1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
29 PP/01 Solaris build fix for Oracle's LDAP libraries.
30 Bugzilla 1109, patch from Stephen Usher.
32 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
34 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
37 TF/02 Fix a couple more cases where we did not log the error message
38 when unlink() failed. See also change 4.74-TF/03.
40 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
41 lock up or crash if it happened to be inside a call to libc when it
42 got a SIGUSR1 from exiwhat.
44 The SIGUSR1 handler appends the current process status to the process
45 log which is later printed by exiwhat. It used to use the general
46 purpose logging code to do this, but several functions it calls are
49 The new output code in the SIGUSR1 handler is specific to the process
50 log, and simple enough that it's easy to inspect for signal safety.
51 Removing some special cases also simplifies the general logging code.
52 Removing the spurious timestamps from the process log simplifies
55 TF/04 Improved ratelimit ACL condition.
57 The /noupdate option has been deprecated in favour of /readonly which
58 has clearer semantics. The /leaky, /strict, and /readonly update modes
59 are mutually exclusive. The update mode is no longer included in the
60 database key; it just determines when the database is updated. (This
61 means that when you upgrde Exim will forget old rate measurements.)
63 Exim now checks that the per_* options are used with an update mode that
64 makes sense for the current ACL. For example, when Exim is processing a
65 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
66 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
67 must specify per_mail/readonly. If you omit the update mode it defaults to
68 /leaky where that makes sense (as before) or /readonly where required.
70 The /noupdate option is now undocumented but still supported for
71 backwards compatibility. It is equivalent to /readonly except that in
72 ACLs where /readonly is required you may specify /leaky/noupdate or
73 /strict/noupdate which are treated the same as /readonly.
75 A useful new feature is the /count= option. This is a generalization
76 of the per_byte option, so that you can measure the throughput of other
77 aggregate values. For example, the per_byte option is now equivalent
78 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
80 The per_rcpt option has been generalized using the /count= mechanism
81 (though it's more complicated than the per_byte equivalence). When it is
82 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
83 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
84 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
85 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
86 Note that using per_rcpt with a non-readonly update mode in more than
87 one ACL will cause the recipients to be double-counted. (The per_mail
88 and per_byte options don't have this problem.)
90 The handling of very low rates has changed slightly. If the computed rate
91 is less than the event's count (usually one) then this event is the first
92 after a long gap. In this case the rate is set to the same as this event's
93 count, so that the first message of a spam run is counted properly.
95 The major new feature is a mechanism for counting the rate of unique
96 events. The new per_addr option counts the number of different
97 recipients that someone has sent messages to in the last time period. It
98 behaves like per_rcpt if all the recipient addresses are different, but
99 duplicate recipient addresses do not increase the measured rate. Like
100 the /count= option this is a general mechanism, so the per_addr option
101 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
102 example, measure the rate that a client uses different sender addresses
103 with the options per_mail/unique=$sender_address. There are further
104 details in the main documentation.
106 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
108 TF/06 Removed a few PCRE remnants.
110 TF/07 Automatically extract Exim's version number from tags in the git
111 repository when doing development or release builds.
113 PP/02 Raise smtp_cmd_buffer_size to 16kB.
114 Bugzilla 879. Patch from Paul Fisher.
116 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
117 Heavily based on revision 40f9a89a from Simon Arlott's tree.
120 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
122 PP/05 Variable $av_failed, true if the AV scanner deferred.
123 Bugzilla 1078. Patch from John Horne.
125 PP/06 Stop make process more reliably on build failure.
126 Bugzilla 1087. Patch from Heiko Schlittermann.
128 PP/07 Make maildir_use_size_file an _expandable_ boolean.
129 Bugzilla 1089. Patch from Heiko Schlittermann.
131 PP/08 Handle ${run} returning more data than OS pipe buffer size.
132 Bugzilla 1131. Patch from Holger Weiß.
134 PP/09 Handle IPv6 addresses with SPF.
135 Bugzilla 860. Patch from Wolfgang Breyha.
137 PP/10 GnuTLS: support TLS 1.2 & 1.1.
139 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
142 PP/11 match_* no longer expand right-hand-side by default.
143 New compile-time build option, EXPAND_LISTMATCH_RHS.
144 New expansion conditions, "inlist", "inlisti".
146 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
148 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
150 PP/14 fix log_write() format string regression from TF/03.
151 Bugzilla 1152. Patch from Dmitry Isaikin.
157 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
159 PP/02 Harmonised TLS library version reporting; only show if debugging.
160 Layout now matches that introduced for other libraries in 4.74 PP/03.
162 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
164 PP/04 New "dns_use_edns0" global option.
166 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
169 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
170 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
172 TK/01 Updated PolarSSL code to 0.14.2.
173 Bugzilla 1097. Patch from Andreas Metzler.
175 PP/07 Catch divide-by-zero in ${eval:...}.
178 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
181 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
182 format-string attack -- SECURITY: remote arbitrary code execution.
184 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
185 time unintentionally subject to list matching rules, letting the header
186 cause arbitrary Exim lookups (of items which can occur in lists, *not*
187 arbitrary string expansion). This allowed for information disclosure.
189 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
190 INT_MIN/-1 -- value coerced to INT_MAX.
196 NM/01 Workround for PCRE version dependency in version reporting
199 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
200 This fixes portability to compilers other than gcc, notably
201 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
203 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
204 makefiles for portability to HP-UX and POSIX correctness.
206 PP/01 Permit LOOKUP_foo enabling on the make command-line.
207 Also via indented variable definition in the Makefile.
208 (Debugging by Oliver Heesakkers).
210 PP/02 Restore caching of spamd results with expanded spamd_address.
211 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
213 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
214 Improves build reliability. Fix from: Frank Elsner
216 NM/02 Fix wide character breakage in the rfc2047 coding
217 Fixes bug 1064. Patch from Andrey N. Oktyabrski
219 NM/03 Allow underscore in dnslist lookups
220 Fixes bug 1026. Patch from Graeme Fowler
222 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
223 Code patches from Adam Ciarcinski of NetBSD.
225 NM/04 Fixed exiqgrep to cope with mailq missing size issue
228 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
229 is logged, to avoid truncation. Patch from John Horne.
231 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
232 Patch from Jakob Hirsch.
234 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
235 SQL string expansion failure details.
236 Patch from Andrey Oktyabrski.
238 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
239 Patch from Simon Arlott.
241 PP/09 New lookups functionality failed to compile on old gcc which rejects
242 extern declarations in function scope.
243 Patch from Oliver Fleischmann
245 PP/10 Use sig_atomic_t for flags set from signal handlers.
246 Check getgroups() return and improve debugging.
247 Fixed developed for diagnosis in bug 927 (which turned out to be
250 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
251 Patch from Mark Zealey.
253 PP/12 Bugzilla 1056: Improved spamd server selection.
254 Patch from Mark Zealey.
256 PP/13 Bugzilla 1086: Deal with maildir quota file races.
257 Based on patch from Heiko Schlittermann.
259 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
260 Patch from Uwe Doering, sign-off by Michael Haardt.
262 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
263 variable declaration deep within a block. Bug and patch from
266 PP/15 lookups-Makefile IRIX compatibilty coercion.
268 PP/16 Make DISABLE_DKIM build knob functional.
270 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
271 Patch by Simon Arlott
273 TF/03 Fix valgrind.h portability to C89 compilers that do not support
274 variable argument macros. Our copy now differs from upstream.
280 TF/01 Failure to get a lock on a hints database can have serious
281 consequences so log it to the panic log.
283 TF/02 Log LMTP confirmation messages in the same way as SMTP,
284 controlled using the smtp_confirmation log selector.
286 TF/03 Include the error message when we fail to unlink a spool file.
288 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
289 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
290 for maintaining out-of-tree patches for some time.
292 PP/01 Bugzilla 139: Documentation and portability issues.
293 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
294 Handle per-OS dynamic-module compilation flags.
296 PP/02 Let /dev/null have normal permissions.
297 The 4.73 fixes were a little too stringent and complained about the
298 permissions on /dev/null. Exempt it from some checks.
299 Reported by Andreas M. Kirchwitz.
301 PP/03 Report version information for many libraries, including
302 Exim version information for dynamically loaded libraries. Created
303 version.h, now support a version extension string for distributors
304 who patch heavily. Dynamic module ABI change.
306 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
307 privilege escalation vulnerability whereby the Exim run-time user
308 can cause root to append content of the attacker's choosing to
311 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
314 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
315 If dropping privileges for untrusted macros, we disabled normal logging
316 on the basis that it would fail; for the Exim run-time user, this is not
317 the case, and it resulted in successful deliveries going unlogged.
318 Fixed. Reported by Andreas Metzler.
324 PP/01 Date: & Message-Id: revert to normally being appended to a message,
325 only prepend for the Resent-* case. Fixes regression introduced in
326 Exim 4.70 by NM/22 for Bugzilla 607.
328 PP/02 Include check_rfc2047_length in configure.default because we're seeing
329 increasing numbers of administrators be bitten by this.
331 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
333 PP/03 Bugzilla 994: added openssl_options main configuration option.
335 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
337 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
339 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
341 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
342 without a peer certificate, leading to a segfault because of an
343 assumption that peers always have certificates. Be a little more
344 paranoid. Problem reported by Martin Tscholak.
346 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
347 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
348 NB: ClamAV planning to remove STREAM in "middle of 2010".
349 CL also introduces -bmalware, various -d+acl logging additions and
350 more caution in buffer sizes.
352 PP/09 Implemented reverse_ip expansion operator.
354 PP/10 Bugzilla 937: provide a "debug" ACL control.
356 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
358 PP/12 Bugzilla 973: Implement --version.
360 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
362 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
364 PP/15 Bugzilla 816: support multiple condition rules on Routers.
366 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
367 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
368 ignore trailing whitespace.
370 JJ/02 prevent non-panic DKIM error from being sent to paniclog
372 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
375 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
376 Notification from Dr Andrew Aitchison.
378 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
379 ExtendedDetectionInfo response format.
380 Notification from John Horne.
382 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
385 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
386 XSL and documented dependency on system catalogs, with examples of how
389 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
392 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
393 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
394 configuration file which is writeable by the Exim user or group.
396 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
397 of configuration files to cover files specified with the -C option if
398 they are going to be used with root privileges, not just the default
401 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
402 option (effectively making it always true).
404 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
405 files to be used while preserving root privileges.
407 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
408 that rogue child processes cannot use them.
410 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
411 run-time user, instead of root.
413 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
414 Exim run-time user without dropping privileges.
416 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
417 result string, instead of calling string_vformat() twice with the same
420 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
421 for other users. Others should always drop root privileges if they use
422 -C on the command line, even for a whitelisted configure file.
424 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
426 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
432 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
433 $data_path, and $header_path variables; fixed documentation bugs and
436 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
437 exipick to access non-standard spools, including the "frozen" queue
440 NM/01 Bugzilla 965: Support mysql stored procedures.
441 Patch from Alain Williams
443 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
445 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
446 Patch from Andreas Metzler
448 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
449 Patch from Kirill Miazine
451 NM/05 Bugzilla 671: Added umask to procmail example.
453 JJ/03 installed exipick 20100323.0, fixing doc bug
455 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
456 directory. Notification and patch from Dan Rosenberg.
458 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
460 TK/02 Improve log output when DKIM signing operation fails.
462 MH/01 Treat the transport option dkim_domain as a colon separated
463 list, not as a single string, and sign the message with each element,
464 omitting multiple occurences of the same signer.
466 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
467 Bugzilla 985, 986. Patch by Simon Arlott
469 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
470 Patch by Simon Arlott
472 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
473 MBX locking. Notification from Dan Rosenberg.
479 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
481 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
483 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
485 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
487 NM/04 Bugzilla 915: Flag broken perl installation during build.
493 TK/01 Added patch by Johannes Berg that expands the main option
494 "spamd_address" if it starts with a dollar sign.
496 TK/02 Write list of recipients to X-Envelope-Sender header when building
497 the mbox-format spool file for content scanning (suggested by Jakob
500 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
501 (http://www.dcc-servers.net/) support via dccifd. Activated by
502 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
504 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
505 by Mark Daniel Reidel <mr@df.eu>.
507 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
508 When building exim an external PCRE library is now needed -
509 PCRE is a system library on the majority of modern systems.
510 See entry on PCRE_LIBS in EDITME file.
512 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
513 conversation. Added nologin parameter to request.
514 Patch contributed by Kirill Miazine.
516 TF/01 Do not log submission mode rewrites if they do not change the address.
518 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
520 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
521 log files in place. Contributed by Roberto Lima.
523 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
525 TF/03 Bugzilla 615: When checking the local_parts router precondition
526 after a local_part_suffix or local_part_prefix option, Exim now
527 does not use the address's named list lookup cache, since this
528 contains cached lookups for the whole local part.
530 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
531 Robert Millan. Documentation is in experimental-spec.txt.
533 TF/04 Bugzilla 668: Fix parallel build (make -j).
535 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
537 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
538 Patch provided by Jan Srzednicki.
540 TF/05 Leading white space used to be stripped from $spam_report which
541 wrecked the formatting. Now it is preserved.
543 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
544 that they are available at delivery time.
546 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
548 TF/08 TLS error reporting now respects the incoming_interface and
549 incoming_port log selectors.
551 TF/09 Produce a more useful error message if an SMTP transport's hosts
552 setting expands to an empty string.
554 NM/06 Bugzilla 744: EXPN did not work under TLS.
555 Patch provided by Phil Pennock.
557 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
558 Patch provided by Richard Godbee.
560 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
561 acl_smtp_notquit, added index entry.
563 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
564 Patch provided by Eugene Bujak.
566 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
567 accept(). Patch provided by Maxim Dounin.
569 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
570 Patch provided by Phil Pennock.
572 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
574 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
575 Patch provided by Brad "anomie" Jorsch.
577 NM/14 Bugzilla 622: Added timeout setting to transport filter.
578 Patch provided by Dean Brooks.
580 TK/05 Add native DKIM support (does not depend on external libraries).
582 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
583 Patch provided by Graeme Fowler.
585 NM/16 Bugzilla 851: Documentation example syntax fix.
587 NM/17 Changed NOTICE file to remove references to embedded PCRE.
589 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
592 NM/19 Bugzilla 745: TLS version reporting.
593 Patch provided by Phil Pennock.
595 NM/20 Bugzilla 167: bool: condition support.
596 Patch provided by Phil Pennock.
598 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
599 clients. Patch provided by Phil Pennock.
601 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
602 Patch provided by Brad "anomie" Jorsch.
604 NM/23 Bugzilla 687: Fix misparses in eximstats.
605 Patch provided by Heiko Schlittermann.
607 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
608 Patch provided by Heiko Schlittermann.
610 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
611 plus update to original patch.
613 NM/26 Bugzilla 799: Documentation correction for ratelimit.
615 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
616 Patch provided by David Brownlee.
618 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
620 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
622 NM/30 Bugzilla 888: TLS documentation bugfixes.
624 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
626 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
627 Unlike the original bugzilla I have changed all shell scripts in src tree.
629 NM/33 Bugzilla 898: Transport filter timeout fix.
630 Patch by Todd Rinaldo.
632 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
633 Patch by Serge Demonchaux.
635 NM/35 Bugzilla 39: Base64 decode bug fixes.
636 Patch by Jakob Hirsch.
638 NM/36 Bugzilla 909: Correct connect() call in dcc code.
640 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
642 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
644 NM/39 Bugzilla 911: Fixed MakeLinks build script.
650 TK/01 Add preliminary DKIM support. Currently requires a forked version of
651 ALT-N's libdkim that I have put here:
652 http://duncanthrax.net/exim-experimental/
654 Note to Michael Haardt: I had to rename some vars in sieve.c. They
655 were called 'true' and it seems that C99 defines that as a reserved
656 keyword to be used with 'bool' variable types. That means you could
657 not include C99-style headers which use bools without triggering
658 build errors in sieve.c.
660 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
661 as mailq or other aliases. Changed the --help handling significantly
662 to do whats expected. exim_usage() emits usage/help information.
664 SC/01 Added the -bylocaldomain option to eximstats.
666 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
668 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
670 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
676 PH/01 Another patch from the Sieve maintainer.
678 PH/02 When an IPv6 address is converted to a string for single-key lookup
679 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
680 dots are used instead of colons so that keys in lsearch files need not
681 contain colons. This was done some time before quoting was made available
682 in lsearch files. However, iplsearch files do require colons in IPv6 keys
683 (notated using the quote facility) so as to distinguish them from IPv4
684 keys. This meant that lookups for IP addresses in host lists did not work
685 for iplsearch lookups.
687 This has been fixed by arranging for IPv6 addresses to be expressed with
688 colons if the lookup type is iplsearch. This is not incompatible, because
689 previously such lookups could never work.
691 The situation is now rather anomolous, since one *can* have colons in
692 ordinary lsearch keys. However, making the change in all cases is
693 incompatible and would probably break a number of configurations.
695 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
698 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
699 conversion specification without a maximum field width, thereby enabling
700 a rogue spamd server to cause a buffer overflow. While nobody in their
701 right mind would setup Exim to query an untrusted spamd server, an
702 attacker that gains access to a server running spamd could potentially
703 exploit this vulnerability to run arbitrary code as the Exim user.
705 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
706 $primary_hostname instead of what libspf2 thinks the hosts name is.
708 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
709 a directory entry by the name of the lookup key. Previously, if a
710 symlink pointed to a non-existing file or a file in a directory that
711 Exim lacked permissions to read, a lookup for a key matching that
712 symlink would fail. Now it is enough that a matching directory entry
713 exists, symlink or not. (Bugzilla 503.)
715 PH/03 The body_linecount and body_zerocount variables are now exported in the
718 PH/04 Added the $dnslist_matched variable.
720 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
721 This means they are set thereafter only if the connection becomes
724 PH/06 Added the client_condition to authenticators so that some can be skipped
725 by clients under certain conditions.
727 PH/07 The error message for a badly-placed control=no_multiline_responses left
728 "_responses" off the end of the name.
730 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
732 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
733 (without spaces) instead of just copying the configuration text.
735 PH/10 Added the /noupdate option to the ratelimit ACL condition.
737 PH/11 Added $max_received_linelength.
739 PH/12 Added +ignore_defer and +include_defer to host lists.
741 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
742 way in which PCRE > 7.0 is built.
744 PH/14 Implemented queue_only_load_latch.
746 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
747 MAIL command. The effect was to mangle the value on 64-bit systems.
749 PH/16 Another patch from the Sieve maintainer.
751 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
753 PH/18 If a system quota error occurred while trying to create the file for
754 a maildir delivery, the message "Mailbox is full" was not appended to the
755 bounce if the delivery eventually timed out. Change 4.67/27 below applied
756 only to a quota excession during the actual writing of the file.
758 PH/19 It seems that peer DN values may contain newlines (and other non-printing
759 characters?) which causes problems in log lines. The DN values are now
760 passed through string_printing() before being added to log lines.
762 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
763 and InterBase are left for another time.)
765 PH/21 Added message_body_newlines option.
767 PH/22 Guard against possible overflow in moan_check_errorcopy().
769 PH/23 POSIX allows open() to be a macro; guard against that.
771 PH/24 If the recipient of an error message contained an @ in the local part
772 (suitably quoted, of course), incorrect values were put in $domain and
773 $local_part during the evaluation of errors_copy.
779 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
780 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
783 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
784 issue a MAIL command.
786 PH/02 In an ACL statement such as
788 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
790 if a client was not listed at all, or was listed with a value other than
791 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
792 the condition was not true (as it should be), so access was not denied.
793 The bug was that the ! inversion was incorrectly passed on to the second
794 item. This has been fixed.
796 PH/03 Added additional dnslists conditions == and =& which are different from
797 = and & when the dns lookup returns more than one IP address.
799 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
800 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
802 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
803 FSYNC, which compiles an option called disable_fsync that allows for
804 bypassing fsync(). The documentation is heavily laced with warnings.
806 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
808 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
809 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
810 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
811 including adding "make clean"; (3) Added -fPIC when compiling the test
812 dynamically loaded module, to get rid of a warning.
814 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
815 message fails, move_frozen_messages = true and ignore_bounce_errors_after
816 = 0s. The bug is otherwise harmless.
818 PH/07 There was a bug in the dovecot authenticator such that the value of
819 $auth1 could be overwritten, and so not correctly preserved, after a
820 successful authentication. This usually meant that the value preserved by
821 the server_setid option was incorrect.
823 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
825 PH/09 Installed PCRE release 7.0.
827 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
828 run for batched SMTP input. It is now run at the start of every message
829 in the batch. While fixing this I discovered that the process information
830 (output by running exiwhat) was not always getting set for -bs and -bS
831 input. This is fixed, and it now also says "batched" for BSMTP.
833 PH/11 Added control=no_pipelining.
835 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
836 patch, slightly modified), and move the expansion of helo_data till after
837 the connection is made in the smtp transport (so it can use these
840 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
842 PH/14 Added log_selector = +pid.
844 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
846 PH/16 Add ${if forany and ${if forall.
848 PH/17 Added dsn_from option to vary the From: line in DSNs.
850 PH/18 Flush SMTP output before performing a callout, unless control =
851 no_callout_flush is set.
853 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
854 was true (the default) a successful delivery failed to delete the retry
855 item, thus causing premature timeout of the address. The bug is now
858 PH/20 Added hosts_avoid_pipelining to the smtp transport.
860 PH/21 Long custom messages for fakedefer and fakereject are now split up
861 into multiline reponses in the same way that messages for "deny" and
862 other ACL rejections are.
864 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
865 with slight modification.
867 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
868 draft, changing the syntax and factoring some duplicate code".
870 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
871 for deliveries of the second and subsequent messages over the same SMTP
874 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
875 ${reduce, with only minor "tidies".
877 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
879 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
880 expansion side effects.
882 PH/27 When a message times out after an over-quota error from an Exim-imposed
883 quota, the bounce message says "mailbox is full". This message was not
884 being given when it was a system quota that was exceeded. It now should
887 MH/03 Made $recipients available in local_scan(). local_scan() already has
888 better access to the recipient list through recipients_list[], but
889 $recipients can be useful in postmaster-provided expansion strings.
891 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
892 in the case of a MAIL command with additional options following the
893 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
894 were accidentally chopped off.
896 PH/29 SMTP synchronization checks are implemented when a command is read -
897 there is a check that no more input is waiting when there shouldn't be
898 any. However, for some commands, a delay in an ACL can mean that it is
899 some time before the response is written. In this time, more input might
900 arrive, invalidly. So now there are extra checks after an ACL has run for
901 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
902 pipelining has not been advertised.
904 PH/30 MH's patch to allow iscntrl() characters to be list separators.
906 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
907 returned in the SMTP response when smtp_return_error_details was false.
910 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
911 instead of the C I/O that was originally supplied, because problems were
914 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
915 Exim which did not show up earlier: it was assuming that a call to
916 SSL_CTX_set_info_callback() might give an error value. In fact, there is
917 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
918 was a macro that became an assignment, so it seemed to work. This has
919 changed to a proper function call with a void return, hence the compile
920 error. Exim's code has been fixed.
922 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
925 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
927 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
929 PH/37 If a message is not accepted after it has had an id assigned (e.g.
930 because it turns out to be too big or there is a timeout) there is no
931 "Completed" line in the log. When some messages of this type were
932 selected by exigrep, they were listed as "not completed". Others were
933 picked up by some special patterns. I have improved the selection
934 criteria to be more general.
936 PH/38 The host_find_failed option in the manualroute router can now be set
937 to "ignore", to completely ignore a host whose IP address cannot be
938 found. If all hosts are ignored, the behaviour is controlled by the new
939 host_all_ignored option.
941 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
942 homing or because of multiple MX records with /mx) generated more than
943 one IP address, and the following item turned out to be the local host,
944 all the secondary addresses of the first item were incorrectly removed
945 from the list, along with the local host and any following hosts (which
946 is what is supposed to happen).
948 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
949 whoever called Exim into the -H file. In the case of the daemon it was
950 behaving confusingly. When first started, it used values for whoever
951 started the daemon, but after a SIGHUP it used the Exim user (because it
952 calls itself on a restart). I have changed the code so that it now always
955 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
956 message are rejected with the same error (e.g. no authentication or bad
957 sender address), and a DATA command is nevertheless sent (as can happen
958 with PIPELINING or a stupid MUA), the error message that was given to the
959 RCPT commands is included in the rejection of the DATA command. This is
960 intended to be helpful for MUAs that show only the final error to their
963 PH/42 Another patch from the Sieve maintainer.
965 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
966 Eximstats - Fixed some broken HTML links and added missing column headers
968 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
969 columns for Rejects, Temp Rejects, Ham, and Spam rows.
971 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
973 PH/43 Yet another patch from the Sieve maintainer.
975 PH/44 I found a way to check for a TCP/IP connection going away before sending
976 the response to the final '.' that terminates a message, but only in the
977 case where the client has not sent further data following the '.'
978 (unfortunately, this is allowed). However, in many cases there won't be
979 any further data because there won't be any more messages to send. A call
980 to select() can be used: if it shows that the input is "ready", there is
981 either input waiting, or the socket has been closed. An attempt to read
982 the next input character can distinguish the two cases. Previously, Exim
983 would have sent an OK response which the client would never have see.
984 This could lead to message repetition. This fix should cure that, at
985 least in a lot of common cases.
987 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
988 advertised in response to EHLO.
994 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
995 fixed by 4.65/MH/01 (is this a record?) are fixed:
997 (i) An empty string was always treated as zero by the numeric comparison
998 operators. This behaviour has been restored.
1000 (ii) It is documented that the numeric comparison operators always treat
1001 their arguments as decimal numbers. This was broken in that numbers
1002 starting with 0 were being interpreted as octal.
1004 While fixing these problems I realized that there was another issue that
1005 hadn't been noticed. Values of message_size_limit (both the global option
1006 and the transport option) were treated as octal if they started with 0.
1007 The documentation was vague. These values are now always treated as
1008 decimal, and I will make that clear in the documentation.
1014 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1015 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1018 MH/01 Don't check that the operands of numeric comparison operators are
1019 integers when their expansion is in "skipping" mode (fixes bug
1020 introduced by 4.64-PH/07).
1022 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1023 child addresses, Exim now panics and dies. Previously, because the count
1024 is held in a short int, deliveries were likely to be lost. As such a
1025 large number of recipients for a single message is ridiculous
1026 (performance will be very, very poor), I have chosen to impose a limit
1027 rather than extend the field.
1033 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1034 leftover -K file (the existence of which was triggered by #402).
1035 While we were at it, introduced process PID as part of the -K
1036 filename. This should rule out race conditions when creating
1039 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1040 processing considerably. Previous code took too long for large mails,
1041 triggering a timeout which in turn triggers #401.
1043 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1044 in the DK code in transports.c. sendfile() is not really portable,
1045 hence the _LINUX specificness.
1047 TF/01 In the add_headers option to the mail command in an Exim filter,
1048 there was a bug that Exim would claim a syntax error in any
1049 header after the first one which had an odd number of characters
1052 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1053 callout verification, Exim cached a "reject" for the entire domain. This
1054 is correct for most verifications, but it is not correct for a recipient
1055 verification with use_sender or use_postmaster set, because in that case
1056 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1057 case of MAIL FROM:<> rejection from other early rejections (e.g.
1058 rejection of HELO). When verifying a recipient using a non-null MAIL
1059 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1060 Whatever the result of the callout, the value of the domain cache is
1061 left unchanged (for any other kind of callout, getting as far as trying
1062 RCPT means that the domain itself is ok).
1064 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1067 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1068 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1069 session, the daemon ignores ECONNECT errors and logs others; it now
1070 ignores EPIPE as well.
1072 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1073 (quoted-printable decoding).
1075 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1076 later the small subsequent patch to fix an introduced bug.
1078 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1080 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1082 PH/08 An error is now given if message_size_limit is specified negative.
1084 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1085 to be given (somewhat) arbitrary names.
1087 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1090 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1091 miscellaneous code fixes
1093 PH/10 Added the log_reject_target ACL modifier to specify where to log
1096 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1097 hostname. This is wrong, because it relates to the incoming message (and
1098 probably the interface on which it is arriving) and not to the outgoing
1099 callout (which could be using a different interface). This has been
1100 changed to use the value of the helo_data option from the smtp transport
1101 instead - this is what is used when a message is actually being sent. If
1102 there is no remote transport (possible with a router that sets up host
1103 addresses), $smtp_active_hostname is used.
1105 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1106 tweaks were necessary in order to get it to work (see also 21 below):
1107 (a) The code assumed that strncpy() returns a negative number on buffer
1108 overflow, which isn't the case. Replaced with Exim's string_format()
1110 (b) There were several signed/unsigned issues. I just did the minimum
1111 hacking in of casts. There is scope for a larger refactoring.
1112 (c) The code used strcasecmp() which is not a standard C function.
1113 Replaced with Exim's strcmpic() function.
1114 (d) The code set only $1; it now sets $auth1 as well.
1115 (e) A simple test gave the error "authentication client didn't specify
1116 service in request". It would seem that Dovecot has changed its
1117 interface. Fortunately there's a specification; I followed it and
1118 changed what the client sends and it appears to be working now.
1120 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1123 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1124 address A is aliased to B and C, where B exists and C does not. Without
1125 -v the output is "A verified" because verification stops after a
1126 successful redirection if more than one address is generated. However,
1127 with -v the child addresses are also verified. Exim was outputting "A
1128 failed to verify" and then showing the successful verification for C,
1129 with its parentage. It now outputs "B failed to verify", showing B's
1130 parentage before showing the successful verification of C.
1132 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1133 look up a TXT record in a specific list after matching in a combined
1136 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1137 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1138 they consult the DNS. I had assumed they would set it the way they
1139 wanted; and indeed my experiments on Linux seem to show that in some
1140 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1141 To be on the safe side, however, I have now made the interface to
1142 host_find_byname() similar to host_find_bydns(), with an argument
1143 containing the DNS resolver options. The host_find_byname() function now
1144 sets these options at its start, just as host_find_bydns() does. The smtp
1145 transport options dns_qualify_single and dns_search_parents are passed to
1146 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1147 of host_find_byname() use the default settings of RES_DEFNAMES
1148 (qualify_single) but not RES_DNSRCH (search_parents).
1150 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1151 spool_read_header() do less string testing, by means of a preliminary
1152 switch on the second character of optional "-foo" lines. (This is
1153 overdue, caused by the large number of possibilities that now exist.
1154 Originally there were few.) While I was there, I also converted the
1155 str(n)cmp tests so they don't re-test the leading "-" and the first
1156 character, in the hope this might squeeze out yet more improvement.
1158 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1159 flag allowing group syntax was set by the header_syntax check but not
1160 turned off, possible causing trouble later; (2) The flag was not being
1161 set at all for the header_verify test, causing "group"-style headers to
1162 be rejected. I have now set it in this case, and also caused header_
1163 verify to ignore an empty address taken from a group. While doing this, I
1164 came across some other cases where the code for allowing group syntax
1165 while scanning a header line wasn't quite right (mostly, not resetting
1166 the flag correctly in the right place). These bugs could have caused
1167 trouble for malformed header lines. I hope it is now all correct.
1169 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1170 with the "reply" argument non-NULL. The code, however (which originally
1171 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1172 but it didn't always do it. This confused somebody who was copying the
1173 code for some other use. I have removed all the tests.
1175 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1176 feature that was used to support insecure browsers during the U.S. crypto
1177 embargo. It requires special client support, and Exim is probably the
1178 only MTA that supported it -- and would never use it because real RSA is
1179 always available. This code has been removed, because it had the bad
1180 effect of slowing Exim down by computing (never used) parameters for the
1181 RSA_EXPORT functionality.
1183 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1184 authenticator to fail if there's a tab character in the incoming data
1185 (there should never be unless someone is messing about, as it's supposed
1186 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1187 if the connection is using TLS or if the remote IP is the same as the
1188 local IP, and the "valid-client-cert option" if a client certificate has
1191 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1192 authenticators. This can be used for authorization after authentication
1193 succeeds. (In the case of plaintext, it servers for both authentication
1196 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1197 if any retry times were supplied.
1199 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1200 connection, where there is no client IP address to check. In this
1201 situation, the verify now always succeeds.
1203 PH/25 Applied John Jetmore's -Mset patch.
1205 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1207 PH/27 In a string expansion for a processed (not raw) header when multiple
1208 headers of the same name were present, leading whitespace was being
1209 removed from all of them, but trailing whitespace was being removed only
1210 from the last one. Now trailing whitespace is removed from each header
1211 before concatenation. Completely empty headers in a concatenation (as
1212 before) are ignored.
1214 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1215 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1217 PH/29 [Removed. This was a change that I later backed out, and forgot to
1218 correct the ChangeLog entry (that I had efficiently created) before
1219 committing the later change.]
1221 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1222 address errors (4xx response to RCPT) over the same connection as other
1223 messages routed to the same hosts. Such deliveries are always "forced",
1224 so retry times are not inspected. This resulted in far too many retries
1225 for the affected addresses. The effect occurred only when there were more
1226 hosts than the hosts_max_try setting in the smtp transport when it had
1227 the 4xx errors. Those hosts that it had tried were not added to the list
1228 of hosts for which the message was waiting, so if all were tried, there
1229 was no problem. Two fixes have been applied:
1231 (i) If there are any address or message errors in an SMTP delivery, none
1232 of the hosts (tried or untried) are now added to the list of hosts
1233 for which the message is waiting, so the message should not be a
1234 candidate for sending over the same connection that was used for a
1235 successful delivery of some other message. This seems entirely
1236 reasonable: after all the message is NOT "waiting for some host".
1237 This is so "obvious" that I'm not sure why it wasn't done
1238 previously. Hope I haven't missed anything, but it can't do any
1239 harm, as the worst effect is to miss an optimization.
1241 (ii) If, despite (i), such a delivery is accidentally attempted, the
1242 routing retry time is respected, so at least it doesn't keep
1243 hammering the server.
1245 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1246 in ${readsocket because some servers need this prod.
1248 PH/32 Added some extra debug output when updating a wait-xxx database.
1250 PH/33 The hint "could be header name not terminated by colon", which has been
1251 given for certain expansion errors for a long time, was not being given
1252 for the ${if def:h_colon_omitted{... case.
1254 PH/34 The spec says: "With one important exception, whenever a domain list is
1255 being scanned, $domain contains the subject domain." There was at least
1256 one case where this was not true.
1258 PH/35 The error "getsockname() failed: connection reset by peer" was being
1259 written to the panic log as well as the main log, but it isn't really
1260 panic-worthy as it just means the connection died rather early on. I have
1261 removed the panic log writing for the ECONNRESET error when getsockname()
1264 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1265 runs only) independently of the message's sender address. This meant
1266 that, if the 4xx error was in fact related to the sender, a different
1267 message to the same recipient with a different sender could confuse
1268 things. In particualar, this can happen when sending to a greylisting
1269 server, but other circumstances could also provoke similar problems.
1270 I have changed the default so that the retry time for these errors is now
1271 based a combination of the sender and recipient addresses. This change
1272 can be overridden by setting address_retry_include_sender=false in the
1275 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1276 remote server are returned as part of bounce messages. This was not
1277 happening for LMTP over a pipe (the lmtp transport), but now it is the
1278 same for both kinds of LMTP.
1280 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1281 in header lines that were in fact CNAMEs. This is no longer the case.
1283 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1284 and queue runs started by the daemon processed all messages. This has
1285 been fixed so that -R and -S can now usefully be given with -q<time>.
1287 PH/40 Import PCRE release 6.7 (fixes some bugs).
1289 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1291 PH/42 Give an error if -q is specified more than once.
1293 PH/43 Renamed the variables $interface_address and $interface_port as
1294 $received_ip_address and $received_port, to make it clear that these
1295 values apply to message reception, and not to the outgoing interface when
1296 a message is delivered. (The old names remain recognized, of course.)
1298 PH/44 There was no timeout on the connect() call when using a Unix domain
1299 socket in the ${readsocket expansion. There now is.
1301 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1302 be meaningful with "accept".
1304 SC/01 Eximstats V1.43
1305 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1307 SC/02 Eximstats V1.44
1308 Use a glob alias rather than an array ref in the generated
1309 parser. This improves both readability and performance.
1311 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1312 Collect SpamAssassin and rejection statistics.
1313 Don't display local sender or destination tables unless
1314 there is data to show.
1315 Added average volumes into the top table text output.
1317 SC/04 Eximstats V1.46
1318 Collect data on the number of addresses (recipients)
1319 as well as the number of messages.
1321 SC/05 Eximstats V1.47
1322 Added 'Message too big' to the list of mail rejection
1323 reasons (thanks to Marco Gaiarin).
1325 SC/06 Eximstats V1.48
1326 Mainlog lines which have GMT offsets and are too short to
1327 have a flag are now skipped.
1329 SC/07 Eximstats V1.49 (Alain Williams)
1330 Added the -emptyok flag.
1332 SC/08 Eximstats V1.50
1333 Fixes for obtaining the IP address from reject messages.
1335 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1336 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1337 whitesspace changes from 4.64-PH/27
1339 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1342 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1343 are found, allow negative numbers in numeric criteria)
1345 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1347 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1348 to match changes made in 4.64-PH/43
1350 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1352 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1353 "domain" is actually an IP address, and force a failure. This locks out
1354 those revolvers/nameservers that support "A-for-A" lookups, in
1355 contravention of the specifications.
1357 PH/48 When a host name was looked up from an IP address, and the subsequent
1358 forward lookup of the name timed out, the host name was left in
1359 $sender_host_name, contrary to the specification.
1361 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1362 restricted to single-key lookups, Exim was not diagnosing an error if
1363 * or *@ was used with a query-style lookup.
1365 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1367 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1368 long ago, but noone interested enough thought of it. Let's just say that
1369 the "1.1" means that there are some new functions that weren't there at
1370 some point in the past.
1372 PH/51 Error processing for expansion failure of helo_data from an smtp
1373 transport during callout processing was broken.
1375 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1376 tested/used via the -bh/-bhc/-bs options.
1378 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1379 bug, fixed in subsequent PCRE releases).
1381 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1382 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1384 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1390 SC/01 Use a glob alias rather than an array ref in eximstats generated
1391 parser. This improves both readability and performance.
1393 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1394 Don't display local sender or destination tables in eximstats unless
1395 there is data to show.
1396 Added average volumes into the eximstats top table text output.
1398 SC/03 Collect data on the number of addresses (recipients) as well
1399 as the number of messages in eximstats.
1401 TF/01 Correct an error in the documentation for the redirect router. Exim
1402 does (usually) call initgroups() when daemonizing.
1404 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1405 with consistent privilege compared to when running as a daemon.
1407 TF/03 Note in the spec that $authenticated_id is not set for local
1408 submissions from trusted users.
1410 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1411 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1413 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1414 by adding some example configuration directives to the default
1415 configuration file. A little bit of work is required to uncomment the
1416 directives and define how usernames and passwords are checked, but
1417 there is now a framework to start from.
1419 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1420 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1421 without this. I don't know how relevant this is to other LDAP libraries.
1423 PH/02 Add the verb name to the "unknown ACL verb" error.
1425 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1427 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1429 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1430 directory not expanded when it should be if an expanded home directory
1431 was set for the address (which is overridden by the transport).
1433 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1436 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1437 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1438 because it is too late at that time, and has no effect.
1440 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1441 security issue with \' (bugzilla #107). I could not use the
1442 PQescapeStringConn() function, because it needs a PGconn value as one of
1445 PH/08 When testing addresses using -bt, indicate those final addresses that
1446 are duplicates that would not cause an additional delivery. At least one
1447 person was confused, thinking that -bt output corresponded to deliveries.
1448 (Suppressing duplicates isn't a good idea as you lose the information
1449 about possibly different redirections that led to the duplicates.)
1451 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1452 systems where poll() doesn't work, in particular OS X.
1454 PH/10 Added more information to debugging output for retry time not reached.
1456 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1457 operations in malware.c.
1459 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1462 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1463 syslog_duplication set false, log lines that would normally be written
1464 both the the main log and to the reject log were not written to syslog at
1467 PH/14 In the default configuration, change the use of "message" in ACL warn
1468 statements to "add_header".
1470 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1471 not followed by a command (e.g. "seen endif").
1473 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1474 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1477 PH/17 Added extra conditions to the default value of delay_warning_condition
1481 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1482 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1483 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1486 The Auto-Submitted: and various List- headers are standardised, whereas I
1487 don't think Precedence: ever was.
1489 PH/18 Refactored debugging code in route_finduser() to show more information,
1490 in particular, the error code if getpwnam() issues one.
1492 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1493 This is apparently needed in addition to the PH/07 change above to avoid
1494 any possible encoding problems.
1496 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1497 but not after initializing Perl.
1499 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1500 output them only if debugging. By default they are written stderr,
1501 apparently, which is not desirable.
1503 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1506 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1509 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1511 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1512 authenticated or an ident call has been made. Suppress the default
1513 values for $authenticated_id and $authenticated_sender (but permit -oMai
1514 and -oMas) when testing with -bh.
1516 PH/24 Re-jigged the order of the tests in the default configuration so that the
1517 tests for valid domains and recipients precede the DNS black list and CSA
1518 tests, on the grounds that those ones are more expensive.
1520 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1521 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1522 This bug exists in every version of Exim that I still have, right back to
1525 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1526 However, an attempt to turn on case-sensitivity in a regex key by
1527 including (?-i) didn't work because the subject string was already
1528 lowercased, and the effects were non-intuitive. It turns out that a
1529 one-line patch can be used to allow (?-i) to work as expected.
1535 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1536 other effects) broke the use of negated acl sub-conditions.
1538 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1541 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1542 "Deny" causes Exim to reject the incoming connection with a 554 error.
1543 Unfortunately, if there is a major crisis, such as a disk failure,
1544 tcp-wrappers gives "deny", whereas what one would like would be some
1545 kind of temporary error. A kludge has been added to help with this.
1546 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1547 554 error is used if errno is still zero or contains ENOENT (which occurs
1548 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1551 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1553 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1554 errors. Otherwise a message that provokes a temporary error (when other
1555 messages do not) can cause a whole host to time out.
1557 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1558 addresses were routed directly to files or pipes from a redirect router.
1559 File deliveries just didn't batch; pipe deliveries might have suffered
1562 PH/06 A failure to get a lock for a hints database would erroneously always say
1563 "Failed to get write lock", even when it was really a read lock.
1565 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1566 of 0600. This has been changed to use the value of the lockfile_mode
1567 option (which defaults to 0600).
1569 PH/08 Applied small patch from the Sieve maintainer.
1571 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1572 folder from quota calculations, a direct delivery into this folder messed
1573 up the contents of the maildirsize file. This was because the regex was
1574 used only to exclude .Trash (or whatever) when the size of the mailbox
1575 was calculated. There was no check that a delivery was happening into an
1576 excluded directory. This bug has been fixed by ignoring all quota
1577 processing for deliveries into excluded directories.
1579 PH/10 Added the maildirfolder_create_regex option to appendfile.
1585 PH/01 The code for finding all the local interface addresses on a FreeBSD
1586 system running IPv6 was broken. This may well have applied to all BSD
1587 systems, as well as to others that have similar system calls. The broken
1588 code found IPv4 interfaces correctly, but gave incorrect values for the
1589 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1590 that it would not match correctly against @[] and not recognize the IPv6
1593 PH/02 The ipliteral router was not recognizing addresses of the form user@
1594 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1596 PH/03 Added disable_ipv6.
1598 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1599 decoded form, because it is most often used to construct To: headers
1600 lines in autoreplies, and the decoded form may well be syntactically
1601 invalid. However, $reply_address has leading white space removed, and all
1602 newlines turned into spaces so that the autoreply transport does not
1605 PH/05 If group was specified without a user on a router, and no group or user
1606 was specified on a transport, the group from the router was ignored.
1608 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1609 for visible compile-time settings that can be used to change these
1610 numbers, for those that want even more. Backwards compatibility with old
1611 spool files has been maintained. However, going back to a previous Exim
1612 release will lost any variables that are in spool files.
1614 PH/07 Two small changes when running in the test harness: increase delay when
1615 passing a TCP/IP connection to a new process, in case the original
1616 process has to generate a bounce, and remove special handling of
1617 127.0.0.2 (sic), which is no longer necessary.
1619 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1620 be the same on different OS.
1622 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1625 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1626 whether --show-vars was specified or not
1628 JJ/02 exipick: Added support for new ACL variable spool format introduced
1631 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1632 syntactically invalid From: or Reply-to: line, and a filter used this to
1633 generate an autoreply, and therefore failed to obtain an address for the
1634 autoreply, Exim could try to deliver to a non-existent relative file
1635 name, causing unrelated and misleading errors. What now happens is that
1636 it logs this as a hard delivery error, but does not attempt to create a
1639 PH/11 The exinext utility has a -C option for testing purposes, but although
1640 the given file was scanned by exinext itself; it wasn't being passed on
1641 when Exim was called.
1643 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1644 an end-of-file indication when reading a command response.
1646 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1647 compiled. In many other places in Exim, IPv6 addresses are always
1648 recognized, so I have changed this. It also means that IPv4 domain
1649 literals of the form [IPV4:n.n.n.n] are now always recognized.
1651 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1652 used if the router is not running as root, for example, when verifying at
1653 ACL time, or when using -bh. The debugging output from this situation was
1654 non-existent - all you got was a failure to exec. I have made two
1657 (a) Failures to set uid/gid, the current directory, or a process leader
1658 in a subprocess such as that created by queryprogram now generate
1659 suitable debugging ouput when -d is set.
1661 (b) The queryprogram router detects when it is not running as root,
1662 outputs suitable debugging information if -d is set, and then runs
1663 the subprocess without attempting to change uid/gid.
1665 PH/15 Minor change to Makefile for building test_host (undocumented testing
1668 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1669 additional section of a DNS packet that returns MX or SRV records.
1670 Instead, it always explicitly searches for A/AAAA records. This avoids
1671 major problems that occur when a DNS server includes only records of one
1672 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1673 fixed another bug: if SRV records were looked up and the corresponding
1674 address records were *not* found in the additional section, the port
1675 values from the SRV records were lost.
1677 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1678 using the correct key (the original address) when searching the retry
1679 rules in order to find which one to use for generating the retry hint.
1681 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1682 adding the default one. Similarly, if it contains a Reply-To: header, the
1683 errors_reply_to option, if set, is not used.
1685 PH/19 When calculating a retry time, Exim used to measure the "time since
1686 failure" by looking at the "first failed" field in the retry record. Now
1687 it does not use this if it is later than than the arrival time of the
1688 message. Instead it uses the arrival time. This makes for better
1689 behaviour in cases where some deliveries succeed, thus re-setting the
1690 "first failed" field. An example is a quota failure for a huge message
1691 when small messages continue to be delivered. Without this change, the
1692 "time since failure" will always be short, possible causing more frequent
1693 delivery attempts for the huge message than are intended.
1694 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1696 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1697 $1, $2, $3) because the numerical variables can be reset during some
1698 expansion items (e.g. "match"), thereby losing the authentication data.
1700 PH/21 Make -bV show the size of off_t variables so that the test suite can
1701 decide whether to run tests for quotas > 2G.
1703 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1704 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1705 filecount value is greater than 2G or if a quota value is greater than 2G
1706 on a system where the size of off_t is not greater than 4, a panic error
1709 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1710 never match. The debug and -bh output now contains an explicit error
1711 message indicating a malformed IPv4 address or mask.
1713 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1714 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1715 PH/23 above applies.
1717 PH/25 Do not write to syslog when running in the test harness. The only
1718 occasion when this arises is a failure to open the main or panic logs
1719 (for which there is an explicit test).
1721 PH/26 Added the /no_tell option to "control=freeze".
1723 PH/27 If a host name lookup failed very early in a connection, for example, if
1724 the IP address matched host_lookup and the reverse lookup yielded a name
1725 that did not have a forward lookup, an error message of the form "no IP
1726 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1727 could be logged. Now it outputs the IP address instead of "NULL".
1729 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1730 allows the sender and the authenticated sender to be set when
1731 submitting a message from within Exim. Since child_open_exim() is
1732 documented for local_scan(), the new function should be too.
1734 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1735 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1736 results in an empty string is now treated as unset.
1738 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1740 PH/31 Added sender_verify_fail logging option.
1742 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1743 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1744 tidied the source and removed it altogether.
1746 PH/33 When a queue run was abandoned because the load average was too high, a
1747 log line was always written; now it is written only if the queue_run log
1748 selector is set. In addition, the log line for abandonment now contains
1749 information about the queue run such as the pid. This is always present
1750 in "start" and "stop" lines but was omitted from the "abandon" line.
1752 PH/34 Omit spaces between a header name and the colon in the error message that
1753 is given when verify = headers_syntax fails (if there are lots of them,
1754 the message gets confusing).
1756 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1757 names, as there are now some PTR records that contain slashes. This check
1758 is only to protect against broken name servers that fall over on strange
1759 characters, so the fact that it applies to all lookups doesn't matter.
1761 PH/36 Now that the new test suite is complete, we can remove some of the
1762 special code in Exim that was needed for the old test suite. For example,
1763 sorting DNS records because real resolvers return them in an arbitrary
1764 order. The new test suite's fake resolver always returns records in the
1767 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1768 bounces) except when queue_only is set, to avoid logging races between
1769 the different processes.
1771 PH/38 Panic-die if .include specifies a non-absolute path.
1773 PH/39 A tweak to the "H" retry rule from its user.
1775 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1776 a label. They prevented compilation on older perls.
1778 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1779 a warning to be raised on newish perls.
1781 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1782 on queue. Changes to match documented behaviour of showing count of
1783 messages matching specified criteria.
1785 PH/40 Changed the default ident timeout from 30s to 5s.
1787 PH/41 Added support for the use of login_cap features, on those BSD systems
1788 that have them, for controlling the resources used by pipe deliveries.
1790 PH/42 The content-scanning code uses fopen() to create files in which to put
1791 message data. Previously it was not paying any attention to the mode of
1792 the files. Exim runs with umask(0) because the rest of the code creates
1793 files with open(), and sets the required mode explicitly. Thus, these
1794 files were ending up world-writeable. This was not a big issue, because,
1795 being within the spool directory, they were not world-accessible. I have
1796 created a function called modefopen, which takes an additional mode
1797 argument. It sets umask(777), creates the file, chmods it to the required
1798 mode, then resets the umask. All the relevant calls to fopen() in the
1799 content scanning code have been changed to use this function.
1801 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1802 to 24 hours. This avoids potential overflow problems when processing G
1803 and H retry rules. I suspect nobody ever tinkers with this value.
1805 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1807 PH/45 When the plaintext authenticator is running as a client, the server's
1808 challenges are checked to ensure they are valid base64 strings. By
1809 default, the authentication attempt is cancelled if an invalid string is
1810 received. Setting client_ignore_invalid_base64 true ignores these errors.
1811 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1812 they are received. Thus, the responses can be made to depend on the
1813 challenges. If an invalid string is ignored, an empty string is placed in
1816 PH/46 Messages that are created by the autoreply transport now contains a
1817 References: header, in accordance with RFCs 2822 and 3834.
1819 PH/47 Added authenticated_sender_force to the smtp transport.
1821 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1823 PH/49 Installed latest patch from the Sieve maintainer.
1825 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1826 was also set, the appendfile transport was unnecessarily scanning a
1827 directory of message files (e.g. for maildir delivery) to find the count
1828 of files (along with the size), even though it did not need this
1829 information. It now does the scan only if it needs to find either the
1830 size of the count of files.
1832 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1834 PH/52 Two bugs concerned with error handling when the smtp transport is
1837 (i) Exim was not creating retry information for temporary errors given
1838 for individual recipients after the DATA command when the smtp transport
1839 was used in LMTP mode. This meant that they could be retried too
1840 frequently, and not timed out correctly.
1842 (ii) Exim was setting the flag that allows error details to be returned
1843 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1844 recipients that were returned after the DATA command.
1846 PH/53 This is related to PH/52, but is more general: for any failing address,
1847 when detailed error information was permitted to be returned to the
1848 sender, but the error was temporary, then after the final timeout, only
1849 "retry timeout exceeded" was returned. Now it returns the full error as
1850 well as "retry timeout exceeded".
1852 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1853 do this, and (what is worse) MTAs that accept it.
1855 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1856 will now be deprecated.
1858 PH/56 New os.c-cygwin from the Cygwin maintainer.
1860 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1861 formats (previously only available in exim formats via -bpr, -bpru,
1862 and -bpra. Now also available in native and exiqgrep formats)
1864 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1865 with very large, slow to parse queues
1867 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1869 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1871 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1872 responses to authentication challenges, though it was showing the
1873 challenges; (ii) I've removed the CR characters from the debug output for
1876 PH/58 Allow for the insertion of a newline as well as a space when a string
1877 is turned into more than one encoded-word during RFC 2047 encoding. The
1878 Sieve code now uses this.
1880 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1881 data_4xx, lost_connection, tls_required.
1883 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1884 message was being sent as an SMTP response.
1886 PH/61 Add -l and -k options to exicyclog.
1888 PH/62 When verifying, if an address was redirected to one new address, so that
1889 verification continued, and the new address failed or deferred after
1890 having set something in $address_data, the value of $address_data was not
1891 passed back to the ACL. This was different to the case when no
1892 redirection occurred. The value is now passed back in both cases.
1894 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1895 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1896 use login_cap.h, so on its own it isn't the distinguishing feature. The
1897 new name refers directly to the setclassresources() function.
1899 PH/65 Added configuration files for NetBSD3.
1901 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1903 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1904 is preferred over IPv4.
1906 PH/68 The bounce_return_message and bounce_return_body options were not being
1907 honoured for bounces generated during the reception of non-SMTP messages.
1908 In particular, this applied to messages rejected by the ACL. This bug has
1909 been fixed. However, if bounce_return_message is true and bounce_return_
1910 body is false, the headers that are returned for a non-SMTP message
1911 include only those that have been read before the error was detected.
1912 (In the case of an ACL rejection, they have all been read.)
1914 PH/69 The HTML version of the specification is now built in a directory called
1915 spec_html instead of spec.html, because the latter looks like a path with
1916 a MIME-type, and this confuses some software.
1918 PH/70 Catch two compiler warnings in sieve.c.
1920 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1921 function verify_get_ident() calls ip_connect() to connect a socket, but
1922 if the "connect()" function timed out, ip_connect() used to close the
1923 socket. However, verify_get_ident() also closes the socket later, and in
1924 between Exim writes to the log, which may get opened at this point. When
1925 the socket was closed in ip_connect(), the log could get the same file
1926 descriptor number as the socket. This naturally causes chaos. The fix is
1927 not to close the socket in ip_connect(); the socket should be closed by
1928 the function that creates it. There was only one place in the code where
1929 this was missing, in the iplookup router, which I don't think anybody now
1930 uses, but I've fixed it anyway.
1932 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1933 well as to direct DNS lookups. Otherwise the handling of names in host
1934 lists is inconsistent and therefore confusing.
1940 PH/01 Two changes to the default runtime configuration:
1942 (1) Move the checks for relay_from_hosts and authenticated clients from
1943 after to before the (commented out) DNS black list checks.
1945 (2) Add control=submission to the relay_from_hosts and authenticated
1946 clients checks, on the grounds that messages accepted by these
1947 statements are most likely to be submissions.
1949 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1951 (1) Generate an error if the third argument for the ${prvs expansion is
1954 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1957 (3) Reset the variables that are obtained from the first argument of
1958 ${prvscheck and used in the second argument before leaving the code,
1959 because their memory is reclaimed, so using them afterwards may do
1962 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1963 one (it's much easier than Tom thought :-).
1965 (5) Because of (4), we can now allow for the use of $prvscheck_result
1966 inside the third argument.
1968 PH/03 For some reason, the default setting of PATH when running a command from
1969 a pipe transport was just "/usr/bin". I have changed it to
1972 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1973 anything to be listed in the output from -bV.
1975 PH/05 When a filter generated an autoreply, the entire To: header line was
1976 quoted in the delivery log line, like this:
1978 => >A.N.Other <ano@some.domain> <original@ddress> ...
1980 This has been changed so that it extracts the operative address. There
1981 may be more than one such address. If so, they are comma-separated, like
1984 => >ano@some.domain,ona@other.domain <original@ddress> ...
1986 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1987 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1988 not being looked up in the rDNS to get a host name, Exim was showing the
1989 IP address twice in Received: lines, even though the IP addresses were
1990 identical. For example:
1992 Received: from [1.2.3.4] (helo=[1.2.3.4])
1994 However, if the real host name was known, it was omitting the HELO data
1995 if it matched the actual IP address. This has been tidied up so that it
1996 doesn't show the same IP address twice.
1998 PH/07 When both +timestamp and +memory debugging was on, the value given by
1999 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2000 called by the debug printing, thereby overwriting the timestamp buffer.
2001 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2003 PH/08 When the original message was included in an autoreply transport, it
2004 always said "this is a copy of the message, including all the headers",
2005 even if body_only or headers_only was set. It now gives an appropriate
2008 PH/09 Applied a patch from the Sieve maintainer which:
2010 o fixes some comments
2011 o adds the (disabled) notify extension core
2012 o adds some debug output for the result of if/elsif tests
2013 o points to the current vacation draft in the documentation
2014 and documents the missing references header update
2018 o fixes a bug in processing the envelope test (when testing
2019 multiple envelope elements, the last element determinted the
2022 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2023 Electronic Mail") by including:
2025 Auto-submitted: auto-generated
2027 in the messages that it generates (bounce messages and others, such as
2028 warnings). In the case of bounce messages for non-SMTP mesages, there was
2029 also a typo: it was using "Auto_submitted" (underscore instead of
2030 hyphen). Since every message generated by Exim is necessarily in response
2031 to another message, thes have all been changed to:
2033 Auto-Submitted: auto-replied
2035 in accordance with these statements in the RFC:
2037 The auto-replied keyword:
2039 - SHOULD be used on messages sent in direct response to another
2040 message by an automatic process,
2042 - MUST NOT be used on manually-generated messages,
2044 - MAY be used on Delivery Status Notifications (DSNs) and Message
2045 Disposition Notifications (MDNs),
2047 - MUST NOT be used on messages generated by automatic or periodic
2048 processes, except for messages which are automatic responses to
2051 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2052 to the default Received: header definition.
2054 PH/12 Added log selector acl_warn_skipped (default on).
2056 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2057 variables because (a) they are in the wrong storage pool and (b) even if
2058 they were copied, it wouldn't work properly because of the caching.
2060 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2061 checking when decoding. Apparently there are clients that generate
2062 overlong encoded strings. Why am I not surprised?
2064 PH/15 If the first argument of "${if match_address" was not empty, but did not
2065 contain an "@" character, Exim crashed. Now it writes a panic log message
2066 and treats the condition as false.
2068 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2070 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2071 extension "envelope-auth". The code is finished and in agreement with
2072 other implementations, but there is no documentation so far and in fact,
2073 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2074 not changing the active code.
2076 Print executed "if" and "elsif" statements when debugging is used. This
2077 helps a great deal to understand what a filter does.
2079 Document more things not specified clearly in RFC3028. I had all this
2080 sorted out, when out of a sudden new issues came to my mind. Oops."
2082 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2085 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2086 canonical form (as documented). However, after a host name lookup from
2087 the IP address, check_host() was doing a simple string comparison with
2088 addresses acquired from the DNS when checking that the found name did
2089 have the original IP as one of its addresses. Since any found IPv6
2090 addresses are likely to be in abbreviated form, the comparison could
2091 fail. Luckily, there already exists a function for doing the comparison
2092 by converting both addresses to binary, so now that is used instead of
2093 the text comparison.
2095 PH/20 There was another similar case to PH/19, when a complete host name was
2096 given in a host list; looking up its IP address could give an abbreviated
2097 form, whereas the current host's name might or might not be abbreviated.
2098 The same fix has been applied.
2104 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2105 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2108 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2109 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2111 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2113 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2114 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2115 Also, for "Auto-Submitted", treat anything other than "no" as
2116 non-personal, in accordance with RFC 3834. (Previously it treated
2117 anything starting "auto-" as non-personal.)
2119 TF/01 The control=submission/name=... option had a problem with syntax
2120 errors if the name included a slash character. The /name= option
2121 now slurps the rest of the string, so it can include any characters
2122 but it must come last in the list of options (after /sender_retain
2125 PH/05 Some modifications to the interface to the fake nameserver for the new
2133 TK/01 Added the "success_on_redirect" address verification option. See
2134 NewStuff for rationale and an example.
2136 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2138 PH/02 Patch to exigrep to allow it to work on syslog lines.
2140 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2141 fread() to skip over the body file's header line, because in Cygwin the
2142 header line is locked and is inaccessible.
2144 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2145 co-exist for some time) to make it clear that it is the Exim ID that is
2146 referenced, not the Message-ID: header line.
2148 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2149 string_format() function, because snprintf() does not exist on all
2152 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2153 ${stat: expansion item.
2155 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2156 protocol synchronization error", to keep the pedants happy.
2158 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2159 well as for IRIX systems, when gcc is being used. See the host.c source
2162 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2164 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2167 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2168 transport and to the smtp transport in LMTP mode.
2170 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2172 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2173 run a filter in a subprocess. This could lead to confusion in subsequent
2174 lookups in the parent process. There should also be a search_tidyup() at
2175 the end of the subprocess.
2177 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2178 only if the host matched helo_try_verify_hosts, which caused the
2179 verification to occur when the EHLO/HELO command was issued. The ACL just
2180 tested the remembered result. Now, if a previous verification attempt has
2181 not happened, "verify = helo" does it there and then.
2183 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2185 TK/03 Fix log output including CR from clamd.
2187 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2188 exist provoked a memory error which could cause a segfault.
2190 PH/15 Installed PCRE 6.2
2192 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2194 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2195 of the problem. Specifically, suggested +O2 rather than +O1 for the
2198 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2200 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2201 record, the port was not correctly logged when the outgoing_port log
2202 selector was set (it logged the transort's default port).
2204 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2205 fallback_hosts, and "hosts" in the smtp transport.
2207 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2208 host errors such as "Connection refused".
2210 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2211 authentication with radiusclient 0.4.9:
2213 - Error returned from rc_read_config was caught wrongly
2214 - Username/password not passed on to radius server due to wrong length.
2216 The presumption is that some radiusclient API changes for 4.51/PH/17
2217 were not taken care of correctly. The code is still untested by me (my
2218 Linux distribution still has 0.3.2 of radiusclient), but it was
2219 contributed by a Radius user.
2221 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2222 expanding the "port" option of the smtp transport.
2224 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2225 while reading a MIME header. Thanks to Tom Hughes for a patch.
2227 PH/24 Include config.h inside local_scan.h so that configuration settings are
2230 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2231 that in an ACL for RCPT (for example), you can examine exactly what was
2234 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2235 commands, but it was not correctly comparing the address with the actual
2236 client host address. Thus, it would show the EHLO address in Received:
2237 header lines when this was not necessary.
2239 PH/27 Added the % operator to ${eval:}.
2241 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2242 it should be ignoring failures (because with -C, for example, it has lost
2243 privilege). It wasn't ignoring creation failures other than "already
2246 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2249 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2250 because an input line was too long, either on its own, or by virtue of
2251 too many continuations, the temporary file was not being removed, and the
2252 return code was incorrect.
2254 PH/31 Missing "BOOL" in function definition in filtertest.c.
2256 PH/32 Applied Sieve patches from the maintainer.
2258 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2260 PH/33 Added "verify = not_blind".
2262 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2263 Local/Makefile (with some defaults set). These are used in built scripts
2264 such as exicyclog, but they have never been used in the exim_install
2265 script (though there are many overriding facilities there). I have
2266 arranged that the exim_install script now takes note of these two
2269 PH/35 Installed configuration files for Dragonfly.
2271 PH/36 When a locally submitted message by a trusted user did not contain a
2272 From: header, and the sender address was obtained from -f or from an SMTP
2273 MAIL command, and the trusted user did not use -F to supply a sender
2274 name, $originator_name was incorrectly used when constructing a From:
2275 header. Furthermore, $originator_name was used for submission mode
2276 messages from external hosts without From: headers in a similar way,
2277 which is clearly wrong.
2279 PH/37 Added control=suppress_local_fixups.
2281 PH/38 When log_selector = +received_sender was set, and the addition of the
2282 sender made the log line's construction buffer exactly full, or one byte
2283 less than full, an overflow happened when the terminating "\n" was
2286 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2287 when the result of a list match is failure because a DNS lookup failed.
2289 PH/40 RM_COMMAND is now used in the building process.
2291 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2292 the "build-* directories that it finds.
2294 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2295 address in a domain literal was a prefix of an interface address.
2297 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2298 when verifying a sender address, unless rewrite_headers is false.
2300 PH/44 Wrote a long comment about why errors_to addresses are verified as
2301 recipients, not senders.
2303 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2304 the ratelimit ACL was added.
2306 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2308 PH/46 Added extra information about PostgreSQL errors to the error string.
2310 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2311 avoiding the need to install special zones in a real server. This is
2312 backwards compatible; if it can't find the fake resolver, it drops back.
2313 Thus, both old and new test suites can be run.
2315 TF/02 Added util/ratelimit.pl
2317 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2318 clock is set back in time.
2320 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2321 Candler <B.Candler@pobox.com>.
2323 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2324 OK for addresses that are the result of redirections.
2326 PH/48 A number of further additions for the benefit of the new test suite,
2327 including a fake gethostbyname() that interfaces to the fake DNS resolver
2330 TF/06 The fix for widen_domains has also been applied to qualify_single and
2331 search_parents which are the other dnslookup options that can cause
2334 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2337 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2339 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2340 service" if it encountered an MX record with an empty target hostname.
2341 The message is now "an MX or SRV record indicated no SMTP service".
2343 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2344 if the DNS of the sending site is misconfigured. This is quite a
2345 common situation. This change restores the behaviour of treating a
2346 helo verification defer as a failure.
2348 PH/51 If self=fail was set on a router, the bounce message did not include the
2349 actual error message.
2355 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2357 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2358 command itself ended in error, the underlying message about the transport
2359 filter timeout was being overwritten with the pipe command error. Now the
2360 underlying error message should be appended to the second error message.
2362 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2364 PH/02 Reduce the amount of output that "make" produces by default. Full output
2365 can still be requested.
2367 PH/03 The warning log line about a condition test deferring for a "warn" verb
2368 was being output only once per connection, rather than after each
2369 occurrence (because it was using the same function as for successful
2370 "warn" verbs). This seems wrong, so I have changed it.
2372 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2373 it should not have, which might have caused a crash in the right
2374 circumstances, but probably never did.
2376 PH/04 Installed a modified version of Tony Finch's patch to make submission
2377 mode fix the return path as well as the Sender: header line, and to
2378 add a /name= option so that you can make the user's friendly name appear
2381 TF/03 Added the control = fakedefer ACL modifier.
2383 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2384 Mark Lowes for thorough testing.
2386 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2388 TK/03 Merged latest SRS patch from Miles Wilton.
2390 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2391 with the definition in sysexits.h (which is #included earlier).
2392 Fortunately, Exim does not actually use EX_OK. The code used to try to
2393 preserve the sysexits.h value, by assumimg that macro definitions were
2394 scanned for macro replacements. I have been disabused of this notion,
2395 so now the code just undefines EX_OK before #including unistd.h.
2397 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2398 in the smtp transport. When a block could not be written in a single
2399 write() function, the timeout was being re-applied to each part-write.
2400 This seems wrong - if the receiver was accepting one byte at a time it
2401 would take for ever. The timeout is now adjusted when this happens. It
2402 doesn't have to be particularly precise.
2404 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2405 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2407 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2408 without a domain if the check to <postmaster@domain> fails.
2410 SC/01 Eximstats: added -xls and the ability to specify output files
2411 (patch written by Frank Heydlauf).
2413 SC/02 Eximstats: use FileHandles for outputing results.
2415 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2417 SC/04 Eximstats: fixed display of large numbers with -nvr option
2419 SC/05 Eximstats: fixed merging of reports with empty tables.
2421 SC/06 Eximstats: added the -include_original_destination flag
2423 SC/07 Eximstats: removed tabs and trailing whitespace.
2425 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2427 TK/06 MBOX spool code: Add real "From " MBOX separator line
2428 so the .eml file is really in mbox format (even though
2429 most programs do not really care). Patch from Alex Miller.
2431 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2432 The latter is generated from $received_to and is only set if the
2433 message has one envelope recipient. SA can use these headers,
2434 obviously out-of-the-box. Patch from Alex Miller.
2436 PH/08 The ${def test on a variable was returning false if the variable's
2437 value was "0", contrary to what the specification has always said!
2438 The result should be true unless the variable is empty.
2440 PH/09 The syntax error of a character other than { following "${if
2441 def:variable_name" (after optional whitespace) was not being diagnosed.
2442 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2443 accidental colon was present, for example, could give incorrect results.
2445 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2446 result is used (not including appendfile, where other changes are about
2449 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2450 This involved changing a lot of size variables from int to off_t. It
2451 should work with maildirs and everything.
2453 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2454 spamd dying while we are connected to it.
2456 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2459 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2460 function for BDB 4.3.
2462 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2464 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2465 the sources for PCRE have been re-arranged and more files are now
2468 PH/15 The code I had for printing potentially long long variables in PH/11
2469 above was not the best (it lost precision). The length of off_t variables
2470 is now inspected at build time, and an appropriate printing format (%ld
2471 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2472 to be "long long int" or "long int". This is needed for the internal
2473 formatting function string_vformat().
2475 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2476 the configuration file to be ":syslog", then the script "guesses" where
2477 the logs files are, rather than using the compiled in default. In our
2478 case the guess is not the same as the compiled default, so the script
2479 suddenly stopped working when I started to use syslog. The patch checks
2480 to see if log_file_path is "". If so, it attempts to read it from exim
2481 with no configuration file to get the compiled in version, before it
2482 falls back to the previous guessing code."
2484 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2485 implementing BATV in an Exim configuration. See NewStuff for the gory
2488 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2489 Makefile that are specific to HP-UX.
2491 PH/18 If the "use_postmaster" option was set for a recipient callout together
2492 with the "random" option, the postmaster address was used as the MAIL
2493 FROM address for the random test, but not for the subsequent recipient
2494 test. It is now used for both.
2496 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2497 patch removes a few documentation additions to RFC 3028, because the
2498 latest draft now contains them. It adds the new en;ascii-case comparator
2499 and a new error check for 8bit text in MIME parts. Comparator and
2500 require names are now matched exactly. I enabled the subaddress
2501 extension, but it is not well tested yet (read: it works for me)."
2503 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2504 rework some of the code of TK/09 above to avoid the hardwired use of
2505 "%lld" and "long long". Replaced the call to snprintf() with a call to
2508 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2509 records point to non-existent hosts", "retry timeout exceeded", and
2510 "retry time not reached for any host after a long failure period".
2512 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2513 experimental DomainKeys support:
2515 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2516 (2) On an error such as an illegally used "control", the wrong name for
2517 the control was given.
2519 These problems did NOT occur unless DomainKeys support was compiled.
2521 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2523 PH/24 Added ${if match_ip condition.
2525 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2526 databases so that it will be absolutely obvious if a crash occurs in the
2527 DB library. This is a regular occurrence (often caused by mis-matched
2530 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2531 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2532 picked up on a user's system that detects such things. There doesn't seem
2533 to be a gcc warning option for this - only an attribute that has to be
2534 put on the function's prototype. It seems that in Fedora Core 4 they have
2535 set this on a number of new functions. No doubt there will be more in due
2538 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2539 specify a transport. However, if an address that was verified by such a
2540 router was the subject of a callout, Exim crashed because it tried to
2541 read the rcpt_include_affixes from the non-existent transport. Now it
2542 just assumes that the setting of that option is false. This bug was
2543 introduced by 4.51/PH/31.
2545 PH/28 Changed -d+all to exclude +memory, because that information is very
2546 rarely of interest, but it makes the output a lot bigger. People tend to
2547 do -d+all out of habit.
2549 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2550 code in os-type was giving problems when libc.so lives in lib64, like on
2553 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2554 aren't the modern standard, and it seems that some systems' include files
2555 don't always have them. Exim was already checking for some of the newer
2556 ones like T_AAAA, and defining it itself. I've added checks for all the
2557 record types that Exim uses.
2559 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2560 not automatically generating a new one, as it is supposed to. This
2561 prevented TLS from working. If the file did exist, but contained invalid
2562 data, a new version was generated, as expected. It was only the case of a
2563 non-existent file that was broken.
2565 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2566 with a change in libdomainkeys > 0.64.
2568 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2569 from DNS. If the selector record carries the flag, it now has
2570 precedence over the domain-wide flag.
2572 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2574 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2575 the use of a port name that isn't defined in /etc/services) occurred, the
2576 message was deferred as in a normal delivery, and thus remained on the
2577 spool, instead of being failed because of the mua_wrapper setting. This
2578 is now fixed, and I tidied up some of the mua_wrapper messages at the
2581 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2582 the messages in a hash of arrays rather than using individual hashes.
2583 This is a bit cleaner and results in dramatic memory savings, albeit
2584 at a slight CPU cost.
2586 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2587 as requested by Marc Sherman.
2589 SC/10 Eximstats: added histograms for user specified patterns as requested
2592 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2594 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2595 fopen() in the content-scanning modules that did not already have it.
2601 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2602 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2604 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2606 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2608 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2609 to test Sieve filters that use "vacation".
2611 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2612 that changes the way the GnuTLS parameters are stored in the cache file.
2613 The new format can be generated externally. For backward compatibility,
2614 if the data in the cache doesn't make sense, Exim assumes it has read an
2615 old-format file, and it generates new data and writes a new file. This
2616 means that you can't go back to an older release without removing the
2619 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2620 work if there are any delivery delays because "one_time" forces the
2621 parent to be marked "delivered", so its unseen clone is never tried
2622 again. For this reason, Exim now forbids the simultaneous setting of
2625 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2626 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2627 entry if you want to know the details. The fix, however, neglected to
2628 consider the case when local delivery batching is involved. The test for
2629 "previously delivered" was not happening when checking to see if an
2630 address could be batched with a previous (undelivered) one; under
2631 certain circumstances this could lead to multiple deliveries to the same
2634 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2635 in its include files, and this causes problems building Exim.
2637 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2638 header_syntax) but Exim was just ignoring anything given after a slash.
2639 In particular, this caused confusion with an attempt to use "verify =
2640 reverse_host_lookup/defer_ok". An error is now given when options are
2641 supplied for verify items that do not have them. (Maybe reverse_host_
2642 lookup should have a defer_ok option, but that's a different point.)
2644 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2645 defined by RFC 821) to 2048, because there were problems with some AUTH
2646 commands, and RFC 1869 says the size should be increased for extended
2647 SMTP commands that take arguments.
2649 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2652 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2653 "unknown" error; now it says that the functionality isn't in the binary.
2655 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2656 an address' error message when a string expansion fails (syntax or
2657 whatever). Otherwise the password may appear in the log. Following change
2658 PH/42 below, there is no longer a chance of it appearing in a bounce
2661 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2663 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2664 of its addresses was ever tried. (Bugzilla bug #2.)
2666 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2667 the result incorrectly in the debug output. (It correctly added a newline
2668 to what was transported.)
2670 TF/01 Added $received_time.
2672 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2673 commented out examples of how to interface to a virus scanner and to
2674 SpamAssassin. Also added commented examples of av_scanner and
2675 spamd_address settings.
2677 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2678 and controls are allowed in which ACLs. There were a couple of minor
2679 errors. Some of the entries in the conditions table (which is a table of
2680 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2681 negation of where the condition IS allowed.
2683 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2685 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2686 header file does not have a version number, so I've had to invent a new
2687 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2688 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2689 radiusclient), but it was contributed by a Radius user.
2691 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2692 files or directories, for OpenSSL.
2694 PH/19 When an Exim process that is running as root has to create an Exim log
2695 file, it does so in a subprocess that runs as exim:exim so as to get the
2696 ownership right at creation (otherwise, other Exim processes might see
2697 the file with the wrong ownership). There was no test for failure of this
2698 fork() call, which would lead to the process getting stuck as it waited
2699 for a non-existent subprocess. Forks do occasionally fail when resources
2700 run out. I reviewed all the other calls to fork(); they all seem to check
2703 PH/20 When checking for unexpected SMTP input at connect time (before writing
2704 the banner), Exim was not dealing correctly with a non-positive return
2705 from the read() function. If the client had disconnected by this time,
2706 the result was a log entry for a synchronization error with an empty
2707 string after "input=" when read() returned zero. If read() returned -1
2708 (an event I could not check), uninitialized data bytes were printed.
2709 There were reports of junk text (parts of files, etc) appearing after
2712 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2714 PH/22 Added support for macro redefinition, and (re)definition in between
2715 driver and ACL definitions.
2717 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2718 forgetting to use the resulting value; it was using the unexpanded value.
2720 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2721 hadn't been configured. The fix is from Juergen Kreileder, who
2722 understands it better than I do:
2724 "Here's what I see happening with three configured cyrus_sasl
2725 authenticators configured (plain, login, cram-md5):
2727 On startup auth_cyrus_sasl_init() gets called for each of these.
2728 This means three calls to sasl_listmech() without a specified mech_list.
2729 => SASL tests which mechs of all available mechs actually work
2730 => three warnings about OTP not working
2731 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2733 With the patch, sasl_listmech() also gets called three times. But now
2734 SASL's mech_list option is set to the server_mech specified in the the
2735 authenticator. Or in other words, the answer from sasl_listmech()
2736 gets limited to just the mech you're testing for (which is different
2738 => the return list contains just 'plain' or 'login', 'cram-md5' or
2739 nothing depending on the value of ob->server_mech.
2741 I've just tested the patch: Authentication still works fine,
2742 unavailable mechs specified in the exim configuration are still
2743 caught, and the auth.log warnings about OTP are gone."
2745 PH/25 When debugging is enabled, the contents of the command line are added
2746 to the debugging output, even when log_selector=+arguments is not
2749 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2750 answer is "GNU", and only if the return is "GNU/something" is the answer
2753 PH/27 $acl_verify_message is now set immediately after the failure of a
2754 verification in an ACL, and so is available in subsequent modifiers. In
2755 particular, the message can be preserved by coding like this:
2757 warn !verify = sender
2758 set acl_m0 = $acl_verify_message
2760 Previously, $acl_verify_message was set only while expanding "message"
2761 and "log_message" when a very denied access.
2763 PH/28 Modified OS/os.c-Linux with
2765 -#ifndef OS_LOAD_AVERAGE
2766 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2768 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2769 nomenclature these days.)
2771 PH/29 Installed patch from the Sieve maintainer that adds the options
2772 sieve_useraddress and sieve_subaddress to the redirect router.
2774 PH/30 In these circumstances:
2775 . Two addresses routed to the same list of hosts;
2776 . First host does not offer TLS;
2777 . First host accepts first address;
2778 . First host gives temporary error to second address;
2779 . Second host offers TLS and a TLS session is established;
2780 . Second host accepts second address.
2781 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2782 and peerdn, if requested) that were in fact used only for the second
2785 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2786 attention to any local part prefix or suffix that was matched by the
2787 router that accepted the address. It now behaves in the same way as it
2788 does for delivery: the affixes are removed from the local part unless
2789 rcpt_include_affixes is set on the transport.
2791 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2792 timeout during the DATA phase of an incoming message.
2794 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2795 applied a patch sanctioned by the Sieve maintainer.
2797 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2798 the uid or gid is negative. A case of a negative gid caused this to be
2799 noticed. The fix allows for either to be negative.
2801 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2802 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2805 PH/36 The change PH/12 above was broken. Fixed it.
2807 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2808 the grounds that routing the same address twice would always produce the
2809 same answer. This might have been true once, but it is certainly no
2810 longer true now. Routing a child address may depend on the previous
2811 routing that produced that child. Some complicated redirection strategies
2812 went wrong when messages had multiple recipients, and made Exim's
2813 behaviour dependent on the order in which the addresses were given.
2815 I have moved the duplicate checking until after the routing is complete.
2816 Exim scans the addresses that are assigned to local and remote
2817 transports, and removes any duplicates. This means that more work will be
2818 done, as duplicates will always all be routed, but duplicates are
2819 presumably rare, so I don't expect this is of any significance.
2821 For deliveries to pipes, files, and autoreplies, the duplicate checking
2822 still happens during the routing process, since they are not going to be
2825 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2826 It corrects a timeout issue with spamd. This is Ian's comment: "The
2827 background is that sometimes spamd either never reads data from a
2828 connection it has accepted, or it never writes response data. The exiscan
2829 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2830 blindly assumes that writes won't block so it may never time out."
2832 PH/39 Allow G after quota size as well as K and M.
2834 PH/40 The value set for $authenticated_id in an authenticator may not contain
2835 binary zeroes or newlines because the value is written to log lines and
2836 to spool files. There was no check on this. Now the value is run through
2837 the string_printing() function so that such characters are converted to
2838 printable escape sequences.
2840 PH/41 $message_linecount is a new variable that contains the total number of
2841 lines in the message. Compare $body_linecount, which is the count for the
2844 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2845 bounce and delay warning messages, except in certain special cases, which
2848 (a) An SMTP error message from a remote host;
2849 (b) A message specified in a :fail: redirection;
2850 (c) A message specified in a "fail" command in a system filter;
2851 (d) A message specified in a FAIL return from the queryprogram router;
2852 (e) A message specified by the cannot_route_message router option.
2854 In these cases only, Exim does include the error details in bounce and
2855 warning messages. There are also a few cases where bland messages such
2856 as "unrouteable address" or "local delivery error" are given.
2858 PH/43 $value is now also set for the "else" part of a ${run expansion.
2860 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2861 being worked on, but at least Exim now implements the latest version to
2864 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2865 process to complete was treated as a delivery failure, a timeout while
2866 writing the message to the pipe was logged, but erroneously treated as a
2867 successful delivery. Such timeouts include transport filter timeouts. For
2868 consistency with the overall process timeout, these timeouts are now
2869 treated as errors, giving rise to delivery failures by default. However,
2870 there is now a new Boolean option for the pipe transport called
2871 timeout_defer, which, if set TRUE, converts the failures into defers for
2872 both kinds of timeout. A transport filter timeout is now identified in
2875 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2876 systems where "make" and "gmake" are different, calling "gmake" at top
2877 level broke things. I've arranged for the value of $(MAKE) to be passed
2878 from the Makefile to this script so that it can call the same version of
2882 A note about Exim versions 4.44 and 4.50
2883 ----------------------------------------
2885 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2886 changes of various kinds. As a consequence, a big documentation update was
2887 needed. This delayed the release for rather longer than seemed good, especially
2888 in the light of a couple of (minor) security issues. Therefore, the changes
2889 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2890 release. So 4.44 and 4.50 are in effect two different branches that both start
2893 I have left the 4.50 change log unchanged; it contains all the changes since
2894 4.43. The change log for 4.44 is below; many of its items are identical to
2895 those for 4.50. This seems to be the most sensible way to preserve the
2896 historical information.
2902 1. Minor wording change to the doc/README.SIEVE file.
2904 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2905 computation of the current number of files was incorrect.
2907 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2908 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2911 4. Give more explanation in the error message when the command for a transport
2912 filter fails to execute.
2914 5. There are several places where Exim runs a non-Exim command in a
2915 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2916 was being done only for the command run by the queryprogram router. It is
2917 now done for all such subprocesses. The other cases are: ${run, transport
2918 filters, and the commands run by the lmtp and pipe transports.
2920 6. Added CONFIGURE_GROUP build-time option.
2922 7. Some older OS have a limit of 256 on the maximum number of file
2923 descriptors. Exim was using setrlimit() to set 1000 as a large value
2924 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2925 systems. I've change it so that if it can't get 1000, it tries for 256.
2927 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2928 was an oversight, and furthermore, ever since the addition of extra
2929 controls (e.g. 4.43/32), the checks on when to allow different forms of
2930 "control" were broken. There should now be diagnostics for all cases when a
2931 control that does not make sense is encountered.
2933 9. Added the /retain_sender option to "control=submission".
2935 10. $recipients is now available in the predata ACL (oversight).
2937 11. Tidy the search cache before the fork to do a delivery from a message
2938 received from the command line. Otherwise the child will trigger a lookup
2939 failure and thereby defer the delivery if it tries to use (for example) a
2940 cached ldap connection that the parent has called unbind on.
2942 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2943 of $address_data from the recipient verification was clobbered by the
2944 sender verification.
2946 13. The value of address_data from a sender verification is now available in
2947 $sender_address_data in subsequent conditions in the ACL statement.
2949 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2951 15. Added a new option "connect=<time>" to callout options, to set a different
2954 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2955 was its contents. (It was OK if the option was not defined at all.)
2957 17. A "Completed" log line is now written for messages that are removed from
2958 the spool by the -Mrm option.
2960 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2961 information about exactly what failed.
2963 19. Added -dd to debug only the daemon process.
2965 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2966 handles timeouts, both on the server side and network timeouts. Renamed the
2967 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2969 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2970 It is now set to "smtps".
2972 22. $host_address is now set to the target address during the checking of
2973 ignore_target_hosts.
2975 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2976 being passed; this would have caused $sender_host_name to have been used if
2977 matching the list had actually called for a host name (not very likely,
2978 since this list is usually IP addresses). A host name is now passed as
2981 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2982 code that specifies a non-restarting handler (typically sigaction() in
2983 modern systems) in an attempt to fix a rare and obscure crash bug.
2985 25. Narrowed the window for a race in the daemon that could cause it to ignore
2986 SIGCHLD signals. This is not a major problem, because they are used only to
2987 wake it up if nothing else does.
2989 26. A malformed maildirsize file could cause Exim to calculate negative values
2990 for the mailbox size or file count. Odd effects could occur as a result.
2991 The maildirsize information is now recalculated if the size or filecount
2994 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2995 support for a long time. Removed HAVE_SYS_VFS_H.
2997 28. Installed the latest version of exipick from John Jetmore.
2999 29. In an address list, if the pattern was not a regular expression, an empty
3000 subject address (from a bounce message) matched only if the pattern was an
3001 empty string. Non-empty patterns were not even tested. This was the wrong
3002 because it is perfectly reasonable to use an empty address as part of a
3003 database query. An empty address is now tested by patterns that are
3004 lookups. However, all the other forms of pattern expect the subject to
3005 contain a local part and a domain, and therefore, for them, an empty
3006 address still always fails if the pattern is not itself empty.
3008 30. Exim went into a mad DNS loop when attempting to do a callout where the
3009 host was specified on an smtp transport, and looking it up yielded more
3010 than one IP address.
3012 31. Re-factored the code for checking spool and log partition space into a
3013 function that finds that data and another that does the check. The former
3014 is then used to implement four new variables: $spool_space, $log_space,
3015 $spool_inodes, and $log_inodes.
3017 32. The RFC2047 encoding function was originally intended for short strings
3018 such as real names; it was not keeping to the 75-character limit for
3019 encoded words that the RFC imposes. It now respects the limit, and
3020 generates multiple encoded words if necessary. To be on the safe side, I
3021 have increased the buffer size for the ${rfc2047: expansion operator from
3024 33. It is now permitted to omit both strings after an "if" condition; if the
3025 condition is true, the result is "true". As before, when the second string
3026 is omitted, a false condition yields an empty string. This makes it less
3027 cumbersome to write custom ACL and router conditions.
3029 34. Failure to deliver a bounce message always caused it to be frozen, even if
3030 there was an errors_to setting on the router. The errors_to setting is now
3033 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3034 canonical form (fully expanded) before being placed in
3035 $sender_host_address.
3037 36. The table in the code that translates DNS record types into text (T_A to
3038 "A" for instance) was missing entries for NS and CNAME. It is just possible
3039 that this could have caused confusion if both these types were looked up
3040 for the same domain, because the text type is used as part of Exim's
3041 per-process caching. But the chance of anyone hitting this buglet seems
3044 37. The dnsdb lookup has been extended in a number of ways.
3046 (1) There is a new type, "zns", which walks up the domain tree until it
3047 finds some nameserver records. It should be used with care.
3049 (2) There is a new type, "mxh", which is like "mx" except that it returns
3050 just the host names, not the priorities.
3052 (3) It is now possible to give a list of domains (or IP addresses) to be
3053 looked up. The behaviour when one of the lookups defers can be
3054 controlled by a keyword.
3056 (4) It is now possible to specify the separator character for use when
3057 multiple records are returned.
3059 38. The dnslists ACL condition has been extended: it is now possible to supply
3060 a list of IP addresses and/or domains to be looked up in a particular DNS
3063 39. Added log_selector=+queue_time_overall.
3065 40. When running the queue in the test harness, wait just a tad after forking a
3066 delivery process, to get repeatability of debugging output.
3068 41. Include certificate and key file names in error message when GnuTLS fails
3069 to set them up, because the GnuTLS error message doesn't include the name
3070 of the failing file when there is a problem reading it.
3072 42. Allow both -bf and -bF in the same test run.
3074 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3076 44. The "Exiscan patch" is now merged into the mainline Exim source.
3078 45. Sometimes the final signoff response after QUIT could fail to get
3079 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3080 < 0 before doing a fflush(). This bug looks as though it goes back to the
3081 introduction of TLS in release 3.20, but "sometimes" must have been rare
3082 because the tests only now provoked it.
3084 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3085 (this can affect the format of dates).
3087 47. exim_tidydb, when checking for the continued existence of a message for
3088 which it has found a message-specific retry record, was not finding
3089 messages that were in split spool directories. Consequently, it was
3090 deleting retry records that should have stayed in existence.
3092 48. Steve fixed some bugs in eximstats.
3094 49. The SPA authentication driver was not abandoning authentication and moving
3095 on to the next authenticator when an expansion was forced to fail,
3096 contradicting the general specification for all authenticators. Instead it
3097 was generating a temporary error. It now behaves as specified.
3099 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3100 (the order specifies the preference for clients). The order is now AES256,
3101 AES128, 3DES, ARCFOUR128.
3103 51. Small patch to Sieve code - explicitly set From: when generating an
3106 52. Exim crashed if a remote delivery caused a very long error message to be
3107 recorded - for instance if somebody sent an entire SpamAssassin report back
3108 as a large number of 550 error lines. This bug was coincidentally fixed by
3109 increasing the size of one of Exim's internal buffers (big_buffer) that
3110 happened as part of the Exiscan merge. However, to be on the safe side, I
3111 have made the code more robust (and fixed the comments that describe what
3114 53. Now that there can be additional text after "Completed" in log lines (if
3115 the queue_time_overall log selector is set), a one-byte patch to exigrep
3116 was needed to allow it to recognize "Completed" as not the last thing in
3119 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3120 patch that reportedly fixes this has been added. I am not expert enough to
3121 create a test for it. This is what the patch creator wrote:
3123 "I found a little strange behaviour of ldap code when working with
3124 Windows 2003 AD Domain, where users was placed in more than one
3125 Organization Units. When I tried to give exim partial DN, the exit code
3126 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3127 But simultaneously result of request was absolutely normal ldap result,
3128 so I produce this patch..."
3130 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3131 REFERENCE, so I have modified the code to exclude the patch when that macro
3134 55. Some experimental protocols are using DNS PTR records for new purposes. The
3135 keys for these records are domain names, not reversed IP addresses. The
3136 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3137 leaves it alone. Component reversal etc. now happens only for IP addresses.
3140 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3142 57. Double the size of the debug message buffer (to 2048) so that more of very
3143 long debug lines gets shown.
3145 58. The exicyclog utility now does better if the number of log files to keep
3146 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3148 59. Two changes related to the smtp_active_hostname option:
3150 (1) $smtp_active_hostname is now available as a variable.
3151 (2) The default for smtp_banner uses $smtp_active_hostname instead
3152 of $primary_hostname.
3154 60. The host_aton() function is supposed to be passed a string that is known
3155 to be a valid IP address. However, in the case of IPv6 addresses, it was
3156 not checking this. This is a hostage to fortune. Exim now panics and dies
3157 if the condition is not met. A case was found where this could be provoked
3158 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3159 components; fortuitously, this particular loophole had already been fixed
3160 by change 4.50/55 above.
3162 If there are any other similar loopholes, the new check in host_aton()
3163 itself should stop them being exploited. The report I received stated that
3164 data on the command line could provoke the exploit when Exim was running as
3165 exim, but did not say which command line option was involved. All I could
3166 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3167 running as the user.
3170 61. There was a buffer overflow vulnerability in the SPA authentication code
3171 (which came originally from the Samba project). I have added a test to the
3172 spa_base64_to_bits() function which I hope fixes it.
3175 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3176 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3178 63. The daemon start-up calls getloadavg() while still root for those OS that
3179 need the first call to be done as root, but it missed one case: when
3180 deliver_queue_load_max is set with deliver_drop_privilege. This is
3181 necessary for the benefit of the queue runner, because there is no re-exec
3182 when deliver_drop_privilege is set.
3184 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3185 This has been fixed.
3187 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3188 was in use, was not putting the data itself into the right store pool;
3189 consequently, it could be overwritten for a subsequent message in the same
3190 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3193 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3195 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3196 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3197 the function were treating the return as a boolean value, which happened to
3198 work because 0=false and not-0=true, but is not correct code.
3200 68. The host_aton() function was not handling scoped IPv6 addresses (those
3201 with, for example, "%eth0" on the end) correctly.
3203 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3204 negated items (that is, ~something) in unsigned ints. Some compilers
3205 apparently mutter when there is no cast.
3207 70. If an address verification called from an ACL failed, and did not produce a
3208 user-specific message (i.e. there was only a "system" message), nothing was
3209 put in $acl_verify_message. In this situation, it now puts the system
3212 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3213 session; change 4.31/43 added the unwanted input to the log line - except
3214 that it did not do this in the start of session case. It now does.
3216 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3217 This is wrong and can cause the other end to generate a synchronization
3218 error if it is another Exim or anything else that does the synchronization
3219 check. A QUIT command is no longer sent after a timeout.
3221 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3222 during host lookups.
3224 74. The defer_ok option of callout verification was not working if it was used
3225 when verifying addresses in header lines, that is, for this case:
3227 verify = header_sender/callout=defer_ok
3229 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3230 those file descriptors could be used for SMTP connections. If anything
3231 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3232 could be sent to the SMTP client, causing chaos. The daemon now opens
3233 stdin, stdout, and stderr to /dev/null when it puts itself into the
3236 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3237 log by default. The user can override this with suitable Perl magic.
3239 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3240 the log message when discard triggers, was not working for the DATA ACL or
3241 for the non-SMTP ACL.
3243 78. Error message wording change in sieve.c.
3245 79. If smtp_accept_max_per_host was set, the number of connections could be
3246 restricted to fewer than expected, because the daemon was trying to set up
3247 a new connection before checking whether the processes handling previous
3248 connections had finished. The check for completed processes is now done
3249 earlier. On busy systems, this bug wouldn't be noticed because something
3250 else would have woken the daemon, and it would have reaped the completed
3253 80. If a message was submitted locally by a user whose login name contained one
3254 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3255 It caused a spool format error. I have fixed the spool reading code. A
3256 related problem was that the "from" clause in the Received: line became
3257 illegal because of the space(s). It is now covered by ${quote_local_part.
3259 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3262 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3264 83. Patch from Sieve maintainer to fix unterminated string problem in
3265 "vacation" handling.
3267 84. Some minor changes to the Linux configuration files to help with other
3268 OS variants using glibc.
3270 85. One more patch for Sieve to update vacation handling to latest spec.
3273 ----------------------------------------------------
3274 See the note above about the 4.44 and 4.50 releases.
3275 ----------------------------------------------------
3281 1. Change 4.43/35 introduced a bug that caused file counts to be
3282 incorrectly computed when quota_filecount was set in an appendfile
3285 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3286 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3289 3. Give more explanation in the error message when the command for a transport
3290 filter fails to execute.
3292 4. There are several places where Exim runs a non-Exim command in a
3293 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3294 was being done only for the command run by the queryprogram router. It is
3295 now done for all such subprocesses. The other cases are: ${run, transport
3296 filters, and the commands run by the lmtp and pipe transports.
3298 5. Some older OS have a limit of 256 on the maximum number of file
3299 descriptors. Exim was using setrlimit() to set 1000 as a large value
3300 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3301 systems. I've change it so that if it can't get 1000, it tries for 256.
3303 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3304 was an oversight, and furthermore, ever since the addition of extra
3305 controls (e.g. 4.43/32), the checks on when to allow different forms of
3306 "control" were broken. There should now be diagnostics for all cases when a
3307 control that does not make sense is encountered.
3309 7. $recipients is now available in the predata ACL (oversight).
3311 8. Tidy the search cache before the fork to do a delivery from a message
3312 received from the command line. Otherwise the child will trigger a lookup
3313 failure and thereby defer the delivery if it tries to use (for example) a
3314 cached ldap connection that the parent has called unbind on.
3316 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3317 of $address_data from the recipient verification was clobbered by the
3318 sender verification.
3320 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3321 was its contents. (It was OK if the option was not defined at all.)
3323 11. A "Completed" log line is now written for messages that are removed from
3324 the spool by the -Mrm option.
3326 12. $host_address is now set to the target address during the checking of
3327 ignore_target_hosts.
3329 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3330 being passed; this would have caused $sender_host_name to have been used if
3331 matching the list had actually called for a host name (not very likely,
3332 since this list is usually IP addresses). A host name is now passed as
3335 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3336 code that specifies a non-restarting handler (typically sigaction() in
3337 modern systems) in an attempt to fix a rare and obscure crash bug.
3339 15. Narrowed the window for a race in the daemon that could cause it to ignore
3340 SIGCHLD signals. This is not a major problem, because they are used only to
3341 wake it up if nothing else does.
3343 16. A malformed maildirsize file could cause Exim to calculate negative values
3344 for the mailbox size or file count. Odd effects could occur as a result.
3345 The maildirsize information is now recalculated if the size or filecount
3348 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3349 support for a long time. Removed HAVE_SYS_VFS_H.
3351 18. Updated exipick to current release from John Jetmore.
3353 19. Allow an empty sender to be matched against a lookup in an address list.
3354 Previously the only cases considered were a regular expression, or an
3357 20. Exim went into a mad DNS lookup loop when doing a callout where the
3358 host was specified on the transport, if the DNS lookup yielded more than
3361 21. The RFC2047 encoding function was originally intended for short strings
3362 such as real names; it was not keeping to the 75-character limit for
3363 encoded words that the RFC imposes. It now respects the limit, and
3364 generates multiple encoded words if necessary. To be on the safe side, I
3365 have increased the buffer size for the ${rfc2047: expansion operator from
3368 22. Failure to deliver a bounce message always caused it to be frozen, even if
3369 there was an errors_to setting on the router. The errors_to setting is now
3372 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3373 canonical form (fully expanded) before being placed in
3374 $sender_host_address.
3376 24. Updated eximstats to version 1.33
3378 25. Include certificate and key file names in error message when GnuTLS fails
3379 to set them up, because the GnuTLS error message doesn't include the name
3380 of the failing file when there is a problem reading it.
3382 26. Expand error message when OpenSSL has problems setting up cert/key files.
3385 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3386 (this can affect the format of dates).
3388 28. exim_tidydb, when checking for the continued existence of a message for
3389 which it has found a message-specific retry record, was not finding
3390 messages that were in split spool directories. Consequently, it was
3391 deleting retry records that should have stayed in existence.
3393 29. eximstats updated to version 1.35
3394 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3395 1.35 - bugfix such that pie charts by volume are generated correctly
3397 30. The SPA authentication driver was not abandoning authentication and moving
3398 on to the next authenticator when an expansion was forced to fail,
3399 contradicting the general specification for all authenticators. Instead it
3400 was generating a temporary error. It now behaves as specified.
3402 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3403 (the order specifies the preference for clients). The order is now AES256,
3404 AES128, 3DES, ARCFOUR128.
3406 31. Small patch to Sieve code - explicitly set From: when generating an
3409 32. Exim crashed if a remote delivery caused a very long error message to be
3410 recorded - for instance if somebody sent an entire SpamAssassin report back
3411 as a large number of 550 error lines. This bug was coincidentally fixed by
3412 increasing the size of one of Exim's internal buffers (big_buffer) that
3413 happened as part of the Exiscan merge. However, to be on the safe side, I
3414 have made the code more robust (and fixed the comments that describe what
3417 33. Some experimental protocols are using DNS PTR records for new purposes. The
3418 keys for these records are domain names, not reversed IP addresses. The
3419 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3420 leaves it alone. Component reversal etc. now happens only for IP addresses.
3423 34. The host_aton() function is supposed to be passed a string that is known
3424 to be a valid IP address. However, in the case of IPv6 addresses, it was
3425 not checking this. This is a hostage to fortune. Exim now panics and dies
3426 if the condition is not met. A case was found where this could be provoked
3427 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3428 components; fortuitously, this particular loophole had already been fixed
3429 by change 4.50/55 or 4.44/33 above.
3431 If there are any other similar loopholes, the new check in host_aton()
3432 itself should stop them being exploited. The report I received stated that
3433 data on the command line could provoke the exploit when Exim was running as
3434 exim, but did not say which command line option was involved. All I could
3435 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3436 running as the user.
3439 35. There was a buffer overflow vulnerability in the SPA authentication code
3440 (which came originally from the Samba project). I have added a test to the
3441 spa_base64_to_bits() function which I hope fixes it.
3444 36. The daemon start-up calls getloadavg() while still root for those OS that
3445 need the first call to be done as root, but it missed one case: when
3446 deliver_queue_load_max is set with deliver_drop_privilege. This is
3447 necessary for the benefit of the queue runner, because there is no re-exec
3448 when deliver_drop_privilege is set.
3450 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3451 was in use, was not putting the data itself into the right store pool;
3452 consequently, it could be overwritten for a subsequent message in the same
3453 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3456 38. Sometimes the final signoff response after QUIT could fail to get
3457 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3458 < 0 before doing a fflush(). This bug looks as though it goes back to the
3459 introduction of TLS in release 3.20, but "sometimes" must have been rare
3460 because the tests only now provoked it.
3466 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3467 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3468 Now it returns an integer. A number of calls were still expecting a T/F
3469 return. Fortuitously, in all cases, the tests worked in OK situations,
3470 which is the norm. However, things would have gone wrong on any write
3471 failures on the smtp file descriptor. This function is used when sending
3472 messages over SMTP and also when doing verify callouts.
3474 2. When Exim is called to do synchronous delivery of a locally submitted
3475 message (the -odf or -odi options), it no longer closes stderr before doing
3478 3. Implemented the mua_wrapper option.
3480 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3482 5. Implemented the functions header_remove(), header_testname(),
3483 header_add_at_position(), and receive_remove_recipient(), and exported them
3486 6. If an ACL "warn" statement specified the addition of headers, Exim already
3487 inserted X-ACL-Warn: at the start if there was no header name. However, it
3488 was not making this test for the second and subsequent header lines if
3489 there were newlines in the string. This meant that an invalid header could
3490 be inserted if Exim was badly configured.
3492 7. Allow an ACL "warn" statement to add header lines at the start or after all
3493 the Received: headers, as well as at the end.
3495 8. Added the rcpt_4xx retry error code.
3497 9. Added postmaster_mailfrom=xxx to callout verification option.
3499 10. Added mailfrom=xxxx to the callout verification option, for verify=
3502 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3503 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3505 12. Inserted some casts to stop certain compilers warning when using pointer
3506 differences as field lengths or precisions in printf-type calls (mostly
3507 affecting debugging statements).
3509 13. Added optional readline() support for -be (dynamically loaded).
3511 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3512 same clock tick as a message's arrival, so that its received time was the
3513 same as the "first fail" time on the retry record, and that message
3514 remained on the queue past the ultimate address timeout, every queue runner
3515 would try a delivery (because it was past the ultimate address timeout) but
3516 after another failure, the ultimate address timeout, which should have then
3517 bounced the address, did not kick in. This was a "< instead of <=" error;
3518 in most cases the first failure would have been in the next clock tick
3519 after the received time, and all would be well.
3521 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3522 being recognized when the domain list was tested by the match_domain
3523 condition in an expansion string.
3525 16. Added the ${str2b64: operator.
3527 17. Exim was always calling setrlimit() to set a large limit for the number of
3528 processes, without checking whether the existing limit was already
3529 adequate. (It did check for the limit on file descriptors.) Furthermore,
3530 errors from getrlimit() and setrlimit() were being ignored. Now they are
3531 logged to the main and panic logs, but Exim does carry on, to try to do its
3532 job under whatever limits there are.
3534 18. Imported PCRE 5.0.
3536 19. Trivial typo in log message " temporarily refused connection" (the leading
3539 20. If the log selector return_path_on_delivery was set and an address was
3540 redirected to /dev/null, the delivery process crashed because it assumed
3541 that a return path would always be set for a "successful" delivery. In this
3542 case, the whole delivery is bypassed as an optimization, and therefore no
3545 21. Internal re-arrangement: the function for sending a challenge and reading
3546 a response while authentication was assuming a zero-terminated challenge
3547 string. It's now changed to take a pointer and a length, to allow for
3548 binary data in such strings.
3550 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3552 23. Exim was not respecting finduser_retries when seeking the login of the
3553 uid under which it was called; it was always trying 10 times. (The default
3554 setting of finduser_retries is zero.) Also, it was sleeping after the final
3555 failure, which is pointless.
3557 24. Implemented tls_on_connect_ports.
3559 25. Implemented acl_smtp_predata.
3561 26. If the domain in control=submission is set empty, Exim assumes that the
3562 authenticated id is a complete email address when it generates From: or
3563 Sender: header lines.
3565 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3566 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3567 chown and chgrp in /bin and hostname in /usr/bin.
3569 28. Exim was keeping the "process log" file open after each use, just as it
3570 does for the main log. This opens the possibility of it remaining open for
3571 long periods when the USR1 signal hits a daemon. Occasional processlog
3572 errors were reported, that could have been caused by this. Anyway, it seems
3573 much more sensible not to leave this file open at all, so that is what now
3576 29. The long-running daemon process does not normally write to the log once it
3577 has entered its main loop, and it closes the log before doing so. This is
3578 so that log files can straightforwardly be renamed and moved. However,
3579 there are a couple of unusual error situations where the daemon does write
3580 log entries, and I had neglected to close the log afterwards.
3582 30. The text of an SMTP error response that was received during a remote
3583 delivery was being truncated at 512 bytes. This is too short for some of
3584 the long messages that one sometimes sees. I've increased the limit to
3587 31. It is now possible to make retry rules that apply only when a message has a
3588 specific sender, in particular, an empty sender.
3590 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3591 it possible to be selective about when SMTP synchronization is enforced.
3593 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3595 32. Implemented hosts_connection_nolog.
3597 33. Added an ACL for QUIT.
3599 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3602 35. Added mailbox_size and mailbox_filecount to appendfile.
3604 36. Added control = no_multiline_responses to ACLs.
3606 37. There was a bug in the logic of the code that waits for the clock to tick
3607 in the case where the clock went backwards by a substantial amount such
3608 that the microsecond fraction of "now" was more than the microsecond
3609 fraction of "then" (but the whole seconds number was less).
3611 38. Added support for the libradius Radius client library this is found on
3612 FreeBSD (previously only the radiusclient library was supported).
3618 1. When certain lookups returned multiple values in the form name=value, the
3619 quoting of the values was not always being done properly. Specifically:
3620 (a) If the value started with a double quote, but contained no whitespace,
3622 (b) If the value contained whitespace other than a space character (i.e.
3623 tabs or newlines or carriage returns) it was not quoted.
3624 This fix has been applied to the mysql and pgsql lookups by writing a
3625 separate quoting function and calling it from the lookup code. The fix
3626 should probably also be applied to nisplus, ibase and oracle lookups, but
3627 since I cannot test any of those, I have not disturbed their existing code.
3629 2. A hit in the callout cache for a specific address caused a log line with no
3630 reason for rejecting RCPT. Now it says "Previous (cached) callout
3631 verification failure".
3633 3. There was an off-by-one bug in the queryprogram router. An over-long
3634 return line was truncated at 256 instead of 255 characters, thereby
3635 overflowing its buffer with the terminating zero. As well as fixing this, I
3636 have increased the buffer size to 1024 (and made a note to document this).
3638 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3639 when Exim is waiting for an SMTP response from a remote server, Exim
3640 restarts its select() call on the socket, thereby resetting its timeout.
3641 This is not a problem when such interrupts are rare. Somebody set up a cron
3642 job to run exiwhat every 2 minutes, which is less than the normal select()
3643 timeout (5 or 10 minutes). This meant that the select() timeout never
3644 kicked in because it was always reset. I have fixed this by comparing the
3645 time when an interrupt arrives with the time at the start of the first call
3646 to select(). If more time than the timeout has elapsed, the interrupt is
3647 treated as a timeout.
3649 5. Some internal re-factoring in preparation for the addition of Sieve
3650 extensions (by MH). In particular, the "personal" test is moved to a
3651 separate function, and given an option for scanning Cc: and Bcc: (which is
3652 not set for Exim filters).
3654 6. When Exim created an email address using the login of the caller as the
3655 local part (e.g. when creating a From: or Sender: header line), it was not
3656 quoting the local part when it contained special characters such as @.
3658 7. Installed new OpenBSD configuration files.
3660 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3661 try to make them clearer.
3663 9. Callout options, other than the timeout value, were being ignored when
3664 verifying sender addresses in header lines. For example, when using
3666 verify = header_sender/callout=no_cache
3668 the cache was (incorrectly) being used.
3670 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3671 only the Cygwin environment.
3673 11. When return_path_on_delivery was set as a log selector, if different remote
3674 addresses in the same message used different return paths and parallel
3675 remote delivery occurred, the wrong values would sometimes be logged.
3676 (Whenever a remote delivery process finished, the return path value from
3677 the most recently started remote delivery process was logged.)
3679 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3680 lines when AUTH and/or TLS are in use. This is the "received protocol"
3681 field. Exim used to use "asmtp" for authenticated SMTP, without any
3682 indication (in the protocol name) for TLS use. Now it follows the RFC and
3683 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3684 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3685 names appear in log lines as well as in Received: header lines.
3687 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3688 extensions, and comparison tests, and to fix some bugs.
3690 14. Changes to the "personal" filter test:
3692 (1) The test was buggy in that it was just doing the equivalent of
3693 "contains" tests on header lines. For example, if a user's address was
3694 anne@some.where, the "personal" test would incorrectly be true for
3696 To: susanne@some.where
3698 This test is now done by extracting each address from the header in turn,
3699 and checking the entire address. Other tests that are part of "personal"
3700 are now done using regular expressions (for example, to check local parts
3701 of addresses in From: header lines).
3703 (2) The list of non-personal local parts in From: addresses has been
3704 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3705 taken from the Sieve specification recommendations.
3707 (3) If the message contains any header line starting with "List-" it is
3708 treated as non-personal.
3710 (4) The test for "circular" in the Subject: header line has been removed
3711 because it now seems ill-conceived.
3713 15. Minor typos in src/EDITME comments corrected.
3715 16. Installed latest exipick from John Jetmore.
3717 17. If headers_add on a router specified a text string that was too long for
3718 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3719 of string_sprintf() is now avoided.
3721 18. $message_body_size was not set (it was always zero) when running the DATA
3722 ACL and the local_scan() function.
3724 19. For the "mail" command in an Exim filter, no default was being set for
3725 the once_repeat time, causing a random time value to be used if "once" was
3726 specified. (If the value happened to be <= 0, no repeat happened.) The
3727 default is now 0s, meaning "never repeat". The "vacation" command was OK
3728 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3729 (I found it when inspecting the code).
3731 20. There is now an overall timeout for performing a callout verification. It
3732 defaults to 4 times the callout timeout, which applies to individual SMTP
3733 commands during the callout. The overall timeout applies when there is more
3734 than one host that can be tried. The timeout is checked before trying the
3735 next host. This prevents very long delays if there are a large number of
3736 hosts and all are timing out (e.g. when the network connections are timing
3737 out). The value of the overall timeout can be changed by specifying an
3738 additional sub-option for "callout", called "maxwait". For example:
3740 verify = sender/callout=5s,maxwait=20s
3742 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3743 to the end before writing, but this should make it even safer).
3745 22. Exim was forgetting that it had advertised PIPELINING for the second and
3746 subsequent messages on an SMTP connection. It was also not resetting its
3747 memory on STARTTLS and an internal HELO.
3749 23. When Exim logs an SMTP synchronization error within a session, it now
3750 records whether PIPELINING has been advertised or not.
3752 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3753 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3754 rather than long int.
3756 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3758 26. Added the never_mail option to autoreply.
3764 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3765 crash if the getsockname() call failed; this can happen if a connection is
3766 closed very soon after it is established. The problem was simply in the
3767 order in which certain operations were done, causing Exim to try to write
3768 to the SMTP stream before it had set up the file descriptor. The bug has
3769 been fixed by making things happen in the correct order.
3775 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3776 before the connection was closed, thus losing the rejection response.
3778 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3779 some early Solaris releases, but causes trouble in current releases where
3780 socklen_t is defined.
3782 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3785 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3788 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3789 that releasing the top of it at the end releases what was used for sub-
3790 expansions (unless the block got too big). However, discard this block if
3791 the first thing is a variable or header, so that we can use its block when
3792 it is dynamic (useful for very large $message_headers, for example).
3794 6. Lookups now cache *every* query, not just the most recent. A new, separate
3795 store pool is used for this. It can be recovered when all lookup caches are
3796 flushed. Lookups now release memory at the end of their result strings.
3797 This has involved some general refactoring of the lookup sources.
3799 7. Some code has been added to the store_xxx() functions to reduce the amount
3800 of flapping under certain conditions.
3802 8. log_incoming_interface used to affect only the <= reception log lines. Now
3803 it causes the local interface and port to be added to several more SMTP log
3804 lines, for example "SMTP connection from", and rejection lines.
3806 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3808 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3810 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3811 could be overwritten at the end of the current message (or the start of a
3812 new message if it was set in a HELO ACL). The value is now preserved for
3813 the duration of the SMTP connection.
3815 12. If a transport had a headers_rewrite setting, and a matching header line
3816 contained an unqualified address, that address was qualified, even if it
3817 did not match any rewriting rules. The underlying bug was that the values
3818 of the flags that permit the existence of unqualified sender and recipient
3819 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3820 non-local messages, and by -bnq for local messages) were not being
3821 preserved with the message after it was received.
3823 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3824 "next input=" as part of the text comprising the host identity instead of
3825 the correct text. The code was using the same buffer for two different
3826 strings. However, depending on which order the printing function evaluated
3827 its arguments, the bug did not always show up. Under Linux, for example, my
3828 test suite worked just fine.
3830 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3831 doesn't work with some older versions of Perl. It has been changed to "my",
3832 which in any case is probably the better facility to use.
3834 15. A really picky compiler found some instances of statements for creating
3835 error messages that either had too many or two few arguments for the format
3838 16. The size of the buffer for calls to the DNS resolver has been increased
3839 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3840 for addresses that have a lot of PTR records. This alleviates a problem; it
3841 does not fully solve it.
3843 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3844 buffer now truncates the list and logs the incident, which is the same
3845 action as happens when Exim is looking up a host name and its aliases.
3846 Previously in this situation something unpredictable would happen;
3847 sometimes it was "internal error: store_reset failed".
3849 18. If a server dropped the connection unexpectedly when an Exim client was
3850 using GnuTLS and trying to read a response, the client delivery process
3851 crashed while trying to generate an error log message.
3853 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3854 string, for example:
3856 warn message = H1: something\nH2: something
3858 the text was added as a single header line from Exim's point of view
3859 though it ended up OK in the delivered message. However, searching for the
3860 second and subsequent header lines using $h_h2: did not work. This has been
3861 fixed. Similarly, if a system filter added multiple headers in this way,
3862 the routers could not see them.
3864 20. Expanded the error message when iplsearch is called with an invalid key to
3865 suggest using net-iplsearch in a host list.
3867 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3868 ACLs are no longer actually imposed (and a message to that effect is
3871 22. If a "gecos" field in a passwd entry contained escaped characters, in
3872 particular, if it contained a \" sequence, Exim got it wrong when building
3873 a From: or a Sender: header from that name. A second bug also caused
3874 incorrect handling when an unquoted " was present following a character
3875 that needed quoting.
3877 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3878 was not being matched caselessly.
3880 24. Arranged for all hyphens in the exim.8 source to be escaped with
3883 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3884 properly. Recipient callout cache records were still being keyed to include
3885 the sender, even when use_sender was set false. This led to far more
3886 callouts that were necessary. The sender is no longer included in the key
3887 when use_sender is false.
3889 26. Added "control = submission" modifier to ACLs.
3891 27. Added the ${base62d: operator to decode base 62 numbers.
3893 28. dnsdb lookups can now access SRV records.
3895 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3896 the configuration file.
3898 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3899 (-v) mode. This makes the output for a verbose queue run more intelligible.
3901 31. Added a use_postmaster feature to recipient callouts.
3903 32. Added the $body_zerocount variable, containing the number of binary zero
3904 bytes in the message body.
3906 33. The time of last modification of the "new" subdirectory is now used as the
3907 "mailbox time last read" when there is a quota error for a maildir
3910 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3912 35. Added +ignore_unknown as a special item in host lists.
3914 36. Code for decoding IPv6 addresses in host lists is now included, even if
3915 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3916 address was recognized as an IP address, but was then not correctly decoded
3917 into binary, causing unexpected and incorrect effects when compared with
3924 1. Very minor rewording of debugging text in manualroute to say "list of
3925 hosts" instead of "hostlist".
3927 2. If verify=header_syntax was set, and a header line with an unqualified
3928 address (no domain) and a large number of spaces between the end of the
3929 name and the colon was received, the reception process suffered a buffer
3930 overflow, and (when I tested it) crashed. This was caused by some obsolete
3931 code that should have been removed. The fix is to remove it!
3933 3. When running in the test harness, delay a bit after writing a bounce
3934 message to get a bit more predictability in the log output.
3936 4. Added a call to search_tidyup() just before forking a reception process. In
3937 theory, someone could use a lookup in the expansion of smtp_accept_max_
3938 per_host which, without the tidyup, could leave open a database connection.
3940 5. Added the variables $recipient_data and $sender_data which get set from a
3941 lookup success in an ACL "recipients" or "senders" condition, or a router
3942 "senders" option, similar to $domain_data and $local_part_data.
3944 6. Moved the writing of debug_print from before to after the "senders" test
3947 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3948 problems for message scanning, either using a data ACL, or using
3949 local_scan() because the Received: header was not generated till after they
3950 were called (in order to set the time as the time of reception completion).
3951 I have revised the way this works. The header is now generated after the
3952 body is received, but before the ACL or local_scan() are called. After they
3953 are run, the timestamp in the header is updated.
3959 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3960 before starting a queue runner without re-exec. This happened only when
3961 deliver_drop_privilege was set or when the Exim user was set to root. The
3962 effect of the bug was that timeouts during subsequent deliveries caused
3963 crashes instead of being properly handled. The handler is now left at its
3964 default (and expected) setting.
3966 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3967 message, again when deliver_drop_privilege is set or Exim is run as root.
3968 The bug described in (1) was not present in this case, but the tidying up
3969 of the other signals was missing. I have made the two cases consistent.
3971 3. The ignore_target_hosts setting on a manualroute router was being ignored
3972 for hosts that were looked up using the /MX notation.
3974 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3977 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3978 operated on the sender address. After changing the $sender_address to <>
3979 for the sender address verify, Exim was re-instated it as the original
3980 (before rewriting) address, but remembering that it had rewritten it, so it
3981 wasn't rewriting it again. This bug also had the effect of breaking the
3982 sender address verification caching when the sender address was rewritten.
3984 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3985 This has been changed so that if the ip literal address matches
3986 ignore_target_hosts, the router declines.
3988 7. Added expansion conditions match_domain, match_address, and match_local_
3989 part (NOT match_host).
3991 8. The placeholder for the Received: header didn't have a length field set.
3993 9. Added code to Exim itself and to exim_lock to test for a specific race
3994 condition that could lead to file corruption when using MBX delivery. The
3995 issue is with the lockfile that is created in /tmp. If this file is removed
3996 after a process has opened it but before that process has acquired a lock,
3997 there is the potential for a second process to recreate the file and also
3998 acquire a lock. This could lead to two Exim processes writing to the file
3999 at the same time. The added code performs the same test as UW imapd; it
4000 checks after acquiring the lock that its file descriptor still refers to
4001 the same named file.
4003 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4004 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4007 11. Added the smtp_active_hostname option. If used, this will typically be made
4008 to depend on the incoming interface address. Because $interface_address is
4009 not set up until the daemon has forked a reception process, error responses
4010 that can happen earlier (such as "too many connections") no longer contain
4013 12. If an expansion in a condition on a "warn" statement fails because a lookup
4014 defers, the "warn" statement is abandoned, and the next ACL statement is
4015 processed. Previously this caused the whole ACL to be aborted.
4017 13. Added the iplsearch lookup type.
4019 14. Added ident_timeout as a log selector.
4021 15. Added tls_certificate_verified as a log selector.
4023 16. Added a global option tls_require_ciphers (compare the smtp transport
4024 option of the same name). This controls incoming TLS connections.
4026 17. I finally figured out how to make tls_require_ciphers do a similar thing
4027 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4028 before starting the TLS session.
4030 18. Tabs are now shown as \t in -bP output.
4032 19. If the log selector return_path_on_delivery was set, Exim crashed when
4033 bouncing a message because it had too many Received: header lines.
4035 20. If two routers both had headers_remove settings, and the first one included
4036 a superfluous trailing colon, the final name in the first list and the
4037 first name in the second list were incorrectly joined into one item (with a
4038 colon in the middle).
4044 1. Added -C and -D options to the exinext utility, mainly to make it easier
4045 to include in the automated testing, but these could be helpful when
4046 multiple configurations are in use.
4048 2. The exinext utility was not formatting the output nicely when there was
4049 an alternate port involved in the retry record key, nor when there was a
4050 message id as well (for retries that were specific to a specific message
4051 and a specific host). It was also confused by IPv6 addresses, because of
4052 the additional colons they contain. I have fixed the IPv4 problem, and
4053 patched it up to do a reasonable job for IPv6.
4055 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4056 delivery, the log line now contains "pipelined" if PIPELINING was used.
4058 4. An SMTP transport process used to panic and die if the bind() call to set
4059 an explicit outgoing interface failed. This has been changed; it is now
4060 treated in the same way as a connect() failure.
4062 5. A reference to $sender_host_name in the part of a conditional expansion
4063 that was being skipped was still causing a DNS lookup. This no longer
4066 6. The def: expansion condition was not recognizing references to header lines
4067 that used bh_ and bheader_.
4069 7. Added the _cache feature to named lists.
4071 8. The code for checking quota_filecount in the appendfile transport was
4072 allowing one more file than it should have been.
4074 9. For compatibility with Sendmail, the command line option
4082 and sets the incoming protocol and host name (for trusted callers). The
4083 host name and its colon can be omitted when only the protocol is to be set.
4084 Note the Exim already has two private options, -pd and -ps, that refer to
4085 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4086 "s", but I don't think that's a major issue.
4088 10. A number of refactoring changes to the code, none of which should affect
4091 (a) The number of logging options was getting close to filling up the
4092 32-bit word that was used as a bit map. I have split them into two classes:
4093 those that are passed in the argument to log_write(), and those that are
4094 only ever tested independently outside of that function. These are now in
4095 separate 32-bit words, so there is plenty of room for expansion again.
4096 There is no change in the user interface or the logging behaviour.
4098 (b) When building, for example, log lines, the code previously used a
4099 macro that called string_cat() twice, in order to add two strings. This is
4100 not really sufficiently general. Furthermore, there was one instance where
4101 it was actually wrong because one of the argument was used twice, and in
4102 one call a function was used. (As it happened, calling the function twice
4103 did not affect the overall behaviour.) The macro has been replaced by a
4104 function that can join an arbitrary number of extra strings onto a growing
4107 (c) The code for expansion conditions now uses a table and a binary chop
4108 instead of a serial search (which was left over from when there were very
4109 few conditions). Also, it now recognizes conditions like "pam" even when
4110 the relevant support is not compiled in: a suitably worded error message is
4111 given if an attempt is made to use such a condition.
4113 11. Added ${time_interval:xxxxx}.
4115 12. A bug was causing one of the ddress fields not to be passed back correctly
4116 from remote delivery subprocesses. The field in question was not being
4117 subsequently used, so this caused to problems in practice.
4119 13. Added new log selectors queue_time and deliver_time.
4121 14. Might have fixed a bug in maildirsizefile handling that threw up
4122 "unexpected character" debug warnings, and recalculated the data
4123 unnecessarily. In any case, I expanded the warning message to give more
4126 15. Added the message "Restricted characters in address" to the statements in
4127 the default ACL that block characters like @ and % in local parts.
4129 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4130 Three changes have been made:
4132 (a) There was a serious bug; a negative response to MAIL caused the whole
4133 recipient domain to be cached as invalid, thereby blocking all messages
4134 to all local parts at the same domain, from all senders. This bug has
4135 been fixed. The domain is no longer cached after a negative response to
4136 MAIL if the sender used is not empty.
4138 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4141 (c) A new callout option, "use_sender" has been added for people who want
4142 the modified behaviour.
4148 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4151 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4152 indeed breaks things for older releases.
4154 3. Added additional logging to the case where there is a problem reading data
4155 from a filter that is running in a subprocess using a pipe, in order to
4156 try to track down a specific problem.
4158 4. Testing facility fudge: when running in the test harness and attempting
4159 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4160 getting "No route to host". Convert this to a timeout.
4162 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4165 6. Some OS don't have socklen_t but use size_t instead. This affects the
4166 fifth argument of getsockopt() amongst other things. This is now
4167 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4168 can be set for individual OS. I have set it for SunOS5, OSF1, and
4169 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4170 some earlier ones do not.
4172 7. Change 4.30/15 was not doing the test caselessly.
4174 8. The standard form for an IPv6 address literal was being rejected by address
4175 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4176 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4177 this, as well as the form without the "IPv6" on the front (but only when
4178 address literals are enabled, of course).
4180 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4182 10. Exim crashed if a message with an empty sender address specified by -f
4183 encountered a router with an errors_to setting. This could be provoked only
4184 by a command such as
4188 where an empty string was supplied; "<>" did not hit this bug.
4190 11. Installed PCRE release 4.5.
4192 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4193 remained set. It is now erased.
4195 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4196 times from message ids (which are base 36 rather than the normal 62).
4198 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4199 were being counted as actual protocol errors, and logged if the log
4200 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4201 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4202 and DATA following a set of rejected RCPTs do not count as protocol errors.
4203 In other words, Exim assumes they were pipelined, though this may not
4204 actually be the case. Of course, in all cases the client gets an
4205 appropriate error code.
4207 15. If a lookup fails in an ACL condition, a message about the failure may
4208 be available; it is used if testing the ACL cannot continue, because most
4209 such messages specify what the cause of the deferral is. However, some
4210 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4211 that caused an old message to be retained and used if a later statement
4212 caused a defer, replacing the real cause of the deferral.
4214 16. If an IP address had so many PTR records that the DNS lookup buffer
4215 was not large enough to hold them, Exim could crash while trying to process
4216 the truncated data. It now detects and logs this case.
4218 17. Further to 4.21/58, another change has been made: if (and only if) the
4219 first line of a message (the first header line) ends with CRLF, a bare LF
4220 in a subsequent header line has a space inserted after it, so as not to
4221 terminate the header.
4223 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4224 unnecessarily, used atoi() instead of strtol(), and didn't check the
4225 termination when getting file sizes from file names by regex.
4227 19. Completely re-implemented the support for maildirsize files, in the light
4228 of a number of problems with the previous contributed implementation
4229 (4.30/29). In particular:
4231 . If the quota is zero, the maildirsize file is maintained, but no quota is
4234 . If the maildir directory does not exist, it is created before any attempt
4235 to write a maildirsize file.
4237 . The quota value in the file is just a cache; if the quota is changed in
4238 the transport, the new value overrides.
4240 . A regular expression is available for excluding directories from the
4243 20. The autoreply transport checks the characters in options that define the
4244 message's headers; it allows continued headers, but it was checking with
4245 isspace() after an embedded newline instead of explicitly looking for a
4248 21. If all the "regular" hosts to which an address was routed had passed their
4249 expiry times, and had not reached their retry times, the address was
4250 bounced, even if fallback hosts were defined. Now Exim should go on to try
4253 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4254 equivalent code in the SMTP transport. Some hosts send humungous responses
4255 to HELO/EHLO, more than 1024 it seems.
4257 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4258 gives compiler warnings in some environments. I've now done it "properly",
4261 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4262 (because of problems with the built-in one) was declared to return uschar *
4263 instead of char *, causing compiler failure.
4265 25. Fixed a file descriptor leak when processing alias/forward files.
4267 26. Fixed a minor format string issue in dbfn.c.
4269 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4271 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4272 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4275 29. When a sender address is verified, it is cached, to save repeating the test
4276 when there is more than one recipient in a message. However, when the
4277 verification involves a callout, it is possible for different callout
4278 options to be set for different recipients. It is too complicated to keep
4279 track of this in the cache, so now Exim always runs a verification when a
4280 callout is required, relying on the callout cache for the optimization.
4281 The overhead is duplication of the address routing, but this should not be
4284 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4285 to be verified with callout=postmaster, and the main callout worked but the
4286 postmaster check failed, the verification correctly failed. However, if a
4287 subsequent RCPT command asked for sender verification *without* the
4288 postmaster check, incorrect caching caused this verification also to fail,
4291 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4292 it was not caching the DNS options (qualify_single, search_parents) that
4293 were used when the lookup failed. A subsequent lookup with different
4294 options therefore always gave the same answer, though there were cases
4295 where it should not have. (Example: a "domains = !$mx_any" option on a
4296 dnslookup router: the "domains" option is always processed without any
4297 widening, but the router might have qualify_single set.) Now Exim uses the
4298 cached value only when the same options are set.
4300 32. Added John Jetmore's "exipick" utility to the distribution.
4302 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4303 than a timeout (e.g. a certificate is required, and is not provided), an
4304 Exim server now closes the connection immediately. Previously it waited for
4305 the client to close - but if the client is SSL, it seems that they each
4306 wait for each other, leading to a delay before one of them times out.
4308 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4309 maintained 0.8.x compatibility because I don't think many are using it, and
4310 it is clearly obsolete.
4312 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4315 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4316 client certificate was expired. A simple patch fixes this, though I don't
4317 understand the full logic of why the verify callback is called multiple
4320 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4321 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4322 which causes problems with some clients (such as the Certicom SSL Plus
4323 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4324 disables the coutermeasure allowing Eudora to connect."
4326 38. Exim was not checking that a write() to a log file succeeded. This could
4327 lead to Bad Things if a log got too big, in particular if it hit a file
4328 size limit. Exim now panics and dies if it cannot write to a log file, just
4329 as it does if it cannot open a log file.
4331 39. Modified OS/Makefile-Linux so that it now contains
4333 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4335 The two -D definitions ensure that Exim is compiled with large file
4336 support, which makes it possible to handle log files that are bigger than
4339 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4340 instance) a domain was checked against a named list that involved a lookup,
4341 causing $domain_data to be set, then another domain was checked against the
4342 same list, then the first domain was re-checked, the value of $domain_data
4343 after the final check could be wrong. In particular, if the second check
4344 failed, it could be set empty. This bug probably also applied to
4347 41. The strip_trailing_dot option was not being applied to the address given
4348 with the -f command-line option.
4350 42. The code for reading a message's header from the spool was incrementing
4351 $received_count, but never initializing it. This meant that the value was
4352 incorrect (doubled) while delivering a message in the same process in which
4353 it was received. In the most common configuration of Exim, this never
4354 happens - a fresh exec is done - but it can happen when
4355 deliver_drop_privilege is set.
4357 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4358 it now includes up to 150 characters of the unexpected data in the log
4361 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4362 and building data strings. The size of both of these buffers was 10 000
4363 bytes - far larger than anybody would *ever* want, thought I. Needless to
4364 say, somebody hit the limit. I have increased the maximum line length to
4365 20 000 and the maximum data length of concatenated lines to 100 000. I have
4366 also fixed two bugs, because there was no checking on these buffers. Tsk,
4367 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4368 buffer is too small.
4370 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4371 lsearch lookups. Now it does.
4373 46. When parsing a route_list item in a manualroute router, a fixed-length
4374 buffer was used for the list of hosts. I made this 1024 bytes long,
4375 thinking that nobody would ever have a list of hosts that long. Wrong.
4376 Somebody had a whole pile of complicated expansion conditions, and the
4377 string was silently truncated, leading to an expansion error. It turns out
4378 that it is easier to change to an unlimited length (owing to other changes
4379 that have happened since this code was originally written) than to build
4380 structure for giving a limitation error. The length of the item that
4381 expands into the list of hosts is now unlimited.
4383 47. The lsearch lookup could not handle data where the length of text line was
4384 more than 4095 characters. Such lines were truncated, leading to shortened
4385 data being returned. It should now handle lines of any length.
4387 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4388 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4391 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4392 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4393 no longer happens in comment lines. A list of replacements is now placed
4394 at the head of all of the source files, except those whose only change is
4395 to replace PERL_COMMAND in the very first #! line.
4397 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4398 messages on the queue, with a bottom-up merge sort, using code contributed
4399 by Michael Haardt. This should make operations like -bp somewhat faster on
4400 large queues. It won't affect queue runners, except when queue_run_in_order
4403 51. Installed eximstats 1.31 in the distribution.
4405 52. Added support for SRV lookups to the dnslookup router.
4407 53. If an ACL referred to $message_body or $message_body_end, the value was not
4408 reset for any messages that followed in the same SMTP session.
4410 54. The store-handling optimization for building very long strings was not
4411 differentiating between the different store pools. I don't think this
4412 actually made any difference in practice, but I've tidied it.
4414 55. While running the routers to verify a sender address, $sender_address
4415 was still set to the sender address. This is wrong, because when routing to
4416 send a bounce to the sender, it would be empty. Therefore, I have changed
4417 it so that, while verifying a sender address, $sender_address is set to <>.
4418 (There is no change to what happens when verifying a recipient address.)
4420 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4421 target A or AAAA records (if not already returned) without resetting the
4422 qualify_single or search_parents options of the DNS resolver. These are
4423 inappropriate in this case because the targets of MX and SRV records must
4424 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4425 target that, when qualified, matched something in the local domain. These
4426 two options are now turned off when doing these lookups.
4428 57. It seems that at least some releases of Reiserfs (which does not have the
4429 concept of a fixed number of inodes) returns zero and not -1 for the
4430 number of available inodes. This interacted badly with check_spool_inodes,
4431 which assumed that -1 was the "no such thing" setting. What I have done is
4432 to check that the total number of inodes is greater than zero before doing
4433 the test of how many are available.
4435 58. When a "warn" ACL statement has a log_message modifier, the message is
4436 remembered, and not repeated. This is to avoid a lot of repetition when a
4437 message has many recipients that cause the same warning to be written.
4438 Howewer, Exim was preserving the list of already written lines for an
4439 entire SMTP session, which doesn't seem right. The memory is now reset if a
4440 new message is started.
4442 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4443 debugging output unless log_rewrite was also set.
4445 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4446 of (int)(handle) when we know that handle contains (void *)(-1).
4448 61. The Exim daemon panic-logs an error return when it closes the incoming
4449 connection. However "connection reset by peer" seems to be common, and
4450 isn't really an error worthy of noting specially, so that particular error
4453 62. When Exim is trying to find all the local interfaces, it used to panic and
4454 die if the ioctl to get the interface flags failed. However, it seems that
4455 on at least one OS (Solaris 9) it is possible to have an interface that is
4456 included in the list of interfaces, but for which you get a failure error
4457 for this call. This happens when the interface is not "plumbed" into a
4458 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4459 failure of the "get flags" call assumes that the interface is down.
4461 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4462 makes life easier for people who are doing arithmetic on fields extracted
4463 from dates, where you often get leading zeros that should not be
4464 interpreted as octal.
4466 64. Added qualify_domain to the redirect router, to override the global
4469 65. If a pathologically long header line contained very many addresses (the
4470 report of this problem mentioned 10 000) and each of them was rewritten,
4471 Exim could use up a very large amount of memory. (It kept on making new
4472 copies of the header line as it rewrote, and never released the old ones.)
4473 At the expense of a bit more processing, the header rewriting function has
4474 been changed so that it no longer eats memory in this way.
4476 66. The generation of the Received: header has been moved from the time that a
4477 message starts to be received, to the time that it finishes. The timestamp
4478 in the Received: header should now be very close to that of the <= log
4479 line. There are two side-effects of this change:
4481 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4482 logged header lines no longer include the local Received: line, because
4483 it has not yet been created. The same applies to a copy of the message
4484 that is returned to a non-SMTP sender when a message is rejected.
4486 (b) When a filter file is tested using -bf, no additional Received: header
4487 is added to the test message. After some thought, I decided that this
4490 This change does not affect the value of $received_for. It is still set
4491 after address rewriting, but before local_scan() is called.
4493 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4495 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4496 gave an unhelpful panic error message, and a defer error. I have managed to
4497 change this behaviour so that it now rejects any supplied certificate,
4498 which seems right, as the list of acceptable certificates is empty.
4500 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4501 gave an unhelpful defer error. I have not managed to make this reject any
4502 supplied certificates, but the error message it gives is "no certificate
4503 supplied", which is not helpful.
4505 70. exigrep's output now also includes lines that are not associated with any
4506 message, but which match the given pattern. Implemented by a patch from
4507 Martin Sluka, which also tidied up the Perl a bit.
4509 71. Recipient callout verification, like sender verification, was using <> in
4510 the MAIL FROM command. This isn't really the right thing, since the actual
4511 sender may affect whether the remote host accepts the recipient or not. I
4512 have changed it to use the actual sender in the callout; this means that
4513 the cache record is now keyed on a recipient/sender pair, not just the
4514 recipient address. There doesn't seem to be a real danger of callout loops,
4515 since a callout by the remote host to check the sender would use <>.
4516 [SEE ABOVE: changed after hitting problems.]
4518 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4519 temporary errors. However, in the case of such a code being given after
4520 the end of a data transmission (i.e. after ".") Exim was failing to write
4521 a retry record for the message. (Yes, there was some broken host that was
4522 actually sending 8xx at this point.)
4524 73. An unknown lookup type in a host list could cause Exim to panic-die when
4525 the list was checked. (An example that provoked this was putting <; in the
4526 middle of a list instead of at the start.) If this happened during a DATA
4527 ACL check, a -D file could be left lying around. This kind of configuration
4528 error no longer causes Exim to die; instead it causes a defer errror. The
4529 incident is still logged to the main and panic logs.
4531 74. Buglet left over from Exim 3 conversion. The message "too many messages
4532 in one connection" was written to the rejectlog but not the mainlog, except
4533 when address rewriting (yes!) was being logged.
4535 75. Added write_rejectlog option.
4537 76. When a system filter was run not as root (that is, when system_filter_user
4538 was set), the values of the $n variables were not being returned to the
4539 main process; thus, they were not subsequently available in the $sn
4542 77. Added +return_path_on_delivery log selector.
4544 78. A connection timeout was being treated differently from recipients deferred
4545 when testing hosts_max_try with a message that was older than the host's
4546 retry timeout. (The host should not be counted, thus allowing all hosts to
4547 be tried at least once before bouncing.) This may have been the cause of an
4548 occasionally reported bug whereby a message would remain on the queue
4549 longer than the retry timeout, but would be bounced if a delivery was
4550 forced. I say "may" because I never totally pinned down the problem;
4551 setting up timeout/retry tests is difficult. See also the next item.
4553 79. The ultimate address timeout was not being applied to errors that involved
4554 a combination of host plus message (for example, a timeout on a MAIL
4555 command). When an address resolved to a number of possible hosts, and they
4556 were not all tried for each delivery (e.g. because of hosts_max_try), a
4557 message could remain on the queue longer than the retry timeout.
4559 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4562 81. Fixed an obscure SMTP outgoing bug which required at least the following
4563 conditions: (a) there was another message waiting for the same server;
4564 (b) the server returned 5xx to all RCPT commands in the first message so
4565 that the message was not completed; (c) the server dropped the connection
4566 or gave a negative response to the RSET that Exim sends to abort the
4567 transaction. The observed case was a dropped connection after DATA that had
4568 been sent in pipelining mode. That is, the server had advertised PIPELINING
4569 but was not implementing it correctly. The effect of the bug was incorrect
4570 behaviour, such as trying another host, and this could lead to a crash.
4576 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4577 and daemon.c were passed as pointers to ints; they should have been
4578 pointers to socklen_t variables (which are typically unsigned ints).
4580 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4583 3. Fixed a really odd bug that affected only the testing scheme; patching a
4584 certain fixed string in the binary changed the value of another string that
4585 happened to be identical to the end of the original first string.
4587 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4588 name", it returns that address as the IP address. On some operating
4589 systems (e.g. Solaris), it also passes back the IP address string as the
4590 "host name". However, on others (e.g. Linux), it passes back an empty
4591 string. Exim wasn't checking for this, and was changing the host name to an
4592 empty string, assuming it had been canonicized.
4594 5. Although rare, it is permitted to have more than one PTR record for a given
4595 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4596 all the names associated with an address, because they do in Solaris.
4597 However, it seems that they do not in Linux for data that comes from the
4598 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4599 I found this out when I moved to a new Linux workstation and tried to run
4600 the Exim test suite.
4602 To get round this problem I have changed the code so that it now does its
4603 own call to the DNS to look up PTR records when searching for a host name.
4604 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4605 addresses that are only in /etc/hosts are still found.
4607 This behaviour is, however, controlled by an option called host_lookup_
4608 order, which defaults to "bydns:byaddr". If people want to use the other
4609 order, or indeed, just use one or the other means of lookup, they can
4610 specify it in this variable.
4612 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4613 some operating systems, this comes back from gethostbyaddr() as an empty
4614 string, and this is what Exim used to test for. However, it seems that in
4615 other systems, "." is yielded. Exim now tests for this case too.
4617 7. The values of check_spool_space and check_log_space are now held internally
4618 as a number of kilobytes instead of an absolute number of bytes. If a
4619 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4620 kilobyte. This means that much larger values can be stored.
4622 8. Exim monitor: an attempt to get the action menu when not actually pointing
4623 at a message produces an empty menu entitled "No message selected". This
4624 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4625 no entries in it ("Shell widget menu has zero width and/or height"). So I
4626 have added a single, blank menu entry in this case.
4628 9. Added ${quote_local_part.
4630 10. MIME decoding is now applied to the contents of Subject: header lines when
4633 11. Now that a reference to $sender_host_address automatically causes a reverse
4634 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4635 host lookup before query-style lookups in lists that might use this
4636 variable. This has therefore been abolished, and the "net-" prefix is no
4637 longer necessary for query-style lookups.
4639 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4640 have been a typo for LFLAGS, so it has been changed.
4642 13. The install script calls Exim with "-C /dev/null" in order to find the
4643 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4644 to be output. Howeve, since Exim outputs its version number before the
4645 error, it didn't break the script. It just looked ugly. I fixed this by
4646 always allowing "-C /dev/null" if the caller is root.
4648 14. Ignore overlarge ACL variable number when reading spool file - insurance
4649 against a later release with more variables having written the file.
4651 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4652 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4653 this, as well as the form without the "IPv6" on the front.
4655 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4656 OS/Makefile-Darwin file.
4658 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4659 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4661 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4662 variable. However, some OS use ints for inodes. I've added cast to long int
4663 to get rid of the compiler warning.
4665 19. I had forgotten to lock out "/../" in configuration file names when
4666 ALT_CONFIG_PREFIX was set.
4668 20. Routers used for verification do not need to specify transports. However,
4669 if such a router generated a host list, and callout was configured, Exim
4670 crashed, because it could not find a port number from the (non-existent)
4671 transport. It now assumes port 25 in this circumstance.
4673 21. Added the -t option to exigrep.
4675 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4676 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4677 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4678 them is set, LOOKUP_LSEARCH is forced.
4680 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4682 24. Added sender and host information to the "rejected by local_scan()" log
4683 line; previously there was no indication of these.
4685 25. Added .include_if_exists.
4687 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4688 Linux. It turns out that not all file systems support this. Apparently some
4689 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4690 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4691 not supported on the file descriptor, is now ignored when Exim is trying to
4692 sync a directory. This applies only to Linux.
4694 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4696 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4697 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4700 29. Added support for maildirsize files from supplied patch (modified a bit).
4702 30. The use of :fail: followed by an empty string could lead Exim to respond to
4703 sender verification failures with (e.g.):
4705 550 Verification failed for <xxx>
4706 550 Sender verify failed
4708 where the first response line was missing the '-' that indicates it is not
4709 the final line of the response.
4711 31. The loop for finding the name of the user that called Exim had a hardwired
4712 limit of 10; it now uses the value of finduser_retries, which is used for
4713 all other user lookups.
4715 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4718 33. Exim was neglecting to zero errno before one call of strtol() when
4719 expanding a string and expecting an integer value. On some systems this
4720 resulted in spurious "integer overflow" errors. Also, it was casting the
4721 result into an int without checking.
4723 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4724 did not work. The code looks as if it has *never* worked, though it appears
4725 to have been documented since at least releast 1.62. I have made it work.
4727 35. The "timeout_DNS" error in retry rules, also documented since at least
4728 1.62, also never worked. As it isn't clear exactly what this means, and
4729 clearly it isn't a major issue, I have abolished the feature by treating it
4730 as "timeout", and writing a warning to the main and panic logs.
4732 36. The display of retry rules for -brt wasn't always showing the error code
4735 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4736 timeout_connect_A, timeout_connect_MX.
4738 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4739 to the empty sender.
4741 39. The daemon was not analysing the content of -oX till after it had closed
4742 stderr and disconnected from the controlling terminal. This meant that any
4743 syntax errors were only noted on the panic log, and the return code from
4744 the command was 0. By re-arranging the code a little, I've made the
4745 decoding happen first, so such errors now appear on stderr, and the return
4746 code is 1. However, the actual setting up of the sockets still happens in
4747 the disconnected process, so errors there are still only recorded on the
4750 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4751 connections (as happens on some IP stacks) was logged at start up time as
4752 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4753 it from "IPv6 and IPv4", which means that two separate sockets are being
4756 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4757 says whether AF_INET or AF_INET6 was passed as an argument.
4759 42. Exiwhat output was messed up when time zones were included in log
4762 43. Exiwhat now gives more information about the daemon's listening ports,
4763 and whether -tls-on-connect was used.
4765 44. The "port" option of the smtp transport is now expanded.
4767 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4768 silently ignored. Now an error message is written to the main and panic
4771 46. There's a new ACL modifier called "logwrite" which writes to a log file
4772 as soon as it is encountered.
4774 47. Added $local_user_uid and $local_user_gid at routing time.
4776 48. Exim crashed when trying to verify a sender address that was being
4779 49. Exim was recognizing only a space character after ".include". It now also
4780 recognizes a tab character.
4782 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4783 extracting the relevant information from the specification. The man page no
4784 longer contains scrambled data for the -d option, and I've added a section
4785 at the front about calling Exim under different names.
4787 51. Added "extra_headers" argument to the "mail" command in filter files.
4789 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4792 53. Installed eximstats 1.29.
4794 54. Added transport_filter_timeout as a generic transport option.
4796 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4797 Cc: header lines. This was required by RFC 822, but it not required by RFC
4800 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4801 incoming messages that did not have them. Now it does so only if the
4802 message originates locally, that is, if there is no associated remote host
4803 address. When Resent- header lines are present, this applies to the Resent-
4804 lines rather than the non-Resent- lines.
4806 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4807 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4809 58. Messages for configuration errors now include the name of the main
4810 configuration file - useful now that there may be more than one file in a
4811 list (.included file names were always shown).
4813 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4814 for those rare installations that do not start the daemon as root or run it
4815 setuid root. I've cut out the call to initgroups() if the daemon is not
4818 60. The Exim user and group can now be bound into the binary as text strings
4819 that are looked up at the start of Exim's processing.
4821 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4823 62. Added $mailstore_basename variable.
4825 63. Installed patch to sieve.c from Michael Haardt.
4827 64. When Exim failed to open the panic log after failing to open the main log,
4828 the original message it was trying to log was written to stderr and debug
4829 output, but if they were not available (the usual case in production), it
4830 was lost. Now it is written to syslog before the two lines that record the
4831 failures to open the logs.
4833 65. Users' Exim filters run in subprocesses under the user's uid. It is
4834 possible for a "deliver" command or an alias in a "personal" command to
4835 provoke an address rewrite. If logging of address rewriting is configured,
4836 this fails because the process is not running as root or exim. There may be
4837 a better way of dealing with this, but for the moment (because 4.30 needs
4838 to be released), I have disabled address rewrite logging when running a
4839 filter in a non-root, non-exim process.
4845 1. The buildconfig auxiliary program wasn't quoting the value set for
4846 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4847 not defined. This bug was masked in 4.22 by the effect that was fixed in
4850 2. Some messages that were rejected after a message id was allocated were
4851 shown as "incomplete" by exigrep. It no longer does this for messages that
4852 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4854 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4855 have allow_domain_literals set, the ID did not get logged in the <= line.
4856 Domain literals are now always recognized in Message-ID: header lines.
4858 4. The first argument for a ${extract expansion item is the key name or field
4859 number. Leading and trailing spaces in this item were not being ignored,
4860 causing some misleading effects.
4862 5. When deliver_drop_privilege was set, single queue runner processes started
4863 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4864 same command in the process it spins off) were not dropping privilege.
4866 6. When the daemon running as "exim" started a queue runner, it always
4867 re-executed Exim in the spun-off process. This is a waste of effort when
4868 deliver_drop_privilege is set. The new process now just calls the
4869 queue-runner function directly.
4875 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4878 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4879 ignored. Though the use of -oP was forcing the writing of a pid file, it
4880 was always written to the default place.
4882 3. If the message "no IP address found for host xxxx" is generated during
4883 incoming verification, it is now followed by identification of the incoming
4884 connection (so you can more easily find what provoked it).
4886 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4888 5. Added some features to "harden" Exim a bit more against certain attacks:
4890 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4891 be put in Local/Makefile. This is like the never_users runtime option,
4892 but it cannot be overridden. The default setting is "root".
4894 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4895 prefix string with which any file named in a -C command line option
4898 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4899 is retained for -C and -D only if the caller of Exim is root. Without
4900 it, the exim user may also use -C and -D and retain privilege.
4902 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4903 command line option is disabled.
4905 6. Macro names set by the -D option must start with an upper case letter, just
4906 like macro names defined in the configuration file.
4908 7. Added "dereference=" facility to LDAP.
4910 8. Two instances of the typo "uknown" in the source files are fixed.
4912 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4913 the Configure-Makefile script screwed up while processing it.
4915 10. Incorporated PCRE 4.4.
4917 11. The SMTP synchronization check was not operating right at the start of an
4918 SMTP session. For example, it could not catch a HELO sent before the client
4919 waited for the greeting. There is now a check for outstanding input at the
4920 point when the greeting is written. Because of the duplex, asynchronous
4921 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4922 way, but not yet received, when the check is performed.
4924 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4925 on TCP/IP sockets, because this apparently causes some broken clients to
4928 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4929 unchanged) from the Cygwin maintainer.
4931 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4932 was supported instead of testing for mbx. Effectively a typo.
4934 15. The spa authenticator server code was not checking that the input it
4935 received was valid base64.
4937 16. The debug output line for the "set" modifier in ACLs was not showing the
4938 name of the variable that was being set.
4940 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4942 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4943 lookup on its own. Something else was needed to trigger the lookup. For
4944 example, a match in host_lookup or the need for a host name in a host list.
4945 Now, if $sender_host_name is referenced and the host name has not yet been
4946 looked up, a lookup is performed. If the lookup fails, the variable remains
4947 empty, and $host_lookup_failed is set to "1".
4949 19. Added "eqi" as a case-independent comparison operator.
4951 20. The saslauthd authentication condition could segfault if neither service
4952 nor realm was specified.
4954 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4955 error message that was logged was misleading, and incoming SMTP
4956 connections were dropped. The message is now more accurate, and temporary
4957 errors are given to SMTP connections.
4959 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4960 (except RSET) with a 421 error, until QUIT is received. However, it was
4961 failing to send a response to QUIT.
4963 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4964 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4965 ACL is now run after the helo_try_verify_hosts code.
4967 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4968 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4969 case-independent, so other case variants are also recognized). Apparently
4970 some systems use these upper case variants.
4972 25. If more than two messages were waiting for the same host, and a transport
4973 filter was specified for the transport, Exim sent two messages over the
4974 same TCP/IP connection, and then failed with "socket operation on non-
4975 socket" when it tried to send the third.
4977 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4979 27. The extern definition of crypt16() in expand.c was not being excluded when
4980 the OS had its own crypt16() function.
4982 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4983 as a preferred synonym for return_size_limit, both as an option and as an
4986 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4988 30. Changed the default configuration ACL to relax the local part checking rule
4989 for addresses that are not in any local domains. For these addresses,
4990 slashes and pipe symbols are allowed within local parts, but the sequence
4991 /../ is explicitly forbidden.
4993 31. SPA server authentication was not clearing the challenge buffer before
4996 32. log_message in a "warn" ACL statement was writing to the reject log as
4997 well as to the main log, which contradicts the documentation and doesn't
4998 seem right (because no rejection is happening). So I have stopped it.
5000 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5001 However, I am unable to do any testing of this.
5003 34. Fixed an infelicity in the appendfile transport. When checking directories
5004 for a mailbox, to see if any needed to be created, it was accidentally
5005 using path names with one or more superfluous leading slashes; tracing
5006 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5008 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5009 contents are added to the log line that is written for every discarded
5010 recipient. (Previously a log_message setting was ignored.)
5012 36. The ${quote: operator now quotes the string if it is empty.
5014 37. The install script runs exim in order to find its version number. If for
5015 some reason other than non-existence or emptiness, which it checks, it
5016 could not run './exim', it was installing it with an empty version number,
5017 i.e. as "exim-". This error state is now caught, and the installation is
5020 38. An argument was missing from the function that creates an error message
5021 when Exim fails to connect to the socket for saslauthd authentication.
5022 This could cause Exim to crash, or give a corrupted message.
5024 39. Added isip, isip4, and isip6 to ${if conditions.
5026 40. The ACL variables $acl_xx are now saved with the message, and can be
5027 accessed later in routers, transports, and filters.
5029 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5030 strings in the file are not string-expanded.
5032 42. If a MAIL command specified a SIZE value that was too large to fit into an
5033 int variable, the check against message_size_limit failed. Such values are
5034 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5035 day this will have to be increased, but I don't think I want to be around
5036 when emails are that large.
5043 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5044 iconv() is not standard in FreeBSD.
5046 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5047 IPv6 enabled. The observed symptom was a segmentation fault on return from
5048 the function os_common_find_running_interfaces() in src/os.c.
5050 3. In the check_special_case() function in daemon.c I had used "errno" as an
5051 argument name, which causes warnings on some systems. This was basically a
5052 typo, since it was named "eno" in the comments!
5054 4. The code that waits for the clock to tick (at a resolution of some fraction
5055 of a second) so as to ensure message-id uniqueness was always waiting for
5056 at least one whole tick, when it could have waited for less. [This is
5057 almost certainly not relevant at current processor speeds, where it is
5058 unlikely to ever wait at all. But we try to future-proof.]
5060 5. The function that sleeps for a time interval that includes fractions of a
5061 second contained a race. It did not block SIGALRM between setting the
5062 timer, and suspending (a couple of lines later). If the interval was short
5063 and the sigsuspend() was delayed until after it had expired, the suspension
5064 never ended. On busy systems this could lead to processes getting stuck for
5067 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5068 process, before it forks any delivery processes. The open lookup caching
5069 mechanism meant that the open file or database connection was passed into
5070 the delivery process. The problem was that delivery processes always tidy
5071 up cached lookup data. This could cause a problem for the next delivery
5072 process started by the queue runner, because the external queue runner
5073 process does not know about the closure. So the next delivery process
5074 still has data in the lookup cache. In the case of a file lookup, there was
5075 no problem because closing a file descriptor in a subprocess doesn't affect
5076 the parent. However, if the lookup was caching a connection to a database,
5077 the connection was closed, and the second delivery process was likely to
5078 see errors such as "PGSQL: query failed: server closed the connection
5079 unexpectedly". The problem has been fixed by closing all cached lookups
5080 in a queue runner before running a delivery process.
5082 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5083 seem to have the "const" qualifier which it has on other OS. I've
5086 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5087 *of the same type* (client or server) with the same public name that an
5088 error should be diagnosed.
5090 9. When Exim looked up a host name for an IP address, but failed to find the
5091 original IP address when looking up the host name (a safety check), it
5092 output the message "<ip address> does not match any IP for NULL", which was
5093 confusing, to say the least. The bug was that the host name should have
5094 appeared instead of "NULL".
5096 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5097 user that is built into the binary, and the -C or -D options is used, root
5098 privilege is dropped before the configuration file is read. In addition,
5099 logging is switched to stderr instead of the normal log files. If the
5100 configuration then re-defines the Exim user, the unprivileged environment
5101 is probably not what is expected, so Exim logs a panic warning message (but
5104 However, if deliver_drop_privilege is set, the unprivileged state may well
5105 be exactly what is intended, so the warning has been cut out in that case,
5106 and Exim is allowed to try to write to its normal log files.
5112 1. smtp_return_error_details was not giving details for temporary sender
5113 or receiver verification errors.
5115 2. Diagnose a configuration error if two authenticators have the same public
5118 3. Exim used not to create the message log file for a message until the first
5119 delivery attempt. This could be confusing when incoming messages were held
5120 for policy or load reasons. The message log file is now created at the time
5121 the message is received, and an initial "Received" line is written to it.
5123 4. The automatically generated man page for command line options had a minor
5124 bug that caused no ill effects; however, a more serious problem was that
5125 the procedure for building the man page automatically didn't always
5126 operate. Consequently, release 4.20 contains an out-of-date version. This
5127 shouldn't happen again.
5129 5. When building Exim with embedded Perl support, the script that builds the
5130 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5131 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5133 6. The freeze_tell option was not being used for messages that were frozen on
5134 arrival, either by an ACL or by local_scan().
5136 7. Added the smtp_incomplete_transaction log selector.
5138 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5139 was accepting AUTH without a new EHLO.
5141 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5142 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5145 10. Logging of TCP/IP connections (when configured) now happens in the main
5146 daemon process instead of the child process, so that the TCP/IP connection
5147 count is more accurate (but it can never be perfect).
5149 11. The use of "drop" in a nested ACL was not being handled correctly in the
5150 outer ACL. Now, if condition failure induced by the nested "drop" causes
5151 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5152 or "require"), the connection is dropped.
5154 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5155 that yield "discard" can now be used with an "accept" or a "discard" verb,
5156 but an error is generated for any others (because I can't see a useful way
5157 to define what should happen).
5159 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5160 are now processed in the same way as lines in the Exim configuration file.
5161 In particular, continuation lines are supported.
5163 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5165 15. Added -ti meaning -t -i.
5167 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5168 domains, and warn by logging if others are found.
5170 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5171 returned from GIFCONF when Exim is trying to find the list of interfaces on
5172 a host. The code in os.c has been modified to copy each ifreq to an aligned
5173 structure in all cases.
5175 Also, in some cases, the returned ifreq's were being copied to a 'struct
5176 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5177 means the last couple of bytes of an IPv6 address could be chopped if the
5178 ifreq contained only a normal sockaddr (14 bytes storage).
5180 18. Named domain lists were not supported in the hosts_treat_as_local option.
5181 An entry such as +xxxx was not recognized, and was treated as a literal
5184 19. Ensure that header lines added by a DATA ACL are included in the reject log
5185 if the ACL subsequently rejects the message.
5187 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5188 MD5 (which is deprecated).
5190 21. When testing a filter file using -bf, Exim was writing a message when it
5191 took the sender from a "From " line in the message, but it was not doing so
5192 when it took $return_path from a Return-Path: header line. It now does.
5194 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5195 with a valid header line field name (a series of printing characters
5196 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5198 23. Changed "disc" in the source to "disk" to conform to the documentation and
5199 the book and for uniformity.
5201 24. Ignore Sendmail's -Ooption=value command line item.
5203 25. When execve() failed while trying to run a command in a pipe transport,
5204 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5205 could be confused with a return value of 69 from the command itself. This
5206 has been changed to 127, the value the shell returns if it is asked to run
5207 a non-existent command. The wording for the related log line suggests a
5208 non-existent command as the problem.
5210 26. If received_header_text expands to an empty string, do not add a Received:
5211 header line to the message. (Well, it adds a token one on the spool, but
5212 marks it "old" so that it doesn't get used or transmitted.)
5214 27. Installed eximstats 1.28 (addition of -nt option).
5216 28. There was no check for failure on the call to getsockname() in the daemon
5217 code. This can fail if there is a shortage of resources on the system, with
5218 ENOMEM, for example. A temporary error is now given on failure.
5220 29. Contrary to the C standard, it seems that in some environments, the
5221 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5222 program. Exim now does this explicitly; it affects the formatting of
5223 timestamps using strftime().
5225 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5226 complaints. I've now initialized all the variables, to avoid this.
5228 32. Header lines added by a system filter were not being "seen" during
5229 transport-time rewrites.
5231 33. The info_callback() function passed to OpenSSL is set up with type void
5232 (*)(SSL *, int, int), as described somewhere. However, when calling the
5233 function (actually a macro) that sets it up, the type void(*)() is
5234 expected. I've put in a cast to prevent warnings from picky compilers.
5236 34. If a DNS black list lookup found a CNAME record, but there were no A
5237 records associated with the domain it pointed at, Exim crashed.
5239 35. If a DNS black list lookup returned more than one A record, Exim ignored
5240 all but the first. It now scans all returned addresses if a particular IP
5241 value is being sought. In this situation, the contents of the
5242 $dnslist_value variable are a list of all the addresses, separated by a
5245 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5246 to accept a host name and all its aliases if the forward lookup for any of
5247 them yielded the IP address of the incoming connection. Now it accepts only
5248 those names whose forward lookup yields the correct IP address. Any other
5249 names are discarded. This closes a loophole whereby a rogue DNS
5250 administrator could create reverse DNS records to break through a
5251 wildcarded host restriction in an ACL.
5253 37. If a user filter or a system filter that ran in a subprocess used any of
5254 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5255 the wrong values were passed to the pipe command ($thisaddress had the
5256 value of $0, $0 had the value of $1, etc). This bug was introduced by
5257 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5259 38. Improved the line breaking for long SMTP error messages from ACLs.
5260 Previously, if there was no break point between 40 and 75 characters, Exim
5261 left the rest of the message alone. Two changes have been made: (a) I've
5262 reduced the minimum length to 35 characters; (b) if it can't find a break
5263 point between 35 and 75 characters, it looks ahead and uses the first one
5264 that it finds. This may give the occasional overlong line, but at least the
5265 remaining text gets split now.
5267 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5268 file descriptors might be low, and that setting 1000 would always raise it.
5269 It turns out that in some environments, the limit is already over 1000 and
5270 that lowering it causes trouble. So now Exim takes care not to decrease it.
5272 40. When delivering a message, the value of $return_path is set to $sender_
5273 address at the start of routing (routers may change the value). By an
5274 oversight, this default was not being set up when an address was tested by
5275 -bt or -bv, which affected the outcome if any router or filter referred to
5278 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5279 log only when "message" or "log_message" are set. However, if one of the
5280 conditions was an address verification, or a call to a nested ACL, the
5281 messages generated by the underlying test were being passed through. This
5282 no longer happens. The underlying message is available in $acl_verify_
5283 message for both "message" and "log_message" expansions, so it can be
5284 passed through if needed.
5286 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5287 new expansion $bh_ to give the encoded byte string without charset
5288 translation. Translation happens only if iconv() is available; HAVE_ICONV
5289 indicates this at build time. HEADERS_CHARSET gives the charset to
5290 translate to; headers_charset can change it in the configuration, and
5291 "headers charset" can change it in an individual filter file.
5293 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5294 that creates RFC 2047 encoded "words" labels them as that charset instead
5295 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5296 expansion operator; (ii) when Exim creates a From: line for a local
5297 message; (iii) when a header line is rewritten to include a "phrase" part.
5299 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5300 buggy, causing it to skip the first lines of messages whose message ID
5301 ended in 'D'. This would not have bitten before Exim release 4.14, because
5302 message IDs were unlikely to end in 'D' before then. The effect was to have
5303 incorrect size information for certain domains.
5305 45. #include "config.h" was missing at the start of the crypt16.c module. This
5306 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5309 46. If there was a timeout during a "random" callout check, Exim treated it as
5310 a failure of the random address, and carried on sending RSET and the real
5311 address. If the delay was just some slowness somewhere, the response to the
5312 original RCPT would be taken as a response to RSET and so on, causing
5313 mayhem of various kinds.
5315 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5316 when I implemented it. It didn't allow for the fact that some option values
5317 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5318 the right test for positive values.
5320 48. Domain names in DNS records are case-independent. Exim always looks them up
5321 in lower case. Some resolvers return domain names in exactly the case they
5322 appear in the zone file, that is, they may contain uppercase letters. Not
5323 all resolvers do this - some return always lower case. Exim was treating a
5324 change of case by a resolver as a change of domain, similar to a widening
5325 of a domain abbreviation. This triggered its re-routing code and so it was
5326 trying to route what was effectively the same domain again. This normally
5327 caused routing to fail (because the router wouldn't handle the domain
5328 twice). Now Exim checks for this case specially, and just changes the
5329 casing of the domain that it ultimately uses when it transmits the message
5332 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5335 50. If a filter generated a file delivery with a non-absolute name (possible if
5336 no home directory exists for the router), the forbid_file option was not
5339 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5340 the existing equality matching.
5342 52. Exim was using ints instead of ino_t variables in some places where it was
5343 dealing with inode numbers.
5345 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5346 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5347 called TMPDIR, and if it finds it is different, it changes its value.
5349 54. The smtp_printf() function is now made available to local_scan() so
5350 additional output lines can be written before returning. There is also an
5351 smtp_fflush() function to enable the detection of a dropped connection.
5352 The variables smtp_input and smtp_batched_input are exported to
5355 55. Changed the default runtime configuration: the message "Unknown user"
5356 has been removed from the ACL, and instead placed on the localuser router,
5357 using the cannot_route_message feature. This means that any verification
5358 failures that generate their own messages won't get overridden. Similarly,
5359 the "Unrouteable address" message that was in the ACL for unverifiable
5360 relay addresses has also been removed.
5362 56. Added hosts_avoid_esmtp to the smtp transport.
5364 57. The exicyclog script was not checking for the esoteric option
5365 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5366 will work only if exicyclog is run under the appropriate euid.
5368 58. Following a discussion on the list, the rules by which Exim recognises line
5369 endings on incoming messages have been changed. The -dropcr and drop_cr
5370 options are now no-ops, retained only for backwards compatibility. The
5371 following line terminators are recognized: LF CRLF CR. However, special
5372 processing applies to CR:
5374 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5375 nor a local message in the state where . is a terminator.
5377 (ii) If a bare CR is encountered in a header line, an extra space is added
5378 after the line terminator so as not to end the header. The reasoning
5379 behind this is that bare CRs in header lines are most likely either
5380 to be mistakes, or people trying to play silly games.
5382 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5383 was being incorrectly given as 18 bytes larger than it should have been.
5384 This is a VOB (very old bug).
5386 60. This may never have affected anything current, but just in case it has:
5387 When the local host is found other than at the start of a list of hosts,
5388 the local host, those with the same MX, and any that follow, are discarded.
5389 When the list in question was part of a longer list of hosts, the following
5390 hosts (not currently being processed) were also being discarded. This no
5391 longer happens. I'm not sure if this situation could ever has previously
5394 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5397 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5398 header. This is something that is recommended in a new Internet Draft, and
5399 is something that is documented as being done by Sendmail. There are two
5400 possible values. For messages generated by the autoreply transport, Exim
5403 Auto-Submitted: auto-replied
5405 whereas for all other generated messages (e.g. bounces) it adds
5407 Auto-Submitted: auto-generated
5409 63. The "personal" condition in filters now includes a test for the
5410 Auto-Submitted: header. If it contains the string "auto-" the message it
5411 not considered personal.
5413 64. Added rcpt_include_affixes as a generic transport option.
5415 65. Added queue_only_override (default true).
5417 66. Added the syslog_duplication option.
5419 67. If what should have been the first header line of a message consisted of
5420 a space followed by a colon, Exim was mis-interpreting it as a header line.
5421 It isn't of course - it is syntactically invalid and should therefore be
5422 treated as the start of the message body. The misbehaviour could have
5423 caused a number of strange effects, including loss of data in subsequent
5424 header lines, and spool format errors.
5426 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5427 client host had authenticated. This control can now be exercised by an ACL
5428 for more flexibility.
5430 69. By default, callouts do not happen when testing with -bh. There is now a
5431 variant, -bhc, which does actually run the callout code, including
5432 consulting and updating the callout cache.
5434 70. Added support for saslauthd authentication, courtesy of Alexander
5437 71. If statvfs() failed on the spool or log directories while checking their
5438 size for availability, Exim confusingly gave the error "space shortage".
5439 Furthermore, in debugging mode it crashed with a floating point exception.
5440 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5441 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5442 serious problem, Exim now writes to the main and panic logs when this
5443 happens, with details of the failure. It then refuses to accept the
5444 incoming message, giving the message "spool directory problem" or "log
5445 directory problem" with a 421 code for SMTP messages.
5447 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5448 0, 1, and 2 exist, because some OS complain for execs without them (see
5449 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5450 descriptors. However, the code omitted to check that the open succeeded,
5451 causing mysterious errors if for some reason the permissions on /dev/null
5452 got screwed. Now Exim writes a message to the main and panic logs, and
5453 bombs out if it can't open /dev/null.
5455 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5456 interact so that it is all more flexible. It is supposed to remain
5457 backwards compatible. Also added extra_local_interfaces.
5459 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5460 to bomb out with an assertion failure - to the client this appears as a
5461 connection drop. This problem occurs in the part of the code that was taken
5462 from the Samba project. Fortunately, the assertion is in a very simple
5463 function, so I have fixed this by reproducing the function inline in the
5464 one place where it is called, and arranging for authentication to fail
5465 instead of killing the process with assert().
5467 75. The SPA client code was not working when the server requested OEM rather
5468 than Unicode encoding.
5470 76. Added code to make require_files with a specific uid setting more usable in
5471 the case where statting the file as root fails - usually a non-root-mounted
5472 NFS file system. When this happens and the failure is EACCES, Exim now
5473 forks a subprocess and does the per-uid checking as the relevant uid.
5475 77. Added process_log_path.
5477 78. If log_file_path was not explicitly set, a setting of check_log_space or
5478 check_log_inodes was ignored.
5480 79. If a space check for the spool or log partitions fails, the incident is now
5481 logged. Of course, in the latter case the data may get lost...
5483 80. Added the %p formatting code to string_format() so that it can be used to
5484 print addresses in debug_print(). Adjusted all the address printing in the
5485 debugging in store.c to use %p rather than %d.
5487 81. There was a concern that a line of code in smtp_in.c could overflow a
5488 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5489 initially expressed, the concern was not well-founded, because trailing
5490 spaces are removed early. However, if the trailing spaces were followed by
5491 a NULL, they did not get removed, so the overflow was possible. Two fixes
5494 (a) I re-wrote the offending code in a cleaner fashion.
5495 (b) If an incoming SMTP command contains a NULL character, it is rejected
5498 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5499 runs initgroups(), so that if the Exim user is in any additional groups,
5500 they will be used during message reception.
5506 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob