option is disabled.
4856 6. Macro names set by the -D option must start with an upper case letter, just
4857 like macro names defined in the configuration file.
4859 7. Added "dereference=" facility to LDAP.
4861 8. Two instances of the typo "uknown" in the source files are fixed.
4863 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4864 the Configure-Makefile script screwed up while processing it.
4866 10. Incorporated PCRE 4.4.
4868 11. The SMTP synchronization check was not operating right at the start of an
4869 SMTP session. For example, it could not catch a HELO sent before the client
4870 waited for the greeting. There is now a check for outstanding input at the
4871 point when the greeting is written. Because of the duplex, asynchronous
4872 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4873 way, but not yet received, when the check is performed.
4875 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4876 on TCP/IP sockets, because this apparently causes some broken clients to
4879 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4880 unchanged) from the Cygwin maintainer.
4882 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4883 was supported instead of testing for mbx. Effectively a typo.
4885 15. The spa authenticator server code was not checking that the input it
4886 received was valid base64.
4888 16. The debug output line for the "set" modifier in ACLs was not showing the
4889 name of the variable that was being set.
4891 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4893 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4894 lookup on its own. Something else was needed to trigger the lookup. For
4895 example, a match in host_lookup or the need for a host name in a host list.
4896 Now, if $sender_host_name is referenced and the host name has not yet been
4897 looked up, a lookup is performed. If the lookup fails, the variable remains
4898 empty, and $host_lookup_failed is set to "1".
4900 19. Added "eqi" as a case-independent comparison operator.
4902 20. The saslauthd authentication condition could segfault if neither service
4903 nor realm was specified.
4905 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4906 error message that was logged was misleading, and incoming SMTP
4907 connections were dropped. The message is now more accurate, and temporary
4908 errors are given to SMTP connections.
4910 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4911 (except RSET) with a 421 error, until QUIT is received. However, it was
4912 failing to send a response to QUIT.
4914 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4915 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4916 ACL is now run after the helo_try_verify_hosts code.
4918 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4919 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4920 case-independent, so other case variants are also recognized). Apparently
4921 some systems use these upper case variants.
4923 25. If more than two messages were waiting for the same host, and a transport
4924 filter was specified for the transport, Exim sent two messages over the
4925 same TCP/IP connection, and then failed with "socket operation on non-
4926 socket" when it tried to send the third.
4928 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4930 27. The extern definition of crypt16() in expand.c was not being excluded when
4931 the OS had its own crypt16() function.
4933 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4934 as a preferred synonym for return_size_limit, both as an option and as an
4937 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4939 30. Changed the default configuration ACL to relax the local part checking rule
4940 for addresses that are not in any local domains. For these addresses,
4941 slashes and pipe symbols are allowed within local parts, but the sequence
4942 /../ is explicitly forbidden.
4944 31. SPA server authentication was not clearing the challenge buffer before
4947 32. log_message in a "warn" ACL statement was writing to the reject log as
4948 well as to the main log, which contradicts the documentation and doesn't
4949 seem right (because no rejection is happening). So I have stopped it.
4951 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4952 However, I am unable to do any testing of this.
4954 34. Fixed an infelicity in the appendfile transport. When checking directories
4955 for a mailbox, to see if any needed to be created, it was accidentally
4956 using path names with one or more superfluous leading slashes; tracing
4957 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4959 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4960 contents are added to the log line that is written for every discarded
4961 recipient. (Previously a log_message setting was ignored.)
4963 36. The ${quote: operator now quotes the string if it is empty.
4965 37. The install script runs exim in order to find its version number. If for
4966 some reason other than non-existence or emptiness, which it checks, it
4967 could not run './exim', it was installing it with an empty version number,
4968 i.e. as "exim-". This error state is now caught, and the installation is
4971 38. An argument was missing from the function that creates an error message
4972 when Exim fails to connect to the socket for saslauthd authentication.
4973 This could cause Exim to crash, or give a corrupted message.
4975 39. Added isip, isip4, and isip6 to ${if conditions.
4977 40. The ACL variables $acl_xx are now saved with the message, and can be
4978 accessed later in routers, transports, and filters.
4980 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4981 strings in the file are not string-expanded.
4983 42. If a MAIL command specified a SIZE value that was too large to fit into an
4984 int variable, the check against message_size_limit failed. Such values are
4985 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4986 day this will have to be increased, but I don't think I want to be around
4987 when emails are that large.
4994 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4995 iconv() is not standard in FreeBSD.
4997 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4998 IPv6 enabled. The observed symptom was a segmentation fault on return from
4999 the function os_common_find_running_interfaces() in src/os.c.
5001 3. In the check_special_case() function in daemon.c I had used "errno" as an
5002 argument name, which causes warnings on some systems. This was basically a
5003 typo, since it was named "eno" in the comments!
5005 4. The code that waits for the clock to tick (at a resolution of some fraction
5006 of a second) so as to ensure message-id uniqueness was always waiting for
5007 at least one whole tick, when it could have waited for less. [This is
5008 almost certainly not relevant at current processor speeds, where it is
5009 unlikely to ever wait at all. But we try to future-proof.]
5011 5. The function that sleeps for a time interval that includes fractions of a
5012 second contained a race. It did not block SIGALRM between setting the
5013 timer, and suspending (a couple of lines later). If the interval was short
5014 and the sigsuspend() was delayed until after it had expired, the suspension
5015 never ended. On busy systems this could lead to processes getting stuck for
5018 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5019 process, before it forks any delivery processes. The open lookup caching
5020 mechanism meant that the open file or database connection was passed into
5021 the delivery process. The problem was that delivery processes always tidy
5022 up cached lookup data. This could cause a problem for the next delivery
5023 process started by the queue runner, because the external queue runner
5024 process does not know about the closure. So the next delivery process
5025 still has data in the lookup cache. In the case of a file lookup, there was
5026 no problem because closing a file descriptor in a subprocess doesn't affect
5027 the parent. However, if the lookup was caching a connection to a database,
5028 the connection was closed, and the second delivery process was likely to
5029 see errors such as "PGSQL: query failed: server closed the connection
5030 unexpectedly". The problem has been fixed by closing all cached lookups
5031 in a queue runner before running a delivery process.
5033 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5034 seem to have the "const" qualifier which it has on other OS. I've
5037 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5038 *of the same type* (client or server) with the same public name that an
5039 error should be diagnosed.
5041 9. When Exim looked up a host name for an IP address, but failed to find the
5042 original IP address when looking up the host name (a safety check), it
5043 output the message "<ip address> does not match any IP for NULL", which was
5044 confusing, to say the least. The bug was that the host name should have
5045 appeared instead of "NULL".
5047 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5048 user that is built into the binary, and the -C or -D options is used, root
5049 privilege is dropped before the configuration file is read. In addition,
5050 logging is switched to stderr instead of the normal log files. If the
5051 configuration then re-defines the Exim user, the unprivileged environment
5052 is probably not what is expected, so Exim logs a panic warning message (but
5055 However, if deliver_drop_privilege is set, the unprivileged state may well
5056 be exactly what is intended, so the warning has been cut out in that case,
5057 and Exim is allowed to try to write to its normal log files.
5063 1. smtp_return_error_details was not giving details for temporary sender
5064 or receiver verification errors.
5066 2. Diagnose a configuration error if two authenticators have the same public
5069 3. Exim used not to create the message log file for a message until the first
5070 delivery attempt. This could be confusing when incoming messages were held
5071 for policy or load reasons. The message log file is now created at the time
5072 the message is received, and an initial "Received" line is written to it.
5074 4. The automatically generated man page for command line options had a minor
5075 bug that caused no ill effects; however, a more serious problem was that
5076 the procedure for building the man page automatically didn't always
5077 operate. Consequently, release 4.20 contains an out-of-date version. This
5078 shouldn't happen again.
5080 5. When building Exim with embedded Perl support, the script that builds the
5081 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5082 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5084 6. The freeze_tell option was not being used for messages that were frozen on
5085 arrival, either by an ACL or by local_scan().
5087 7. Added the smtp_incomplete_transaction log selector.
5089 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5090 was accepting AUTH without a new EHLO.
5092 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5093 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5096 10. Logging of TCP/IP connections (when configured) now happens in the main
5097 daemon process instead of the child process, so that the TCP/IP connection
5098 count is more accurate (but it can never be perfect).
5100 11. The use of "drop" in a nested ACL was not being handled correctly in the
5101 outer ACL. Now, if condition failure induced by the nested "drop" causes
5102 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5103 or "require"), the connection is dropped.
5105 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5106 that yield "discard" can now be used with an "accept" or a "discard" verb,
5107 but an error is generated for any others (because I can't see a useful way
5108 to define what should happen).
5110 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5111 are now processed in the same way as lines in the Exim configuration file.
5112 In particular, continuation lines are supported.
5114 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5116 15. Added -ti meaning -t -i.
5118 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5119 domains, and warn by logging if others are found.
5121 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5122 returned from GIFCONF when Exim is trying to find the list of interfaces on
5123 a host. The code in os.c has been modified to copy each ifreq to an aligned
5124 structure in all cases.
5126 Also, in some cases, the returned ifreq's were being copied to a 'struct
5127 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5128 means the last couple of bytes of an IPv6 address could be chopped if the
5129 ifreq contained only a normal sockaddr (14 bytes storage).
5131 18. Named domain lists were not supported in the hosts_treat_as_local option.
5132 An entry such as +xxxx was not recognized, and was treated as a literal
5135 19. Ensure that header lines added by a DATA ACL are included in the reject log
5136 if the ACL subsequently rejects the message.
5138 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5139 MD5 (which is deprecated).
5141 21. When testing a filter file using -bf, Exim was writing a message when it
5142 took the sender from a "From " line in the message, but it was not doing so
5143 when it took $return_path from a Return-Path: header line. It now does.
5145 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5146 with a valid header line field name (a series of printing characters
5147 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5149 23. Changed "disc" in the source to "disk" to conform to the documentation and
5150 the book and for uniformity.
5152 24. Ignore Sendmail's -Ooption=value command line item.
5154 25. When execve() failed while trying to run a command in a pipe transport,
5155 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5156 could be confused with a return value of 69 from the command itself. This
5157 has been changed to 127, the value the shell returns if it is asked to run
5158 a non-existent command. The wording for the related log line suggests a
5159 non-existent command as the problem.
5161 26. If received_header_text expands to an empty string, do not add a Received:
5162 header line to the message. (Well, it adds a token one on the spool, but
5163 marks it "old" so that it doesn't get used or transmitted.)
5165 27. Installed eximstats 1.28 (addition of -nt option).
5167 28. There was no check for failure on the call to getsockname() in the daemon
5168 code. This can fail if there is a shortage of resources on the system, with
5169 ENOMEM, for example. A temporary error is now given on failure.
5171 29. Contrary to the C standard, it seems that in some environments, the
5172 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5173 program. Exim now does this explicitly; it affects the formatting of
5174 timestamps using strftime().
5176 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5177 complaints. I've now initialized all the variables, to avoid this.
5179 32. Header lines added by a system filter were not being "seen" during
5180 transport-time rewrites.
5182 33. The info_callback() function passed to OpenSSL is set up with type void
5183 (*)(SSL *, int, int), as described somewhere. However, when calling the
5184 function (actually a macro) that sets it up, the type void(*)() is
5185 expected. I've put in a cast to prevent warnings from picky compilers.
5187 34. If a DNS black list lookup found a CNAME record, but there were no A
5188 records associated with the domain it pointed at, Exim crashed.
5190 35. If a DNS black list lookup returned more than one A record, Exim ignored
5191 all but the first. It now scans all returned addresses if a particular IP
5192 value is being sought. In this situation, the contents of the
5193 $dnslist_value variable are a list of all the addresses, separated by a
5196 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5197 to accept a host name and all its aliases if the forward lookup for any of
5198 them yielded the IP address of the incoming connection. Now it accepts only
5199 those names whose forward lookup yields the correct IP address. Any other
5200 names are discarded. This closes a loophole whereby a rogue DNS
5201 administrator could create reverse DNS records to break through a
5202 wildcarded host restriction in an ACL.
5204 37. If a user filter or a system filter that ran in a subprocess used any of
5205 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5206 the wrong values were passed to the pipe command ($thisaddress had the
5207 value of $0, $0 had the value of $1, etc). This bug was introduced by
5208 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5210 38. Improved the line breaking for long SMTP error messages from ACLs.
5211 Previously, if there was no break point between 40 and 75 characters, Exim
5212 left the rest of the message alone. Two changes have been made: (a) I've
5213 reduced the minimum length to 35 characters; (b) if it can't find a break
5214 point between 35 and 75 characters, it looks ahead and uses the first one
5215 that it finds. This may give the occasional overlong line, but at least the
5216 remaining text gets split now.
5218 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5219 file descriptors might be low, and that setting 1000 would always raise it.
5220 It turns out that in some environments, the limit is already over 1000 and
5221 that lowering it causes trouble. So now Exim takes care not to decrease it.
5223 40. When delivering a message, the value of $return_path is set to $sender_
5224 address at the start of routing (routers may change the value). By an
5225 oversight, this default was not being set up when an address was tested by
5226 -bt or -bv, which affected the outcome if any router or filter referred to
5229 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5230 log only when "message" or "log_message" are set. However, if one of the
5231 conditions was an address verification, or a call to a nested ACL, the
5232 messages generated by the underlying test were being passed through. This
5233 no longer happens. The underlying message is available in $acl_verify_
5234 message for both "message" and "log_message" expansions, so it can be
5235 passed through if needed.
5237 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5238 new expansion $bh_ to give the encoded byte string without charset
5239 translation. Translation happens only if iconv() is available; HAVE_ICONV
5240 indicates this at build time. HEADERS_CHARSET gives the charset to
5241 translate to; headers_charset can change it in the configuration, and
5242 "headers charset" can change it in an individual filter file.
5244 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5245 that creates RFC 2047 encoded "words" labels them as that charset instead
5246 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5247 expansion operator; (ii) when Exim creates a From: line for a local
5248 message; (iii) when a header line is rewritten to include a "phrase" part.
5250 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5251 buggy, causing it to skip the first lines of messages whose message ID
5252 ended in 'D'. This would not have bitten before Exim release 4.14, because
5253 message IDs were unlikely to end in 'D' before then. The effect was to have
5254 incorrect size information for certain domains.
5256 45. #include "config.h" was missing at the start of the crypt16.c module. This
5257 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5260 46. If there was a timeout during a "random" callout check, Exim treated it as
5261 a failure of the random address, and carried on sending RSET and the real
5262 address. If the delay was just some slowness somewhere, the response to the
5263 original RCPT would be taken as a response to RSET and so on, causing
5264 mayhem of various kinds.
5266 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5267 when I implemented it. It didn't allow for the fact that some option values
5268 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5269 the right test for positive values.
5271 48. Domain names in DNS records are case-independent. Exim always looks them up
5272 in lower case. Some resolvers return domain names in exactly the case they
5273 appear in the zone file, that is, they may contain uppercase letters. Not
5274 all resolvers do this - some return always lower case. Exim was treating a
5275 change of case by a resolver as a change of domain, similar to a widening
5276 of a domain abbreviation. This triggered its re-routing code and so it was
5277 trying to route what was effectively the same domain again. This normally
5278 caused routing to fail (because the router wouldn't handle the domain
5279 twice). Now Exim checks for this case specially, and just changes the
5280 casing of the domain that it ultimately uses when it transmits the message
5283 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5286 50. If a filter generated a file delivery with a non-absolute name (possible if
5287 no home directory exists for the router), the forbid_file option was not
5290 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5291 the existing equality matching.
5293 52. Exim was using ints instead of ino_t variables in some places where it was
5294 dealing with inode numbers.
5296 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5297 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5298 called TMPDIR, and if it finds it is different, it changes its value.
5300 54. The smtp_printf() function is now made available to local_scan() so
5301 additional output lines can be written before returning. There is also an
5302 smtp_fflush() function to enable the detection of a dropped connection.
5303 The variables smtp_input and smtp_batched_input are exported to
5306 55. Changed the default runtime configuration: the message "Unknown user"
5307 has been removed from the ACL, and instead placed on the localuser router,
5308 using the cannot_route_message feature. This means that any verification
5309 failures that generate their own messages won't get overridden. Similarly,
5310 the "Unrouteable address" message that was in the ACL for unverifiable
5311 relay addresses has also been removed.
5313 56. Added hosts_avoid_esmtp to the smtp transport.
5315 57. The exicyclog script was not checking for the esoteric option
5316 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5317 will work only if exicyclog is run under the appropriate euid.
5319 58. Following a discussion on the list, the rules by which Exim recognises line
5320 endings on incoming messages have been changed. The -dropcr and drop_cr
5321 options are now no-ops, retained only for backwards compatibility. The
5322 following line terminators are recognized: LF CRLF CR. However, special
5323 processing applies to CR:
5325 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5326 nor a local message in the state where . is a terminator.
5328 (ii) If a bare CR is encountered in a header line, an extra space is added
5329 after the line terminator so as not to end the header. The reasoning
5330 behind this is that bare CRs in header lines are most likely either
5331 to be mistakes, or people trying to play silly games.
5333 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5334 was being incorrectly given as 18 bytes larger than it should have been.
5335 This is a VOB (very old bug).
5337 60. This may never have affected anything current, but just in case it has:
5338 When the local host is found other than at the start of a list of hosts,
5339 the local host, those with the same MX, and any that follow, are discarded.
5340 When the list in question was part of a longer list of hosts, the following
5341 hosts (not currently being processed) were also being discarded. This no
5342 longer happens. I'm not sure if this situation could ever has previously
5345 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5348 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5349 header. This is something that is recommended in a new Internet Draft, and
5350 is something that is documented as being done by Sendmail. There are two
5351 possible values. For messages generated by the autoreply transport, Exim
5354 Auto-Submitted: auto-replied
5356 whereas for all other generated messages (e.g. bounces) it adds
5358 Auto-Submitted: auto-generated
5360 63. The "personal" condition in filters now includes a test for the
5361 Auto-Submitted: header. If it contains the string "auto-" the message it
5362 not considered personal.
5364 64. Added rcpt_include_affixes as a generic transport option.
5366 65. Added queue_only_override (default true).
5368 66. Added the syslog_duplication option.
5370 67. If what should have been the first header line of a message consisted of
5371 a space followed by a colon, Exim was mis-interpreting it as a header line.
5372 It isn't of course - it is syntactically invalid and should therefore be
5373 treated as the start of the message body. The misbehaviour could have
5374 caused a number of strange effects, including loss of data in subsequent
5375 header lines, and spool format errors.
5377 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5378 client host had authenticated. This control can now be exercised by an ACL
5379 for more flexibility.
5381 69. By default, callouts do not happen when testing with -bh. There is now a
5382 variant, -bhc, which does actually run the callout code, including
5383 consulting and updating the callout cache.
5385 70. Added support for saslauthd authentication, courtesy of Alexander
5388 71. If statvfs() failed on the spool or log directories while checking their
5389 size for availability, Exim confusingly gave the error "space shortage".
5390 Furthermore, in debugging mode it crashed with a floating point exception.
5391 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5392 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5393 serious problem, Exim now writes to the main and panic logs when this
5394 happens, with details of the failure. It then refuses to accept the
5395 incoming message, giving the message "spool directory problem" or "log
5396 directory problem" with a 421 code for SMTP messages.
5398 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5399 0, 1, and 2 exist, because some OS complain for execs without them (see
5400 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5401 descriptors. However, the code omitted to check that the open succeeded,
5402 causing mysterious errors if for some reason the permissions on /dev/null
5403 got screwed. Now Exim writes a message to the main and panic logs, and
5404 bombs out if it can't open /dev/null.
5406 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5407 interact so that it is all more flexible. It is supposed to remain
5408 backwards compatible. Also added extra_local_interfaces.
5410 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5411 to bomb out with an assertion failure - to the client this appears as a
5412 connection drop. This problem occurs in the part of the code that was taken
5413 from the Samba project. Fortunately, the assertion is in a very simple
5414 function, so I have fixed this by reproducing the function inline in the
5415 one place where it is called, and arranging for authentication to fail
5416 instead of killing the process with assert().
5418 75. The SPA client code was not working when the server requested OEM rather
5419 than Unicode encoding.
5421 76. Added code to make require_files with a specific uid setting more usable in
5422 the case where statting the file as root fails - usually a non-root-mounted
5423 NFS file system. When this happens and the failure is EACCES, Exim now
5424 forks a subprocess and does the per-uid checking as the relevant uid.
5426 77. Added process_log_path.
5428 78. If log_file_path was not explicitly set, a setting of check_log_space or
5429 check_log_inodes was ignored.
5431 79. If a space check for the spool or log partitions fails, the incident is now
5432 logged. Of course, in the latter case the data may get lost...
5434 80. Added the %p formatting code to string_format() so that it can be used to
5435 print addresses in debug_print(). Adjusted all the address printing in the
5436 debugging in store.c to use %p rather than %d.
5438 81. There was a concern that a line of code in smtp_in.c could overflow a
5439 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5440 initially expressed, the concern was not well-founded, because trailing
5441 spaces are removed early. However, if the trailing spaces were followed by
5442 a NULL, they did not get removed, so the overflow was possible. Two fixes
5445 (a) I re-wrote the offending code in a cleaner fashion.
5446 (b) If an incoming SMTP command contains a NULL character, it is rejected
5449 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5450 runs initgroups(), so that if the Exim user is in any additional groups,
5451 they will be used during message reception.
5457 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob