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.
92 Bugzilla 879. 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 Bugzilla 1078. Patch from John Horne.
103 PP/06 Stop make process more reliably on build failure.
104 Bugzilla 1087. Patch from Heiko Schlittermann.
106 PP/07 Make maildir_use_size_file an _expandable_ boolean.
107 Bugzilla 1089. Patch from Heiko Schlittermann.
109 PP/08 Handle ${run} returning more data than OS pipe buffer size.
110 Bugzilla 1131. Patch from Holger Weiß.
112 PP/09 Handle IPv6 addresses with SPF.
113 Bugzilla 860. Patch from Wolfgang Breyha.
115 PP/10 GnuTLS: support TLS 1.2 & 1.1.
117 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
120 PP/11 match_* no longer expand right-hand-side by default.
121 New compile-time build option, EXPAND_LISTMATCH_RHS.
122 New expansion conditions, "inlist", "inlisti".
124 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
126 PP/13 shell and compiler warnings fixes for RC1-RC3 changes.
128 PP/14 fix log_write() format string regression from TF/03.
129 Bugzilla 1152. Patch from Dmitry Isaikin.
135 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
137 PP/02 Harmonised TLS library version reporting; only show if debugging.
138 Layout now matches that introduced for other libraries in 4.74 PP/03.
140 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
142 PP/04 New "dns_use_edns0" global option.
144 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
147 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
148 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
150 TK/01 Updated PolarSSL code to 0.14.2.
151 Bugzilla 1097. Patch from Andreas Metzler.
153 PP/07 Catch divide-by-zero in ${eval:...}.
156 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
159 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
160 format-string attack -- SECURITY: remote arbitrary code execution.
162 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
163 time unintentionally subject to list matching rules, letting the header
164 cause arbitrary Exim lookups (of items which can occur in lists, *not*
165 arbitrary string expansion). This allowed for information disclosure.
167 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
168 INT_MIN/-1 -- value coerced to INT_MAX.
174 NM/01 Workround for PCRE version dependency in version reporting
177 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
178 This fixes portability to compilers other than gcc, notably
179 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
181 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
182 makefiles for portability to HP-UX and POSIX correctness.
184 PP/01 Permit LOOKUP_foo enabling on the make command-line.
185 Also via indented variable definition in the Makefile.
186 (Debugging by Oliver Heesakkers).
188 PP/02 Restore caching of spamd results with expanded spamd_address.
189 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
191 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
192 Improves build reliability. Fix from: Frank Elsner
194 NM/02 Fix wide character breakage in the rfc2047 coding
195 Fixes bug 1064. Patch from Andrey N. Oktyabrski
197 NM/03 Allow underscore in dnslist lookups
198 Fixes bug 1026. Patch from Graeme Fowler
200 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
201 Code patches from Adam Ciarcinski of NetBSD.
203 NM/04 Fixed exiqgrep to cope with mailq missing size issue
206 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
207 is logged, to avoid truncation. Patch from John Horne.
209 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
210 Patch from Jakob Hirsch.
212 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
213 SQL string expansion failure details.
214 Patch from Andrey Oktyabrski.
216 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
217 Patch from Simon Arlott.
219 PP/09 New lookups functionality failed to compile on old gcc which rejects
220 extern declarations in function scope.
221 Patch from Oliver Fleischmann
223 PP/10 Use sig_atomic_t for flags set from signal handlers.
224 Check getgroups() return and improve debugging.
225 Fixed developed for diagnosis in bug 927 (which turned out to be
228 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
229 Patch from Mark Zealey.
231 PP/12 Bugzilla 1056: Improved spamd server selection.
232 Patch from Mark Zealey.
234 PP/13 Bugzilla 1086: Deal with maildir quota file races.
235 Based on patch from Heiko Schlittermann.
237 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
238 Patch from Uwe Doering, sign-off by Michael Haardt.
240 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
241 variable declaration deep within a block. Bug and patch from
244 PP/15 lookups-Makefile IRIX compatibilty coercion.
246 PP/16 Make DISABLE_DKIM build knob functional.
248 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
249 Patch by Simon Arlott
251 TF/03 Fix valgrind.h portability to C89 compilers that do not support
252 variable argument macros. Our copy now differs from upstream.
258 TF/01 Failure to get a lock on a hints database can have serious
259 consequences so log it to the panic log.
261 TF/02 Log LMTP confirmation messages in the same way as SMTP,
262 controlled using the smtp_confirmation log selector.
264 TF/03 Include the error message when we fail to unlink a spool file.
266 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
267 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
268 for maintaining out-of-tree patches for some time.
270 PP/01 Bugzilla 139: Documentation and portability issues.
271 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
272 Handle per-OS dynamic-module compilation flags.
274 PP/02 Let /dev/null have normal permissions.
275 The 4.73 fixes were a little too stringent and complained about the
276 permissions on /dev/null. Exempt it from some checks.
277 Reported by Andreas M. Kirchwitz.
279 PP/03 Report version information for many libraries, including
280 Exim version information for dynamically loaded libraries. Created
281 version.h, now support a version extension string for distributors
282 who patch heavily. Dynamic module ABI change.
284 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
285 privilege escalation vulnerability whereby the Exim run-time user
286 can cause root to append content of the attacker's choosing to
289 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
292 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
293 If dropping privileges for untrusted macros, we disabled normal logging
294 on the basis that it would fail; for the Exim run-time user, this is not
295 the case, and it resulted in successful deliveries going unlogged.
296 Fixed. Reported by Andreas Metzler.
302 PP/01 Date: & Message-Id: revert to normally being appended to a message,
303 only prepend for the Resent-* case. Fixes regression introduced in
304 Exim 4.70 by NM/22 for Bugzilla 607.
306 PP/02 Include check_rfc2047_length in configure.default because we're seeing
307 increasing numbers of administrators be bitten by this.
309 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
311 PP/03 Bugzilla 994: added openssl_options main configuration option.
313 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
315 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
317 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
319 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
320 without a peer certificate, leading to a segfault because of an
321 assumption that peers always have certificates. Be a little more
322 paranoid. Problem reported by Martin Tscholak.
324 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
325 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
326 NB: ClamAV planning to remove STREAM in "middle of 2010".
327 CL also introduces -bmalware, various -d+acl logging additions and
328 more caution in buffer sizes.
330 PP/09 Implemented reverse_ip expansion operator.
332 PP/10 Bugzilla 937: provide a "debug" ACL control.
334 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
336 PP/12 Bugzilla 973: Implement --version.
338 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
340 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
342 PP/15 Bugzilla 816: support multiple condition rules on Routers.
344 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
345 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
346 ignore trailing whitespace.
348 JJ/02 prevent non-panic DKIM error from being sent to paniclog
350 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
353 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
354 Notification from Dr Andrew Aitchison.
356 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
357 ExtendedDetectionInfo response format.
358 Notification from John Horne.
360 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
363 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
364 XSL and documented dependency on system catalogs, with examples of how
367 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
370 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
371 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
372 configuration file which is writeable by the Exim user or group.
374 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
375 of configuration files to cover files specified with the -C option if
376 they are going to be used with root privileges, not just the default
379 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
380 option (effectively making it always true).
382 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
383 files to be used while preserving root privileges.
385 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
386 that rogue child processes cannot use them.
388 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
389 run-time user, instead of root.
391 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
392 Exim run-time user without dropping privileges.
394 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
395 result string, instead of calling string_vformat() twice with the same
398 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
399 for other users. Others should always drop root privileges if they use
400 -C on the command line, even for a whitelisted configure file.
402 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
404 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
410 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
411 $data_path, and $header_path variables; fixed documentation bugs and
414 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
415 exipick to access non-standard spools, including the "frozen" queue
418 NM/01 Bugzilla 965: Support mysql stored procedures.
419 Patch from Alain Williams
421 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
423 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
424 Patch from Andreas Metzler
426 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
427 Patch from Kirill Miazine
429 NM/05 Bugzilla 671: Added umask to procmail example.
431 JJ/03 installed exipick 20100323.0, fixing doc bug
433 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
434 directory. Notification and patch from Dan Rosenberg.
436 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
438 TK/02 Improve log output when DKIM signing operation fails.
440 MH/01 Treat the transport option dkim_domain as a colon separated
441 list, not as a single string, and sign the message with each element,
442 omitting multiple occurences of the same signer.
444 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
445 Bugzilla 985, 986. Patch by Simon Arlott
447 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
448 Patch by Simon Arlott
450 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
451 MBX locking. Notification from Dan Rosenberg.
457 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
459 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
461 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
463 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
465 NM/04 Bugzilla 915: Flag broken perl installation during build.
471 TK/01 Added patch by Johannes Berg that expands the main option
472 "spamd_address" if it starts with a dollar sign.
474 TK/02 Write list of recipients to X-Envelope-Sender header when building
475 the mbox-format spool file for content scanning (suggested by Jakob
478 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
479 (http://www.dcc-servers.net/) support via dccifd. Activated by
480 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
482 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
483 by Mark Daniel Reidel <mr@df.eu>.
485 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
486 When building exim an external PCRE library is now needed -
487 PCRE is a system library on the majority of modern systems.
488 See entry on PCRE_LIBS in EDITME file.
490 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
491 conversation. Added nologin parameter to request.
492 Patch contributed by Kirill Miazine.
494 TF/01 Do not log submission mode rewrites if they do not change the address.
496 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
498 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
499 log files in place. Contributed by Roberto Lima.
501 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
503 TF/03 Bugzilla 615: When checking the local_parts router precondition
504 after a local_part_suffix or local_part_prefix option, Exim now
505 does not use the address's named list lookup cache, since this
506 contains cached lookups for the whole local part.
508 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
509 Robert Millan. Documentation is in experimental-spec.txt.
511 TF/04 Bugzilla 668: Fix parallel build (make -j).
513 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
515 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
516 Patch provided by Jan Srzednicki.
518 TF/05 Leading white space used to be stripped from $spam_report which
519 wrecked the formatting. Now it is preserved.
521 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
522 that they are available at delivery time.
524 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
526 TF/08 TLS error reporting now respects the incoming_interface and
527 incoming_port log selectors.
529 TF/09 Produce a more useful error message if an SMTP transport's hosts
530 setting expands to an empty string.
532 NM/06 Bugzilla 744: EXPN did not work under TLS.
533 Patch provided by Phil Pennock.
535 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
536 Patch provided by Richard Godbee.
538 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
539 acl_smtp_notquit, added index entry.
541 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
542 Patch provided by Eugene Bujak.
544 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
545 accept(). Patch provided by Maxim Dounin.
547 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
548 Patch provided by Phil Pennock.
550 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
552 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
553 Patch provided by Brad "anomie" Jorsch.
555 NM/14 Bugzilla 622: Added timeout setting to transport filter.
556 Patch provided by Dean Brooks.
558 TK/05 Add native DKIM support (does not depend on external libraries).
560 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
561 Patch provided by Graeme Fowler.
563 NM/16 Bugzilla 851: Documentation example syntax fix.
565 NM/17 Changed NOTICE file to remove references to embedded PCRE.
567 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
570 NM/19 Bugzilla 745: TLS version reporting.
571 Patch provided by Phil Pennock.
573 NM/20 Bugzilla 167: bool: condition support.
574 Patch provided by Phil Pennock.
576 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
577 clients. Patch provided by Phil Pennock.
579 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
580 Patch provided by Brad "anomie" Jorsch.
582 NM/23 Bugzilla 687: Fix misparses in eximstats.
583 Patch provided by Heiko Schlittermann.
585 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
586 Patch provided by Heiko Schlittermann.
588 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
589 plus update to original patch.
591 NM/26 Bugzilla 799: Documentation correction for ratelimit.
593 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
594 Patch provided by David Brownlee.
596 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
598 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
600 NM/30 Bugzilla 888: TLS documentation bugfixes.
602 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
604 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
605 Unlike the original bugzilla I have changed all shell scripts in src tree.
607 NM/33 Bugzilla 898: Transport filter timeout fix.
608 Patch by Todd Rinaldo.
610 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
611 Patch by Serge Demonchaux.
613 NM/35 Bugzilla 39: Base64 decode bug fixes.
614 Patch by Jakob Hirsch.
616 NM/36 Bugzilla 909: Correct connect() call in dcc code.
618 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
620 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
622 NM/39 Bugzilla 911: Fixed MakeLinks build script.
628 TK/01 Add preliminary DKIM support. Currently requires a forked version of
629 ALT-N's libdkim that I have put here:
630 http://duncanthrax.net/exim-experimental/
632 Note to Michael Haardt: I had to rename some vars in sieve.c. They
633 were called 'true' and it seems that C99 defines that as a reserved
634 keyword to be used with 'bool' variable types. That means you could
635 not include C99-style headers which use bools without triggering
636 build errors in sieve.c.
638 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
639 as mailq or other aliases. Changed the --help handling significantly
640 to do whats expected. exim_usage() emits usage/help information.
642 SC/01 Added the -bylocaldomain option to eximstats.
644 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
646 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
648 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
654 PH/01 Another patch from the Sieve maintainer.
656 PH/02 When an IPv6 address is converted to a string for single-key lookup
657 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
658 dots are used instead of colons so that keys in lsearch files need not
659 contain colons. This was done some time before quoting was made available
660 in lsearch files. However, iplsearch files do require colons in IPv6 keys
661 (notated using the quote facility) so as to distinguish them from IPv4
662 keys. This meant that lookups for IP addresses in host lists did not work
663 for iplsearch lookups.
665 This has been fixed by arranging for IPv6 addresses to be expressed with
666 colons if the lookup type is iplsearch. This is not incompatible, because
667 previously such lookups could never work.
669 The situation is now rather anomolous, since one *can* have colons in
670 ordinary lsearch keys. However, making the change in all cases is
671 incompatible and would probably break a number of configurations.
673 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
676 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
677 conversion specification without a maximum field width, thereby enabling
678 a rogue spamd server to cause a buffer overflow. While nobody in their
679 right mind would setup Exim to query an untrusted spamd server, an
680 attacker that gains access to a server running spamd could potentially
681 exploit this vulnerability to run arbitrary code as the Exim user.
683 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
684 $primary_hostname instead of what libspf2 thinks the hosts name is.
686 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
687 a directory entry by the name of the lookup key. Previously, if a
688 symlink pointed to a non-existing file or a file in a directory that
689 Exim lacked permissions to read, a lookup for a key matching that
690 symlink would fail. Now it is enough that a matching directory entry
691 exists, symlink or not. (Bugzilla 503.)
693 PH/03 The body_linecount and body_zerocount variables are now exported in the
696 PH/04 Added the $dnslist_matched variable.
698 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
699 This means they are set thereafter only if the connection becomes
702 PH/06 Added the client_condition to authenticators so that some can be skipped
703 by clients under certain conditions.
705 PH/07 The error message for a badly-placed control=no_multiline_responses left
706 "_responses" off the end of the name.
708 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
710 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
711 (without spaces) instead of just copying the configuration text.
713 PH/10 Added the /noupdate option to the ratelimit ACL condition.
715 PH/11 Added $max_received_linelength.
717 PH/12 Added +ignore_defer and +include_defer to host lists.
719 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
720 way in which PCRE > 7.0 is built.
722 PH/14 Implemented queue_only_load_latch.
724 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
725 MAIL command. The effect was to mangle the value on 64-bit systems.
727 PH/16 Another patch from the Sieve maintainer.
729 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
731 PH/18 If a system quota error occurred while trying to create the file for
732 a maildir delivery, the message "Mailbox is full" was not appended to the
733 bounce if the delivery eventually timed out. Change 4.67/27 below applied
734 only to a quota excession during the actual writing of the file.
736 PH/19 It seems that peer DN values may contain newlines (and other non-printing
737 characters?) which causes problems in log lines. The DN values are now
738 passed through string_printing() before being added to log lines.
740 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
741 and InterBase are left for another time.)
743 PH/21 Added message_body_newlines option.
745 PH/22 Guard against possible overflow in moan_check_errorcopy().
747 PH/23 POSIX allows open() to be a macro; guard against that.
749 PH/24 If the recipient of an error message contained an @ in the local part
750 (suitably quoted, of course), incorrect values were put in $domain and
751 $local_part during the evaluation of errors_copy.
757 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
758 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
761 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
762 issue a MAIL command.
764 PH/02 In an ACL statement such as
766 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
768 if a client was not listed at all, or was listed with a value other than
769 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
770 the condition was not true (as it should be), so access was not denied.
771 The bug was that the ! inversion was incorrectly passed on to the second
772 item. This has been fixed.
774 PH/03 Added additional dnslists conditions == and =& which are different from
775 = and & when the dns lookup returns more than one IP address.
777 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
778 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
780 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
781 FSYNC, which compiles an option called disable_fsync that allows for
782 bypassing fsync(). The documentation is heavily laced with warnings.
784 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
786 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
787 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
788 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
789 including adding "make clean"; (3) Added -fPIC when compiling the test
790 dynamically loaded module, to get rid of a warning.
792 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
793 message fails, move_frozen_messages = true and ignore_bounce_errors_after
794 = 0s. The bug is otherwise harmless.
796 PH/07 There was a bug in the dovecot authenticator such that the value of
797 $auth1 could be overwritten, and so not correctly preserved, after a
798 successful authentication. This usually meant that the value preserved by
799 the server_setid option was incorrect.
801 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
803 PH/09 Installed PCRE release 7.0.
805 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
806 run for batched SMTP input. It is now run at the start of every message
807 in the batch. While fixing this I discovered that the process information
808 (output by running exiwhat) was not always getting set for -bs and -bS
809 input. This is fixed, and it now also says "batched" for BSMTP.
811 PH/11 Added control=no_pipelining.
813 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
814 patch, slightly modified), and move the expansion of helo_data till after
815 the connection is made in the smtp transport (so it can use these
818 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
820 PH/14 Added log_selector = +pid.
822 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
824 PH/16 Add ${if forany and ${if forall.
826 PH/17 Added dsn_from option to vary the From: line in DSNs.
828 PH/18 Flush SMTP output before performing a callout, unless control =
829 no_callout_flush is set.
831 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
832 was true (the default) a successful delivery failed to delete the retry
833 item, thus causing premature timeout of the address. The bug is now
836 PH/20 Added hosts_avoid_pipelining to the smtp transport.
838 PH/21 Long custom messages for fakedefer and fakereject are now split up
839 into multiline reponses in the same way that messages for "deny" and
840 other ACL rejections are.
842 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
843 with slight modification.
845 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
846 draft, changing the syntax and factoring some duplicate code".
848 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
849 for deliveries of the second and subsequent messages over the same SMTP
852 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
853 ${reduce, with only minor "tidies".
855 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
857 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
858 expansion side effects.
860 PH/27 When a message times out after an over-quota error from an Exim-imposed
861 quota, the bounce message says "mailbox is full". This message was not
862 being given when it was a system quota that was exceeded. It now should
865 MH/03 Made $recipients available in local_scan(). local_scan() already has
866 better access to the recipient list through recipients_list[], but
867 $recipients can be useful in postmaster-provided expansion strings.
869 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
870 in the case of a MAIL command with additional options following the
871 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
872 were accidentally chopped off.
874 PH/29 SMTP synchronization checks are implemented when a command is read -
875 there is a check that no more input is waiting when there shouldn't be
876 any. However, for some commands, a delay in an ACL can mean that it is
877 some time before the response is written. In this time, more input might
878 arrive, invalidly. So now there are extra checks after an ACL has run for
879 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
880 pipelining has not been advertised.
882 PH/30 MH's patch to allow iscntrl() characters to be list separators.
884 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
885 returned in the SMTP response when smtp_return_error_details was false.
888 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
889 instead of the C I/O that was originally supplied, because problems were
892 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
893 Exim which did not show up earlier: it was assuming that a call to
894 SSL_CTX_set_info_callback() might give an error value. In fact, there is
895 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
896 was a macro that became an assignment, so it seemed to work. This has
897 changed to a proper function call with a void return, hence the compile
898 error. Exim's code has been fixed.
900 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
903 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
905 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
907 PH/37 If a message is not accepted after it has had an id assigned (e.g.
908 because it turns out to be too big or there is a timeout) there is no
909 "Completed" line in the log. When some messages of this type were
910 selected by exigrep, they were listed as "not completed". Others were
911 picked up by some special patterns. I have improved the selection
912 criteria to be more general.
914 PH/38 The host_find_failed option in the manualroute router can now be set
915 to "ignore", to completely ignore a host whose IP address cannot be
916 found. If all hosts are ignored, the behaviour is controlled by the new
917 host_all_ignored option.
919 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
920 homing or because of multiple MX records with /mx) generated more than
921 one IP address, and the following item turned out to be the local host,
922 all the secondary addresses of the first item were incorrectly removed
923 from the list, along with the local host and any following hosts (which
924 is what is supposed to happen).
926 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
927 whoever called Exim into the -H file. In the case of the daemon it was
928 behaving confusingly. When first started, it used values for whoever
929 started the daemon, but after a SIGHUP it used the Exim user (because it
930 calls itself on a restart). I have changed the code so that it now always
933 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
934 message are rejected with the same error (e.g. no authentication or bad
935 sender address), and a DATA command is nevertheless sent (as can happen
936 with PIPELINING or a stupid MUA), the error message that was given to the
937 RCPT commands is included in the rejection of the DATA command. This is
938 intended to be helpful for MUAs that show only the final error to their
941 PH/42 Another patch from the Sieve maintainer.
943 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
944 Eximstats - Fixed some broken HTML links and added missing column headers
946 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
947 columns for Rejects, Temp Rejects, Ham, and Spam rows.
949 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
951 PH/43 Yet another patch from the Sieve maintainer.
953 PH/44 I found a way to check for a TCP/IP connection going away before sending
954 the response to the final '.' that terminates a message, but only in the
955 case where the client has not sent further data following the '.'
956 (unfortunately, this is allowed). However, in many cases there won't be
957 any further data because there won't be any more messages to send. A call
958 to select() can be used: if it shows that the input is "ready", there is
959 either input waiting, or the socket has been closed. An attempt to read
960 the next input character can distinguish the two cases. Previously, Exim
961 would have sent an OK response which the client would never have see.
962 This could lead to message repetition. This fix should cure that, at
963 least in a lot of common cases.
965 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
966 advertised in response to EHLO.
972 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
973 fixed by 4.65/MH/01 (is this a record?) are fixed:
975 (i) An empty string was always treated as zero by the numeric comparison
976 operators. This behaviour has been restored.
978 (ii) It is documented that the numeric comparison operators always treat
979 their arguments as decimal numbers. This was broken in that numbers
980 starting with 0 were being interpreted as octal.
982 While fixing these problems I realized that there was another issue that
983 hadn't been noticed. Values of message_size_limit (both the global option
984 and the transport option) were treated as octal if they started with 0.
985 The documentation was vague. These values are now always treated as
986 decimal, and I will make that clear in the documentation.
992 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
993 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
996 MH/01 Don't check that the operands of numeric comparison operators are
997 integers when their expansion is in "skipping" mode (fixes bug
998 introduced by 4.64-PH/07).
1000 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1001 child addresses, Exim now panics and dies. Previously, because the count
1002 is held in a short int, deliveries were likely to be lost. As such a
1003 large number of recipients for a single message is ridiculous
1004 (performance will be very, very poor), I have chosen to impose a limit
1005 rather than extend the field.
1011 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1012 leftover -K file (the existence of which was triggered by #402).
1013 While we were at it, introduced process PID as part of the -K
1014 filename. This should rule out race conditions when creating
1017 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1018 processing considerably. Previous code took too long for large mails,
1019 triggering a timeout which in turn triggers #401.
1021 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1022 in the DK code in transports.c. sendfile() is not really portable,
1023 hence the _LINUX specificness.
1025 TF/01 In the add_headers option to the mail command in an Exim filter,
1026 there was a bug that Exim would claim a syntax error in any
1027 header after the first one which had an odd number of characters
1030 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1031 callout verification, Exim cached a "reject" for the entire domain. This
1032 is correct for most verifications, but it is not correct for a recipient
1033 verification with use_sender or use_postmaster set, because in that case
1034 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1035 case of MAIL FROM:<> rejection from other early rejections (e.g.
1036 rejection of HELO). When verifying a recipient using a non-null MAIL
1037 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1038 Whatever the result of the callout, the value of the domain cache is
1039 left unchanged (for any other kind of callout, getting as far as trying
1040 RCPT means that the domain itself is ok).
1042 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1045 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1046 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1047 session, the daemon ignores ECONNECT errors and logs others; it now
1048 ignores EPIPE as well.
1050 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1051 (quoted-printable decoding).
1053 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1054 later the small subsequent patch to fix an introduced bug.
1056 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1058 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1060 PH/08 An error is now given if message_size_limit is specified negative.
1062 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1063 to be given (somewhat) arbitrary names.
1065 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1068 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1069 miscellaneous code fixes
1071 PH/10 Added the log_reject_target ACL modifier to specify where to log
1074 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1075 hostname. This is wrong, because it relates to the incoming message (and
1076 probably the interface on which it is arriving) and not to the outgoing
1077 callout (which could be using a different interface). This has been
1078 changed to use the value of the helo_data option from the smtp transport
1079 instead - this is what is used when a message is actually being sent. If
1080 there is no remote transport (possible with a router that sets up host
1081 addresses), $smtp_active_hostname is used.
1083 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1084 tweaks were necessary in order to get it to work (see also 21 below):
1085 (a) The code assumed that strncpy() returns a negative number on buffer
1086 overflow, which isn't the case. Replaced with Exim's string_format()
1088 (b) There were several signed/unsigned issues. I just did the minimum
1089 hacking in of casts. There is scope for a larger refactoring.
1090 (c) The code used strcasecmp() which is not a standard C function.
1091 Replaced with Exim's strcmpic() function.
1092 (d) The code set only $1; it now sets $auth1 as well.
1093 (e) A simple test gave the error "authentication client didn't specify
1094 service in request". It would seem that Dovecot has changed its
1095 interface. Fortunately there's a specification; I followed it and
1096 changed what the client sends and it appears to be working now.
1098 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1101 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1102 address A is aliased to B and C, where B exists and C does not. Without
1103 -v the output is "A verified" because verification stops after a
1104 successful redirection if more than one address is generated. However,
1105 with -v the child addresses are also verified. Exim was outputting "A
1106 failed to verify" and then showing the successful verification for C,
1107 with its parentage. It now outputs "B failed to verify", showing B's
1108 parentage before showing the successful verification of C.
1110 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1111 look up a TXT record in a specific list after matching in a combined
1114 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1115 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1116 they consult the DNS. I had assumed they would set it the way they
1117 wanted; and indeed my experiments on Linux seem to show that in some
1118 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1119 To be on the safe side, however, I have now made the interface to
1120 host_find_byname() similar to host_find_bydns(), with an argument
1121 containing the DNS resolver options. The host_find_byname() function now
1122 sets these options at its start, just as host_find_bydns() does. The smtp
1123 transport options dns_qualify_single and dns_search_parents are passed to
1124 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1125 of host_find_byname() use the default settings of RES_DEFNAMES
1126 (qualify_single) but not RES_DNSRCH (search_parents).
1128 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1129 spool_read_header() do less string testing, by means of a preliminary
1130 switch on the second character of optional "-foo" lines. (This is
1131 overdue, caused by the large number of possibilities that now exist.
1132 Originally there were few.) While I was there, I also converted the
1133 str(n)cmp tests so they don't re-test the leading "-" and the first
1134 character, in the hope this might squeeze out yet more improvement.
1136 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1137 flag allowing group syntax was set by the header_syntax check but not
1138 turned off, possible causing trouble later; (2) The flag was not being
1139 set at all for the header_verify test, causing "group"-style headers to
1140 be rejected. I have now set it in this case, and also caused header_
1141 verify to ignore an empty address taken from a group. While doing this, I
1142 came across some other cases where the code for allowing group syntax
1143 while scanning a header line wasn't quite right (mostly, not resetting
1144 the flag correctly in the right place). These bugs could have caused
1145 trouble for malformed header lines. I hope it is now all correct.
1147 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1148 with the "reply" argument non-NULL. The code, however (which originally
1149 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1150 but it didn't always do it. This confused somebody who was copying the
1151 code for some other use. I have removed all the tests.
1153 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1154 feature that was used to support insecure browsers during the U.S. crypto
1155 embargo. It requires special client support, and Exim is probably the
1156 only MTA that supported it -- and would never use it because real RSA is
1157 always available. This code has been removed, because it had the bad
1158 effect of slowing Exim down by computing (never used) parameters for the
1159 RSA_EXPORT functionality.
1161 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1162 authenticator to fail if there's a tab character in the incoming data
1163 (there should never be unless someone is messing about, as it's supposed
1164 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1165 if the connection is using TLS or if the remote IP is the same as the
1166 local IP, and the "valid-client-cert option" if a client certificate has
1169 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1170 authenticators. This can be used for authorization after authentication
1171 succeeds. (In the case of plaintext, it servers for both authentication
1174 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1175 if any retry times were supplied.
1177 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1178 connection, where there is no client IP address to check. In this
1179 situation, the verify now always succeeds.
1181 PH/25 Applied John Jetmore's -Mset patch.
1183 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1185 PH/27 In a string expansion for a processed (not raw) header when multiple
1186 headers of the same name were present, leading whitespace was being
1187 removed from all of them, but trailing whitespace was being removed only
1188 from the last one. Now trailing whitespace is removed from each header
1189 before concatenation. Completely empty headers in a concatenation (as
1190 before) are ignored.
1192 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1193 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1195 PH/29 [Removed. This was a change that I later backed out, and forgot to
1196 correct the ChangeLog entry (that I had efficiently created) before
1197 committing the later change.]
1199 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1200 address errors (4xx response to RCPT) over the same connection as other
1201 messages routed to the same hosts. Such deliveries are always "forced",
1202 so retry times are not inspected. This resulted in far too many retries
1203 for the affected addresses. The effect occurred only when there were more
1204 hosts than the hosts_max_try setting in the smtp transport when it had
1205 the 4xx errors. Those hosts that it had tried were not added to the list
1206 of hosts for which the message was waiting, so if all were tried, there
1207 was no problem. Two fixes have been applied:
1209 (i) If there are any address or message errors in an SMTP delivery, none
1210 of the hosts (tried or untried) are now added to the list of hosts
1211 for which the message is waiting, so the message should not be a
1212 candidate for sending over the same connection that was used for a
1213 successful delivery of some other message. This seems entirely
1214 reasonable: after all the message is NOT "waiting for some host".
1215 This is so "obvious" that I'm not sure why it wasn't done
1216 previously. Hope I haven't missed anything, but it can't do any
1217 harm, as the worst effect is to miss an optimization.
1219 (ii) If, despite (i), such a delivery is accidentally attempted, the
1220 routing retry time is respected, so at least it doesn't keep
1221 hammering the server.
1223 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1224 in ${readsocket because some servers need this prod.
1226 PH/32 Added some extra debug output when updating a wait-xxx database.
1228 PH/33 The hint "could be header name not terminated by colon", which has been
1229 given for certain expansion errors for a long time, was not being given
1230 for the ${if def:h_colon_omitted{... case.
1232 PH/34 The spec says: "With one important exception, whenever a domain list is
1233 being scanned, $domain contains the subject domain." There was at least
1234 one case where this was not true.
1236 PH/35 The error "getsockname() failed: connection reset by peer" was being
1237 written to the panic log as well as the main log, but it isn't really
1238 panic-worthy as it just means the connection died rather early on. I have
1239 removed the panic log writing for the ECONNRESET error when getsockname()
1242 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1243 runs only) independently of the message's sender address. This meant
1244 that, if the 4xx error was in fact related to the sender, a different
1245 message to the same recipient with a different sender could confuse
1246 things. In particualar, this can happen when sending to a greylisting
1247 server, but other circumstances could also provoke similar problems.
1248 I have changed the default so that the retry time for these errors is now
1249 based a combination of the sender and recipient addresses. This change
1250 can be overridden by setting address_retry_include_sender=false in the
1253 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1254 remote server are returned as part of bounce messages. This was not
1255 happening for LMTP over a pipe (the lmtp transport), but now it is the
1256 same for both kinds of LMTP.
1258 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1259 in header lines that were in fact CNAMEs. This is no longer the case.
1261 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1262 and queue runs started by the daemon processed all messages. This has
1263 been fixed so that -R and -S can now usefully be given with -q<time>.
1265 PH/40 Import PCRE release 6.7 (fixes some bugs).
1267 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1269 PH/42 Give an error if -q is specified more than once.
1271 PH/43 Renamed the variables $interface_address and $interface_port as
1272 $received_ip_address and $received_port, to make it clear that these
1273 values apply to message reception, and not to the outgoing interface when
1274 a message is delivered. (The old names remain recognized, of course.)
1276 PH/44 There was no timeout on the connect() call when using a Unix domain
1277 socket in the ${readsocket expansion. There now is.
1279 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1280 be meaningful with "accept".
1282 SC/01 Eximstats V1.43
1283 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1285 SC/02 Eximstats V1.44
1286 Use a glob alias rather than an array ref in the generated
1287 parser. This improves both readability and performance.
1289 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1290 Collect SpamAssassin and rejection statistics.
1291 Don't display local sender or destination tables unless
1292 there is data to show.
1293 Added average volumes into the top table text output.
1295 SC/04 Eximstats V1.46
1296 Collect data on the number of addresses (recipients)
1297 as well as the number of messages.
1299 SC/05 Eximstats V1.47
1300 Added 'Message too big' to the list of mail rejection
1301 reasons (thanks to Marco Gaiarin).
1303 SC/06 Eximstats V1.48
1304 Mainlog lines which have GMT offsets and are too short to
1305 have a flag are now skipped.
1307 SC/07 Eximstats V1.49 (Alain Williams)
1308 Added the -emptyok flag.
1310 SC/08 Eximstats V1.50
1311 Fixes for obtaining the IP address from reject messages.
1313 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1314 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1315 whitesspace changes from 4.64-PH/27
1317 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1320 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1321 are found, allow negative numbers in numeric criteria)
1323 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1325 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1326 to match changes made in 4.64-PH/43
1328 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1330 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1331 "domain" is actually an IP address, and force a failure. This locks out
1332 those revolvers/nameservers that support "A-for-A" lookups, in
1333 contravention of the specifications.
1335 PH/48 When a host name was looked up from an IP address, and the subsequent
1336 forward lookup of the name timed out, the host name was left in
1337 $sender_host_name, contrary to the specification.
1339 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1340 restricted to single-key lookups, Exim was not diagnosing an error if
1341 * or *@ was used with a query-style lookup.
1343 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1345 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1346 long ago, but noone interested enough thought of it. Let's just say that
1347 the "1.1" means that there are some new functions that weren't there at
1348 some point in the past.
1350 PH/51 Error processing for expansion failure of helo_data from an smtp
1351 transport during callout processing was broken.
1353 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1354 tested/used via the -bh/-bhc/-bs options.
1356 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1357 bug, fixed in subsequent PCRE releases).
1359 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1360 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1362 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1368 SC/01 Use a glob alias rather than an array ref in eximstats generated
1369 parser. This improves both readability and performance.
1371 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1372 Don't display local sender or destination tables in eximstats unless
1373 there is data to show.
1374 Added average volumes into the eximstats top table text output.
1376 SC/03 Collect data on the number of addresses (recipients) as well
1377 as the number of messages in eximstats.
1379 TF/01 Correct an error in the documentation for the redirect router. Exim
1380 does (usually) call initgroups() when daemonizing.
1382 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1383 with consistent privilege compared to when running as a daemon.
1385 TF/03 Note in the spec that $authenticated_id is not set for local
1386 submissions from trusted users.
1388 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1389 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1391 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1392 by adding some example configuration directives to the default
1393 configuration file. A little bit of work is required to uncomment the
1394 directives and define how usernames and passwords are checked, but
1395 there is now a framework to start from.
1397 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1398 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1399 without this. I don't know how relevant this is to other LDAP libraries.
1401 PH/02 Add the verb name to the "unknown ACL verb" error.
1403 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1405 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1407 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1408 directory not expanded when it should be if an expanded home directory
1409 was set for the address (which is overridden by the transport).
1411 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1414 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1415 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1416 because it is too late at that time, and has no effect.
1418 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1419 security issue with \' (bugzilla #107). I could not use the
1420 PQescapeStringConn() function, because it needs a PGconn value as one of
1423 PH/08 When testing addresses using -bt, indicate those final addresses that
1424 are duplicates that would not cause an additional delivery. At least one
1425 person was confused, thinking that -bt output corresponded to deliveries.
1426 (Suppressing duplicates isn't a good idea as you lose the information
1427 about possibly different redirections that led to the duplicates.)
1429 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1430 systems where poll() doesn't work, in particular OS X.
1432 PH/10 Added more information to debugging output for retry time not reached.
1434 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1435 operations in malware.c.
1437 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1440 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1441 syslog_duplication set false, log lines that would normally be written
1442 both the the main log and to the reject log were not written to syslog at
1445 PH/14 In the default configuration, change the use of "message" in ACL warn
1446 statements to "add_header".
1448 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1449 not followed by a command (e.g. "seen endif").
1451 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1452 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1455 PH/17 Added extra conditions to the default value of delay_warning_condition
1459 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1460 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1461 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1464 The Auto-Submitted: and various List- headers are standardised, whereas I
1465 don't think Precedence: ever was.
1467 PH/18 Refactored debugging code in route_finduser() to show more information,
1468 in particular, the error code if getpwnam() issues one.
1470 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1471 This is apparently needed in addition to the PH/07 change above to avoid
1472 any possible encoding problems.
1474 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1475 but not after initializing Perl.
1477 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1478 output them only if debugging. By default they are written stderr,
1479 apparently, which is not desirable.
1481 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1484 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1487 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1489 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1490 authenticated or an ident call has been made. Suppress the default
1491 values for $authenticated_id and $authenticated_sender (but permit -oMai
1492 and -oMas) when testing with -bh.
1494 PH/24 Re-jigged the order of the tests in the default configuration so that the
1495 tests for valid domains and recipients precede the DNS black list and CSA
1496 tests, on the grounds that those ones are more expensive.
1498 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1499 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1500 This bug exists in every version of Exim that I still have, right back to
1503 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1504 However, an attempt to turn on case-sensitivity in a regex key by
1505 including (?-i) didn't work because the subject string was already
1506 lowercased, and the effects were non-intuitive. It turns out that a
1507 one-line patch can be used to allow (?-i) to work as expected.
1513 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1514 other effects) broke the use of negated acl sub-conditions.
1516 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1519 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1520 "Deny" causes Exim to reject the incoming connection with a 554 error.
1521 Unfortunately, if there is a major crisis, such as a disk failure,
1522 tcp-wrappers gives "deny", whereas what one would like would be some
1523 kind of temporary error. A kludge has been added to help with this.
1524 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1525 554 error is used if errno is still zero or contains ENOENT (which occurs
1526 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1529 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1531 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1532 errors. Otherwise a message that provokes a temporary error (when other
1533 messages do not) can cause a whole host to time out.
1535 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1536 addresses were routed directly to files or pipes from a redirect router.
1537 File deliveries just didn't batch; pipe deliveries might have suffered
1540 PH/06 A failure to get a lock for a hints database would erroneously always say
1541 "Failed to get write lock", even when it was really a read lock.
1543 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1544 of 0600. This has been changed to use the value of the lockfile_mode
1545 option (which defaults to 0600).
1547 PH/08 Applied small patch from the Sieve maintainer.
1549 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1550 folder from quota calculations, a direct delivery into this folder messed
1551 up the contents of the maildirsize file. This was because the regex was
1552 used only to exclude .Trash (or whatever) when the size of the mailbox
1553 was calculated. There was no check that a delivery was happening into an
1554 excluded directory. This bug has been fixed by ignoring all quota
1555 processing for deliveries into excluded directories.
1557 PH/10 Added the maildirfolder_create_regex option to appendfile.
1563 PH/01 The code for finding all the local interface addresses on a FreeBSD
1564 system running IPv6 was broken. This may well have applied to all BSD
1565 systems, as well as to others that have similar system calls. The broken
1566 code found IPv4 interfaces correctly, but gave incorrect values for the
1567 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1568 that it would not match correctly against @[] and not recognize the IPv6
1571 PH/02 The ipliteral router was not recognizing addresses of the form user@
1572 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1574 PH/03 Added disable_ipv6.
1576 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1577 decoded form, because it is most often used to construct To: headers
1578 lines in autoreplies, and the decoded form may well be syntactically
1579 invalid. However, $reply_address has leading white space removed, and all
1580 newlines turned into spaces so that the autoreply transport does not
1583 PH/05 If group was specified without a user on a router, and no group or user
1584 was specified on a transport, the group from the router was ignored.
1586 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1587 for visible compile-time settings that can be used to change these
1588 numbers, for those that want even more. Backwards compatibility with old
1589 spool files has been maintained. However, going back to a previous Exim
1590 release will lost any variables that are in spool files.
1592 PH/07 Two small changes when running in the test harness: increase delay when
1593 passing a TCP/IP connection to a new process, in case the original
1594 process has to generate a bounce, and remove special handling of
1595 127.0.0.2 (sic), which is no longer necessary.
1597 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1598 be the same on different OS.
1600 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1603 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1604 whether --show-vars was specified or not
1606 JJ/02 exipick: Added support for new ACL variable spool format introduced
1609 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1610 syntactically invalid From: or Reply-to: line, and a filter used this to
1611 generate an autoreply, and therefore failed to obtain an address for the
1612 autoreply, Exim could try to deliver to a non-existent relative file
1613 name, causing unrelated and misleading errors. What now happens is that
1614 it logs this as a hard delivery error, but does not attempt to create a
1617 PH/11 The exinext utility has a -C option for testing purposes, but although
1618 the given file was scanned by exinext itself; it wasn't being passed on
1619 when Exim was called.
1621 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1622 an end-of-file indication when reading a command response.
1624 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1625 compiled. In many other places in Exim, IPv6 addresses are always
1626 recognized, so I have changed this. It also means that IPv4 domain
1627 literals of the form [IPV4:n.n.n.n] are now always recognized.
1629 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1630 used if the router is not running as root, for example, when verifying at
1631 ACL time, or when using -bh. The debugging output from this situation was
1632 non-existent - all you got was a failure to exec. I have made two
1635 (a) Failures to set uid/gid, the current directory, or a process leader
1636 in a subprocess such as that created by queryprogram now generate
1637 suitable debugging ouput when -d is set.
1639 (b) The queryprogram router detects when it is not running as root,
1640 outputs suitable debugging information if -d is set, and then runs
1641 the subprocess without attempting to change uid/gid.
1643 PH/15 Minor change to Makefile for building test_host (undocumented testing
1646 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1647 additional section of a DNS packet that returns MX or SRV records.
1648 Instead, it always explicitly searches for A/AAAA records. This avoids
1649 major problems that occur when a DNS server includes only records of one
1650 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1651 fixed another bug: if SRV records were looked up and the corresponding
1652 address records were *not* found in the additional section, the port
1653 values from the SRV records were lost.
1655 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1656 using the correct key (the original address) when searching the retry
1657 rules in order to find which one to use for generating the retry hint.
1659 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1660 adding the default one. Similarly, if it contains a Reply-To: header, the
1661 errors_reply_to option, if set, is not used.
1663 PH/19 When calculating a retry time, Exim used to measure the "time since
1664 failure" by looking at the "first failed" field in the retry record. Now
1665 it does not use this if it is later than than the arrival time of the
1666 message. Instead it uses the arrival time. This makes for better
1667 behaviour in cases where some deliveries succeed, thus re-setting the
1668 "first failed" field. An example is a quota failure for a huge message
1669 when small messages continue to be delivered. Without this change, the
1670 "time since failure" will always be short, possible causing more frequent
1671 delivery attempts for the huge message than are intended.
1672 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1674 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1675 $1, $2, $3) because the numerical variables can be reset during some
1676 expansion items (e.g. "match"), thereby losing the authentication data.
1678 PH/21 Make -bV show the size of off_t variables so that the test suite can
1679 decide whether to run tests for quotas > 2G.
1681 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1682 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1683 filecount value is greater than 2G or if a quota value is greater than 2G
1684 on a system where the size of off_t is not greater than 4, a panic error
1687 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1688 never match. The debug and -bh output now contains an explicit error
1689 message indicating a malformed IPv4 address or mask.
1691 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1692 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1693 PH/23 above applies.
1695 PH/25 Do not write to syslog when running in the test harness. The only
1696 occasion when this arises is a failure to open the main or panic logs
1697 (for which there is an explicit test).
1699 PH/26 Added the /no_tell option to "control=freeze".
1701 PH/27 If a host name lookup failed very early in a connection, for example, if
1702 the IP address matched host_lookup and the reverse lookup yielded a name
1703 that did not have a forward lookup, an error message of the form "no IP
1704 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1705 could be logged. Now it outputs the IP address instead of "NULL".
1707 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1708 allows the sender and the authenticated sender to be set when
1709 submitting a message from within Exim. Since child_open_exim() is
1710 documented for local_scan(), the new function should be too.
1712 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1713 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1714 results in an empty string is now treated as unset.
1716 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1718 PH/31 Added sender_verify_fail logging option.
1720 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1721 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1722 tidied the source and removed it altogether.
1724 PH/33 When a queue run was abandoned because the load average was too high, a
1725 log line was always written; now it is written only if the queue_run log
1726 selector is set. In addition, the log line for abandonment now contains
1727 information about the queue run such as the pid. This is always present
1728 in "start" and "stop" lines but was omitted from the "abandon" line.
1730 PH/34 Omit spaces between a header name and the colon in the error message that
1731 is given when verify = headers_syntax fails (if there are lots of them,
1732 the message gets confusing).
1734 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1735 names, as there are now some PTR records that contain slashes. This check
1736 is only to protect against broken name servers that fall over on strange
1737 characters, so the fact that it applies to all lookups doesn't matter.
1739 PH/36 Now that the new test suite is complete, we can remove some of the
1740 special code in Exim that was needed for the old test suite. For example,
1741 sorting DNS records because real resolvers return them in an arbitrary
1742 order. The new test suite's fake resolver always returns records in the
1745 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1746 bounces) except when queue_only is set, to avoid logging races between
1747 the different processes.
1749 PH/38 Panic-die if .include specifies a non-absolute path.
1751 PH/39 A tweak to the "H" retry rule from its user.
1753 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1754 a label. They prevented compilation on older perls.
1756 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1757 a warning to be raised on newish perls.
1759 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1760 on queue. Changes to match documented behaviour of showing count of
1761 messages matching specified criteria.
1763 PH/40 Changed the default ident timeout from 30s to 5s.
1765 PH/41 Added support for the use of login_cap features, on those BSD systems
1766 that have them, for controlling the resources used by pipe deliveries.
1768 PH/42 The content-scanning code uses fopen() to create files in which to put
1769 message data. Previously it was not paying any attention to the mode of
1770 the files. Exim runs with umask(0) because the rest of the code creates
1771 files with open(), and sets the required mode explicitly. Thus, these
1772 files were ending up world-writeable. This was not a big issue, because,
1773 being within the spool directory, they were not world-accessible. I have
1774 created a function called modefopen, which takes an additional mode
1775 argument. It sets umask(777), creates the file, chmods it to the required
1776 mode, then resets the umask. All the relevant calls to fopen() in the
1777 content scanning code have been changed to use this function.
1779 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1780 to 24 hours. This avoids potential overflow problems when processing G
1781 and H retry rules. I suspect nobody ever tinkers with this value.
1783 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1785 PH/45 When the plaintext authenticator is running as a client, the server's
1786 challenges are checked to ensure they are valid base64 strings. By
1787 default, the authentication attempt is cancelled if an invalid string is
1788 received. Setting client_ignore_invalid_base64 true ignores these errors.
1789 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1790 they are received. Thus, the responses can be made to depend on the
1791 challenges. If an invalid string is ignored, an empty string is placed in
1794 PH/46 Messages that are created by the autoreply transport now contains a
1795 References: header, in accordance with RFCs 2822 and 3834.
1797 PH/47 Added authenticated_sender_force to the smtp transport.
1799 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1801 PH/49 Installed latest patch from the Sieve maintainer.
1803 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1804 was also set, the appendfile transport was unnecessarily scanning a
1805 directory of message files (e.g. for maildir delivery) to find the count
1806 of files (along with the size), even though it did not need this
1807 information. It now does the scan only if it needs to find either the
1808 size of the count of files.
1810 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1812 PH/52 Two bugs concerned with error handling when the smtp transport is
1815 (i) Exim was not creating retry information for temporary errors given
1816 for individual recipients after the DATA command when the smtp transport
1817 was used in LMTP mode. This meant that they could be retried too
1818 frequently, and not timed out correctly.
1820 (ii) Exim was setting the flag that allows error details to be returned
1821 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1822 recipients that were returned after the DATA command.
1824 PH/53 This is related to PH/52, but is more general: for any failing address,
1825 when detailed error information was permitted to be returned to the
1826 sender, but the error was temporary, then after the final timeout, only
1827 "retry timeout exceeded" was returned. Now it returns the full error as
1828 well as "retry timeout exceeded".
1830 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1831 do this, and (what is worse) MTAs that accept it.
1833 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1834 will now be deprecated.
1836 PH/56 New os.c-cygwin from the Cygwin maintainer.
1838 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1839 formats (previously only available in exim formats via -bpr, -bpru,
1840 and -bpra. Now also available in native and exiqgrep formats)
1842 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1843 with very large, slow to parse queues
1845 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1847 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1849 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1850 responses to authentication challenges, though it was showing the
1851 challenges; (ii) I've removed the CR characters from the debug output for
1854 PH/58 Allow for the insertion of a newline as well as a space when a string
1855 is turned into more than one encoded-word during RFC 2047 encoding. The
1856 Sieve code now uses this.
1858 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1859 data_4xx, lost_connection, tls_required.
1861 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1862 message was being sent as an SMTP response.
1864 PH/61 Add -l and -k options to exicyclog.
1866 PH/62 When verifying, if an address was redirected to one new address, so that
1867 verification continued, and the new address failed or deferred after
1868 having set something in $address_data, the value of $address_data was not
1869 passed back to the ACL. This was different to the case when no
1870 redirection occurred. The value is now passed back in both cases.
1872 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1873 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1874 use login_cap.h, so on its own it isn't the distinguishing feature. The
1875 new name refers directly to the setclassresources() function.
1877 PH/65 Added configuration files for NetBSD3.
1879 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1881 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1882 is preferred over IPv4.
1884 PH/68 The bounce_return_message and bounce_return_body options were not being
1885 honoured for bounces generated during the reception of non-SMTP messages.
1886 In particular, this applied to messages rejected by the ACL. This bug has
1887 been fixed. However, if bounce_return_message is true and bounce_return_
1888 body is false, the headers that are returned for a non-SMTP message
1889 include only those that have been read before the error was detected.
1890 (In the case of an ACL rejection, they have all been read.)
1892 PH/69 The HTML version of the specification is now built in a directory called
1893 spec_html instead of spec.html, because the latter looks like a path with
1894 a MIME-type, and this confuses some software.
1896 PH/70 Catch two compiler warnings in sieve.c.
1898 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1899 function verify_get_ident() calls ip_connect() to connect a socket, but
1900 if the "connect()" function timed out, ip_connect() used to close the
1901 socket. However, verify_get_ident() also closes the socket later, and in
1902 between Exim writes to the log, which may get opened at this point. When
1903 the socket was closed in ip_connect(), the log could get the same file
1904 descriptor number as the socket. This naturally causes chaos. The fix is
1905 not to close the socket in ip_connect(); the socket should be closed by
1906 the function that creates it. There was only one place in the code where
1907 this was missing, in the iplookup router, which I don't think anybody now
1908 uses, but I've fixed it anyway.
1910 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1911 well as to direct DNS lookups. Otherwise the handling of names in host
1912 lists is inconsistent and therefore confusing.
1918 PH/01 Two changes to the default runtime configuration:
1920 (1) Move the checks for relay_from_hosts and authenticated clients from
1921 after to before the (commented out) DNS black list checks.
1923 (2) Add control=submission to the relay_from_hosts and authenticated
1924 clients checks, on the grounds that messages accepted by these
1925 statements are most likely to be submissions.
1927 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1929 (1) Generate an error if the third argument for the ${prvs expansion is
1932 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1935 (3) Reset the variables that are obtained from the first argument of
1936 ${prvscheck and used in the second argument before leaving the code,
1937 because their memory is reclaimed, so using them afterwards may do
1940 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1941 one (it's much easier than Tom thought :-).
1943 (5) Because of (4), we can now allow for the use of $prvscheck_result
1944 inside the third argument.
1946 PH/03 For some reason, the default setting of PATH when running a command from
1947 a pipe transport was just "/usr/bin". I have changed it to
1950 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1951 anything to be listed in the output from -bV.
1953 PH/05 When a filter generated an autoreply, the entire To: header line was
1954 quoted in the delivery log line, like this:
1956 => >A.N.Other <ano@some.domain> <original@ddress> ...
1958 This has been changed so that it extracts the operative address. There
1959 may be more than one such address. If so, they are comma-separated, like
1962 => >ano@some.domain,ona@other.domain <original@ddress> ...
1964 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1965 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1966 not being looked up in the rDNS to get a host name, Exim was showing the
1967 IP address twice in Received: lines, even though the IP addresses were
1968 identical. For example:
1970 Received: from [1.2.3.4] (helo=[1.2.3.4])
1972 However, if the real host name was known, it was omitting the HELO data
1973 if it matched the actual IP address. This has been tidied up so that it
1974 doesn't show the same IP address twice.
1976 PH/07 When both +timestamp and +memory debugging was on, the value given by
1977 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1978 called by the debug printing, thereby overwriting the timestamp buffer.
1979 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1981 PH/08 When the original message was included in an autoreply transport, it
1982 always said "this is a copy of the message, including all the headers",
1983 even if body_only or headers_only was set. It now gives an appropriate
1986 PH/09 Applied a patch from the Sieve maintainer which:
1988 o fixes some comments
1989 o adds the (disabled) notify extension core
1990 o adds some debug output for the result of if/elsif tests
1991 o points to the current vacation draft in the documentation
1992 and documents the missing references header update
1996 o fixes a bug in processing the envelope test (when testing
1997 multiple envelope elements, the last element determinted the
2000 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2001 Electronic Mail") by including:
2003 Auto-submitted: auto-generated
2005 in the messages that it generates (bounce messages and others, such as
2006 warnings). In the case of bounce messages for non-SMTP mesages, there was
2007 also a typo: it was using "Auto_submitted" (underscore instead of
2008 hyphen). Since every message generated by Exim is necessarily in response
2009 to another message, thes have all been changed to:
2011 Auto-Submitted: auto-replied
2013 in accordance with these statements in the RFC:
2015 The auto-replied keyword:
2017 - SHOULD be used on messages sent in direct response to another
2018 message by an automatic process,
2020 - MUST NOT be used on manually-generated messages,
2022 - MAY be used on Delivery Status Notifications (DSNs) and Message
2023 Disposition Notifications (MDNs),
2025 - MUST NOT be used on messages generated by automatic or periodic
2026 processes, except for messages which are automatic responses to
2029 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2030 to the default Received: header definition.
2032 PH/12 Added log selector acl_warn_skipped (default on).
2034 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2035 variables because (a) they are in the wrong storage pool and (b) even if
2036 they were copied, it wouldn't work properly because of the caching.
2038 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2039 checking when decoding. Apparently there are clients that generate
2040 overlong encoded strings. Why am I not surprised?
2042 PH/15 If the first argument of "${if match_address" was not empty, but did not
2043 contain an "@" character, Exim crashed. Now it writes a panic log message
2044 and treats the condition as false.
2046 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2048 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2049 extension "envelope-auth". The code is finished and in agreement with
2050 other implementations, but there is no documentation so far and in fact,
2051 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2052 not changing the active code.
2054 Print executed "if" and "elsif" statements when debugging is used. This
2055 helps a great deal to understand what a filter does.
2057 Document more things not specified clearly in RFC3028. I had all this
2058 sorted out, when out of a sudden new issues came to my mind. Oops."
2060 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2063 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2064 canonical form (as documented). However, after a host name lookup from
2065 the IP address, check_host() was doing a simple string comparison with
2066 addresses acquired from the DNS when checking that the found name did
2067 have the original IP as one of its addresses. Since any found IPv6
2068 addresses are likely to be in abbreviated form, the comparison could
2069 fail. Luckily, there already exists a function for doing the comparison
2070 by converting both addresses to binary, so now that is used instead of
2071 the text comparison.
2073 PH/20 There was another similar case to PH/19, when a complete host name was
2074 given in a host list; looking up its IP address could give an abbreviated
2075 form, whereas the current host's name might or might not be abbreviated.
2076 The same fix has been applied.
2082 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2083 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2086 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2087 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2089 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2091 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2092 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2093 Also, for "Auto-Submitted", treat anything other than "no" as
2094 non-personal, in accordance with RFC 3834. (Previously it treated
2095 anything starting "auto-" as non-personal.)
2097 TF/01 The control=submission/name=... option had a problem with syntax
2098 errors if the name included a slash character. The /name= option
2099 now slurps the rest of the string, so it can include any characters
2100 but it must come last in the list of options (after /sender_retain
2103 PH/05 Some modifications to the interface to the fake nameserver for the new
2111 TK/01 Added the "success_on_redirect" address verification option. See
2112 NewStuff for rationale and an example.
2114 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2116 PH/02 Patch to exigrep to allow it to work on syslog lines.
2118 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2119 fread() to skip over the body file's header line, because in Cygwin the
2120 header line is locked and is inaccessible.
2122 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2123 co-exist for some time) to make it clear that it is the Exim ID that is
2124 referenced, not the Message-ID: header line.
2126 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2127 string_format() function, because snprintf() does not exist on all
2130 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2131 ${stat: expansion item.
2133 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2134 protocol synchronization error", to keep the pedants happy.
2136 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2137 well as for IRIX systems, when gcc is being used. See the host.c source
2140 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2142 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2145 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2146 transport and to the smtp transport in LMTP mode.
2148 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2150 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2151 run a filter in a subprocess. This could lead to confusion in subsequent
2152 lookups in the parent process. There should also be a search_tidyup() at
2153 the end of the subprocess.
2155 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2156 only if the host matched helo_try_verify_hosts, which caused the
2157 verification to occur when the EHLO/HELO command was issued. The ACL just
2158 tested the remembered result. Now, if a previous verification attempt has
2159 not happened, "verify = helo" does it there and then.
2161 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2163 TK/03 Fix log output including CR from clamd.
2165 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2166 exist provoked a memory error which could cause a segfault.
2168 PH/15 Installed PCRE 6.2
2170 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2172 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2173 of the problem. Specifically, suggested +O2 rather than +O1 for the
2176 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2178 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2179 record, the port was not correctly logged when the outgoing_port log
2180 selector was set (it logged the transort's default port).
2182 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2183 fallback_hosts, and "hosts" in the smtp transport.
2185 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2186 host errors such as "Connection refused".
2188 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2189 authentication with radiusclient 0.4.9:
2191 - Error returned from rc_read_config was caught wrongly
2192 - Username/password not passed on to radius server due to wrong length.
2194 The presumption is that some radiusclient API changes for 4.51/PH/17
2195 were not taken care of correctly. The code is still untested by me (my
2196 Linux distribution still has 0.3.2 of radiusclient), but it was
2197 contributed by a Radius user.
2199 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2200 expanding the "port" option of the smtp transport.
2202 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2203 while reading a MIME header. Thanks to Tom Hughes for a patch.
2205 PH/24 Include config.h inside local_scan.h so that configuration settings are
2208 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2209 that in an ACL for RCPT (for example), you can examine exactly what was
2212 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2213 commands, but it was not correctly comparing the address with the actual
2214 client host address. Thus, it would show the EHLO address in Received:
2215 header lines when this was not necessary.
2217 PH/27 Added the % operator to ${eval:}.
2219 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2220 it should be ignoring failures (because with -C, for example, it has lost
2221 privilege). It wasn't ignoring creation failures other than "already
2224 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2227 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2228 because an input line was too long, either on its own, or by virtue of
2229 too many continuations, the temporary file was not being removed, and the
2230 return code was incorrect.
2232 PH/31 Missing "BOOL" in function definition in filtertest.c.
2234 PH/32 Applied Sieve patches from the maintainer.
2236 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2238 PH/33 Added "verify = not_blind".
2240 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2241 Local/Makefile (with some defaults set). These are used in built scripts
2242 such as exicyclog, but they have never been used in the exim_install
2243 script (though there are many overriding facilities there). I have
2244 arranged that the exim_install script now takes note of these two
2247 PH/35 Installed configuration files for Dragonfly.
2249 PH/36 When a locally submitted message by a trusted user did not contain a
2250 From: header, and the sender address was obtained from -f or from an SMTP
2251 MAIL command, and the trusted user did not use -F to supply a sender
2252 name, $originator_name was incorrectly used when constructing a From:
2253 header. Furthermore, $originator_name was used for submission mode
2254 messages from external hosts without From: headers in a similar way,
2255 which is clearly wrong.
2257 PH/37 Added control=suppress_local_fixups.
2259 PH/38 When log_selector = +received_sender was set, and the addition of the
2260 sender made the log line's construction buffer exactly full, or one byte
2261 less than full, an overflow happened when the terminating "\n" was
2264 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2265 when the result of a list match is failure because a DNS lookup failed.
2267 PH/40 RM_COMMAND is now used in the building process.
2269 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2270 the "build-* directories that it finds.
2272 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2273 address in a domain literal was a prefix of an interface address.
2275 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2276 when verifying a sender address, unless rewrite_headers is false.
2278 PH/44 Wrote a long comment about why errors_to addresses are verified as
2279 recipients, not senders.
2281 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2282 the ratelimit ACL was added.
2284 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2286 PH/46 Added extra information about PostgreSQL errors to the error string.
2288 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2289 avoiding the need to install special zones in a real server. This is
2290 backwards compatible; if it can't find the fake resolver, it drops back.
2291 Thus, both old and new test suites can be run.
2293 TF/02 Added util/ratelimit.pl
2295 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2296 clock is set back in time.
2298 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2299 Candler <B.Candler@pobox.com>.
2301 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2302 OK for addresses that are the result of redirections.
2304 PH/48 A number of further additions for the benefit of the new test suite,
2305 including a fake gethostbyname() that interfaces to the fake DNS resolver
2308 TF/06 The fix for widen_domains has also been applied to qualify_single and
2309 search_parents which are the other dnslookup options that can cause
2312 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2315 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2317 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2318 service" if it encountered an MX record with an empty target hostname.
2319 The message is now "an MX or SRV record indicated no SMTP service".
2321 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2322 if the DNS of the sending site is misconfigured. This is quite a
2323 common situation. This change restores the behaviour of treating a
2324 helo verification defer as a failure.
2326 PH/51 If self=fail was set on a router, the bounce message did not include the
2327 actual error message.
2333 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2335 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2336 command itself ended in error, the underlying message about the transport
2337 filter timeout was being overwritten with the pipe command error. Now the
2338 underlying error message should be appended to the second error message.
2340 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2342 PH/02 Reduce the amount of output that "make" produces by default. Full output
2343 can still be requested.
2345 PH/03 The warning log line about a condition test deferring for a "warn" verb
2346 was being output only once per connection, rather than after each
2347 occurrence (because it was using the same function as for successful
2348 "warn" verbs). This seems wrong, so I have changed it.
2350 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2351 it should not have, which might have caused a crash in the right
2352 circumstances, but probably never did.
2354 PH/04 Installed a modified version of Tony Finch's patch to make submission
2355 mode fix the return path as well as the Sender: header line, and to
2356 add a /name= option so that you can make the user's friendly name appear
2359 TF/03 Added the control = fakedefer ACL modifier.
2361 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2362 Mark Lowes for thorough testing.
2364 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2366 TK/03 Merged latest SRS patch from Miles Wilton.
2368 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2369 with the definition in sysexits.h (which is #included earlier).
2370 Fortunately, Exim does not actually use EX_OK. The code used to try to
2371 preserve the sysexits.h value, by assumimg that macro definitions were
2372 scanned for macro replacements. I have been disabused of this notion,
2373 so now the code just undefines EX_OK before #including unistd.h.
2375 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2376 in the smtp transport. When a block could not be written in a single
2377 write() function, the timeout was being re-applied to each part-write.
2378 This seems wrong - if the receiver was accepting one byte at a time it
2379 would take for ever. The timeout is now adjusted when this happens. It
2380 doesn't have to be particularly precise.
2382 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2383 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2385 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2386 without a domain if the check to <postmaster@domain> fails.
2388 SC/01 Eximstats: added -xls and the ability to specify output files
2389 (patch written by Frank Heydlauf).
2391 SC/02 Eximstats: use FileHandles for outputing results.
2393 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2395 SC/04 Eximstats: fixed display of large numbers with -nvr option
2397 SC/05 Eximstats: fixed merging of reports with empty tables.
2399 SC/06 Eximstats: added the -include_original_destination flag
2401 SC/07 Eximstats: removed tabs and trailing whitespace.
2403 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2405 TK/06 MBOX spool code: Add real "From " MBOX separator line
2406 so the .eml file is really in mbox format (even though
2407 most programs do not really care). Patch from Alex Miller.
2409 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2410 The latter is generated from $received_to and is only set if the
2411 message has one envelope recipient. SA can use these headers,
2412 obviously out-of-the-box. Patch from Alex Miller.
2414 PH/08 The ${def test on a variable was returning false if the variable's
2415 value was "0", contrary to what the specification has always said!
2416 The result should be true unless the variable is empty.
2418 PH/09 The syntax error of a character other than { following "${if
2419 def:variable_name" (after optional whitespace) was not being diagnosed.
2420 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2421 accidental colon was present, for example, could give incorrect results.
2423 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2424 result is used (not including appendfile, where other changes are about
2427 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2428 This involved changing a lot of size variables from int to off_t. It
2429 should work with maildirs and everything.
2431 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2432 spamd dying while we are connected to it.
2434 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2437 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2438 function for BDB 4.3.
2440 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2442 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2443 the sources for PCRE have been re-arranged and more files are now
2446 PH/15 The code I had for printing potentially long long variables in PH/11
2447 above was not the best (it lost precision). The length of off_t variables
2448 is now inspected at build time, and an appropriate printing format (%ld
2449 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2450 to be "long long int" or "long int". This is needed for the internal
2451 formatting function string_vformat().
2453 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2454 the configuration file to be ":syslog", then the script "guesses" where
2455 the logs files are, rather than using the compiled in default. In our
2456 case the guess is not the same as the compiled default, so the script
2457 suddenly stopped working when I started to use syslog. The patch checks
2458 to see if log_file_path is "". If so, it attempts to read it from exim
2459 with no configuration file to get the compiled in version, before it
2460 falls back to the previous guessing code."
2462 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2463 implementing BATV in an Exim configuration. See NewStuff for the gory
2466 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2467 Makefile that are specific to HP-UX.
2469 PH/18 If the "use_postmaster" option was set for a recipient callout together
2470 with the "random" option, the postmaster address was used as the MAIL
2471 FROM address for the random test, but not for the subsequent recipient
2472 test. It is now used for both.
2474 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2475 patch removes a few documentation additions to RFC 3028, because the
2476 latest draft now contains them. It adds the new en;ascii-case comparator
2477 and a new error check for 8bit text in MIME parts. Comparator and
2478 require names are now matched exactly. I enabled the subaddress
2479 extension, but it is not well tested yet (read: it works for me)."
2481 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2482 rework some of the code of TK/09 above to avoid the hardwired use of
2483 "%lld" and "long long". Replaced the call to snprintf() with a call to
2486 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2487 records point to non-existent hosts", "retry timeout exceeded", and
2488 "retry time not reached for any host after a long failure period".
2490 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2491 experimental DomainKeys support:
2493 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2494 (2) On an error such as an illegally used "control", the wrong name for
2495 the control was given.
2497 These problems did NOT occur unless DomainKeys support was compiled.
2499 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2501 PH/24 Added ${if match_ip condition.
2503 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2504 databases so that it will be absolutely obvious if a crash occurs in the
2505 DB library. This is a regular occurrence (often caused by mis-matched
2508 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2509 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2510 picked up on a user's system that detects such things. There doesn't seem
2511 to be a gcc warning option for this - only an attribute that has to be
2512 put on the function's prototype. It seems that in Fedora Core 4 they have
2513 set this on a number of new functions. No doubt there will be more in due
2516 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2517 specify a transport. However, if an address that was verified by such a
2518 router was the subject of a callout, Exim crashed because it tried to
2519 read the rcpt_include_affixes from the non-existent transport. Now it
2520 just assumes that the setting of that option is false. This bug was
2521 introduced by 4.51/PH/31.
2523 PH/28 Changed -d+all to exclude +memory, because that information is very
2524 rarely of interest, but it makes the output a lot bigger. People tend to
2525 do -d+all out of habit.
2527 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2528 code in os-type was giving problems when libc.so lives in lib64, like on
2531 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2532 aren't the modern standard, and it seems that some systems' include files
2533 don't always have them. Exim was already checking for some of the newer
2534 ones like T_AAAA, and defining it itself. I've added checks for all the
2535 record types that Exim uses.
2537 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2538 not automatically generating a new one, as it is supposed to. This
2539 prevented TLS from working. If the file did exist, but contained invalid
2540 data, a new version was generated, as expected. It was only the case of a
2541 non-existent file that was broken.
2543 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2544 with a change in libdomainkeys > 0.64.
2546 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2547 from DNS. If the selector record carries the flag, it now has
2548 precedence over the domain-wide flag.
2550 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2552 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2553 the use of a port name that isn't defined in /etc/services) occurred, the
2554 message was deferred as in a normal delivery, and thus remained on the
2555 spool, instead of being failed because of the mua_wrapper setting. This
2556 is now fixed, and I tidied up some of the mua_wrapper messages at the
2559 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2560 the messages in a hash of arrays rather than using individual hashes.
2561 This is a bit cleaner and results in dramatic memory savings, albeit
2562 at a slight CPU cost.
2564 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2565 as requested by Marc Sherman.
2567 SC/10 Eximstats: added histograms for user specified patterns as requested
2570 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2572 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2573 fopen() in the content-scanning modules that did not already have it.
2579 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2580 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2582 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2584 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2586 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2587 to test Sieve filters that use "vacation".
2589 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2590 that changes the way the GnuTLS parameters are stored in the cache file.
2591 The new format can be generated externally. For backward compatibility,
2592 if the data in the cache doesn't make sense, Exim assumes it has read an
2593 old-format file, and it generates new data and writes a new file. This
2594 means that you can't go back to an older release without removing the
2597 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2598 work if there are any delivery delays because "one_time" forces the
2599 parent to be marked "delivered", so its unseen clone is never tried
2600 again. For this reason, Exim now forbids the simultaneous setting of
2603 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2604 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2605 entry if you want to know the details. The fix, however, neglected to
2606 consider the case when local delivery batching is involved. The test for
2607 "previously delivered" was not happening when checking to see if an
2608 address could be batched with a previous (undelivered) one; under
2609 certain circumstances this could lead to multiple deliveries to the same
2612 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2613 in its include files, and this causes problems building Exim.
2615 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2616 header_syntax) but Exim was just ignoring anything given after a slash.
2617 In particular, this caused confusion with an attempt to use "verify =
2618 reverse_host_lookup/defer_ok". An error is now given when options are
2619 supplied for verify items that do not have them. (Maybe reverse_host_
2620 lookup should have a defer_ok option, but that's a different point.)
2622 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2623 defined by RFC 821) to 2048, because there were problems with some AUTH
2624 commands, and RFC 1869 says the size should be increased for extended
2625 SMTP commands that take arguments.
2627 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2630 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2631 "unknown" error; now it says that the functionality isn't in the binary.
2633 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2634 an address' error message when a string expansion fails (syntax or
2635 whatever). Otherwise the password may appear in the log. Following change
2636 PH/42 below, there is no longer a chance of it appearing in a bounce
2639 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2641 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2642 of its addresses was ever tried. (Bugzilla bug #2.)
2644 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2645 the result incorrectly in the debug output. (It correctly added a newline
2646 to what was transported.)
2648 TF/01 Added $received_time.
2650 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2651 commented out examples of how to interface to a virus scanner and to
2652 SpamAssassin. Also added commented examples of av_scanner and
2653 spamd_address settings.
2655 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2656 and controls are allowed in which ACLs. There were a couple of minor
2657 errors. Some of the entries in the conditions table (which is a table of
2658 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2659 negation of where the condition IS allowed.
2661 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2663 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2664 header file does not have a version number, so I've had to invent a new
2665 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2666 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2667 radiusclient), but it was contributed by a Radius user.
2669 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2670 files or directories, for OpenSSL.
2672 PH/19 When an Exim process that is running as root has to create an Exim log
2673 file, it does so in a subprocess that runs as exim:exim so as to get the
2674 ownership right at creation (otherwise, other Exim processes might see
2675 the file with the wrong ownership). There was no test for failure of this
2676 fork() call, which would lead to the process getting stuck as it waited
2677 for a non-existent subprocess. Forks do occasionally fail when resources
2678 run out. I reviewed all the other calls to fork(); they all seem to check
2681 PH/20 When checking for unexpected SMTP input at connect time (before writing
2682 the banner), Exim was not dealing correctly with a non-positive return
2683 from the read() function. If the client had disconnected by this time,
2684 the result was a log entry for a synchronization error with an empty
2685 string after "input=" when read() returned zero. If read() returned -1
2686 (an event I could not check), uninitialized data bytes were printed.
2687 There were reports of junk text (parts of files, etc) appearing after
2690 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2692 PH/22 Added support for macro redefinition, and (re)definition in between
2693 driver and ACL definitions.
2695 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2696 forgetting to use the resulting value; it was using the unexpanded value.
2698 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2699 hadn't been configured. The fix is from Juergen Kreileder, who
2700 understands it better than I do:
2702 "Here's what I see happening with three configured cyrus_sasl
2703 authenticators configured (plain, login, cram-md5):
2705 On startup auth_cyrus_sasl_init() gets called for each of these.
2706 This means three calls to sasl_listmech() without a specified mech_list.
2707 => SASL tests which mechs of all available mechs actually work
2708 => three warnings about OTP not working
2709 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2711 With the patch, sasl_listmech() also gets called three times. But now
2712 SASL's mech_list option is set to the server_mech specified in the the
2713 authenticator. Or in other words, the answer from sasl_listmech()
2714 gets limited to just the mech you're testing for (which is different
2716 => the return list contains just 'plain' or 'login', 'cram-md5' or
2717 nothing depending on the value of ob->server_mech.
2719 I've just tested the patch: Authentication still works fine,
2720 unavailable mechs specified in the exim configuration are still
2721 caught, and the auth.log warnings about OTP are gone."
2723 PH/25 When debugging is enabled, the contents of the command line are added
2724 to the debugging output, even when log_selector=+arguments is not
2727 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2728 answer is "GNU", and only if the return is "GNU/something" is the answer
2731 PH/27 $acl_verify_message is now set immediately after the failure of a
2732 verification in an ACL, and so is available in subsequent modifiers. In
2733 particular, the message can be preserved by coding like this:
2735 warn !verify = sender
2736 set acl_m0 = $acl_verify_message
2738 Previously, $acl_verify_message was set only while expanding "message"
2739 and "log_message" when a very denied access.
2741 PH/28 Modified OS/os.c-Linux with
2743 -#ifndef OS_LOAD_AVERAGE
2744 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2746 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2747 nomenclature these days.)
2749 PH/29 Installed patch from the Sieve maintainer that adds the options
2750 sieve_useraddress and sieve_subaddress to the redirect router.
2752 PH/30 In these circumstances:
2753 . Two addresses routed to the same list of hosts;
2754 . First host does not offer TLS;
2755 . First host accepts first address;
2756 . First host gives temporary error to second address;
2757 . Second host offers TLS and a TLS session is established;
2758 . Second host accepts second address.
2759 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2760 and peerdn, if requested) that were in fact used only for the second
2763 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2764 attention to any local part prefix or suffix that was matched by the
2765 router that accepted the address. It now behaves in the same way as it
2766 does for delivery: the affixes are removed from the local part unless
2767 rcpt_include_affixes is set on the transport.
2769 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2770 timeout during the DATA phase of an incoming message.
2772 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2773 applied a patch sanctioned by the Sieve maintainer.
2775 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2776 the uid or gid is negative. A case of a negative gid caused this to be
2777 noticed. The fix allows for either to be negative.
2779 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2780 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2783 PH/36 The change PH/12 above was broken. Fixed it.
2785 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2786 the grounds that routing the same address twice would always produce the
2787 same answer. This might have been true once, but it is certainly no
2788 longer true now. Routing a child address may depend on the previous
2789 routing that produced that child. Some complicated redirection strategies
2790 went wrong when messages had multiple recipients, and made Exim's
2791 behaviour dependent on the order in which the addresses were given.
2793 I have moved the duplicate checking until after the routing is complete.
2794 Exim scans the addresses that are assigned to local and remote
2795 transports, and removes any duplicates. This means that more work will be
2796 done, as duplicates will always all be routed, but duplicates are
2797 presumably rare, so I don't expect this is of any significance.
2799 For deliveries to pipes, files, and autoreplies, the duplicate checking
2800 still happens during the routing process, since they are not going to be
2803 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2804 It corrects a timeout issue with spamd. This is Ian's comment: "The
2805 background is that sometimes spamd either never reads data from a
2806 connection it has accepted, or it never writes response data. The exiscan
2807 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2808 blindly assumes that writes won't block so it may never time out."
2810 PH/39 Allow G after quota size as well as K and M.
2812 PH/40 The value set for $authenticated_id in an authenticator may not contain
2813 binary zeroes or newlines because the value is written to log lines and
2814 to spool files. There was no check on this. Now the value is run through
2815 the string_printing() function so that such characters are converted to
2816 printable escape sequences.
2818 PH/41 $message_linecount is a new variable that contains the total number of
2819 lines in the message. Compare $body_linecount, which is the count for the
2822 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2823 bounce and delay warning messages, except in certain special cases, which
2826 (a) An SMTP error message from a remote host;
2827 (b) A message specified in a :fail: redirection;
2828 (c) A message specified in a "fail" command in a system filter;
2829 (d) A message specified in a FAIL return from the queryprogram router;
2830 (e) A message specified by the cannot_route_message router option.
2832 In these cases only, Exim does include the error details in bounce and
2833 warning messages. There are also a few cases where bland messages such
2834 as "unrouteable address" or "local delivery error" are given.
2836 PH/43 $value is now also set for the "else" part of a ${run expansion.
2838 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2839 being worked on, but at least Exim now implements the latest version to
2842 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2843 process to complete was treated as a delivery failure, a timeout while
2844 writing the message to the pipe was logged, but erroneously treated as a
2845 successful delivery. Such timeouts include transport filter timeouts. For
2846 consistency with the overall process timeout, these timeouts are now
2847 treated as errors, giving rise to delivery failures by default. However,
2848 there is now a new Boolean option for the pipe transport called
2849 timeout_defer, which, if set TRUE, converts the failures into defers for
2850 both kinds of timeout. A transport filter timeout is now identified in
2853 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2854 systems where "make" and "gmake" are different, calling "gmake" at top
2855 level broke things. I've arranged for the value of $(MAKE) to be passed
2856 from the Makefile to this script so that it can call the same version of
2860 A note about Exim versions 4.44 and 4.50
2861 ----------------------------------------
2863 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2864 changes of various kinds. As a consequence, a big documentation update was
2865 needed. This delayed the release for rather longer than seemed good, especially
2866 in the light of a couple of (minor) security issues. Therefore, the changes
2867 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2868 release. So 4.44 and 4.50 are in effect two different branches that both start
2871 I have left the 4.50 change log unchanged; it contains all the changes since
2872 4.43. The change log for 4.44 is below; many of its items are identical to
2873 those for 4.50. This seems to be the most sensible way to preserve the
2874 historical information.
2880 1. Minor wording change to the doc/README.SIEVE file.
2882 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2883 computation of the current number of files was incorrect.
2885 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2886 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2889 4. Give more explanation in the error message when the command for a transport
2890 filter fails to execute.
2892 5. There are several places where Exim runs a non-Exim command in a
2893 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2894 was being done only for the command run by the queryprogram router. It is
2895 now done for all such subprocesses. The other cases are: ${run, transport
2896 filters, and the commands run by the lmtp and pipe transports.
2898 6. Added CONFIGURE_GROUP build-time option.
2900 7. Some older OS have a limit of 256 on the maximum number of file
2901 descriptors. Exim was using setrlimit() to set 1000 as a large value
2902 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2903 systems. I've change it so that if it can't get 1000, it tries for 256.
2905 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2906 was an oversight, and furthermore, ever since the addition of extra
2907 controls (e.g. 4.43/32), the checks on when to allow different forms of
2908 "control" were broken. There should now be diagnostics for all cases when a
2909 control that does not make sense is encountered.
2911 9. Added the /retain_sender option to "control=submission".
2913 10. $recipients is now available in the predata ACL (oversight).
2915 11. Tidy the search cache before the fork to do a delivery from a message
2916 received from the command line. Otherwise the child will trigger a lookup
2917 failure and thereby defer the delivery if it tries to use (for example) a
2918 cached ldap connection that the parent has called unbind on.
2920 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2921 of $address_data from the recipient verification was clobbered by the
2922 sender verification.
2924 13. The value of address_data from a sender verification is now available in
2925 $sender_address_data in subsequent conditions in the ACL statement.
2927 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2929 15. Added a new option "connect=<time>" to callout options, to set a different
2932 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2933 was its contents. (It was OK if the option was not defined at all.)
2935 17. A "Completed" log line is now written for messages that are removed from
2936 the spool by the -Mrm option.
2938 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2939 information about exactly what failed.
2941 19. Added -dd to debug only the daemon process.
2943 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2944 handles timeouts, both on the server side and network timeouts. Renamed the
2945 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2947 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2948 It is now set to "smtps".
2950 22. $host_address is now set to the target address during the checking of
2951 ignore_target_hosts.
2953 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2954 being passed; this would have caused $sender_host_name to have been used if
2955 matching the list had actually called for a host name (not very likely,
2956 since this list is usually IP addresses). A host name is now passed as
2959 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2960 code that specifies a non-restarting handler (typically sigaction() in
2961 modern systems) in an attempt to fix a rare and obscure crash bug.
2963 25. Narrowed the window for a race in the daemon that could cause it to ignore
2964 SIGCHLD signals. This is not a major problem, because they are used only to
2965 wake it up if nothing else does.
2967 26. A malformed maildirsize file could cause Exim to calculate negative values
2968 for the mailbox size or file count. Odd effects could occur as a result.
2969 The maildirsize information is now recalculated if the size or filecount
2972 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2973 support for a long time. Removed HAVE_SYS_VFS_H.
2975 28. Installed the latest version of exipick from John Jetmore.
2977 29. In an address list, if the pattern was not a regular expression, an empty
2978 subject address (from a bounce message) matched only if the pattern was an
2979 empty string. Non-empty patterns were not even tested. This was the wrong
2980 because it is perfectly reasonable to use an empty address as part of a
2981 database query. An empty address is now tested by patterns that are
2982 lookups. However, all the other forms of pattern expect the subject to
2983 contain a local part and a domain, and therefore, for them, an empty
2984 address still always fails if the pattern is not itself empty.
2986 30. Exim went into a mad DNS loop when attempting to do a callout where the
2987 host was specified on an smtp transport, and looking it up yielded more
2988 than one IP address.
2990 31. Re-factored the code for checking spool and log partition space into a
2991 function that finds that data and another that does the check. The former
2992 is then used to implement four new variables: $spool_space, $log_space,
2993 $spool_inodes, and $log_inodes.
2995 32. The RFC2047 encoding function was originally intended for short strings
2996 such as real names; it was not keeping to the 75-character limit for
2997 encoded words that the RFC imposes. It now respects the limit, and
2998 generates multiple encoded words if necessary. To be on the safe side, I
2999 have increased the buffer size for the ${rfc2047: expansion operator from
3002 33. It is now permitted to omit both strings after an "if" condition; if the
3003 condition is true, the result is "true". As before, when the second string
3004 is omitted, a false condition yields an empty string. This makes it less
3005 cumbersome to write custom ACL and router conditions.
3007 34. Failure to deliver a bounce message always caused it to be frozen, even if
3008 there was an errors_to setting on the router. The errors_to setting is now
3011 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3012 canonical form (fully expanded) before being placed in
3013 $sender_host_address.
3015 36. The table in the code that translates DNS record types into text (T_A to
3016 "A" for instance) was missing entries for NS and CNAME. It is just possible
3017 that this could have caused confusion if both these types were looked up
3018 for the same domain, because the text type is used as part of Exim's
3019 per-process caching. But the chance of anyone hitting this buglet seems
3022 37. The dnsdb lookup has been extended in a number of ways.
3024 (1) There is a new type, "zns", which walks up the domain tree until it
3025 finds some nameserver records. It should be used with care.
3027 (2) There is a new type, "mxh", which is like "mx" except that it returns
3028 just the host names, not the priorities.
3030 (3) It is now possible to give a list of domains (or IP addresses) to be
3031 looked up. The behaviour when one of the lookups defers can be
3032 controlled by a keyword.
3034 (4) It is now possible to specify the separator character for use when
3035 multiple records are returned.
3037 38. The dnslists ACL condition has been extended: it is now possible to supply
3038 a list of IP addresses and/or domains to be looked up in a particular DNS
3041 39. Added log_selector=+queue_time_overall.
3043 40. When running the queue in the test harness, wait just a tad after forking a
3044 delivery process, to get repeatability of debugging output.
3046 41. Include certificate and key file names in error message when GnuTLS fails
3047 to set them up, because the GnuTLS error message doesn't include the name
3048 of the failing file when there is a problem reading it.
3050 42. Allow both -bf and -bF in the same test run.
3052 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3054 44. The "Exiscan patch" is now merged into the mainline Exim source.
3056 45. Sometimes the final signoff response after QUIT could fail to get
3057 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3058 < 0 before doing a fflush(). This bug looks as though it goes back to the
3059 introduction of TLS in release 3.20, but "sometimes" must have been rare
3060 because the tests only now provoked it.
3062 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3063 (this can affect the format of dates).
3065 47. exim_tidydb, when checking for the continued existence of a message for
3066 which it has found a message-specific retry record, was not finding
3067 messages that were in split spool directories. Consequently, it was
3068 deleting retry records that should have stayed in existence.
3070 48. Steve fixed some bugs in eximstats.
3072 49. The SPA authentication driver was not abandoning authentication and moving
3073 on to the next authenticator when an expansion was forced to fail,
3074 contradicting the general specification for all authenticators. Instead it
3075 was generating a temporary error. It now behaves as specified.
3077 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3078 (the order specifies the preference for clients). The order is now AES256,
3079 AES128, 3DES, ARCFOUR128.
3081 51. Small patch to Sieve code - explicitly set From: when generating an
3084 52. Exim crashed if a remote delivery caused a very long error message to be
3085 recorded - for instance if somebody sent an entire SpamAssassin report back
3086 as a large number of 550 error lines. This bug was coincidentally fixed by
3087 increasing the size of one of Exim's internal buffers (big_buffer) that
3088 happened as part of the Exiscan merge. However, to be on the safe side, I
3089 have made the code more robust (and fixed the comments that describe what
3092 53. Now that there can be additional text after "Completed" in log lines (if
3093 the queue_time_overall log selector is set), a one-byte patch to exigrep
3094 was needed to allow it to recognize "Completed" as not the last thing in
3097 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3098 patch that reportedly fixes this has been added. I am not expert enough to
3099 create a test for it. This is what the patch creator wrote:
3101 "I found a little strange behaviour of ldap code when working with
3102 Windows 2003 AD Domain, where users was placed in more than one
3103 Organization Units. When I tried to give exim partial DN, the exit code
3104 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3105 But simultaneously result of request was absolutely normal ldap result,
3106 so I produce this patch..."
3108 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3109 REFERENCE, so I have modified the code to exclude the patch when that macro
3112 55. Some experimental protocols are using DNS PTR records for new purposes. The
3113 keys for these records are domain names, not reversed IP addresses. The
3114 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3115 leaves it alone. Component reversal etc. now happens only for IP addresses.
3118 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3120 57. Double the size of the debug message buffer (to 2048) so that more of very
3121 long debug lines gets shown.
3123 58. The exicyclog utility now does better if the number of log files to keep
3124 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3126 59. Two changes related to the smtp_active_hostname option:
3128 (1) $smtp_active_hostname is now available as a variable.
3129 (2) The default for smtp_banner uses $smtp_active_hostname instead
3130 of $primary_hostname.
3132 60. The host_aton() function is supposed to be passed a string that is known
3133 to be a valid IP address. However, in the case of IPv6 addresses, it was
3134 not checking this. This is a hostage to fortune. Exim now panics and dies
3135 if the condition is not met. A case was found where this could be provoked
3136 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3137 components; fortuitously, this particular loophole had already been fixed
3138 by change 4.50/55 above.
3140 If there are any other similar loopholes, the new check in host_aton()
3141 itself should stop them being exploited. The report I received stated that
3142 data on the command line could provoke the exploit when Exim was running as
3143 exim, but did not say which command line option was involved. All I could
3144 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3145 running as the user.
3148 61. There was a buffer overflow vulnerability in the SPA authentication code
3149 (which came originally from the Samba project). I have added a test to the
3150 spa_base64_to_bits() function which I hope fixes it.
3153 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3154 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3156 63. The daemon start-up calls getloadavg() while still root for those OS that
3157 need the first call to be done as root, but it missed one case: when
3158 deliver_queue_load_max is set with deliver_drop_privilege. This is
3159 necessary for the benefit of the queue runner, because there is no re-exec
3160 when deliver_drop_privilege is set.
3162 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3163 This has been fixed.
3165 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3166 was in use, was not putting the data itself into the right store pool;
3167 consequently, it could be overwritten for a subsequent message in the same
3168 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3171 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3173 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3174 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3175 the function were treating the return as a boolean value, which happened to
3176 work because 0=false and not-0=true, but is not correct code.
3178 68. The host_aton() function was not handling scoped IPv6 addresses (those
3179 with, for example, "%eth0" on the end) correctly.
3181 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3182 negated items (that is, ~something) in unsigned ints. Some compilers
3183 apparently mutter when there is no cast.
3185 70. If an address verification called from an ACL failed, and did not produce a
3186 user-specific message (i.e. there was only a "system" message), nothing was
3187 put in $acl_verify_message. In this situation, it now puts the system
3190 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3191 session; change 4.31/43 added the unwanted input to the log line - except
3192 that it did not do this in the start of session case. It now does.
3194 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3195 This is wrong and can cause the other end to generate a synchronization
3196 error if it is another Exim or anything else that does the synchronization
3197 check. A QUIT command is no longer sent after a timeout.
3199 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3200 during host lookups.
3202 74. The defer_ok option of callout verification was not working if it was used
3203 when verifying addresses in header lines, that is, for this case:
3205 verify = header_sender/callout=defer_ok
3207 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3208 those file descriptors could be used for SMTP connections. If anything
3209 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3210 could be sent to the SMTP client, causing chaos. The daemon now opens
3211 stdin, stdout, and stderr to /dev/null when it puts itself into the
3214 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3215 log by default. The user can override this with suitable Perl magic.
3217 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3218 the log message when discard triggers, was not working for the DATA ACL or
3219 for the non-SMTP ACL.
3221 78. Error message wording change in sieve.c.
3223 79. If smtp_accept_max_per_host was set, the number of connections could be
3224 restricted to fewer than expected, because the daemon was trying to set up
3225 a new connection before checking whether the processes handling previous
3226 connections had finished. The check for completed processes is now done
3227 earlier. On busy systems, this bug wouldn't be noticed because something
3228 else would have woken the daemon, and it would have reaped the completed
3231 80. If a message was submitted locally by a user whose login name contained one
3232 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3233 It caused a spool format error. I have fixed the spool reading code. A
3234 related problem was that the "from" clause in the Received: line became
3235 illegal because of the space(s). It is now covered by ${quote_local_part.
3237 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3240 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3242 83. Patch from Sieve maintainer to fix unterminated string problem in
3243 "vacation" handling.
3245 84. Some minor changes to the Linux configuration files to help with other
3246 OS variants using glibc.
3248 85. One more patch for Sieve to update vacation handling to latest spec.
3251 ----------------------------------------------------
3252 See the note above about the 4.44 and 4.50 releases.
3253 ----------------------------------------------------
3259 1. Change 4.43/35 introduced a bug that caused file counts to be
3260 incorrectly computed when quota_filecount was set in an appendfile
3263 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3264 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3267 3. Give more explanation in the error message when the command for a transport
3268 filter fails to execute.
3270 4. There are several places where Exim runs a non-Exim command in a
3271 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3272 was being done only for the command run by the queryprogram router. It is
3273 now done for all such subprocesses. The other cases are: ${run, transport
3274 filters, and the commands run by the lmtp and pipe transports.
3276 5. Some older OS have a limit of 256 on the maximum number of file
3277 descriptors. Exim was using setrlimit() to set 1000 as a large value
3278 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3279 systems. I've change it so that if it can't get 1000, it tries for 256.
3281 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3282 was an oversight, and furthermore, ever since the addition of extra
3283 controls (e.g. 4.43/32), the checks on when to allow different forms of
3284 "control" were broken. There should now be diagnostics for all cases when a
3285 control that does not make sense is encountered.
3287 7. $recipients is now available in the predata ACL (oversight).
3289 8. Tidy the search cache before the fork to do a delivery from a message
3290 received from the command line. Otherwise the child will trigger a lookup
3291 failure and thereby defer the delivery if it tries to use (for example) a
3292 cached ldap connection that the parent has called unbind on.
3294 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3295 of $address_data from the recipient verification was clobbered by the
3296 sender verification.
3298 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3299 was its contents. (It was OK if the option was not defined at all.)
3301 11. A "Completed" log line is now written for messages that are removed from
3302 the spool by the -Mrm option.
3304 12. $host_address is now set to the target address during the checking of
3305 ignore_target_hosts.
3307 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3308 being passed; this would have caused $sender_host_name to have been used if
3309 matching the list had actually called for a host name (not very likely,
3310 since this list is usually IP addresses). A host name is now passed as
3313 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3314 code that specifies a non-restarting handler (typically sigaction() in
3315 modern systems) in an attempt to fix a rare and obscure crash bug.
3317 15. Narrowed the window for a race in the daemon that could cause it to ignore
3318 SIGCHLD signals. This is not a major problem, because they are used only to
3319 wake it up if nothing else does.
3321 16. A malformed maildirsize file could cause Exim to calculate negative values
3322 for the mailbox size or file count. Odd effects could occur as a result.
3323 The maildirsize information is now recalculated if the size or filecount
3326 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3327 support for a long time. Removed HAVE_SYS_VFS_H.
3329 18. Updated exipick to current release from John Jetmore.
3331 19. Allow an empty sender to be matched against a lookup in an address list.
3332 Previously the only cases considered were a regular expression, or an
3335 20. Exim went into a mad DNS lookup loop when doing a callout where the
3336 host was specified on the transport, if the DNS lookup yielded more than
3339 21. The RFC2047 encoding function was originally intended for short strings
3340 such as real names; it was not keeping to the 75-character limit for
3341 encoded words that the RFC imposes. It now respects the limit, and
3342 generates multiple encoded words if necessary. To be on the safe side, I
3343 have increased the buffer size for the ${rfc2047: expansion operator from
3346 22. Failure to deliver a bounce message always caused it to be frozen, even if
3347 there was an errors_to setting on the router. The errors_to setting is now
3350 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3351 canonical form (fully expanded) before being placed in
3352 $sender_host_address.
3354 24. Updated eximstats to version 1.33
3356 25. Include certificate and key file names in error message when GnuTLS fails
3357 to set them up, because the GnuTLS error message doesn't include the name
3358 of the failing file when there is a problem reading it.
3360 26. Expand error message when OpenSSL has problems setting up cert/key files.
3363 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3364 (this can affect the format of dates).
3366 28. exim_tidydb, when checking for the continued existence of a message for
3367 which it has found a message-specific retry record, was not finding
3368 messages that were in split spool directories. Consequently, it was
3369 deleting retry records that should have stayed in existence.
3371 29. eximstats updated to version 1.35
3372 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3373 1.35 - bugfix such that pie charts by volume are generated correctly
3375 30. The SPA authentication driver was not abandoning authentication and moving
3376 on to the next authenticator when an expansion was forced to fail,
3377 contradicting the general specification for all authenticators. Instead it
3378 was generating a temporary error. It now behaves as specified.
3380 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3381 (the order specifies the preference for clients). The order is now AES256,
3382 AES128, 3DES, ARCFOUR128.
3384 31. Small patch to Sieve code - explicitly set From: when generating an
3387 32. Exim crashed if a remote delivery caused a very long error message to be
3388 recorded - for instance if somebody sent an entire SpamAssassin report back
3389 as a large number of 550 error lines. This bug was coincidentally fixed by
3390 increasing the size of one of Exim's internal buffers (big_buffer) that
3391 happened as part of the Exiscan merge. However, to be on the safe side, I
3392 have made the code more robust (and fixed the comments that describe what
3395 33. Some experimental protocols are using DNS PTR records for new purposes. The
3396 keys for these records are domain names, not reversed IP addresses. The
3397 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3398 leaves it alone. Component reversal etc. now happens only for IP addresses.
3401 34. The host_aton() function is supposed to be passed a string that is known
3402 to be a valid IP address. However, in the case of IPv6 addresses, it was
3403 not checking this. This is a hostage to fortune. Exim now panics and dies
3404 if the condition is not met. A case was found where this could be provoked
3405 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3406 components; fortuitously, this particular loophole had already been fixed
3407 by change 4.50/55 or 4.44/33 above.
3409 If there are any other similar loopholes, the new check in host_aton()
3410 itself should stop them being exploited. The report I received stated that
3411 data on the command line could provoke the exploit when Exim was running as
3412 exim, but did not say which command line option was involved. All I could
3413 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3414 running as the user.
3417 35. There was a buffer overflow vulnerability in the SPA authentication code
3418 (which came originally from the Samba project). I have added a test to the
3419 spa_base64_to_bits() function which I hope fixes it.
3422 36. The daemon start-up calls getloadavg() while still root for those OS that
3423 need the first call to be done as root, but it missed one case: when
3424 deliver_queue_load_max is set with deliver_drop_privilege. This is
3425 necessary for the benefit of the queue runner, because there is no re-exec
3426 when deliver_drop_privilege is set.
3428 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3429 was in use, was not putting the data itself into the right store pool;
3430 consequently, it could be overwritten for a subsequent message in the same
3431 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3434 38. Sometimes the final signoff response after QUIT could fail to get
3435 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3436 < 0 before doing a fflush(). This bug looks as though it goes back to the
3437 introduction of TLS in release 3.20, but "sometimes" must have been rare
3438 because the tests only now provoked it.
3444 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3445 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3446 Now it returns an integer. A number of calls were still expecting a T/F
3447 return. Fortuitously, in all cases, the tests worked in OK situations,
3448 which is the norm. However, things would have gone wrong on any write
3449 failures on the smtp file descriptor. This function is used when sending
3450 messages over SMTP and also when doing verify callouts.
3452 2. When Exim is called to do synchronous delivery of a locally submitted
3453 message (the -odf or -odi options), it no longer closes stderr before doing
3456 3. Implemented the mua_wrapper option.
3458 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3460 5. Implemented the functions header_remove(), header_testname(),
3461 header_add_at_position(), and receive_remove_recipient(), and exported them
3464 6. If an ACL "warn" statement specified the addition of headers, Exim already
3465 inserted X-ACL-Warn: at the start if there was no header name. However, it
3466 was not making this test for the second and subsequent header lines if
3467 there were newlines in the string. This meant that an invalid header could
3468 be inserted if Exim was badly configured.
3470 7. Allow an ACL "warn" statement to add header lines at the start or after all
3471 the Received: headers, as well as at the end.
3473 8. Added the rcpt_4xx retry error code.
3475 9. Added postmaster_mailfrom=xxx to callout verification option.
3477 10. Added mailfrom=xxxx to the callout verification option, for verify=
3480 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3481 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3483 12. Inserted some casts to stop certain compilers warning when using pointer
3484 differences as field lengths or precisions in printf-type calls (mostly
3485 affecting debugging statements).
3487 13. Added optional readline() support for -be (dynamically loaded).
3489 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3490 same clock tick as a message's arrival, so that its received time was the
3491 same as the "first fail" time on the retry record, and that message
3492 remained on the queue past the ultimate address timeout, every queue runner
3493 would try a delivery (because it was past the ultimate address timeout) but
3494 after another failure, the ultimate address timeout, which should have then
3495 bounced the address, did not kick in. This was a "< instead of <=" error;
3496 in most cases the first failure would have been in the next clock tick
3497 after the received time, and all would be well.
3499 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3500 being recognized when the domain list was tested by the match_domain
3501 condition in an expansion string.
3503 16. Added the ${str2b64: operator.
3505 17. Exim was always calling setrlimit() to set a large limit for the number of
3506 processes, without checking whether the existing limit was already
3507 adequate. (It did check for the limit on file descriptors.) Furthermore,
3508 errors from getrlimit() and setrlimit() were being ignored. Now they are
3509 logged to the main and panic logs, but Exim does carry on, to try to do its
3510 job under whatever limits there are.
3512 18. Imported PCRE 5.0.
3514 19. Trivial typo in log message " temporarily refused connection" (the leading
3517 20. If the log selector return_path_on_delivery was set and an address was
3518 redirected to /dev/null, the delivery process crashed because it assumed
3519 that a return path would always be set for a "successful" delivery. In this
3520 case, the whole delivery is bypassed as an optimization, and therefore no
3523 21. Internal re-arrangement: the function for sending a challenge and reading
3524 a response while authentication was assuming a zero-terminated challenge
3525 string. It's now changed to take a pointer and a length, to allow for
3526 binary data in such strings.
3528 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3530 23. Exim was not respecting finduser_retries when seeking the login of the
3531 uid under which it was called; it was always trying 10 times. (The default
3532 setting of finduser_retries is zero.) Also, it was sleeping after the final
3533 failure, which is pointless.
3535 24. Implemented tls_on_connect_ports.
3537 25. Implemented acl_smtp_predata.
3539 26. If the domain in control=submission is set empty, Exim assumes that the
3540 authenticated id is a complete email address when it generates From: or
3541 Sender: header lines.
3543 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3544 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3545 chown and chgrp in /bin and hostname in /usr/bin.
3547 28. Exim was keeping the "process log" file open after each use, just as it
3548 does for the main log. This opens the possibility of it remaining open for
3549 long periods when the USR1 signal hits a daemon. Occasional processlog
3550 errors were reported, that could have been caused by this. Anyway, it seems
3551 much more sensible not to leave this file open at all, so that is what now
3554 29. The long-running daemon process does not normally write to the log once it
3555 has entered its main loop, and it closes the log before doing so. This is
3556 so that log files can straightforwardly be renamed and moved. However,
3557 there are a couple of unusual error situations where the daemon does write
3558 log entries, and I had neglected to close the log afterwards.
3560 30. The text of an SMTP error response that was received during a remote
3561 delivery was being truncated at 512 bytes. This is too short for some of
3562 the long messages that one sometimes sees. I've increased the limit to
3565 31. It is now possible to make retry rules that apply only when a message has a
3566 specific sender, in particular, an empty sender.
3568 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3569 it possible to be selective about when SMTP synchronization is enforced.
3571 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3573 32. Implemented hosts_connection_nolog.
3575 33. Added an ACL for QUIT.
3577 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3580 35. Added mailbox_size and mailbox_filecount to appendfile.
3582 36. Added control = no_multiline_responses to ACLs.
3584 37. There was a bug in the logic of the code that waits for the clock to tick
3585 in the case where the clock went backwards by a substantial amount such
3586 that the microsecond fraction of "now" was more than the microsecond
3587 fraction of "then" (but the whole seconds number was less).
3589 38. Added support for the libradius Radius client library this is found on
3590 FreeBSD (previously only the radiusclient library was supported).
3596 1. When certain lookups returned multiple values in the form name=value, the
3597 quoting of the values was not always being done properly. Specifically:
3598 (a) If the value started with a double quote, but contained no whitespace,
3600 (b) If the value contained whitespace other than a space character (i.e.
3601 tabs or newlines or carriage returns) it was not quoted.
3602 This fix has been applied to the mysql and pgsql lookups by writing a
3603 separate quoting function and calling it from the lookup code. The fix
3604 should probably also be applied to nisplus, ibase and oracle lookups, but
3605 since I cannot test any of those, I have not disturbed their existing code.
3607 2. A hit in the callout cache for a specific address caused a log line with no
3608 reason for rejecting RCPT. Now it says "Previous (cached) callout
3609 verification failure".
3611 3. There was an off-by-one bug in the queryprogram router. An over-long
3612 return line was truncated at 256 instead of 255 characters, thereby
3613 overflowing its buffer with the terminating zero. As well as fixing this, I
3614 have increased the buffer size to 1024 (and made a note to document this).
3616 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3617 when Exim is waiting for an SMTP response from a remote server, Exim
3618 restarts its select() call on the socket, thereby resetting its timeout.
3619 This is not a problem when such interrupts are rare. Somebody set up a cron
3620 job to run exiwhat every 2 minutes, which is less than the normal select()
3621 timeout (5 or 10 minutes). This meant that the select() timeout never
3622 kicked in because it was always reset. I have fixed this by comparing the
3623 time when an interrupt arrives with the time at the start of the first call
3624 to select(). If more time than the timeout has elapsed, the interrupt is
3625 treated as a timeout.
3627 5. Some internal re-factoring in preparation for the addition of Sieve
3628 extensions (by MH). In particular, the "personal" test is moved to a
3629 separate function, and given an option for scanning Cc: and Bcc: (which is
3630 not set for Exim filters).
3632 6. When Exim created an email address using the login of the caller as the
3633 local part (e.g. when creating a From: or Sender: header line), it was not
3634 quoting the local part when it contained special characters such as @.
3636 7. Installed new OpenBSD configuration files.
3638 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3639 try to make them clearer.
3641 9. Callout options, other than the timeout value, were being ignored when
3642 verifying sender addresses in header lines. For example, when using
3644 verify = header_sender/callout=no_cache
3646 the cache was (incorrectly) being used.
3648 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3649 only the Cygwin environment.
3651 11. When return_path_on_delivery was set as a log selector, if different remote
3652 addresses in the same message used different return paths and parallel
3653 remote delivery occurred, the wrong values would sometimes be logged.
3654 (Whenever a remote delivery process finished, the return path value from
3655 the most recently started remote delivery process was logged.)
3657 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3658 lines when AUTH and/or TLS are in use. This is the "received protocol"
3659 field. Exim used to use "asmtp" for authenticated SMTP, without any
3660 indication (in the protocol name) for TLS use. Now it follows the RFC and
3661 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3662 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3663 names appear in log lines as well as in Received: header lines.
3665 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3666 extensions, and comparison tests, and to fix some bugs.
3668 14. Changes to the "personal" filter test:
3670 (1) The test was buggy in that it was just doing the equivalent of
3671 "contains" tests on header lines. For example, if a user's address was
3672 anne@some.where, the "personal" test would incorrectly be true for
3674 To: susanne@some.where
3676 This test is now done by extracting each address from the header in turn,
3677 and checking the entire address. Other tests that are part of "personal"
3678 are now done using regular expressions (for example, to check local parts
3679 of addresses in From: header lines).
3681 (2) The list of non-personal local parts in From: addresses has been
3682 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3683 taken from the Sieve specification recommendations.
3685 (3) If the message contains any header line starting with "List-" it is
3686 treated as non-personal.
3688 (4) The test for "circular" in the Subject: header line has been removed
3689 because it now seems ill-conceived.
3691 15. Minor typos in src/EDITME comments corrected.
3693 16. Installed latest exipick from John Jetmore.
3695 17. If headers_add on a router specified a text string that was too long for
3696 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3697 of string_sprintf() is now avoided.
3699 18. $message_body_size was not set (it was always zero) when running the DATA
3700 ACL and the local_scan() function.
3702 19. For the "mail" command in an Exim filter, no default was being set for
3703 the once_repeat time, causing a random time value to be used if "once" was
3704 specified. (If the value happened to be <= 0, no repeat happened.) The
3705 default is now 0s, meaning "never repeat". The "vacation" command was OK
3706 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3707 (I found it when inspecting the code).
3709 20. There is now an overall timeout for performing a callout verification. It
3710 defaults to 4 times the callout timeout, which applies to individual SMTP
3711 commands during the callout. The overall timeout applies when there is more
3712 than one host that can be tried. The timeout is checked before trying the
3713 next host. This prevents very long delays if there are a large number of
3714 hosts and all are timing out (e.g. when the network connections are timing
3715 out). The value of the overall timeout can be changed by specifying an
3716 additional sub-option for "callout", called "maxwait". For example:
3718 verify = sender/callout=5s,maxwait=20s
3720 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3721 to the end before writing, but this should make it even safer).
3723 22. Exim was forgetting that it had advertised PIPELINING for the second and
3724 subsequent messages on an SMTP connection. It was also not resetting its
3725 memory on STARTTLS and an internal HELO.
3727 23. When Exim logs an SMTP synchronization error within a session, it now
3728 records whether PIPELINING has been advertised or not.
3730 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3731 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3732 rather than long int.
3734 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3736 26. Added the never_mail option to autoreply.
3742 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3743 crash if the getsockname() call failed; this can happen if a connection is
3744 closed very soon after it is established. The problem was simply in the
3745 order in which certain operations were done, causing Exim to try to write
3746 to the SMTP stream before it had set up the file descriptor. The bug has
3747 been fixed by making things happen in the correct order.
3753 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3754 before the connection was closed, thus losing the rejection response.
3756 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3757 some early Solaris releases, but causes trouble in current releases where
3758 socklen_t is defined.
3760 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3763 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3766 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3767 that releasing the top of it at the end releases what was used for sub-
3768 expansions (unless the block got too big). However, discard this block if
3769 the first thing is a variable or header, so that we can use its block when
3770 it is dynamic (useful for very large $message_headers, for example).
3772 6. Lookups now cache *every* query, not just the most recent. A new, separate
3773 store pool is used for this. It can be recovered when all lookup caches are
3774 flushed. Lookups now release memory at the end of their result strings.
3775 This has involved some general refactoring of the lookup sources.
3777 7. Some code has been added to the store_xxx() functions to reduce the amount
3778 of flapping under certain conditions.
3780 8. log_incoming_interface used to affect only the <= reception log lines. Now
3781 it causes the local interface and port to be added to several more SMTP log
3782 lines, for example "SMTP connection from", and rejection lines.
3784 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3786 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3788 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3789 could be overwritten at the end of the current message (or the start of a
3790 new message if it was set in a HELO ACL). The value is now preserved for
3791 the duration of the SMTP connection.
3793 12. If a transport had a headers_rewrite setting, and a matching header line
3794 contained an unqualified address, that address was qualified, even if it
3795 did not match any rewriting rules. The underlying bug was that the values
3796 of the flags that permit the existence of unqualified sender and recipient
3797 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3798 non-local messages, and by -bnq for local messages) were not being
3799 preserved with the message after it was received.
3801 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3802 "next input=" as part of the text comprising the host identity instead of
3803 the correct text. The code was using the same buffer for two different
3804 strings. However, depending on which order the printing function evaluated
3805 its arguments, the bug did not always show up. Under Linux, for example, my
3806 test suite worked just fine.
3808 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3809 doesn't work with some older versions of Perl. It has been changed to "my",
3810 which in any case is probably the better facility to use.
3812 15. A really picky compiler found some instances of statements for creating
3813 error messages that either had too many or two few arguments for the format
3816 16. The size of the buffer for calls to the DNS resolver has been increased
3817 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3818 for addresses that have a lot of PTR records. This alleviates a problem; it
3819 does not fully solve it.
3821 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3822 buffer now truncates the list and logs the incident, which is the same
3823 action as happens when Exim is looking up a host name and its aliases.
3824 Previously in this situation something unpredictable would happen;
3825 sometimes it was "internal error: store_reset failed".
3827 18. If a server dropped the connection unexpectedly when an Exim client was
3828 using GnuTLS and trying to read a response, the client delivery process
3829 crashed while trying to generate an error log message.
3831 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3832 string, for example:
3834 warn message = H1: something\nH2: something
3836 the text was added as a single header line from Exim's point of view
3837 though it ended up OK in the delivered message. However, searching for the
3838 second and subsequent header lines using $h_h2: did not work. This has been
3839 fixed. Similarly, if a system filter added multiple headers in this way,
3840 the routers could not see them.
3842 20. Expanded the error message when iplsearch is called with an invalid key to
3843 suggest using net-iplsearch in a host list.
3845 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3846 ACLs are no longer actually imposed (and a message to that effect is
3849 22. If a "gecos" field in a passwd entry contained escaped characters, in
3850 particular, if it contained a \" sequence, Exim got it wrong when building
3851 a From: or a Sender: header from that name. A second bug also caused
3852 incorrect handling when an unquoted " was present following a character
3853 that needed quoting.
3855 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3856 was not being matched caselessly.
3858 24. Arranged for all hyphens in the exim.8 source to be escaped with
3861 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3862 properly. Recipient callout cache records were still being keyed to include
3863 the sender, even when use_sender was set false. This led to far more
3864 callouts that were necessary. The sender is no longer included in the key
3865 when use_sender is false.
3867 26. Added "control = submission" modifier to ACLs.
3869 27. Added the ${base62d: operator to decode base 62 numbers.
3871 28. dnsdb lookups can now access SRV records.
3873 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3874 the configuration file.
3876 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3877 (-v) mode. This makes the output for a verbose queue run more intelligible.
3879 31. Added a use_postmaster feature to recipient callouts.
3881 32. Added the $body_zerocount variable, containing the number of binary zero
3882 bytes in the message body.
3884 33. The time of last modification of the "new" subdirectory is now used as the
3885 "mailbox time last read" when there is a quota error for a maildir
3888 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3890 35. Added +ignore_unknown as a special item in host lists.
3892 36. Code for decoding IPv6 addresses in host lists is now included, even if
3893 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3894 address was recognized as an IP address, but was then not correctly decoded
3895 into binary, causing unexpected and incorrect effects when compared with
3902 1. Very minor rewording of debugging text in manualroute to say "list of
3903 hosts" instead of "hostlist".
3905 2. If verify=header_syntax was set, and a header line with an unqualified
3906 address (no domain) and a large number of spaces between the end of the
3907 name and the colon was received, the reception process suffered a buffer
3908 overflow, and (when I tested it) crashed. This was caused by some obsolete
3909 code that should have been removed. The fix is to remove it!
3911 3. When running in the test harness, delay a bit after writing a bounce
3912 message to get a bit more predictability in the log output.
3914 4. Added a call to search_tidyup() just before forking a reception process. In
3915 theory, someone could use a lookup in the expansion of smtp_accept_max_
3916 per_host which, without the tidyup, could leave open a database connection.
3918 5. Added the variables $recipient_data and $sender_data which get set from a
3919 lookup success in an ACL "recipients" or "senders" condition, or a router
3920 "senders" option, similar to $domain_data and $local_part_data.
3922 6. Moved the writing of debug_print from before to after the "senders" test
3925 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3926 problems for message scanning, either using a data ACL, or using
3927 local_scan() because the Received: header was not generated till after they
3928 were called (in order to set the time as the time of reception completion).
3929 I have revised the way this works. The header is now generated after the
3930 body is received, but before the ACL or local_scan() are called. After they
3931 are run, the timestamp in the header is updated.
3937 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3938 before starting a queue runner without re-exec. This happened only when
3939 deliver_drop_privilege was set or when the Exim user was set to root. The
3940 effect of the bug was that timeouts during subsequent deliveries caused
3941 crashes instead of being properly handled. The handler is now left at its
3942 default (and expected) setting.
3944 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3945 message, again when deliver_drop_privilege is set or Exim is run as root.
3946 The bug described in (1) was not present in this case, but the tidying up
3947 of the other signals was missing. I have made the two cases consistent.
3949 3. The ignore_target_hosts setting on a manualroute router was being ignored
3950 for hosts that were looked up using the /MX notation.
3952 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3955 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3956 operated on the sender address. After changing the $sender_address to <>
3957 for the sender address verify, Exim was re-instated it as the original
3958 (before rewriting) address, but remembering that it had rewritten it, so it
3959 wasn't rewriting it again. This bug also had the effect of breaking the
3960 sender address verification caching when the sender address was rewritten.
3962 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3963 This has been changed so that if the ip literal address matches
3964 ignore_target_hosts, the router declines.
3966 7. Added expansion conditions match_domain, match_address, and match_local_
3967 part (NOT match_host).
3969 8. The placeholder for the Received: header didn't have a length field set.
3971 9. Added code to Exim itself and to exim_lock to test for a specific race
3972 condition that could lead to file corruption when using MBX delivery. The
3973 issue is with the lockfile that is created in /tmp. If this file is removed
3974 after a process has opened it but before that process has acquired a lock,
3975 there is the potential for a second process to recreate the file and also
3976 acquire a lock. This could lead to two Exim processes writing to the file
3977 at the same time. The added code performs the same test as UW imapd; it
3978 checks after acquiring the lock that its file descriptor still refers to
3979 the same named file.
3981 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3982 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3985 11. Added the smtp_active_hostname option. If used, this will typically be made
3986 to depend on the incoming interface address. Because $interface_address is
3987 not set up until the daemon has forked a reception process, error responses
3988 that can happen earlier (such as "too many connections") no longer contain
3991 12. If an expansion in a condition on a "warn" statement fails because a lookup
3992 defers, the "warn" statement is abandoned, and the next ACL statement is
3993 processed. Previously this caused the whole ACL to be aborted.
3995 13. Added the iplsearch lookup type.
3997 14. Added ident_timeout as a log selector.
3999 15. Added tls_certificate_verified as a log selector.
4001 16. Added a global option tls_require_ciphers (compare the smtp transport
4002 option of the same name). This controls incoming TLS connections.
4004 17. I finally figured out how to make tls_require_ciphers do a similar thing
4005 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4006 before starting the TLS session.
4008 18. Tabs are now shown as \t in -bP output.
4010 19. If the log selector return_path_on_delivery was set, Exim crashed when
4011 bouncing a message because it had too many Received: header lines.
4013 20. If two routers both had headers_remove settings, and the first one included
4014 a superfluous trailing colon, the final name in the first list and the
4015 first name in the second list were incorrectly joined into one item (with a
4016 colon in the middle).
4022 1. Added -C and -D options to the exinext utility, mainly to make it easier
4023 to include in the automated testing, but these could be helpful when
4024 multiple configurations are in use.
4026 2. The exinext utility was not formatting the output nicely when there was
4027 an alternate port involved in the retry record key, nor when there was a
4028 message id as well (for retries that were specific to a specific message
4029 and a specific host). It was also confused by IPv6 addresses, because of
4030 the additional colons they contain. I have fixed the IPv4 problem, and
4031 patched it up to do a reasonable job for IPv6.
4033 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4034 delivery, the log line now contains "pipelined" if PIPELINING was used.
4036 4. An SMTP transport process used to panic and die if the bind() call to set
4037 an explicit outgoing interface failed. This has been changed; it is now
4038 treated in the same way as a connect() failure.
4040 5. A reference to $sender_host_name in the part of a conditional expansion
4041 that was being skipped was still causing a DNS lookup. This no longer
4044 6. The def: expansion condition was not recognizing references to header lines
4045 that used bh_ and bheader_.
4047 7. Added the _cache feature to named lists.
4049 8. The code for checking quota_filecount in the appendfile transport was
4050 allowing one more file than it should have been.
4052 9. For compatibility with Sendmail, the command line option
4060 and sets the incoming protocol and host name (for trusted callers). The
4061 host name and its colon can be omitted when only the protocol is to be set.
4062 Note the Exim already has two private options, -pd and -ps, that refer to
4063 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4064 "s", but I don't think that's a major issue.
4066 10. A number of refactoring changes to the code, none of which should affect
4069 (a) The number of logging options was getting close to filling up the
4070 32-bit word that was used as a bit map. I have split them into two classes:
4071 those that are passed in the argument to log_write(), and those that are
4072 only ever tested independently outside of that function. These are now in
4073 separate 32-bit words, so there is plenty of room for expansion again.
4074 There is no change in the user interface or the logging behaviour.
4076 (b) When building, for example, log lines, the code previously used a
4077 macro that called string_cat() twice, in order to add two strings. This is
4078 not really sufficiently general. Furthermore, there was one instance where
4079 it was actually wrong because one of the argument was used twice, and in
4080 one call a function was used. (As it happened, calling the function twice
4081 did not affect the overall behaviour.) The macro has been replaced by a
4082 function that can join an arbitrary number of extra strings onto a growing
4085 (c) The code for expansion conditions now uses a table and a binary chop
4086 instead of a serial search (which was left over from when there were very
4087 few conditions). Also, it now recognizes conditions like "pam" even when
4088 the relevant support is not compiled in: a suitably worded error message is
4089 given if an attempt is made to use such a condition.
4091 11. Added ${time_interval:xxxxx}.
4093 12. A bug was causing one of the ddress fields not to be passed back correctly
4094 from remote delivery subprocesses. The field in question was not being
4095 subsequently used, so this caused to problems in practice.
4097 13. Added new log selectors queue_time and deliver_time.
4099 14. Might have fixed a bug in maildirsizefile handling that threw up
4100 "unexpected character" debug warnings, and recalculated the data
4101 unnecessarily. In any case, I expanded the warning message to give more
4104 15. Added the message "Restricted characters in address" to the statements in
4105 the default ACL that block characters like @ and % in local parts.
4107 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4108 Three changes have been made:
4110 (a) There was a serious bug; a negative response to MAIL caused the whole
4111 recipient domain to be cached as invalid, thereby blocking all messages
4112 to all local parts at the same domain, from all senders. This bug has
4113 been fixed. The domain is no longer cached after a negative response to
4114 MAIL if the sender used is not empty.
4116 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4119 (c) A new callout option, "use_sender" has been added for people who want
4120 the modified behaviour.
4126 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4129 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4130 indeed breaks things for older releases.
4132 3. Added additional logging to the case where there is a problem reading data
4133 from a filter that is running in a subprocess using a pipe, in order to
4134 try to track down a specific problem.
4136 4. Testing facility fudge: when running in the test harness and attempting
4137 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4138 getting "No route to host". Convert this to a timeout.
4140 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4143 6. Some OS don't have socklen_t but use size_t instead. This affects the
4144 fifth argument of getsockopt() amongst other things. This is now
4145 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4146 can be set for individual OS. I have set it for SunOS5, OSF1, and
4147 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4148 some earlier ones do not.
4150 7. Change 4.30/15 was not doing the test caselessly.
4152 8. The standard form for an IPv6 address literal was being rejected by address
4153 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4154 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4155 this, as well as the form without the "IPv6" on the front (but only when
4156 address literals are enabled, of course).
4158 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4160 10. Exim crashed if a message with an empty sender address specified by -f
4161 encountered a router with an errors_to setting. This could be provoked only
4162 by a command such as
4166 where an empty string was supplied; "<>" did not hit this bug.
4168 11. Installed PCRE release 4.5.
4170 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4171 remained set. It is now erased.
4173 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4174 times from message ids (which are base 36 rather than the normal 62).
4176 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4177 were being counted as actual protocol errors, and logged if the log
4178 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4179 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4180 and DATA following a set of rejected RCPTs do not count as protocol errors.
4181 In other words, Exim assumes they were pipelined, though this may not
4182 actually be the case. Of course, in all cases the client gets an
4183 appropriate error code.
4185 15. If a lookup fails in an ACL condition, a message about the failure may
4186 be available; it is used if testing the ACL cannot continue, because most
4187 such messages specify what the cause of the deferral is. However, some
4188 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4189 that caused an old message to be retained and used if a later statement
4190 caused a defer, replacing the real cause of the deferral.
4192 16. If an IP address had so many PTR records that the DNS lookup buffer
4193 was not large enough to hold them, Exim could crash while trying to process
4194 the truncated data. It now detects and logs this case.
4196 17. Further to 4.21/58, another change has been made: if (and only if) the
4197 first line of a message (the first header line) ends with CRLF, a bare LF
4198 in a subsequent header line has a space inserted after it, so as not to
4199 terminate the header.
4201 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4202 unnecessarily, used atoi() instead of strtol(), and didn't check the
4203 termination when getting file sizes from file names by regex.
4205 19. Completely re-implemented the support for maildirsize files, in the light
4206 of a number of problems with the previous contributed implementation
4207 (4.30/29). In particular:
4209 . If the quota is zero, the maildirsize file is maintained, but no quota is
4212 . If the maildir directory does not exist, it is created before any attempt
4213 to write a maildirsize file.
4215 . The quota value in the file is just a cache; if the quota is changed in
4216 the transport, the new value overrides.
4218 . A regular expression is available for excluding directories from the
4221 20. The autoreply transport checks the characters in options that define the
4222 message's headers; it allows continued headers, but it was checking with
4223 isspace() after an embedded newline instead of explicitly looking for a
4226 21. If all the "regular" hosts to which an address was routed had passed their
4227 expiry times, and had not reached their retry times, the address was
4228 bounced, even if fallback hosts were defined. Now Exim should go on to try
4231 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4232 equivalent code in the SMTP transport. Some hosts send humungous responses
4233 to HELO/EHLO, more than 1024 it seems.
4235 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4236 gives compiler warnings in some environments. I've now done it "properly",
4239 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4240 (because of problems with the built-in one) was declared to return uschar *
4241 instead of char *, causing compiler failure.
4243 25. Fixed a file descriptor leak when processing alias/forward files.
4245 26. Fixed a minor format string issue in dbfn.c.
4247 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4249 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4250 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4253 29. When a sender address is verified, it is cached, to save repeating the test
4254 when there is more than one recipient in a message. However, when the
4255 verification involves a callout, it is possible for different callout
4256 options to be set for different recipients. It is too complicated to keep
4257 track of this in the cache, so now Exim always runs a verification when a
4258 callout is required, relying on the callout cache for the optimization.
4259 The overhead is duplication of the address routing, but this should not be
4262 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4263 to be verified with callout=postmaster, and the main callout worked but the
4264 postmaster check failed, the verification correctly failed. However, if a
4265 subsequent RCPT command asked for sender verification *without* the
4266 postmaster check, incorrect caching caused this verification also to fail,
4269 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4270 it was not caching the DNS options (qualify_single, search_parents) that
4271 were used when the lookup failed. A subsequent lookup with different
4272 options therefore always gave the same answer, though there were cases
4273 where it should not have. (Example: a "domains = !$mx_any" option on a
4274 dnslookup router: the "domains" option is always processed without any
4275 widening, but the router might have qualify_single set.) Now Exim uses the
4276 cached value only when the same options are set.
4278 32. Added John Jetmore's "exipick" utility to the distribution.
4280 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4281 than a timeout (e.g. a certificate is required, and is not provided), an
4282 Exim server now closes the connection immediately. Previously it waited for
4283 the client to close - but if the client is SSL, it seems that they each
4284 wait for each other, leading to a delay before one of them times out.
4286 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4287 maintained 0.8.x compatibility because I don't think many are using it, and
4288 it is clearly obsolete.
4290 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4293 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4294 client certificate was expired. A simple patch fixes this, though I don't
4295 understand the full logic of why the verify callback is called multiple
4298 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4299 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4300 which causes problems with some clients (such as the Certicom SSL Plus
4301 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4302 disables the coutermeasure allowing Eudora to connect."
4304 38. Exim was not checking that a write() to a log file succeeded. This could
4305 lead to Bad Things if a log got too big, in particular if it hit a file
4306 size limit. Exim now panics and dies if it cannot write to a log file, just
4307 as it does if it cannot open a log file.
4309 39. Modified OS/Makefile-Linux so that it now contains
4311 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4313 The two -D definitions ensure that Exim is compiled with large file
4314 support, which makes it possible to handle log files that are bigger than
4317 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4318 instance) a domain was checked against a named list that involved a lookup,
4319 causing $domain_data to be set, then another domain was checked against the
4320 same list, then the first domain was re-checked, the value of $domain_data
4321 after the final check could be wrong. In particular, if the second check
4322 failed, it could be set empty. This bug probably also applied to
4325 41. The strip_trailing_dot option was not being applied to the address given
4326 with the -f command-line option.
4328 42. The code for reading a message's header from the spool was incrementing
4329 $received_count, but never initializing it. This meant that the value was
4330 incorrect (doubled) while delivering a message in the same process in which
4331 it was received. In the most common configuration of Exim, this never
4332 happens - a fresh exec is done - but it can happen when
4333 deliver_drop_privilege is set.
4335 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4336 it now includes up to 150 characters of the unexpected data in the log
4339 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4340 and building data strings. The size of both of these buffers was 10 000
4341 bytes - far larger than anybody would *ever* want, thought I. Needless to
4342 say, somebody hit the limit. I have increased the maximum line length to
4343 20 000 and the maximum data length of concatenated lines to 100 000. I have
4344 also fixed two bugs, because there was no checking on these buffers. Tsk,
4345 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4346 buffer is too small.
4348 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4349 lsearch lookups. Now it does.
4351 46. When parsing a route_list item in a manualroute router, a fixed-length
4352 buffer was used for the list of hosts. I made this 1024 bytes long,
4353 thinking that nobody would ever have a list of hosts that long. Wrong.
4354 Somebody had a whole pile of complicated expansion conditions, and the
4355 string was silently truncated, leading to an expansion error. It turns out
4356 that it is easier to change to an unlimited length (owing to other changes
4357 that have happened since this code was originally written) than to build
4358 structure for giving a limitation error. The length of the item that
4359 expands into the list of hosts is now unlimited.
4361 47. The lsearch lookup could not handle data where the length of text line was
4362 more than 4095 characters. Such lines were truncated, leading to shortened
4363 data being returned. It should now handle lines of any length.
4365 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4366 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4369 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4370 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4371 no longer happens in comment lines. A list of replacements is now placed
4372 at the head of all of the source files, except those whose only change is
4373 to replace PERL_COMMAND in the very first #! line.
4375 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4376 messages on the queue, with a bottom-up merge sort, using code contributed
4377 by Michael Haardt. This should make operations like -bp somewhat faster on
4378 large queues. It won't affect queue runners, except when queue_run_in_order
4381 51. Installed eximstats 1.31 in the distribution.
4383 52. Added support for SRV lookups to the dnslookup router.
4385 53. If an ACL referred to $message_body or $message_body_end, the value was not
4386 reset for any messages that followed in the same SMTP session.
4388 54. The store-handling optimization for building very long strings was not
4389 differentiating between the different store pools. I don't think this
4390 actually made any difference in practice, but I've tidied it.
4392 55. While running the routers to verify a sender address, $sender_address
4393 was still set to the sender address. This is wrong, because when routing to
4394 send a bounce to the sender, it would be empty. Therefore, I have changed
4395 it so that, while verifying a sender address, $sender_address is set to <>.
4396 (There is no change to what happens when verifying a recipient address.)
4398 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4399 target A or AAAA records (if not already returned) without resetting the
4400 qualify_single or search_parents options of the DNS resolver. These are
4401 inappropriate in this case because the targets of MX and SRV records must
4402 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4403 target that, when qualified, matched something in the local domain. These
4404 two options are now turned off when doing these lookups.
4406 57. It seems that at least some releases of Reiserfs (which does not have the
4407 concept of a fixed number of inodes) returns zero and not -1 for the
4408 number of available inodes. This interacted badly with check_spool_inodes,
4409 which assumed that -1 was the "no such thing" setting. What I have done is
4410 to check that the total number of inodes is greater than zero before doing
4411 the test of how many are available.
4413 58. When a "warn" ACL statement has a log_message modifier, the message is
4414 remembered, and not repeated. This is to avoid a lot of repetition when a
4415 message has many recipients that cause the same warning to be written.
4416 Howewer, Exim was preserving the list of already written lines for an
4417 entire SMTP session, which doesn't seem right. The memory is now reset if a
4418 new message is started.
4420 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4421 debugging output unless log_rewrite was also set.
4423 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4424 of (int)(handle) when we know that handle contains (void *)(-1).
4426 61. The Exim daemon panic-logs an error return when it closes the incoming
4427 connection. However "connection reset by peer" seems to be common, and
4428 isn't really an error worthy of noting specially, so that particular error
4431 62. When Exim is trying to find all the local interfaces, it used to panic and
4432 die if the ioctl to get the interface flags failed. However, it seems that
4433 on at least one OS (Solaris 9) it is possible to have an interface that is
4434 included in the list of interfaces, but for which you get a failure error
4435 for this call. This happens when the interface is not "plumbed" into a
4436 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4437 failure of the "get flags" call assumes that the interface is down.
4439 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4440 makes life easier for people who are doing arithmetic on fields extracted
4441 from dates, where you often get leading zeros that should not be
4442 interpreted as octal.
4444 64. Added qualify_domain to the redirect router, to override the global
4447 65. If a pathologically long header line contained very many addresses (the
4448 report of this problem mentioned 10 000) and each of them was rewritten,
4449 Exim could use up a very large amount of memory. (It kept on making new
4450 copies of the header line as it rewrote, and never released the old ones.)
4451 At the expense of a bit more processing, the header rewriting function has
4452 been changed so that it no longer eats memory in this way.
4454 66. The generation of the Received: header has been moved from the time that a
4455 message starts to be received, to the time that it finishes. The timestamp
4456 in the Received: header should now be very close to that of the <= log
4457 line. There are two side-effects of this change:
4459 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4460 logged header lines no longer include the local Received: line, because
4461 it has not yet been created. The same applies to a copy of the message
4462 that is returned to a non-SMTP sender when a message is rejected.
4464 (b) When a filter file is tested using -bf, no additional Received: header
4465 is added to the test message. After some thought, I decided that this
4468 This change does not affect the value of $received_for. It is still set
4469 after address rewriting, but before local_scan() is called.
4471 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4473 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4474 gave an unhelpful panic error message, and a defer error. I have managed to
4475 change this behaviour so that it now rejects any supplied certificate,
4476 which seems right, as the list of acceptable certificates is empty.
4478 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4479 gave an unhelpful defer error. I have not managed to make this reject any
4480 supplied certificates, but the error message it gives is "no certificate
4481 supplied", which is not helpful.
4483 70. exigrep's output now also includes lines that are not associated with any
4484 message, but which match the given pattern. Implemented by a patch from
4485 Martin Sluka, which also tidied up the Perl a bit.
4487 71. Recipient callout verification, like sender verification, was using <> in
4488 the MAIL FROM command. This isn't really the right thing, since the actual
4489 sender may affect whether the remote host accepts the recipient or not. I
4490 have changed it to use the actual sender in the callout; this means that
4491 the cache record is now keyed on a recipient/sender pair, not just the
4492 recipient address. There doesn't seem to be a real danger of callout loops,
4493 since a callout by the remote host to check the sender would use <>.
4494 [SEE ABOVE: changed after hitting problems.]
4496 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4497 temporary errors. However, in the case of such a code being given after
4498 the end of a data transmission (i.e. after ".") Exim was failing to write
4499 a retry record for the message. (Yes, there was some broken host that was
4500 actually sending 8xx at this point.)
4502 73. An unknown lookup type in a host list could cause Exim to panic-die when
4503 the list was checked. (An example that provoked this was putting <; in the
4504 middle of a list instead of at the start.) If this happened during a DATA
4505 ACL check, a -D file could be left lying around. This kind of configuration
4506 error no longer causes Exim to die; instead it causes a defer errror. The
4507 incident is still logged to the main and panic logs.
4509 74. Buglet left over from Exim 3 conversion. The message "too many messages
4510 in one connection" was written to the rejectlog but not the mainlog, except
4511 when address rewriting (yes!) was being logged.
4513 75. Added write_rejectlog option.
4515 76. When a system filter was run not as root (that is, when system_filter_user
4516 was set), the values of the $n variables were not being returned to the
4517 main process; thus, they were not subsequently available in the $sn
4520 77. Added +return_path_on_delivery log selector.
4522 78. A connection timeout was being treated differently from recipients deferred
4523 when testing hosts_max_try with a message that was older than the host's
4524 retry timeout. (The host should not be counted, thus allowing all hosts to
4525 be tried at least once before bouncing.) This may have been the cause of an
4526 occasionally reported bug whereby a message would remain on the queue
4527 longer than the retry timeout, but would be bounced if a delivery was
4528 forced. I say "may" because I never totally pinned down the problem;
4529 setting up timeout/retry tests is difficult. See also the next item.
4531 79. The ultimate address timeout was not being applied to errors that involved
4532 a combination of host plus message (for example, a timeout on a MAIL
4533 command). When an address resolved to a number of possible hosts, and they
4534 were not all tried for each delivery (e.g. because of hosts_max_try), a
4535 message could remain on the queue longer than the retry timeout.
4537 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4540 81. Fixed an obscure SMTP outgoing bug which required at least the following
4541 conditions: (a) there was another message waiting for the same server;
4542 (b) the server returned 5xx to all RCPT commands in the first message so
4543 that the message was not completed; (c) the server dropped the connection
4544 or gave a negative response to the RSET that Exim sends to abort the
4545 transaction. The observed case was a dropped connection after DATA that had
4546 been sent in pipelining mode. That is, the server had advertised PIPELINING
4547 but was not implementing it correctly. The effect of the bug was incorrect
4548 behaviour, such as trying another host, and this could lead to a crash.
4554 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4555 and daemon.c were passed as pointers to ints; they should have been
4556 pointers to socklen_t variables (which are typically unsigned ints).
4558 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4561 3. Fixed a really odd bug that affected only the testing scheme; patching a
4562 certain fixed string in the binary changed the value of another string that
4563 happened to be identical to the end of the original first string.
4565 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4566 name", it returns that address as the IP address. On some operating
4567 systems (e.g. Solaris), it also passes back the IP address string as the
4568 "host name". However, on others (e.g. Linux), it passes back an empty
4569 string. Exim wasn't checking for this, and was changing the host name to an
4570 empty string, assuming it had been canonicized.
4572 5. Although rare, it is permitted to have more than one PTR record for a given
4573 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4574 all the names associated with an address, because they do in Solaris.
4575 However, it seems that they do not in Linux for data that comes from the
4576 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4577 I found this out when I moved to a new Linux workstation and tried to run
4578 the Exim test suite.
4580 To get round this problem I have changed the code so that it now does its
4581 own call to the DNS to look up PTR records when searching for a host name.
4582 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4583 addresses that are only in /etc/hosts are still found.
4585 This behaviour is, however, controlled by an option called host_lookup_
4586 order, which defaults to "bydns:byaddr". If people want to use the other
4587 order, or indeed, just use one or the other means of lookup, they can
4588 specify it in this variable.
4590 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4591 some operating systems, this comes back from gethostbyaddr() as an empty
4592 string, and this is what Exim used to test for. However, it seems that in
4593 other systems, "." is yielded. Exim now tests for this case too.
4595 7. The values of check_spool_space and check_log_space are now held internally
4596 as a number of kilobytes instead of an absolute number of bytes. If a
4597 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4598 kilobyte. This means that much larger values can be stored.
4600 8. Exim monitor: an attempt to get the action menu when not actually pointing
4601 at a message produces an empty menu entitled "No message selected". This
4602 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4603 no entries in it ("Shell widget menu has zero width and/or height"). So I
4604 have added a single, blank menu entry in this case.
4606 9. Added ${quote_local_part.
4608 10. MIME decoding is now applied to the contents of Subject: header lines when
4611 11. Now that a reference to $sender_host_address automatically causes a reverse
4612 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4613 host lookup before query-style lookups in lists that might use this
4614 variable. This has therefore been abolished, and the "net-" prefix is no
4615 longer necessary for query-style lookups.
4617 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4618 have been a typo for LFLAGS, so it has been changed.
4620 13. The install script calls Exim with "-C /dev/null" in order to find the
4621 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4622 to be output. Howeve, since Exim outputs its version number before the
4623 error, it didn't break the script. It just looked ugly. I fixed this by
4624 always allowing "-C /dev/null" if the caller is root.
4626 14. Ignore overlarge ACL variable number when reading spool file - insurance
4627 against a later release with more variables having written the file.
4629 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4630 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4631 this, as well as the form without the "IPv6" on the front.
4633 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4634 OS/Makefile-Darwin file.
4636 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4637 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4639 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4640 variable. However, some OS use ints for inodes. I've added cast to long int
4641 to get rid of the compiler warning.
4643 19. I had forgotten to lock out "/../" in configuration file names when
4644 ALT_CONFIG_PREFIX was set.
4646 20. Routers used for verification do not need to specify transports. However,
4647 if such a router generated a host list, and callout was configured, Exim
4648 crashed, because it could not find a port number from the (non-existent)
4649 transport. It now assumes port 25 in this circumstance.
4651 21. Added the -t option to exigrep.
4653 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4654 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4655 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4656 them is set, LOOKUP_LSEARCH is forced.
4658 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4660 24. Added sender and host information to the "rejected by local_scan()" log
4661 line; previously there was no indication of these.
4663 25. Added .include_if_exists.
4665 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4666 Linux. It turns out that not all file systems support this. Apparently some
4667 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4668 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4669 not supported on the file descriptor, is now ignored when Exim is trying to
4670 sync a directory. This applies only to Linux.
4672 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4674 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4675 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4678 29. Added support for maildirsize files from supplied patch (modified a bit).
4680 30. The use of :fail: followed by an empty string could lead Exim to respond to
4681 sender verification failures with (e.g.):
4683 550 Verification failed for <xxx>
4684 550 Sender verify failed
4686 where the first response line was missing the '-' that indicates it is not
4687 the final line of the response.
4689 31. The loop for finding the name of the user that called Exim had a hardwired
4690 limit of 10; it now uses the value of finduser_retries, which is used for
4691 all other user lookups.
4693 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4696 33. Exim was neglecting to zero errno before one call of strtol() when
4697 expanding a string and expecting an integer value. On some systems this
4698 resulted in spurious "integer overflow" errors. Also, it was casting the
4699 result into an int without checking.
4701 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4702 did not work. The code looks as if it has *never* worked, though it appears
4703 to have been documented since at least releast 1.62. I have made it work.
4705 35. The "timeout_DNS" error in retry rules, also documented since at least
4706 1.62, also never worked. As it isn't clear exactly what this means, and
4707 clearly it isn't a major issue, I have abolished the feature by treating it
4708 as "timeout", and writing a warning to the main and panic logs.
4710 36. The display of retry rules for -brt wasn't always showing the error code
4713 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4714 timeout_connect_A, timeout_connect_MX.
4716 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4717 to the empty sender.
4719 39. The daemon was not analysing the content of -oX till after it had closed
4720 stderr and disconnected from the controlling terminal. This meant that any
4721 syntax errors were only noted on the panic log, and the return code from
4722 the command was 0. By re-arranging the code a little, I've made the
4723 decoding happen first, so such errors now appear on stderr, and the return
4724 code is 1. However, the actual setting up of the sockets still happens in
4725 the disconnected process, so errors there are still only recorded on the
4728 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4729 connections (as happens on some IP stacks) was logged at start up time as
4730 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4731 it from "IPv6 and IPv4", which means that two separate sockets are being
4734 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4735 says whether AF_INET or AF_INET6 was passed as an argument.
4737 42. Exiwhat output was messed up when time zones were included in log
4740 43. Exiwhat now gives more information about the daemon's listening ports,
4741 and whether -tls-on-connect was used.
4743 44. The "port" option of the smtp transport is now expanded.
4745 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4746 silently ignored. Now an error message is written to the main and panic
4749 46. There's a new ACL modifier called "logwrite" which writes to a log file
4750 as soon as it is encountered.
4752 47. Added $local_user_uid and $local_user_gid at routing time.
4754 48. Exim crashed when trying to verify a sender address that was being
4757 49. Exim was recognizing only a space character after ".include". It now also
4758 recognizes a tab character.
4760 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4761 extracting the relevant information from the specification. The man page no
4762 longer contains scrambled data for the -d option, and I've added a section
4763 at the front about calling Exim under different names.
4765 51. Added "extra_headers" argument to the "mail" command in filter files.
4767 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4770 53. Installed eximstats 1.29.
4772 54. Added transport_filter_timeout as a generic transport option.
4774 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4775 Cc: header lines. This was required by RFC 822, but it not required by RFC
4778 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4779 incoming messages that did not have them. Now it does so only if the
4780 message originates locally, that is, if there is no associated remote host
4781 address. When Resent- header lines are present, this applies to the Resent-
4782 lines rather than the non-Resent- lines.
4784 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4785 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4787 58. Messages for configuration errors now include the name of the main
4788 configuration file - useful now that there may be more than one file in a
4789 list (.included file names were always shown).
4791 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4792 for those rare installations that do not start the daemon as root or run it
4793 setuid root. I've cut out the call to initgroups() if the daemon is not
4796 60. The Exim user and group can now be bound into the binary as text strings
4797 that are looked up at the start of Exim's processing.
4799 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4801 62. Added $mailstore_basename variable.
4803 63. Installed patch to sieve.c from Michael Haardt.
4805 64. When Exim failed to open the panic log after failing to open the main log,
4806 the original message it was trying to log was written to stderr and debug
4807 output, but if they were not available (the usual case in production), it
4808 was lost. Now it is written to syslog before the two lines that record the
4809 failures to open the logs.
4811 65. Users' Exim filters run in subprocesses under the user's uid. It is
4812 possible for a "deliver" command or an alias in a "personal" command to
4813 provoke an address rewrite. If logging of address rewriting is configured,
4814 this fails because the process is not running as root or exim. There may be
4815 a better way of dealing with this, but for the moment (because 4.30 needs
4816 to be released), I have disabled address rewrite logging when running a
4817 filter in a non-root, non-exim process.
4823 1. The buildconfig auxiliary program wasn't quoting the value set for
4824 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4825 not defined. This bug was masked in 4.22 by the effect that was fixed in
4828 2. Some messages that were rejected after a message id was allocated were
4829 shown as "incomplete" by exigrep. It no longer does this for messages that
4830 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4832 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4833 have allow_domain_literals set, the ID did not get logged in the <= line.
4834 Domain literals are now always recognized in Message-ID: header lines.
4836 4. The first argument for a ${extract expansion item is the key name or field
4837 number. Leading and trailing spaces in this item were not being ignored,
4838 causing some misleading effects.
4840 5. When deliver_drop_privilege was set, single queue runner processes started
4841 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4842 same command in the process it spins off) were not dropping privilege.
4844 6. When the daemon running as "exim" started a queue runner, it always
4845 re-executed Exim in the spun-off process. This is a waste of effort when
4846 deliver_drop_privilege is set. The new process now just calls the
4847 queue-runner function directly.
4853 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4856 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4857 ignored. Though the use of -oP was forcing the writing of a pid file, it
4858 was always written to the default place.
4860 3. If the message "no IP address found for host xxxx" is generated during
4861 incoming verification, it is now followed by identification of the incoming
4862 connection (so you can more easily find what provoked it).
4864 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4866 5. Added some features to "harden" Exim a bit more against certain attacks:
4868 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4869 be put in Local/Makefile. This is like the never_users runtime option,
4870 but it cannot be overridden. The default setting is "root".
4872 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4873 prefix string with which any file named in a -C command line option
4876 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4877 is retained for -C and -D only if the caller of Exim is root. Without
4878 it, the exim user may also use -C and -D and retain privilege.
4880 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4881 command line option is disabled.
4883 6. Macro names set by the -D option must start with an upper case letter, just
4884 like macro names defined in the configuration file.
4886 7. Added "dereference=" facility to LDAP.
4888 8. Two instances of the typo "uknown" in the source files are fixed.
4890 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4891 the Configure-Makefile script screwed up while processing it.
4893 10. Incorporated PCRE 4.4.
4895 11. The SMTP synchronization check was not operating right at the start of an
4896 SMTP session. For example, it could not catch a HELO sent before the client
4897 waited for the greeting. There is now a check for outstanding input at the
4898 point when the greeting is written. Because of the duplex, asynchronous
4899 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4900 way, but not yet received, when the check is performed.
4902 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4903 on TCP/IP sockets, because this apparently causes some broken clients to
4906 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4907 unchanged) from the Cygwin maintainer.
4909 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4910 was supported instead of testing for mbx. Effectively a typo.
4912 15. The spa authenticator server code was not checking that the input it
4913 received was valid base64.
4915 16. The debug output line for the "set" modifier in ACLs was not showing the
4916 name of the variable that was being set.
4918 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4920 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4921 lookup on its own. Something else was needed to trigger the lookup. For
4922 example, a match in host_lookup or the need for a host name in a host list.
4923 Now, if $sender_host_name is referenced and the host name has not yet been
4924 looked up, a lookup is performed. If the lookup fails, the variable remains
4925 empty, and $host_lookup_failed is set to "1".
4927 19. Added "eqi" as a case-independent comparison operator.
4929 20. The saslauthd authentication condition could segfault if neither service
4930 nor realm was specified.
4932 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4933 error message that was logged was misleading, and incoming SMTP
4934 connections were dropped. The message is now more accurate, and temporary
4935 errors are given to SMTP connections.
4937 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4938 (except RSET) with a 421 error, until QUIT is received. However, it was
4939 failing to send a response to QUIT.
4941 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4942 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4943 ACL is now run after the helo_try_verify_hosts code.
4945 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4946 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4947 case-independent, so other case variants are also recognized). Apparently
4948 some systems use these upper case variants.
4950 25. If more than two messages were waiting for the same host, and a transport
4951 filter was specified for the transport, Exim sent two messages over the
4952 same TCP/IP connection, and then failed with "socket operation on non-
4953 socket" when it tried to send the third.
4955 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4957 27. The extern definition of crypt16() in expand.c was not being excluded when
4958 the OS had its own crypt16() function.
4960 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4961 as a preferred synonym for return_size_limit, both as an option and as an
4964 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4966 30. Changed the default configuration ACL to relax the local part checking rule
4967 for addresses that are not in any local domains. For these addresses,
4968 slashes and pipe symbols are allowed within local parts, but the sequence
4969 /../ is explicitly forbidden.
4971 31. SPA server authentication was not clearing the challenge buffer before
4974 32. log_message in a "warn" ACL statement was writing to the reject log as
4975 well as to the main log, which contradicts the documentation and doesn't
4976 seem right (because no rejection is happening). So I have stopped it.
4978 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4979 However, I am unable to do any testing of this.
4981 34. Fixed an infelicity in the appendfile transport. When checking directories
4982 for a mailbox, to see if any needed to be created, it was accidentally
4983 using path names with one or more superfluous leading slashes; tracing
4984 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4986 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4987 contents are added to the log line that is written for every discarded
4988 recipient. (Previously a log_message setting was ignored.)
4990 36. The ${quote: operator now quotes the string if it is empty.
4992 37. The install script runs exim in order to find its version number. If for
4993 some reason other than non-existence or emptiness, which it checks, it
4994 could not run './exim', it was installing it with an empty version number,
4995 i.e. as "exim-". This error state is now caught, and the installation is
4998 38. An argument was missing from the function that creates an error message
4999 when Exim fails to connect to the socket for saslauthd authentication.
5000 This could cause Exim to crash, or give a corrupted message.
5002 39. Added isip, isip4, and isip6 to ${if conditions.
5004 40. The ACL variables $acl_xx are now saved with the message, and can be
5005 accessed later in routers, transports, and filters.
5007 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5008 strings in the file are not string-expanded.
5010 42. If a MAIL command specified a SIZE value that was too large to fit into an
5011 int variable, the check against message_size_limit failed. Such values are
5012 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5013 day this will have to be increased, but I don't think I want to be around
5014 when emails are that large.
5021 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5022 iconv() is not standard in FreeBSD.
5024 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5025 IPv6 enabled. The observed symptom was a segmentation fault on return from
5026 the function os_common_find_running_interfaces() in src/os.c.
5028 3. In the check_special_case() function in daemon.c I had used "errno" as an
5029 argument name, which causes warnings on some systems. This was basically a
5030 typo, since it was named "eno" in the comments!
5032 4. The code that waits for the clock to tick (at a resolution of some fraction
5033 of a second) so as to ensure message-id uniqueness was always waiting for
5034 at least one whole tick, when it could have waited for less. [This is
5035 almost certainly not relevant at current processor speeds, where it is
5036 unlikely to ever wait at all. But we try to future-proof.]
5038 5. The function that sleeps for a time interval that includes fractions of a
5039 second contained a race. It did not block SIGALRM between setting the
5040 timer, and suspending (a couple of lines later). If the interval was short
5041 and the sigsuspend() was delayed until after it had expired, the suspension
5042 never ended. On busy systems this could lead to processes getting stuck for
5045 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5046 process, before it forks any delivery processes. The open lookup caching
5047 mechanism meant that the open file or database connection was passed into
5048 the delivery process. The problem was that delivery processes always tidy
5049 up cached lookup data. This could cause a problem for the next delivery
5050 process started by the queue runner, because the external queue runner
5051 process does not know about the closure. So the next delivery process
5052 still has data in the lookup cache. In the case of a file lookup, there was
5053 no problem because closing a file descriptor in a subprocess doesn't affect
5054 the parent. However, if the lookup was caching a connection to a database,
5055 the connection was closed, and the second delivery process was likely to
5056 see errors such as "PGSQL: query failed: server closed the connection
5057 unexpectedly". The problem has been fixed by closing all cached lookups
5058 in a queue runner before running a delivery process.
5060 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5061 seem to have the "const" qualifier which it has on other OS. I've
5064 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5065 *of the same type* (client or server) with the same public name that an
5066 error should be diagnosed.
5068 9. When Exim looked up a host name for an IP address, but failed to find the
5069 original IP address when looking up the host name (a safety check), it
5070 output the message "<ip address> does not match any IP for NULL", which was
5071 confusing, to say the least. The bug was that the host name should have
5072 appeared instead of "NULL".
5074 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5075 user that is built into the binary, and the -C or -D options is used, root
5076 privilege is dropped before the configuration file is read. In addition,
5077 logging is switched to stderr instead of the normal log files. If the
5078 configuration then re-defines the Exim user, the unprivileged environment
5079 is probably not what is expected, so Exim logs a panic warning message (but
5082 However, if deliver_drop_privilege is set, the unprivileged state may well
5083 be exactly what is intended, so the warning has been cut out in that case,
5084 and Exim is allowed to try to write to its normal log files.
5090 1. smtp_return_error_details was not giving details for temporary sender
5091 or receiver verification errors.
5093 2. Diagnose a configuration error if two authenticators have the same public
5096 3. Exim used not to create the message log file for a message until the first
5097 delivery attempt. This could be confusing when incoming messages were held
5098 for policy or load reasons. The message log file is now created at the time
5099 the message is received, and an initial "Received" line is written to it.
5101 4. The automatically generated man page for command line options had a minor
5102 bug that caused no ill effects; however, a more serious problem was that
5103 the procedure for building the man page automatically didn't always
5104 operate. Consequently, release 4.20 contains an out-of-date version. This
5105 shouldn't happen again.
5107 5. When building Exim with embedded Perl support, the script that builds the
5108 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5109 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5111 6. The freeze_tell option was not being used for messages that were frozen on
5112 arrival, either by an ACL or by local_scan().
5114 7. Added the smtp_incomplete_transaction log selector.
5116 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5117 was accepting AUTH without a new EHLO.
5119 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5120 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5123 10. Logging of TCP/IP connections (when configured) now happens in the main
5124 daemon process instead of the child process, so that the TCP/IP connection
5125 count is more accurate (but it can never be perfect).
5127 11. The use of "drop" in a nested ACL was not being handled correctly in the
5128 outer ACL. Now, if condition failure induced by the nested "drop" causes
5129 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5130 or "require"), the connection is dropped.
5132 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5133 that yield "discard" can now be used with an "accept" or a "discard" verb,
5134 but an error is generated for any others (because I can't see a useful way
5135 to define what should happen).
5137 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5138 are now processed in the same way as lines in the Exim configuration file.
5139 In particular, continuation lines are supported.
5141 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5143 15. Added -ti meaning -t -i.
5145 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5146 domains, and warn by logging if others are found.
5148 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5149 returned from GIFCONF when Exim is trying to find the list of interfaces on
5150 a host. The code in os.c has been modified to copy each ifreq to an aligned
5151 structure in all cases.
5153 Also, in some cases, the returned ifreq's were being copied to a 'struct
5154 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5155 means the last couple of bytes of an IPv6 address could be chopped if the
5156 ifreq contained only a normal sockaddr (14 bytes storage).
5158 18. Named domain lists were not supported in the hosts_treat_as_local option.
5159 An entry such as +xxxx was not recognized, and was treated as a literal
5162 19. Ensure that header lines added by a DATA ACL are included in the reject log
5163 if the ACL subsequently rejects the message.
5165 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5166 MD5 (which is deprecated).
5168 21. When testing a filter file using -bf, Exim was writing a message when it
5169 took the sender from a "From " line in the message, but it was not doing so
5170 when it took $return_path from a Return-Path: header line. It now does.
5172 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5173 with a valid header line field name (a series of printing characters
5174 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5176 23. Changed "disc" in the source to "disk" to conform to the documentation and
5177 the book and for uniformity.
5179 24. Ignore Sendmail's -Ooption=value command line item.
5181 25. When execve() failed while trying to run a command in a pipe transport,
5182 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5183 could be confused with a return value of 69 from the command itself. This
5184 has been changed to 127, the value the shell returns if it is asked to run
5185 a non-existent command. The wording for the related log line suggests a
5186 non-existent command as the problem.
5188 26. If received_header_text expands to an empty string, do not add a Received:
5189 header line to the message. (Well, it adds a token one on the spool, but
5190 marks it "old" so that it doesn't get used or transmitted.)
5192 27. Installed eximstats 1.28 (addition of -nt option).
5194 28. There was no check for failure on the call to getsockname() in the daemon
5195 code. This can fail if there is a shortage of resources on the system, with
5196 ENOMEM, for example. A temporary error is now given on failure.
5198 29. Contrary to the C standard, it seems that in some environments, the
5199 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5200 program. Exim now does this explicitly; it affects the formatting of
5201 timestamps using strftime().
5203 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5204 complaints. I've now initialized all the variables, to avoid this.
5206 32. Header lines added by a system filter were not being "seen" during
5207 transport-time rewrites.
5209 33. The info_callback() function passed to OpenSSL is set up with type void
5210 (*)(SSL *, int, int), as described somewhere. However, when calling the
5211 function (actually a macro) that sets it up, the type void(*)() is
5212 expected. I've put in a cast to prevent warnings from picky compilers.
5214 34. If a DNS black list lookup found a CNAME record, but there were no A
5215 records associated with the domain it pointed at, Exim crashed.
5217 35. If a DNS black list lookup returned more than one A record, Exim ignored
5218 all but the first. It now scans all returned addresses if a particular IP
5219 value is being sought. In this situation, the contents of the
5220 $dnslist_value variable are a list of all the addresses, separated by a
5223 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5224 to accept a host name and all its aliases if the forward lookup for any of
5225 them yielded the IP address of the incoming connection. Now it accepts only
5226 those names whose forward lookup yields the correct IP address. Any other
5227 names are discarded. This closes a loophole whereby a rogue DNS
5228 administrator could create reverse DNS records to break through a
5229 wildcarded host restriction in an ACL.
5231 37. If a user filter or a system filter that ran in a subprocess used any of
5232 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5233 the wrong values were passed to the pipe command ($thisaddress had the
5234 value of $0, $0 had the value of $1, etc). This bug was introduced by
5235 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5237 38. Improved the line breaking for long SMTP error messages from ACLs.
5238 Previously, if there was no break point between 40 and 75 characters, Exim
5239 left the rest of the message alone. Two changes have been made: (a) I've
5240 reduced the minimum length to 35 characters; (b) if it can't find a break
5241 point between 35 and 75 characters, it looks ahead and uses the first one
5242 that it finds. This may give the occasional overlong line, but at least the
5243 remaining text gets split now.
5245 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5246 file descriptors might be low, and that setting 1000 would always raise it.
5247 It turns out that in some environments, the limit is already over 1000 and
5248 that lowering it causes trouble. So now Exim takes care not to decrease it.
5250 40. When delivering a message, the value of $return_path is set to $sender_
5251 address at the start of routing (routers may change the value). By an
5252 oversight, this default was not being set up when an address was tested by
5253 -bt or -bv, which affected the outcome if any router or filter referred to
5256 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5257 log only when "message" or "log_message" are set. However, if one of the
5258 conditions was an address verification, or a call to a nested ACL, the
5259 messages generated by the underlying test were being passed through. This
5260 no longer happens. The underlying message is available in $acl_verify_
5261 message for both "message" and "log_message" expansions, so it can be
5262 passed through if needed.
5264 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5265 new expansion $bh_ to give the encoded byte string without charset
5266 translation. Translation happens only if iconv() is available; HAVE_ICONV
5267 indicates this at build time. HEADERS_CHARSET gives the charset to
5268 translate to; headers_charset can change it in the configuration, and
5269 "headers charset" can change it in an individual filter file.
5271 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5272 that creates RFC 2047 encoded "words" labels them as that charset instead
5273 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5274 expansion operator; (ii) when Exim creates a From: line for a local
5275 message; (iii) when a header line is rewritten to include a "phrase" part.
5277 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5278 buggy, causing it to skip the first lines of messages whose message ID
5279 ended in 'D'. This would not have bitten before Exim release 4.14, because
5280 message IDs were unlikely to end in 'D' before then. The effect was to have
5281 incorrect size information for certain domains.
5283 45. #include "config.h" was missing at the start of the crypt16.c module. This
5284 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5287 46. If there was a timeout during a "random" callout check, Exim treated it as
5288 a failure of the random address, and carried on sending RSET and the real
5289 address. If the delay was just some slowness somewhere, the response to the
5290 original RCPT would be taken as a response to RSET and so on, causing
5291 mayhem of various kinds.
5293 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5294 when I implemented it. It didn't allow for the fact that some option values
5295 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5296 the right test for positive values.
5298 48. Domain names in DNS records are case-independent. Exim always looks them up
5299 in lower case. Some resolvers return domain names in exactly the case they
5300 appear in the zone file, that is, they may contain uppercase letters. Not
5301 all resolvers do this - some return always lower case. Exim was treating a
5302 change of case by a resolver as a change of domain, similar to a widening
5303 of a domain abbreviation. This triggered its re-routing code and so it was
5304 trying to route what was effectively the same domain again. This normally
5305 caused routing to fail (because the router wouldn't handle the domain
5306 twice). Now Exim checks for this case specially, and just changes the
5307 casing of the domain that it ultimately uses when it transmits the message
5310 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5313 50. If a filter generated a file delivery with a non-absolute name (possible if
5314 no home directory exists for the router), the forbid_file option was not
5317 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5318 the existing equality matching.
5320 52. Exim was using ints instead of ino_t variables in some places where it was
5321 dealing with inode numbers.
5323 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5324 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5325 called TMPDIR, and if it finds it is different, it changes its value.
5327 54. The smtp_printf() function is now made available to local_scan() so
5328 additional output lines can be written before returning. There is also an
5329 smtp_fflush() function to enable the detection of a dropped connection.
5330 The variables smtp_input and smtp_batched_input are exported to
5333 55. Changed the default runtime configuration: the message "Unknown user"
5334 has been removed from the ACL, and instead placed on the localuser router,
5335 using the cannot_route_message feature. This means that any verification
5336 failures that generate their own messages won't get overridden. Similarly,
5337 the "Unrouteable address" message that was in the ACL for unverifiable
5338 relay addresses has also been removed.
5340 56. Added hosts_avoid_esmtp to the smtp transport.
5342 57. The exicyclog script was not checking for the esoteric option
5343 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5344 will work only if exicyclog is run under the appropriate euid.
5346 58. Following a discussion on the list, the rules by which Exim recognises line
5347 endings on incoming messages have been changed. The -dropcr and drop_cr
5348 options are now no-ops, retained only for backwards compatibility. The
5349 following line terminators are recognized: LF CRLF CR. However, special
5350 processing applies to CR:
5352 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5353 nor a local message in the state where . is a terminator.
5355 (ii) If a bare CR is encountered in a header line, an extra space is added
5356 after the line terminator so as not to end the header. The reasoning
5357 behind this is that bare CRs in header lines are most likely either
5358 to be mistakes, or people trying to play silly games.
5360 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5361 was being incorrectly given as 18 bytes larger than it should have been.
5362 This is a VOB (very old bug).
5364 60. This may never have affected anything current, but just in case it has:
5365 When the local host is found other than at the start of a list of hosts,
5366 the local host, those with the same MX, and any that follow, are discarded.
5367 When the list in question was part of a longer list of hosts, the following
5368 hosts (not currently being processed) were also being discarded. This no
5369 longer happens. I'm not sure if this situation could ever has previously
5372 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5375 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5376 header. This is something that is recommended in a new Internet Draft, and
5377 is something that is documented as being done by Sendmail. There are two
5378 possible values. For messages generated by the autoreply transport, Exim
5381 Auto-Submitted: auto-replied
5383 whereas for all other generated messages (e.g. bounces) it adds
5385 Auto-Submitted: auto-generated
5387 63. The "personal" condition in filters now includes a test for the
5388 Auto-Submitted: header. If it contains the string "auto-" the message it
5389 not considered personal.
5391 64. Added rcpt_include_affixes as a generic transport option.
5393 65. Added queue_only_override (default true).
5395 66. Added the syslog_duplication option.
5397 67. If what should have been the first header line of a message consisted of
5398 a space followed by a colon, Exim was mis-interpreting it as a header line.
5399 It isn't of course - it is syntactically invalid and should therefore be
5400 treated as the start of the message body. The misbehaviour could have
5401 caused a number of strange effects, including loss of data in subsequent
5402 header lines, and spool format errors.
5404 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5405 client host had authenticated. This control can now be exercised by an ACL
5406 for more flexibility.
5408 69. By default, callouts do not happen when testing with -bh. There is now a
5409 variant, -bhc, which does actually run the callout code, including
5410 consulting and updating the callout cache.
5412 70. Added support for saslauthd authentication, courtesy of Alexander
5415 71. If statvfs() failed on the spool or log directories while checking their
5416 size for availability, Exim confusingly gave the error "space shortage".
5417 Furthermore, in debugging mode it crashed with a floating point exception.
5418 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5419 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5420 serious problem, Exim now writes to the main and panic logs when this
5421 happens, with details of the failure. It then refuses to accept the
5422 incoming message, giving the message "spool directory problem" or "log
5423 directory problem" with a 421 code for SMTP messages.
5425 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5426 0, 1, and 2 exist, because some OS complain for execs without them (see
5427 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5428 descriptors. However, the code omitted to check that the open succeeded,
5429 causing mysterious errors if for some reason the permissions on /dev/null
5430 got screwed. Now Exim writes a message to the main and panic logs, and
5431 bombs out if it can't open /dev/null.
5433 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5434 interact so that it is all more flexible. It is supposed to remain
5435 backwards compatible. Also added extra_local_interfaces.
5437 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5438 to bomb out with an assertion failure - to the client this appears as a
5439 connection drop. This problem occurs in the part of the code that was taken
5440 from the Samba project. Fortunately, the assertion is in a very simple
5441 function, so I have fixed this by reproducing the function inline in the
5442 one place where it is called, and arranging for authentication to fail
5443 instead of killing the process with assert().
5445 75. The SPA client code was not working when the server requested OEM rather
5446 than Unicode encoding.
5448 76. Added code to make require_files with a specific uid setting more usable in
5449 the case where statting the file as root fails - usually a non-root-mounted
5450 NFS file system. When this happens and the failure is EACCES, Exim now
5451 forks a subprocess and does the per-uid checking as the relevant uid.
5453 77. Added process_log_path.
5455 78. If log_file_path was not explicitly set, a setting of check_log_space or
5456 check_log_inodes was ignored.
5458 79. If a space check for the spool or log partitions fails, the incident is now
5459 logged. Of course, in the latter case the data may get lost...
5461 80. Added the %p formatting code to string_format() so that it can be used to
5462 print addresses in debug_print(). Adjusted all the address printing in the
5463 debugging in store.c to use %p rather than %d.
5465 81. There was a concern that a line of code in smtp_in.c could overflow a
5466 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5467 initially expressed, the concern was not well-founded, because trailing
5468 spaces are removed early. However, if the trailing spaces were followed by
5469 a NULL, they did not get removed, so the overflow was possible. Two fixes
5472 (a) I re-wrote the offending code in a cleaner fashion.
5473 (b) If an incoming SMTP command contains a NULL character, it is rejected
5476 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5477 runs initgroups(), so that if the Exim user is in any additional groups,
5478 they will be used during message reception.
5484 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob