1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
35 PP/01 Solaris build fix for Oracle's LDAP libraries.
36 Bugzilla 1109, patch from Stephen Usher.
38 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
40 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
43 TF/02 Fix a couple more cases where we did not log the error message
44 when unlink() failed. See also change 4.74-TF/03.
46 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
47 lock up or crash if it happened to be inside a call to libc when it
48 got a SIGUSR1 from exiwhat.
50 The SIGUSR1 handler appends the current process status to the process
51 log which is later printed by exiwhat. It used to use the general
52 purpose logging code to do this, but several functions it calls are
55 The new output code in the SIGUSR1 handler is specific to the process
56 log, and simple enough that it's easy to inspect for signal safety.
57 Removing some special cases also simplifies the general logging code.
58 Removing the spurious timestamps from the process log simplifies
61 TF/04 Improved ratelimit ACL condition.
63 The /noupdate option has been deprecated in favour of /readonly which
64 has clearer semantics. The /leaky, /strict, and /readonly update modes
65 are mutually exclusive. The update mode is no longer included in the
66 database key; it just determines when the database is updated. (This
67 means that when you upgrde Exim will forget old rate measurements.)
69 Exim now checks that the per_* options are used with an update mode that
70 makes sense for the current ACL. For example, when Exim is processing a
71 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
72 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
73 must specify per_mail/readonly. If you omit the update mode it defaults to
74 /leaky where that makes sense (as before) or /readonly where required.
76 The /noupdate option is now undocumented but still supported for
77 backwards compatibility. It is equivalent to /readonly except that in
78 ACLs where /readonly is required you may specify /leaky/noupdate or
79 /strict/noupdate which are treated the same as /readonly.
81 A useful new feature is the /count= option. This is a generalization
82 of the per_byte option, so that you can measure the throughput of other
83 aggregate values. For example, the per_byte option is now equivalent
84 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
86 The per_rcpt option has been generalized using the /count= mechanism
87 (though it's more complicated than the per_byte equivalence). When it is
88 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
89 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
90 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
91 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
92 Note that using per_rcpt with a non-readonly update mode in more than
93 one ACL will cause the recipients to be double-counted. (The per_mail
94 and per_byte options don't have this problem.)
96 The handling of very low rates has changed slightly. If the computed rate
97 is less than the event's count (usually one) then this event is the first
98 after a long gap. In this case the rate is set to the same as this event's
99 count, so that the first message of a spam run is counted properly.
101 The major new feature is a mechanism for counting the rate of unique
102 events. The new per_addr option counts the number of different
103 recipients that someone has sent messages to in the last time period. It
104 behaves like per_rcpt if all the recipient addresses are different, but
105 duplicate recipient addresses do not increase the measured rate. Like
106 the /count= option this is a general mechanism, so the per_addr option
107 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
108 example, measure the rate that a client uses different sender addresses
109 with the options per_mail/unique=$sender_address. There are further
110 details in the main documentation.
112 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
114 TF/06 Removed a few PCRE remnants.
116 TF/07 Automatically extract Exim's version number from tags in the git
117 repository when doing development or release builds.
119 PP/02 Raise smtp_cmd_buffer_size to 16kB.
120 Bugzilla 879. Patch from Paul Fisher.
122 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
123 Heavily based on revision 40f9a89a from Simon Arlott's tree.
126 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
128 PP/05 Variable $av_failed, true if the AV scanner deferred.
129 Bugzilla 1078. Patch from John Horne.
131 PP/06 Stop make process more reliably on build failure.
132 Bugzilla 1087. Patch from Heiko Schlittermann.
134 PP/07 Make maildir_use_size_file an _expandable_ boolean.
135 Bugzilla 1089. Patch from Heiko Schlittermann.
137 PP/08 Handle ${run} returning more data than OS pipe buffer size.
138 Bugzilla 1131. Patch from Holger Weiß.
140 PP/09 Handle IPv6 addresses with SPF.
141 Bugzilla 860. Patch from Wolfgang Breyha.
143 PP/10 GnuTLS: support TLS 1.2 & 1.1.
145 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
148 PP/11 match_* no longer expand right-hand-side by default.
149 New compile-time build option, EXPAND_LISTMATCH_RHS.
150 New expansion conditions, "inlist", "inlisti".
152 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
154 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
156 PP/14 fix log_write() format string regression from TF/03.
157 Bugzilla 1152. Patch from Dmitry Isaikin.
163 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
165 PP/02 Harmonised TLS library version reporting; only show if debugging.
166 Layout now matches that introduced for other libraries in 4.74 PP/03.
168 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
170 PP/04 New "dns_use_edns0" global option.
172 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
175 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
176 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
178 TK/01 Updated PolarSSL code to 0.14.2.
179 Bugzilla 1097. Patch from Andreas Metzler.
181 PP/07 Catch divide-by-zero in ${eval:...}.
184 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
187 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
188 format-string attack -- SECURITY: remote arbitrary code execution.
190 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
191 time unintentionally subject to list matching rules, letting the header
192 cause arbitrary Exim lookups (of items which can occur in lists, *not*
193 arbitrary string expansion). This allowed for information disclosure.
195 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
196 INT_MIN/-1 -- value coerced to INT_MAX.
202 NM/01 Workround for PCRE version dependency in version reporting
205 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
206 This fixes portability to compilers other than gcc, notably
207 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
209 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
210 makefiles for portability to HP-UX and POSIX correctness.
212 PP/01 Permit LOOKUP_foo enabling on the make command-line.
213 Also via indented variable definition in the Makefile.
214 (Debugging by Oliver Heesakkers).
216 PP/02 Restore caching of spamd results with expanded spamd_address.
217 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
219 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
220 Improves build reliability. Fix from: Frank Elsner
222 NM/02 Fix wide character breakage in the rfc2047 coding
223 Fixes bug 1064. Patch from Andrey N. Oktyabrski
225 NM/03 Allow underscore in dnslist lookups
226 Fixes bug 1026. Patch from Graeme Fowler
228 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
229 Code patches from Adam Ciarcinski of NetBSD.
231 NM/04 Fixed exiqgrep to cope with mailq missing size issue
234 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
235 is logged, to avoid truncation. Patch from John Horne.
237 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
238 Patch from Jakob Hirsch.
240 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
241 SQL string expansion failure details.
242 Patch from Andrey Oktyabrski.
244 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
245 Patch from Simon Arlott.
247 PP/09 New lookups functionality failed to compile on old gcc which rejects
248 extern declarations in function scope.
249 Patch from Oliver Fleischmann
251 PP/10 Use sig_atomic_t for flags set from signal handlers.
252 Check getgroups() return and improve debugging.
253 Fixed developed for diagnosis in bug 927 (which turned out to be
256 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
257 Patch from Mark Zealey.
259 PP/12 Bugzilla 1056: Improved spamd server selection.
260 Patch from Mark Zealey.
262 PP/13 Bugzilla 1086: Deal with maildir quota file races.
263 Based on patch from Heiko Schlittermann.
265 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
266 Patch from Uwe Doering, sign-off by Michael Haardt.
268 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
269 variable declaration deep within a block. Bug and patch from
272 PP/15 lookups-Makefile IRIX compatibilty coercion.
274 PP/16 Make DISABLE_DKIM build knob functional.
276 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
277 Patch by Simon Arlott
279 TF/03 Fix valgrind.h portability to C89 compilers that do not support
280 variable argument macros. Our copy now differs from upstream.
286 TF/01 Failure to get a lock on a hints database can have serious
287 consequences so log it to the panic log.
289 TF/02 Log LMTP confirmation messages in the same way as SMTP,
290 controlled using the smtp_confirmation log selector.
292 TF/03 Include the error message when we fail to unlink a spool file.
294 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
295 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
296 for maintaining out-of-tree patches for some time.
298 PP/01 Bugzilla 139: Documentation and portability issues.
299 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
300 Handle per-OS dynamic-module compilation flags.
302 PP/02 Let /dev/null have normal permissions.
303 The 4.73 fixes were a little too stringent and complained about the
304 permissions on /dev/null. Exempt it from some checks.
305 Reported by Andreas M. Kirchwitz.
307 PP/03 Report version information for many libraries, including
308 Exim version information for dynamically loaded libraries. Created
309 version.h, now support a version extension string for distributors
310 who patch heavily. Dynamic module ABI change.
312 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
313 privilege escalation vulnerability whereby the Exim run-time user
314 can cause root to append content of the attacker's choosing to
317 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
320 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
321 If dropping privileges for untrusted macros, we disabled normal logging
322 on the basis that it would fail; for the Exim run-time user, this is not
323 the case, and it resulted in successful deliveries going unlogged.
324 Fixed. Reported by Andreas Metzler.
330 PP/01 Date: & Message-Id: revert to normally being appended to a message,
331 only prepend for the Resent-* case. Fixes regression introduced in
332 Exim 4.70 by NM/22 for Bugzilla 607.
334 PP/02 Include check_rfc2047_length in configure.default because we're seeing
335 increasing numbers of administrators be bitten by this.
337 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
339 PP/03 Bugzilla 994: added openssl_options main configuration option.
341 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
343 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
345 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
347 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
348 without a peer certificate, leading to a segfault because of an
349 assumption that peers always have certificates. Be a little more
350 paranoid. Problem reported by Martin Tscholak.
352 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
353 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
354 NB: ClamAV planning to remove STREAM in "middle of 2010".
355 CL also introduces -bmalware, various -d+acl logging additions and
356 more caution in buffer sizes.
358 PP/09 Implemented reverse_ip expansion operator.
360 PP/10 Bugzilla 937: provide a "debug" ACL control.
362 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
364 PP/12 Bugzilla 973: Implement --version.
366 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
368 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
370 PP/15 Bugzilla 816: support multiple condition rules on Routers.
372 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
373 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
374 ignore trailing whitespace.
376 JJ/02 prevent non-panic DKIM error from being sent to paniclog
378 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
381 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
382 Notification from Dr Andrew Aitchison.
384 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
385 ExtendedDetectionInfo response format.
386 Notification from John Horne.
388 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
391 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
392 XSL and documented dependency on system catalogs, with examples of how
395 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
398 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
399 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
400 configuration file which is writeable by the Exim user or group.
402 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
403 of configuration files to cover files specified with the -C option if
404 they are going to be used with root privileges, not just the default
407 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
408 option (effectively making it always true).
410 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
411 files to be used while preserving root privileges.
413 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
414 that rogue child processes cannot use them.
416 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
417 run-time user, instead of root.
419 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
420 Exim run-time user without dropping privileges.
422 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
423 result string, instead of calling string_vformat() twice with the same
426 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
427 for other users. Others should always drop root privileges if they use
428 -C on the command line, even for a whitelisted configure file.
430 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
432 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
438 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
439 $data_path, and $header_path variables; fixed documentation bugs and
442 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
443 exipick to access non-standard spools, including the "frozen" queue
446 NM/01 Bugzilla 965: Support mysql stored procedures.
447 Patch from Alain Williams
449 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
451 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
452 Patch from Andreas Metzler
454 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
455 Patch from Kirill Miazine
457 NM/05 Bugzilla 671: Added umask to procmail example.
459 JJ/03 installed exipick 20100323.0, fixing doc bug
461 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
462 directory. Notification and patch from Dan Rosenberg.
464 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
466 TK/02 Improve log output when DKIM signing operation fails.
468 MH/01 Treat the transport option dkim_domain as a colon separated
469 list, not as a single string, and sign the message with each element,
470 omitting multiple occurences of the same signer.
472 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
473 Bugzilla 985, 986. Patch by Simon Arlott
475 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
476 Patch by Simon Arlott
478 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
479 MBX locking. Notification from Dan Rosenberg.
485 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
487 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
489 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
491 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
493 NM/04 Bugzilla 915: Flag broken perl installation during build.
499 TK/01 Added patch by Johannes Berg that expands the main option
500 "spamd_address" if it starts with a dollar sign.
502 TK/02 Write list of recipients to X-Envelope-Sender header when building
503 the mbox-format spool file for content scanning (suggested by Jakob
506 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
507 (http://www.dcc-servers.net/) support via dccifd. Activated by
508 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
510 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
511 by Mark Daniel Reidel <mr@df.eu>.
513 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
514 When building exim an external PCRE library is now needed -
515 PCRE is a system library on the majority of modern systems.
516 See entry on PCRE_LIBS in EDITME file.
518 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
519 conversation. Added nologin parameter to request.
520 Patch contributed by Kirill Miazine.
522 TF/01 Do not log submission mode rewrites if they do not change the address.
524 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
526 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
527 log files in place. Contributed by Roberto Lima.
529 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
531 TF/03 Bugzilla 615: When checking the local_parts router precondition
532 after a local_part_suffix or local_part_prefix option, Exim now
533 does not use the address's named list lookup cache, since this
534 contains cached lookups for the whole local part.
536 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
537 Robert Millan. Documentation is in experimental-spec.txt.
539 TF/04 Bugzilla 668: Fix parallel build (make -j).
541 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
543 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
544 Patch provided by Jan Srzednicki.
546 TF/05 Leading white space used to be stripped from $spam_report which
547 wrecked the formatting. Now it is preserved.
549 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
550 that they are available at delivery time.
552 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
554 TF/08 TLS error reporting now respects the incoming_interface and
555 incoming_port log selectors.
557 TF/09 Produce a more useful error message if an SMTP transport's hosts
558 setting expands to an empty string.
560 NM/06 Bugzilla 744: EXPN did not work under TLS.
561 Patch provided by Phil Pennock.
563 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
564 Patch provided by Richard Godbee.
566 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
567 acl_smtp_notquit, added index entry.
569 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
570 Patch provided by Eugene Bujak.
572 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
573 accept(). Patch provided by Maxim Dounin.
575 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
576 Patch provided by Phil Pennock.
578 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
580 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
581 Patch provided by Brad "anomie" Jorsch.
583 NM/14 Bugzilla 622: Added timeout setting to transport filter.
584 Patch provided by Dean Brooks.
586 TK/05 Add native DKIM support (does not depend on external libraries).
588 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
589 Patch provided by Graeme Fowler.
591 NM/16 Bugzilla 851: Documentation example syntax fix.
593 NM/17 Changed NOTICE file to remove references to embedded PCRE.
595 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
598 NM/19 Bugzilla 745: TLS version reporting.
599 Patch provided by Phil Pennock.
601 NM/20 Bugzilla 167: bool: condition support.
602 Patch provided by Phil Pennock.
604 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
605 clients. Patch provided by Phil Pennock.
607 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
608 Patch provided by Brad "anomie" Jorsch.
610 NM/23 Bugzilla 687: Fix misparses in eximstats.
611 Patch provided by Heiko Schlittermann.
613 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
614 Patch provided by Heiko Schlittermann.
616 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
617 plus update to original patch.
619 NM/26 Bugzilla 799: Documentation correction for ratelimit.
621 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
622 Patch provided by David Brownlee.
624 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
626 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
628 NM/30 Bugzilla 888: TLS documentation bugfixes.
630 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
632 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
633 Unlike the original bugzilla I have changed all shell scripts in src tree.
635 NM/33 Bugzilla 898: Transport filter timeout fix.
636 Patch by Todd Rinaldo.
638 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
639 Patch by Serge Demonchaux.
641 NM/35 Bugzilla 39: Base64 decode bug fixes.
642 Patch by Jakob Hirsch.
644 NM/36 Bugzilla 909: Correct connect() call in dcc code.
646 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
648 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
650 NM/39 Bugzilla 911: Fixed MakeLinks build script.
656 TK/01 Add preliminary DKIM support. Currently requires a forked version of
657 ALT-N's libdkim that I have put here:
658 http://duncanthrax.net/exim-experimental/
660 Note to Michael Haardt: I had to rename some vars in sieve.c. They
661 were called 'true' and it seems that C99 defines that as a reserved
662 keyword to be used with 'bool' variable types. That means you could
663 not include C99-style headers which use bools without triggering
664 build errors in sieve.c.
666 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
667 as mailq or other aliases. Changed the --help handling significantly
668 to do whats expected. exim_usage() emits usage/help information.
670 SC/01 Added the -bylocaldomain option to eximstats.
672 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
674 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
676 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
682 PH/01 Another patch from the Sieve maintainer.
684 PH/02 When an IPv6 address is converted to a string for single-key lookup
685 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
686 dots are used instead of colons so that keys in lsearch files need not
687 contain colons. This was done some time before quoting was made available
688 in lsearch files. However, iplsearch files do require colons in IPv6 keys
689 (notated using the quote facility) so as to distinguish them from IPv4
690 keys. This meant that lookups for IP addresses in host lists did not work
691 for iplsearch lookups.
693 This has been fixed by arranging for IPv6 addresses to be expressed with
694 colons if the lookup type is iplsearch. This is not incompatible, because
695 previously such lookups could never work.
697 The situation is now rather anomolous, since one *can* have colons in
698 ordinary lsearch keys. However, making the change in all cases is
699 incompatible and would probably break a number of configurations.
701 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
704 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
705 conversion specification without a maximum field width, thereby enabling
706 a rogue spamd server to cause a buffer overflow. While nobody in their
707 right mind would setup Exim to query an untrusted spamd server, an
708 attacker that gains access to a server running spamd could potentially
709 exploit this vulnerability to run arbitrary code as the Exim user.
711 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
712 $primary_hostname instead of what libspf2 thinks the hosts name is.
714 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
715 a directory entry by the name of the lookup key. Previously, if a
716 symlink pointed to a non-existing file or a file in a directory that
717 Exim lacked permissions to read, a lookup for a key matching that
718 symlink would fail. Now it is enough that a matching directory entry
719 exists, symlink or not. (Bugzilla 503.)
721 PH/03 The body_linecount and body_zerocount variables are now exported in the
724 PH/04 Added the $dnslist_matched variable.
726 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
727 This means they are set thereafter only if the connection becomes
730 PH/06 Added the client_condition to authenticators so that some can be skipped
731 by clients under certain conditions.
733 PH/07 The error message for a badly-placed control=no_multiline_responses left
734 "_responses" off the end of the name.
736 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
738 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
739 (without spaces) instead of just copying the configuration text.
741 PH/10 Added the /noupdate option to the ratelimit ACL condition.
743 PH/11 Added $max_received_linelength.
745 PH/12 Added +ignore_defer and +include_defer to host lists.
747 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
748 way in which PCRE > 7.0 is built.
750 PH/14 Implemented queue_only_load_latch.
752 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
753 MAIL command. The effect was to mangle the value on 64-bit systems.
755 PH/16 Another patch from the Sieve maintainer.
757 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
759 PH/18 If a system quota error occurred while trying to create the file for
760 a maildir delivery, the message "Mailbox is full" was not appended to the
761 bounce if the delivery eventually timed out. Change 4.67/27 below applied
762 only to a quota excession during the actual writing of the file.
764 PH/19 It seems that peer DN values may contain newlines (and other non-printing
765 characters?) which causes problems in log lines. The DN values are now
766 passed through string_printing() before being added to log lines.
768 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
769 and InterBase are left for another time.)
771 PH/21 Added message_body_newlines option.
773 PH/22 Guard against possible overflow in moan_check_errorcopy().
775 PH/23 POSIX allows open() to be a macro; guard against that.
777 PH/24 If the recipient of an error message contained an @ in the local part
778 (suitably quoted, of course), incorrect values were put in $domain and
779 $local_part during the evaluation of errors_copy.
785 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
786 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
789 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
790 issue a MAIL command.
792 PH/02 In an ACL statement such as
794 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
796 if a client was not listed at all, or was listed with a value other than
797 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
798 the condition was not true (as it should be), so access was not denied.
799 The bug was that the ! inversion was incorrectly passed on to the second
800 item. This has been fixed.
802 PH/03 Added additional dnslists conditions == and =& which are different from
803 = and & when the dns lookup returns more than one IP address.
805 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
806 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
808 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
809 FSYNC, which compiles an option called disable_fsync that allows for
810 bypassing fsync(). The documentation is heavily laced with warnings.
812 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
814 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
815 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
816 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
817 including adding "make clean"; (3) Added -fPIC when compiling the test
818 dynamically loaded module, to get rid of a warning.
820 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
821 message fails, move_frozen_messages = true and ignore_bounce_errors_after
822 = 0s. The bug is otherwise harmless.
824 PH/07 There was a bug in the dovecot authenticator such that the value of
825 $auth1 could be overwritten, and so not correctly preserved, after a
826 successful authentication. This usually meant that the value preserved by
827 the server_setid option was incorrect.
829 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
831 PH/09 Installed PCRE release 7.0.
833 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
834 run for batched SMTP input. It is now run at the start of every message
835 in the batch. While fixing this I discovered that the process information
836 (output by running exiwhat) was not always getting set for -bs and -bS
837 input. This is fixed, and it now also says "batched" for BSMTP.
839 PH/11 Added control=no_pipelining.
841 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
842 patch, slightly modified), and move the expansion of helo_data till after
843 the connection is made in the smtp transport (so it can use these
846 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
848 PH/14 Added log_selector = +pid.
850 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
852 PH/16 Add ${if forany and ${if forall.
854 PH/17 Added dsn_from option to vary the From: line in DSNs.
856 PH/18 Flush SMTP output before performing a callout, unless control =
857 no_callout_flush is set.
859 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
860 was true (the default) a successful delivery failed to delete the retry
861 item, thus causing premature timeout of the address. The bug is now
864 PH/20 Added hosts_avoid_pipelining to the smtp transport.
866 PH/21 Long custom messages for fakedefer and fakereject are now split up
867 into multiline reponses in the same way that messages for "deny" and
868 other ACL rejections are.
870 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
871 with slight modification.
873 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
874 draft, changing the syntax and factoring some duplicate code".
876 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
877 for deliveries of the second and subsequent messages over the same SMTP
880 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
881 ${reduce, with only minor "tidies".
883 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
885 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
886 expansion side effects.
888 PH/27 When a message times out after an over-quota error from an Exim-imposed
889 quota, the bounce message says "mailbox is full". This message was not
890 being given when it was a system quota that was exceeded. It now should
893 MH/03 Made $recipients available in local_scan(). local_scan() already has
894 better access to the recipient list through recipients_list[], but
895 $recipients can be useful in postmaster-provided expansion strings.
897 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
898 in the case of a MAIL command with additional options following the
899 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
900 were accidentally chopped off.
902 PH/29 SMTP synchronization checks are implemented when a command is read -
903 there is a check that no more input is waiting when there shouldn't be
904 any. However, for some commands, a delay in an ACL can mean that it is
905 some time before the response is written. In this time, more input might
906 arrive, invalidly. So now there are extra checks after an ACL has run for
907 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
908 pipelining has not been advertised.
910 PH/30 MH's patch to allow iscntrl() characters to be list separators.
912 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
913 returned in the SMTP response when smtp_return_error_details was false.
916 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
917 instead of the C I/O that was originally supplied, because problems were
920 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
921 Exim which did not show up earlier: it was assuming that a call to
922 SSL_CTX_set_info_callback() might give an error value. In fact, there is
923 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
924 was a macro that became an assignment, so it seemed to work. This has
925 changed to a proper function call with a void return, hence the compile
926 error. Exim's code has been fixed.
928 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
931 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
933 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
935 PH/37 If a message is not accepted after it has had an id assigned (e.g.
936 because it turns out to be too big or there is a timeout) there is no
937 "Completed" line in the log. When some messages of this type were
938 selected by exigrep, they were listed as "not completed". Others were
939 picked up by some special patterns. I have improved the selection
940 criteria to be more general.
942 PH/38 The host_find_failed option in the manualroute router can now be set
943 to "ignore", to completely ignore a host whose IP address cannot be
944 found. If all hosts are ignored, the behaviour is controlled by the new
945 host_all_ignored option.
947 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
948 homing or because of multiple MX records with /mx) generated more than
949 one IP address, and the following item turned out to be the local host,
950 all the secondary addresses of the first item were incorrectly removed
951 from the list, along with the local host and any following hosts (which
952 is what is supposed to happen).
954 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
955 whoever called Exim into the -H file. In the case of the daemon it was
956 behaving confusingly. When first started, it used values for whoever
957 started the daemon, but after a SIGHUP it used the Exim user (because it
958 calls itself on a restart). I have changed the code so that it now always
961 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
962 message are rejected with the same error (e.g. no authentication or bad
963 sender address), and a DATA command is nevertheless sent (as can happen
964 with PIPELINING or a stupid MUA), the error message that was given to the
965 RCPT commands is included in the rejection of the DATA command. This is
966 intended to be helpful for MUAs that show only the final error to their
969 PH/42 Another patch from the Sieve maintainer.
971 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
972 Eximstats - Fixed some broken HTML links and added missing column headers
974 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
975 columns for Rejects, Temp Rejects, Ham, and Spam rows.
977 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
979 PH/43 Yet another patch from the Sieve maintainer.
981 PH/44 I found a way to check for a TCP/IP connection going away before sending
982 the response to the final '.' that terminates a message, but only in the
983 case where the client has not sent further data following the '.'
984 (unfortunately, this is allowed). However, in many cases there won't be
985 any further data because there won't be any more messages to send. A call
986 to select() can be used: if it shows that the input is "ready", there is
987 either input waiting, or the socket has been closed. An attempt to read
988 the next input character can distinguish the two cases. Previously, Exim
989 would have sent an OK response which the client would never have see.
990 This could lead to message repetition. This fix should cure that, at
991 least in a lot of common cases.
993 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
994 advertised in response to EHLO.
1000 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1001 fixed by 4.65/MH/01 (is this a record?) are fixed:
1003 (i) An empty string was always treated as zero by the numeric comparison
1004 operators. This behaviour has been restored.
1006 (ii) It is documented that the numeric comparison operators always treat
1007 their arguments as decimal numbers. This was broken in that numbers
1008 starting with 0 were being interpreted as octal.
1010 While fixing these problems I realized that there was another issue that
1011 hadn't been noticed. Values of message_size_limit (both the global option
1012 and the transport option) were treated as octal if they started with 0.
1013 The documentation was vague. These values are now always treated as
1014 decimal, and I will make that clear in the documentation.
1020 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1021 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1024 MH/01 Don't check that the operands of numeric comparison operators are
1025 integers when their expansion is in "skipping" mode (fixes bug
1026 introduced by 4.64-PH/07).
1028 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1029 child addresses, Exim now panics and dies. Previously, because the count
1030 is held in a short int, deliveries were likely to be lost. As such a
1031 large number of recipients for a single message is ridiculous
1032 (performance will be very, very poor), I have chosen to impose a limit
1033 rather than extend the field.
1039 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1040 leftover -K file (the existence of which was triggered by #402).
1041 While we were at it, introduced process PID as part of the -K
1042 filename. This should rule out race conditions when creating
1045 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1046 processing considerably. Previous code took too long for large mails,
1047 triggering a timeout which in turn triggers #401.
1049 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1050 in the DK code in transports.c. sendfile() is not really portable,
1051 hence the _LINUX specificness.
1053 TF/01 In the add_headers option to the mail command in an Exim filter,
1054 there was a bug that Exim would claim a syntax error in any
1055 header after the first one which had an odd number of characters
1058 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1059 callout verification, Exim cached a "reject" for the entire domain. This
1060 is correct for most verifications, but it is not correct for a recipient
1061 verification with use_sender or use_postmaster set, because in that case
1062 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1063 case of MAIL FROM:<> rejection from other early rejections (e.g.
1064 rejection of HELO). When verifying a recipient using a non-null MAIL
1065 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1066 Whatever the result of the callout, the value of the domain cache is
1067 left unchanged (for any other kind of callout, getting as far as trying
1068 RCPT means that the domain itself is ok).
1070 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1073 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1074 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1075 session, the daemon ignores ECONNECT errors and logs others; it now
1076 ignores EPIPE as well.
1078 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1079 (quoted-printable decoding).
1081 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1082 later the small subsequent patch to fix an introduced bug.
1084 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1086 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1088 PH/08 An error is now given if message_size_limit is specified negative.
1090 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1091 to be given (somewhat) arbitrary names.
1093 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1096 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1097 miscellaneous code fixes
1099 PH/10 Added the log_reject_target ACL modifier to specify where to log
1102 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1103 hostname. This is wrong, because it relates to the incoming message (and
1104 probably the interface on which it is arriving) and not to the outgoing
1105 callout (which could be using a different interface). This has been
1106 changed to use the value of the helo_data option from the smtp transport
1107 instead - this is what is used when a message is actually being sent. If
1108 there is no remote transport (possible with a router that sets up host
1109 addresses), $smtp_active_hostname is used.
1111 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1112 tweaks were necessary in order to get it to work (see also 21 below):
1113 (a) The code assumed that strncpy() returns a negative number on buffer
1114 overflow, which isn't the case. Replaced with Exim's string_format()
1116 (b) There were several signed/unsigned issues. I just did the minimum
1117 hacking in of casts. There is scope for a larger refactoring.
1118 (c) The code used strcasecmp() which is not a standard C function.
1119 Replaced with Exim's strcmpic() function.
1120 (d) The code set only $1; it now sets $auth1 as well.
1121 (e) A simple test gave the error "authentication client didn't specify
1122 service in request". It would seem that Dovecot has changed its
1123 interface. Fortunately there's a specification; I followed it and
1124 changed what the client sends and it appears to be working now.
1126 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1129 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1130 address A is aliased to B and C, where B exists and C does not. Without
1131 -v the output is "A verified" because verification stops after a
1132 successful redirection if more than one address is generated. However,
1133 with -v the child addresses are also verified. Exim was outputting "A
1134 failed to verify" and then showing the successful verification for C,
1135 with its parentage. It now outputs "B failed to verify", showing B's
1136 parentage before showing the successful verification of C.
1138 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1139 look up a TXT record in a specific list after matching in a combined
1142 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1143 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1144 they consult the DNS. I had assumed they would set it the way they
1145 wanted; and indeed my experiments on Linux seem to show that in some
1146 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1147 To be on the safe side, however, I have now made the interface to
1148 host_find_byname() similar to host_find_bydns(), with an argument
1149 containing the DNS resolver options. The host_find_byname() function now
1150 sets these options at its start, just as host_find_bydns() does. The smtp
1151 transport options dns_qualify_single and dns_search_parents are passed to
1152 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1153 of host_find_byname() use the default settings of RES_DEFNAMES
1154 (qualify_single) but not RES_DNSRCH (search_parents).
1156 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1157 spool_read_header() do less string testing, by means of a preliminary
1158 switch on the second character of optional "-foo" lines. (This is
1159 overdue, caused by the large number of possibilities that now exist.
1160 Originally there were few.) While I was there, I also converted the
1161 str(n)cmp tests so they don't re-test the leading "-" and the first
1162 character, in the hope this might squeeze out yet more improvement.
1164 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1165 flag allowing group syntax was set by the header_syntax check but not
1166 turned off, possible causing trouble later; (2) The flag was not being
1167 set at all for the header_verify test, causing "group"-style headers to
1168 be rejected. I have now set it in this case, and also caused header_
1169 verify to ignore an empty address taken from a group. While doing this, I
1170 came across some other cases where the code for allowing group syntax
1171 while scanning a header line wasn't quite right (mostly, not resetting
1172 the flag correctly in the right place). These bugs could have caused
1173 trouble for malformed header lines. I hope it is now all correct.
1175 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1176 with the "reply" argument non-NULL. The code, however (which originally
1177 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1178 but it didn't always do it. This confused somebody who was copying the
1179 code for some other use. I have removed all the tests.
1181 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1182 feature that was used to support insecure browsers during the U.S. crypto
1183 embargo. It requires special client support, and Exim is probably the
1184 only MTA that supported it -- and would never use it because real RSA is
1185 always available. This code has been removed, because it had the bad
1186 effect of slowing Exim down by computing (never used) parameters for the
1187 RSA_EXPORT functionality.
1189 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1190 authenticator to fail if there's a tab character in the incoming data
1191 (there should never be unless someone is messing about, as it's supposed
1192 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1193 if the connection is using TLS or if the remote IP is the same as the
1194 local IP, and the "valid-client-cert option" if a client certificate has
1197 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1198 authenticators. This can be used for authorization after authentication
1199 succeeds. (In the case of plaintext, it servers for both authentication
1202 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1203 if any retry times were supplied.
1205 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1206 connection, where there is no client IP address to check. In this
1207 situation, the verify now always succeeds.
1209 PH/25 Applied John Jetmore's -Mset patch.
1211 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1213 PH/27 In a string expansion for a processed (not raw) header when multiple
1214 headers of the same name were present, leading whitespace was being
1215 removed from all of them, but trailing whitespace was being removed only
1216 from the last one. Now trailing whitespace is removed from each header
1217 before concatenation. Completely empty headers in a concatenation (as
1218 before) are ignored.
1220 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1221 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1223 PH/29 [Removed. This was a change that I later backed out, and forgot to
1224 correct the ChangeLog entry (that I had efficiently created) before
1225 committing the later change.]
1227 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1228 address errors (4xx response to RCPT) over the same connection as other
1229 messages routed to the same hosts. Such deliveries are always "forced",
1230 so retry times are not inspected. This resulted in far too many retries
1231 for the affected addresses. The effect occurred only when there were more
1232 hosts than the hosts_max_try setting in the smtp transport when it had
1233 the 4xx errors. Those hosts that it had tried were not added to the list
1234 of hosts for which the message was waiting, so if all were tried, there
1235 was no problem. Two fixes have been applied:
1237 (i) If there are any address or message errors in an SMTP delivery, none
1238 of the hosts (tried or untried) are now added to the list of hosts
1239 for which the message is waiting, so the message should not be a
1240 candidate for sending over the same connection that was used for a
1241 successful delivery of some other message. This seems entirely
1242 reasonable: after all the message is NOT "waiting for some host".
1243 This is so "obvious" that I'm not sure why it wasn't done
1244 previously. Hope I haven't missed anything, but it can't do any
1245 harm, as the worst effect is to miss an optimization.
1247 (ii) If, despite (i), such a delivery is accidentally attempted, the
1248 routing retry time is respected, so at least it doesn't keep
1249 hammering the server.
1251 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1252 in ${readsocket because some servers need this prod.
1254 PH/32 Added some extra debug output when updating a wait-xxx database.
1256 PH/33 The hint "could be header name not terminated by colon", which has been
1257 given for certain expansion errors for a long time, was not being given
1258 for the ${if def:h_colon_omitted{... case.
1260 PH/34 The spec says: "With one important exception, whenever a domain list is
1261 being scanned, $domain contains the subject domain." There was at least
1262 one case where this was not true.
1264 PH/35 The error "getsockname() failed: connection reset by peer" was being
1265 written to the panic log as well as the main log, but it isn't really
1266 panic-worthy as it just means the connection died rather early on. I have
1267 removed the panic log writing for the ECONNRESET error when getsockname()
1270 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1271 runs only) independently of the message's sender address. This meant
1272 that, if the 4xx error was in fact related to the sender, a different
1273 message to the same recipient with a different sender could confuse
1274 things. In particualar, this can happen when sending to a greylisting
1275 server, but other circumstances could also provoke similar problems.
1276 I have changed the default so that the retry time for these errors is now
1277 based a combination of the sender and recipient addresses. This change
1278 can be overridden by setting address_retry_include_sender=false in the
1281 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1282 remote server are returned as part of bounce messages. This was not
1283 happening for LMTP over a pipe (the lmtp transport), but now it is the
1284 same for both kinds of LMTP.
1286 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1287 in header lines that were in fact CNAMEs. This is no longer the case.
1289 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1290 and queue runs started by the daemon processed all messages. This has
1291 been fixed so that -R and -S can now usefully be given with -q<time>.
1293 PH/40 Import PCRE release 6.7 (fixes some bugs).
1295 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1297 PH/42 Give an error if -q is specified more than once.
1299 PH/43 Renamed the variables $interface_address and $interface_port as
1300 $received_ip_address and $received_port, to make it clear that these
1301 values apply to message reception, and not to the outgoing interface when
1302 a message is delivered. (The old names remain recognized, of course.)
1304 PH/44 There was no timeout on the connect() call when using a Unix domain
1305 socket in the ${readsocket expansion. There now is.
1307 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1308 be meaningful with "accept".
1310 SC/01 Eximstats V1.43
1311 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1313 SC/02 Eximstats V1.44
1314 Use a glob alias rather than an array ref in the generated
1315 parser. This improves both readability and performance.
1317 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1318 Collect SpamAssassin and rejection statistics.
1319 Don't display local sender or destination tables unless
1320 there is data to show.
1321 Added average volumes into the top table text output.
1323 SC/04 Eximstats V1.46
1324 Collect data on the number of addresses (recipients)
1325 as well as the number of messages.
1327 SC/05 Eximstats V1.47
1328 Added 'Message too big' to the list of mail rejection
1329 reasons (thanks to Marco Gaiarin).
1331 SC/06 Eximstats V1.48
1332 Mainlog lines which have GMT offsets and are too short to
1333 have a flag are now skipped.
1335 SC/07 Eximstats V1.49 (Alain Williams)
1336 Added the -emptyok flag.
1338 SC/08 Eximstats V1.50
1339 Fixes for obtaining the IP address from reject messages.
1341 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1342 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1343 whitesspace changes from 4.64-PH/27
1345 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1348 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1349 are found, allow negative numbers in numeric criteria)
1351 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1353 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1354 to match changes made in 4.64-PH/43
1356 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1358 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1359 "domain" is actually an IP address, and force a failure. This locks out
1360 those revolvers/nameservers that support "A-for-A" lookups, in
1361 contravention of the specifications.
1363 PH/48 When a host name was looked up from an IP address, and the subsequent
1364 forward lookup of the name timed out, the host name was left in
1365 $sender_host_name, contrary to the specification.
1367 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1368 restricted to single-key lookups, Exim was not diagnosing an error if
1369 * or *@ was used with a query-style lookup.
1371 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1373 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1374 long ago, but noone interested enough thought of it. Let's just say that
1375 the "1.1" means that there are some new functions that weren't there at
1376 some point in the past.
1378 PH/51 Error processing for expansion failure of helo_data from an smtp
1379 transport during callout processing was broken.
1381 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1382 tested/used via the -bh/-bhc/-bs options.
1384 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1385 bug, fixed in subsequent PCRE releases).
1387 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1388 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1390 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1396 SC/01 Use a glob alias rather than an array ref in eximstats generated
1397 parser. This improves both readability and performance.
1399 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1400 Don't display local sender or destination tables in eximstats unless
1401 there is data to show.
1402 Added average volumes into the eximstats top table text output.
1404 SC/03 Collect data on the number of addresses (recipients) as well
1405 as the number of messages in eximstats.
1407 TF/01 Correct an error in the documentation for the redirect router. Exim
1408 does (usually) call initgroups() when daemonizing.
1410 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1411 with consistent privilege compared to when running as a daemon.
1413 TF/03 Note in the spec that $authenticated_id is not set for local
1414 submissions from trusted users.
1416 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1417 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1419 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1420 by adding some example configuration directives to the default
1421 configuration file. A little bit of work is required to uncomment the
1422 directives and define how usernames and passwords are checked, but
1423 there is now a framework to start from.
1425 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1426 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1427 without this. I don't know how relevant this is to other LDAP libraries.
1429 PH/02 Add the verb name to the "unknown ACL verb" error.
1431 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1433 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1435 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1436 directory not expanded when it should be if an expanded home directory
1437 was set for the address (which is overridden by the transport).
1439 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1442 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1443 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1444 because it is too late at that time, and has no effect.
1446 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1447 security issue with \' (bugzilla #107). I could not use the
1448 PQescapeStringConn() function, because it needs a PGconn value as one of
1451 PH/08 When testing addresses using -bt, indicate those final addresses that
1452 are duplicates that would not cause an additional delivery. At least one
1453 person was confused, thinking that -bt output corresponded to deliveries.
1454 (Suppressing duplicates isn't a good idea as you lose the information
1455 about possibly different redirections that led to the duplicates.)
1457 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1458 systems where poll() doesn't work, in particular OS X.
1460 PH/10 Added more information to debugging output for retry time not reached.
1462 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1463 operations in malware.c.
1465 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1468 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1469 syslog_duplication set false, log lines that would normally be written
1470 both the the main log and to the reject log were not written to syslog at
1473 PH/14 In the default configuration, change the use of "message" in ACL warn
1474 statements to "add_header".
1476 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1477 not followed by a command (e.g. "seen endif").
1479 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1480 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1483 PH/17 Added extra conditions to the default value of delay_warning_condition
1487 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1488 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1489 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1492 The Auto-Submitted: and various List- headers are standardised, whereas I
1493 don't think Precedence: ever was.
1495 PH/18 Refactored debugging code in route_finduser() to show more information,
1496 in particular, the error code if getpwnam() issues one.
1498 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1499 This is apparently needed in addition to the PH/07 change above to avoid
1500 any possible encoding problems.
1502 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1503 but not after initializing Perl.
1505 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1506 output them only if debugging. By default they are written stderr,
1507 apparently, which is not desirable.
1509 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1512 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1515 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1517 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1518 authenticated or an ident call has been made. Suppress the default
1519 values for $authenticated_id and $authenticated_sender (but permit -oMai
1520 and -oMas) when testing with -bh.
1522 PH/24 Re-jigged the order of the tests in the default configuration so that the
1523 tests for valid domains and recipients precede the DNS black list and CSA
1524 tests, on the grounds that those ones are more expensive.
1526 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1527 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1528 This bug exists in every version of Exim that I still have, right back to
1531 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1532 However, an attempt to turn on case-sensitivity in a regex key by
1533 including (?-i) didn't work because the subject string was already
1534 lowercased, and the effects were non-intuitive. It turns out that a
1535 one-line patch can be used to allow (?-i) to work as expected.
1541 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1542 other effects) broke the use of negated acl sub-conditions.
1544 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1547 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1548 "Deny" causes Exim to reject the incoming connection with a 554 error.
1549 Unfortunately, if there is a major crisis, such as a disk failure,
1550 tcp-wrappers gives "deny", whereas what one would like would be some
1551 kind of temporary error. A kludge has been added to help with this.
1552 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1553 554 error is used if errno is still zero or contains ENOENT (which occurs
1554 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1557 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1559 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1560 errors. Otherwise a message that provokes a temporary error (when other
1561 messages do not) can cause a whole host to time out.
1563 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1564 addresses were routed directly to files or pipes from a redirect router.
1565 File deliveries just didn't batch; pipe deliveries might have suffered
1568 PH/06 A failure to get a lock for a hints database would erroneously always say
1569 "Failed to get write lock", even when it was really a read lock.
1571 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1572 of 0600. This has been changed to use the value of the lockfile_mode
1573 option (which defaults to 0600).
1575 PH/08 Applied small patch from the Sieve maintainer.
1577 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1578 folder from quota calculations, a direct delivery into this folder messed
1579 up the contents of the maildirsize file. This was because the regex was
1580 used only to exclude .Trash (or whatever) when the size of the mailbox
1581 was calculated. There was no check that a delivery was happening into an
1582 excluded directory. This bug has been fixed by ignoring all quota
1583 processing for deliveries into excluded directories.
1585 PH/10 Added the maildirfolder_create_regex option to appendfile.
1591 PH/01 The code for finding all the local interface addresses on a FreeBSD
1592 system running IPv6 was broken. This may well have applied to all BSD
1593 systems, as well as to others that have similar system calls. The broken
1594 code found IPv4 interfaces correctly, but gave incorrect values for the
1595 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1596 that it would not match correctly against @[] and not recognize the IPv6
1599 PH/02 The ipliteral router was not recognizing addresses of the form user@
1600 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1602 PH/03 Added disable_ipv6.
1604 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1605 decoded form, because it is most often used to construct To: headers
1606 lines in autoreplies, and the decoded form may well be syntactically
1607 invalid. However, $reply_address has leading white space removed, and all
1608 newlines turned into spaces so that the autoreply transport does not
1611 PH/05 If group was specified without a user on a router, and no group or user
1612 was specified on a transport, the group from the router was ignored.
1614 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1615 for visible compile-time settings that can be used to change these
1616 numbers, for those that want even more. Backwards compatibility with old
1617 spool files has been maintained. However, going back to a previous Exim
1618 release will lost any variables that are in spool files.
1620 PH/07 Two small changes when running in the test harness: increase delay when
1621 passing a TCP/IP connection to a new process, in case the original
1622 process has to generate a bounce, and remove special handling of
1623 127.0.0.2 (sic), which is no longer necessary.
1625 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1626 be the same on different OS.
1628 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1631 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1632 whether --show-vars was specified or not
1634 JJ/02 exipick: Added support for new ACL variable spool format introduced
1637 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1638 syntactically invalid From: or Reply-to: line, and a filter used this to
1639 generate an autoreply, and therefore failed to obtain an address for the
1640 autoreply, Exim could try to deliver to a non-existent relative file
1641 name, causing unrelated and misleading errors. What now happens is that
1642 it logs this as a hard delivery error, but does not attempt to create a
1645 PH/11 The exinext utility has a -C option for testing purposes, but although
1646 the given file was scanned by exinext itself; it wasn't being passed on
1647 when Exim was called.
1649 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1650 an end-of-file indication when reading a command response.
1652 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1653 compiled. In many other places in Exim, IPv6 addresses are always
1654 recognized, so I have changed this. It also means that IPv4 domain
1655 literals of the form [IPV4:n.n.n.n] are now always recognized.
1657 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1658 used if the router is not running as root, for example, when verifying at
1659 ACL time, or when using -bh. The debugging output from this situation was
1660 non-existent - all you got was a failure to exec. I have made two
1663 (a) Failures to set uid/gid, the current directory, or a process leader
1664 in a subprocess such as that created by queryprogram now generate
1665 suitable debugging ouput when -d is set.
1667 (b) The queryprogram router detects when it is not running as root,
1668 outputs suitable debugging information if -d is set, and then runs
1669 the subprocess without attempting to change uid/gid.
1671 PH/15 Minor change to Makefile for building test_host (undocumented testing
1674 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1675 additional section of a DNS packet that returns MX or SRV records.
1676 Instead, it always explicitly searches for A/AAAA records. This avoids
1677 major problems that occur when a DNS server includes only records of one
1678 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1679 fixed another bug: if SRV records were looked up and the corresponding
1680 address records were *not* found in the additional section, the port
1681 values from the SRV records were lost.
1683 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1684 using the correct key (the original address) when searching the retry
1685 rules in order to find which one to use for generating the retry hint.
1687 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1688 adding the default one. Similarly, if it contains a Reply-To: header, the
1689 errors_reply_to option, if set, is not used.
1691 PH/19 When calculating a retry time, Exim used to measure the "time since
1692 failure" by looking at the "first failed" field in the retry record. Now
1693 it does not use this if it is later than than the arrival time of the
1694 message. Instead it uses the arrival time. This makes for better
1695 behaviour in cases where some deliveries succeed, thus re-setting the
1696 "first failed" field. An example is a quota failure for a huge message
1697 when small messages continue to be delivered. Without this change, the
1698 "time since failure" will always be short, possible causing more frequent
1699 delivery attempts for the huge message than are intended.
1700 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1702 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1703 $1, $2, $3) because the numerical variables can be reset during some
1704 expansion items (e.g. "match"), thereby losing the authentication data.
1706 PH/21 Make -bV show the size of off_t variables so that the test suite can
1707 decide whether to run tests for quotas > 2G.
1709 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1710 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1711 filecount value is greater than 2G or if a quota value is greater than 2G
1712 on a system where the size of off_t is not greater than 4, a panic error
1715 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1716 never match. The debug and -bh output now contains an explicit error
1717 message indicating a malformed IPv4 address or mask.
1719 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1720 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1721 PH/23 above applies.
1723 PH/25 Do not write to syslog when running in the test harness. The only
1724 occasion when this arises is a failure to open the main or panic logs
1725 (for which there is an explicit test).
1727 PH/26 Added the /no_tell option to "control=freeze".
1729 PH/27 If a host name lookup failed very early in a connection, for example, if
1730 the IP address matched host_lookup and the reverse lookup yielded a name
1731 that did not have a forward lookup, an error message of the form "no IP
1732 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1733 could be logged. Now it outputs the IP address instead of "NULL".
1735 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1736 allows the sender and the authenticated sender to be set when
1737 submitting a message from within Exim. Since child_open_exim() is
1738 documented for local_scan(), the new function should be too.
1740 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1741 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1742 results in an empty string is now treated as unset.
1744 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1746 PH/31 Added sender_verify_fail logging option.
1748 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1749 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1750 tidied the source and removed it altogether.
1752 PH/33 When a queue run was abandoned because the load average was too high, a
1753 log line was always written; now it is written only if the queue_run log
1754 selector is set. In addition, the log line for abandonment now contains
1755 information about the queue run such as the pid. This is always present
1756 in "start" and "stop" lines but was omitted from the "abandon" line.
1758 PH/34 Omit spaces between a header name and the colon in the error message that
1759 is given when verify = headers_syntax fails (if there are lots of them,
1760 the message gets confusing).
1762 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1763 names, as there are now some PTR records that contain slashes. This check
1764 is only to protect against broken name servers that fall over on strange
1765 characters, so the fact that it applies to all lookups doesn't matter.
1767 PH/36 Now that the new test suite is complete, we can remove some of the
1768 special code in Exim that was needed for the old test suite. For example,
1769 sorting DNS records because real resolvers return them in an arbitrary
1770 order. The new test suite's fake resolver always returns records in the
1773 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1774 bounces) except when queue_only is set, to avoid logging races between
1775 the different processes.
1777 PH/38 Panic-die if .include specifies a non-absolute path.
1779 PH/39 A tweak to the "H" retry rule from its user.
1781 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1782 a label. They prevented compilation on older perls.
1784 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1785 a warning to be raised on newish perls.
1787 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1788 on queue. Changes to match documented behaviour of showing count of
1789 messages matching specified criteria.
1791 PH/40 Changed the default ident timeout from 30s to 5s.
1793 PH/41 Added support for the use of login_cap features, on those BSD systems
1794 that have them, for controlling the resources used by pipe deliveries.
1796 PH/42 The content-scanning code uses fopen() to create files in which to put
1797 message data. Previously it was not paying any attention to the mode of
1798 the files. Exim runs with umask(0) because the rest of the code creates
1799 files with open(), and sets the required mode explicitly. Thus, these
1800 files were ending up world-writeable. This was not a big issue, because,
1801 being within the spool directory, they were not world-accessible. I have
1802 created a function called modefopen, which takes an additional mode
1803 argument. It sets umask(777), creates the file, chmods it to the required
1804 mode, then resets the umask. All the relevant calls to fopen() in the
1805 content scanning code have been changed to use this function.
1807 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1808 to 24 hours. This avoids potential overflow problems when processing G
1809 and H retry rules. I suspect nobody ever tinkers with this value.
1811 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1813 PH/45 When the plaintext authenticator is running as a client, the server's
1814 challenges are checked to ensure they are valid base64 strings. By
1815 default, the authentication attempt is cancelled if an invalid string is
1816 received. Setting client_ignore_invalid_base64 true ignores these errors.
1817 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1818 they are received. Thus, the responses can be made to depend on the
1819 challenges. If an invalid string is ignored, an empty string is placed in
1822 PH/46 Messages that are created by the autoreply transport now contains a
1823 References: header, in accordance with RFCs 2822 and 3834.
1825 PH/47 Added authenticated_sender_force to the smtp transport.
1827 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1829 PH/49 Installed latest patch from the Sieve maintainer.
1831 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1832 was also set, the appendfile transport was unnecessarily scanning a
1833 directory of message files (e.g. for maildir delivery) to find the count
1834 of files (along with the size), even though it did not need this
1835 information. It now does the scan only if it needs to find either the
1836 size of the count of files.
1838 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1840 PH/52 Two bugs concerned with error handling when the smtp transport is
1843 (i) Exim was not creating retry information for temporary errors given
1844 for individual recipients after the DATA command when the smtp transport
1845 was used in LMTP mode. This meant that they could be retried too
1846 frequently, and not timed out correctly.
1848 (ii) Exim was setting the flag that allows error details to be returned
1849 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1850 recipients that were returned after the DATA command.
1852 PH/53 This is related to PH/52, but is more general: for any failing address,
1853 when detailed error information was permitted to be returned to the
1854 sender, but the error was temporary, then after the final timeout, only
1855 "retry timeout exceeded" was returned. Now it returns the full error as
1856 well as "retry timeout exceeded".
1858 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1859 do this, and (what is worse) MTAs that accept it.
1861 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1862 will now be deprecated.
1864 PH/56 New os.c-cygwin from the Cygwin maintainer.
1866 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1867 formats (previously only available in exim formats via -bpr, -bpru,
1868 and -bpra. Now also available in native and exiqgrep formats)
1870 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1871 with very large, slow to parse queues
1873 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1875 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1877 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1878 responses to authentication challenges, though it was showing the
1879 challenges; (ii) I've removed the CR characters from the debug output for
1882 PH/58 Allow for the insertion of a newline as well as a space when a string
1883 is turned into more than one encoded-word during RFC 2047 encoding. The
1884 Sieve code now uses this.
1886 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1887 data_4xx, lost_connection, tls_required.
1889 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1890 message was being sent as an SMTP response.
1892 PH/61 Add -l and -k options to exicyclog.
1894 PH/62 When verifying, if an address was redirected to one new address, so that
1895 verification continued, and the new address failed or deferred after
1896 having set something in $address_data, the value of $address_data was not
1897 passed back to the ACL. This was different to the case when no
1898 redirection occurred. The value is now passed back in both cases.
1900 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1901 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1902 use login_cap.h, so on its own it isn't the distinguishing feature. The
1903 new name refers directly to the setclassresources() function.
1905 PH/65 Added configuration files for NetBSD3.
1907 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1909 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1910 is preferred over IPv4.
1912 PH/68 The bounce_return_message and bounce_return_body options were not being
1913 honoured for bounces generated during the reception of non-SMTP messages.
1914 In particular, this applied to messages rejected by the ACL. This bug has
1915 been fixed. However, if bounce_return_message is true and bounce_return_
1916 body is false, the headers that are returned for a non-SMTP message
1917 include only those that have been read before the error was detected.
1918 (In the case of an ACL rejection, they have all been read.)
1920 PH/69 The HTML version of the specification is now built in a directory called
1921 spec_html instead of spec.html, because the latter looks like a path with
1922 a MIME-type, and this confuses some software.
1924 PH/70 Catch two compiler warnings in sieve.c.
1926 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1927 function verify_get_ident() calls ip_connect() to connect a socket, but
1928 if the "connect()" function timed out, ip_connect() used to close the
1929 socket. However, verify_get_ident() also closes the socket later, and in
1930 between Exim writes to the log, which may get opened at this point. When
1931 the socket was closed in ip_connect(), the log could get the same file
1932 descriptor number as the socket. This naturally causes chaos. The fix is
1933 not to close the socket in ip_connect(); the socket should be closed by
1934 the function that creates it. There was only one place in the code where
1935 this was missing, in the iplookup router, which I don't think anybody now
1936 uses, but I've fixed it anyway.
1938 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1939 well as to direct DNS lookups. Otherwise the handling of names in host
1940 lists is inconsistent and therefore confusing.
1946 PH/01 Two changes to the default runtime configuration:
1948 (1) Move the checks for relay_from_hosts and authenticated clients from
1949 after to before the (commented out) DNS black list checks.
1951 (2) Add control=submission to the relay_from_hosts and authenticated
1952 clients checks, on the grounds that messages accepted by these
1953 statements are most likely to be submissions.
1955 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1957 (1) Generate an error if the third argument for the ${prvs expansion is
1960 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1963 (3) Reset the variables that are obtained from the first argument of
1964 ${prvscheck and used in the second argument before leaving the code,
1965 because their memory is reclaimed, so using them afterwards may do
1968 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1969 one (it's much easier than Tom thought :-).
1971 (5) Because of (4), we can now allow for the use of $prvscheck_result
1972 inside the third argument.
1974 PH/03 For some reason, the default setting of PATH when running a command from
1975 a pipe transport was just "/usr/bin". I have changed it to
1978 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1979 anything to be listed in the output from -bV.
1981 PH/05 When a filter generated an autoreply, the entire To: header line was
1982 quoted in the delivery log line, like this:
1984 => >A.N.Other <ano@some.domain> <original@ddress> ...
1986 This has been changed so that it extracts the operative address. There
1987 may be more than one such address. If so, they are comma-separated, like
1990 => >ano@some.domain,ona@other.domain <original@ddress> ...
1992 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1993 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1994 not being looked up in the rDNS to get a host name, Exim was showing the
1995 IP address twice in Received: lines, even though the IP addresses were
1996 identical. For example:
1998 Received: from [1.2.3.4] (helo=[1.2.3.4])
2000 However, if the real host name was known, it was omitting the HELO data
2001 if it matched the actual IP address. This has been tidied up so that it
2002 doesn't show the same IP address twice.
2004 PH/07 When both +timestamp and +memory debugging was on, the value given by
2005 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2006 called by the debug printing, thereby overwriting the timestamp buffer.
2007 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2009 PH/08 When the original message was included in an autoreply transport, it
2010 always said "this is a copy of the message, including all the headers",
2011 even if body_only or headers_only was set. It now gives an appropriate
2014 PH/09 Applied a patch from the Sieve maintainer which:
2016 o fixes some comments
2017 o adds the (disabled) notify extension core
2018 o adds some debug output for the result of if/elsif tests
2019 o points to the current vacation draft in the documentation
2020 and documents the missing references header update
2024 o fixes a bug in processing the envelope test (when testing
2025 multiple envelope elements, the last element determinted the
2028 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2029 Electronic Mail") by including:
2031 Auto-submitted: auto-generated
2033 in the messages that it generates (bounce messages and others, such as
2034 warnings). In the case of bounce messages for non-SMTP mesages, there was
2035 also a typo: it was using "Auto_submitted" (underscore instead of
2036 hyphen). Since every message generated by Exim is necessarily in response
2037 to another message, thes have all been changed to:
2039 Auto-Submitted: auto-replied
2041 in accordance with these statements in the RFC:
2043 The auto-replied keyword:
2045 - SHOULD be used on messages sent in direct response to another
2046 message by an automatic process,
2048 - MUST NOT be used on manually-generated messages,
2050 - MAY be used on Delivery Status Notifications (DSNs) and Message
2051 Disposition Notifications (MDNs),
2053 - MUST NOT be used on messages generated by automatic or periodic
2054 processes, except for messages which are automatic responses to
2057 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2058 to the default Received: header definition.
2060 PH/12 Added log selector acl_warn_skipped (default on).
2062 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2063 variables because (a) they are in the wrong storage pool and (b) even if
2064 they were copied, it wouldn't work properly because of the caching.
2066 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2067 checking when decoding. Apparently there are clients that generate
2068 overlong encoded strings. Why am I not surprised?
2070 PH/15 If the first argument of "${if match_address" was not empty, but did not
2071 contain an "@" character, Exim crashed. Now it writes a panic log message
2072 and treats the condition as false.
2074 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2076 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2077 extension "envelope-auth". The code is finished and in agreement with
2078 other implementations, but there is no documentation so far and in fact,
2079 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2080 not changing the active code.
2082 Print executed "if" and "elsif" statements when debugging is used. This
2083 helps a great deal to understand what a filter does.
2085 Document more things not specified clearly in RFC3028. I had all this
2086 sorted out, when out of a sudden new issues came to my mind. Oops."
2088 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2091 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2092 canonical form (as documented). However, after a host name lookup from
2093 the IP address, check_host() was doing a simple string comparison with
2094 addresses acquired from the DNS when checking that the found name did
2095 have the original IP as one of its addresses. Since any found IPv6
2096 addresses are likely to be in abbreviated form, the comparison could
2097 fail. Luckily, there already exists a function for doing the comparison
2098 by converting both addresses to binary, so now that is used instead of
2099 the text comparison.
2101 PH/20 There was another similar case to PH/19, when a complete host name was
2102 given in a host list; looking up its IP address could give an abbreviated
2103 form, whereas the current host's name might or might not be abbreviated.
2104 The same fix has been applied.
2110 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2111 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2114 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2115 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2117 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2119 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2120 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2121 Also, for "Auto-Submitted", treat anything other than "no" as
2122 non-personal, in accordance with RFC 3834. (Previously it treated
2123 anything starting "auto-" as non-personal.)
2125 TF/01 The control=submission/name=... option had a problem with syntax
2126 errors if the name included a slash character. The /name= option
2127 now slurps the rest of the string, so it can include any characters
2128 but it must come last in the list of options (after /sender_retain
2131 PH/05 Some modifications to the interface to the fake nameserver for the new
2139 TK/01 Added the "success_on_redirect" address verification option. See
2140 NewStuff for rationale and an example.
2142 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2144 PH/02 Patch to exigrep to allow it to work on syslog lines.
2146 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2147 fread() to skip over the body file's header line, because in Cygwin the
2148 header line is locked and is inaccessible.
2150 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2151 co-exist for some time) to make it clear that it is the Exim ID that is
2152 referenced, not the Message-ID: header line.
2154 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2155 string_format() function, because snprintf() does not exist on all
2158 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2159 ${stat: expansion item.
2161 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2162 protocol synchronization error", to keep the pedants happy.
2164 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2165 well as for IRIX systems, when gcc is being used. See the host.c source
2168 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2170 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2173 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2174 transport and to the smtp transport in LMTP mode.
2176 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2178 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2179 run a filter in a subprocess. This could lead to confusion in subsequent
2180 lookups in the parent process. There should also be a search_tidyup() at
2181 the end of the subprocess.
2183 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2184 only if the host matched helo_try_verify_hosts, which caused the
2185 verification to occur when the EHLO/HELO command was issued. The ACL just
2186 tested the remembered result. Now, if a previous verification attempt has
2187 not happened, "verify = helo" does it there and then.
2189 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2191 TK/03 Fix log output including CR from clamd.
2193 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2194 exist provoked a memory error which could cause a segfault.
2196 PH/15 Installed PCRE 6.2
2198 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2200 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2201 of the problem. Specifically, suggested +O2 rather than +O1 for the
2204 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2206 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2207 record, the port was not correctly logged when the outgoing_port log
2208 selector was set (it logged the transort's default port).
2210 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2211 fallback_hosts, and "hosts" in the smtp transport.
2213 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2214 host errors such as "Connection refused".
2216 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2217 authentication with radiusclient 0.4.9:
2219 - Error returned from rc_read_config was caught wrongly
2220 - Username/password not passed on to radius server due to wrong length.
2222 The presumption is that some radiusclient API changes for 4.51/PH/17
2223 were not taken care of correctly. The code is still untested by me (my
2224 Linux distribution still has 0.3.2 of radiusclient), but it was
2225 contributed by a Radius user.
2227 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2228 expanding the "port" option of the smtp transport.
2230 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2231 while reading a MIME header. Thanks to Tom Hughes for a patch.
2233 PH/24 Include config.h inside local_scan.h so that configuration settings are
2236 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2237 that in an ACL for RCPT (for example), you can examine exactly what was
2240 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2241 commands, but it was not correctly comparing the address with the actual
2242 client host address. Thus, it would show the EHLO address in Received:
2243 header lines when this was not necessary.
2245 PH/27 Added the % operator to ${eval:}.
2247 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2248 it should be ignoring failures (because with -C, for example, it has lost
2249 privilege). It wasn't ignoring creation failures other than "already
2252 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2255 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2256 because an input line was too long, either on its own, or by virtue of
2257 too many continuations, the temporary file was not being removed, and the
2258 return code was incorrect.
2260 PH/31 Missing "BOOL" in function definition in filtertest.c.
2262 PH/32 Applied Sieve patches from the maintainer.
2264 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2266 PH/33 Added "verify = not_blind".
2268 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2269 Local/Makefile (with some defaults set). These are used in built scripts
2270 such as exicyclog, but they have never been used in the exim_install
2271 script (though there are many overriding facilities there). I have
2272 arranged that the exim_install script now takes note of these two
2275 PH/35 Installed configuration files for Dragonfly.
2277 PH/36 When a locally submitted message by a trusted user did not contain a
2278 From: header, and the sender address was obtained from -f or from an SMTP
2279 MAIL command, and the trusted user did not use -F to supply a sender
2280 name, $originator_name was incorrectly used when constructing a From:
2281 header. Furthermore, $originator_name was used for submission mode
2282 messages from external hosts without From: headers in a similar way,
2283 which is clearly wrong.
2285 PH/37 Added control=suppress_local_fixups.
2287 PH/38 When log_selector = +received_sender was set, and the addition of the
2288 sender made the log line's construction buffer exactly full, or one byte
2289 less than full, an overflow happened when the terminating "\n" was
2292 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2293 when the result of a list match is failure because a DNS lookup failed.
2295 PH/40 RM_COMMAND is now used in the building process.
2297 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2298 the "build-* directories that it finds.
2300 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2301 address in a domain literal was a prefix of an interface address.
2303 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2304 when verifying a sender address, unless rewrite_headers is false.
2306 PH/44 Wrote a long comment about why errors_to addresses are verified as
2307 recipients, not senders.
2309 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2310 the ratelimit ACL was added.
2312 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2314 PH/46 Added extra information about PostgreSQL errors to the error string.
2316 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2317 avoiding the need to install special zones in a real server. This is
2318 backwards compatible; if it can't find the fake resolver, it drops back.
2319 Thus, both old and new test suites can be run.
2321 TF/02 Added util/ratelimit.pl
2323 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2324 clock is set back in time.
2326 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2327 Candler <B.Candler@pobox.com>.
2329 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2330 OK for addresses that are the result of redirections.
2332 PH/48 A number of further additions for the benefit of the new test suite,
2333 including a fake gethostbyname() that interfaces to the fake DNS resolver
2336 TF/06 The fix for widen_domains has also been applied to qualify_single and
2337 search_parents which are the other dnslookup options that can cause
2340 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2343 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2345 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2346 service" if it encountered an MX record with an empty target hostname.
2347 The message is now "an MX or SRV record indicated no SMTP service".
2349 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2350 if the DNS of the sending site is misconfigured. This is quite a
2351 common situation. This change restores the behaviour of treating a
2352 helo verification defer as a failure.
2354 PH/51 If self=fail was set on a router, the bounce message did not include the
2355 actual error message.
2361 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2363 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2364 command itself ended in error, the underlying message about the transport
2365 filter timeout was being overwritten with the pipe command error. Now the
2366 underlying error message should be appended to the second error message.
2368 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2370 PH/02 Reduce the amount of output that "make" produces by default. Full output
2371 can still be requested.
2373 PH/03 The warning log line about a condition test deferring for a "warn" verb
2374 was being output only once per connection, rather than after each
2375 occurrence (because it was using the same function as for successful
2376 "warn" verbs). This seems wrong, so I have changed it.
2378 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2379 it should not have, which might have caused a crash in the right
2380 circumstances, but probably never did.
2382 PH/04 Installed a modified version of Tony Finch's patch to make submission
2383 mode fix the return path as well as the Sender: header line, and to
2384 add a /name= option so that you can make the user's friendly name appear
2387 TF/03 Added the control = fakedefer ACL modifier.
2389 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2390 Mark Lowes for thorough testing.
2392 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2394 TK/03 Merged latest SRS patch from Miles Wilton.
2396 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2397 with the definition in sysexits.h (which is #included earlier).
2398 Fortunately, Exim does not actually use EX_OK. The code used to try to
2399 preserve the sysexits.h value, by assumimg that macro definitions were
2400 scanned for macro replacements. I have been disabused of this notion,
2401 so now the code just undefines EX_OK before #including unistd.h.
2403 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2404 in the smtp transport. When a block could not be written in a single
2405 write() function, the timeout was being re-applied to each part-write.
2406 This seems wrong - if the receiver was accepting one byte at a time it
2407 would take for ever. The timeout is now adjusted when this happens. It
2408 doesn't have to be particularly precise.
2410 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2411 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2413 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2414 without a domain if the check to <postmaster@domain> fails.
2416 SC/01 Eximstats: added -xls and the ability to specify output files
2417 (patch written by Frank Heydlauf).
2419 SC/02 Eximstats: use FileHandles for outputing results.
2421 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2423 SC/04 Eximstats: fixed display of large numbers with -nvr option
2425 SC/05 Eximstats: fixed merging of reports with empty tables.
2427 SC/06 Eximstats: added the -include_original_destination flag
2429 SC/07 Eximstats: removed tabs and trailing whitespace.
2431 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2433 TK/06 MBOX spool code: Add real "From " MBOX separator line
2434 so the .eml file is really in mbox format (even though
2435 most programs do not really care). Patch from Alex Miller.
2437 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2438 The latter is generated from $received_to and is only set if the
2439 message has one envelope recipient. SA can use these headers,
2440 obviously out-of-the-box. Patch from Alex Miller.
2442 PH/08 The ${def test on a variable was returning false if the variable's
2443 value was "0", contrary to what the specification has always said!
2444 The result should be true unless the variable is empty.
2446 PH/09 The syntax error of a character other than { following "${if
2447 def:variable_name" (after optional whitespace) was not being diagnosed.
2448 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2449 accidental colon was present, for example, could give incorrect results.
2451 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2452 result is used (not including appendfile, where other changes are about
2455 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2456 This involved changing a lot of size variables from int to off_t. It
2457 should work with maildirs and everything.
2459 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2460 spamd dying while we are connected to it.
2462 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2465 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2466 function for BDB 4.3.
2468 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2470 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2471 the sources for PCRE have been re-arranged and more files are now
2474 PH/15 The code I had for printing potentially long long variables in PH/11
2475 above was not the best (it lost precision). The length of off_t variables
2476 is now inspected at build time, and an appropriate printing format (%ld
2477 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2478 to be "long long int" or "long int". This is needed for the internal
2479 formatting function string_vformat().
2481 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2482 the configuration file to be ":syslog", then the script "guesses" where
2483 the logs files are, rather than using the compiled in default. In our
2484 case the guess is not the same as the compiled default, so the script
2485 suddenly stopped working when I started to use syslog. The patch checks
2486 to see if log_file_path is "". If so, it attempts to read it from exim
2487 with no configuration file to get the compiled in version, before it
2488 falls back to the previous guessing code."
2490 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2491 implementing BATV in an Exim configuration. See NewStuff for the gory
2494 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2495 Makefile that are specific to HP-UX.
2497 PH/18 If the "use_postmaster" option was set for a recipient callout together
2498 with the "random" option, the postmaster address was used as the MAIL
2499 FROM address for the random test, but not for the subsequent recipient
2500 test. It is now used for both.
2502 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2503 patch removes a few documentation additions to RFC 3028, because the
2504 latest draft now contains them. It adds the new en;ascii-case comparator
2505 and a new error check for 8bit text in MIME parts. Comparator and
2506 require names are now matched exactly. I enabled the subaddress
2507 extension, but it is not well tested yet (read: it works for me)."
2509 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2510 rework some of the code of TK/09 above to avoid the hardwired use of
2511 "%lld" and "long long". Replaced the call to snprintf() with a call to
2514 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2515 records point to non-existent hosts", "retry timeout exceeded", and
2516 "retry time not reached for any host after a long failure period".
2518 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2519 experimental DomainKeys support:
2521 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2522 (2) On an error such as an illegally used "control", the wrong name for
2523 the control was given.
2525 These problems did NOT occur unless DomainKeys support was compiled.
2527 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2529 PH/24 Added ${if match_ip condition.
2531 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2532 databases so that it will be absolutely obvious if a crash occurs in the
2533 DB library. This is a regular occurrence (often caused by mis-matched
2536 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2537 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2538 picked up on a user's system that detects such things. There doesn't seem
2539 to be a gcc warning option for this - only an attribute that has to be
2540 put on the function's prototype. It seems that in Fedora Core 4 they have
2541 set this on a number of new functions. No doubt there will be more in due
2544 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2545 specify a transport. However, if an address that was verified by such a
2546 router was the subject of a callout, Exim crashed because it tried to
2547 read the rcpt_include_affixes from the non-existent transport. Now it
2548 just assumes that the setting of that option is false. This bug was
2549 introduced by 4.51/PH/31.
2551 PH/28 Changed -d+all to exclude +memory, because that information is very
2552 rarely of interest, but it makes the output a lot bigger. People tend to
2553 do -d+all out of habit.
2555 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2556 code in os-type was giving problems when libc.so lives in lib64, like on
2559 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2560 aren't the modern standard, and it seems that some systems' include files
2561 don't always have them. Exim was already checking for some of the newer
2562 ones like T_AAAA, and defining it itself. I've added checks for all the
2563 record types that Exim uses.
2565 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2566 not automatically generating a new one, as it is supposed to. This
2567 prevented TLS from working. If the file did exist, but contained invalid
2568 data, a new version was generated, as expected. It was only the case of a
2569 non-existent file that was broken.
2571 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2572 with a change in libdomainkeys > 0.64.
2574 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2575 from DNS. If the selector record carries the flag, it now has
2576 precedence over the domain-wide flag.
2578 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2580 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2581 the use of a port name that isn't defined in /etc/services) occurred, the
2582 message was deferred as in a normal delivery, and thus remained on the
2583 spool, instead of being failed because of the mua_wrapper setting. This
2584 is now fixed, and I tidied up some of the mua_wrapper messages at the
2587 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2588 the messages in a hash of arrays rather than using individual hashes.
2589 This is a bit cleaner and results in dramatic memory savings, albeit
2590 at a slight CPU cost.
2592 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2593 as requested by Marc Sherman.
2595 SC/10 Eximstats: added histograms for user specified patterns as requested
2598 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2600 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2601 fopen() in the content-scanning modules that did not already have it.
2607 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2608 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2610 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2612 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2614 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2615 to test Sieve filters that use "vacation".
2617 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2618 that changes the way the GnuTLS parameters are stored in the cache file.
2619 The new format can be generated externally. For backward compatibility,
2620 if the data in the cache doesn't make sense, Exim assumes it has read an
2621 old-format file, and it generates new data and writes a new file. This
2622 means that you can't go back to an older release without removing the
2625 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2626 work if there are any delivery delays because "one_time" forces the
2627 parent to be marked "delivered", so its unseen clone is never tried
2628 again. For this reason, Exim now forbids the simultaneous setting of
2631 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2632 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2633 entry if you want to know the details. The fix, however, neglected to
2634 consider the case when local delivery batching is involved. The test for
2635 "previously delivered" was not happening when checking to see if an
2636 address could be batched with a previous (undelivered) one; under
2637 certain circumstances this could lead to multiple deliveries to the same
2640 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2641 in its include files, and this causes problems building Exim.
2643 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2644 header_syntax) but Exim was just ignoring anything given after a slash.
2645 In particular, this caused confusion with an attempt to use "verify =
2646 reverse_host_lookup/defer_ok". An error is now given when options are
2647 supplied for verify items that do not have them. (Maybe reverse_host_
2648 lookup should have a defer_ok option, but that's a different point.)
2650 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2651 defined by RFC 821) to 2048, because there were problems with some AUTH
2652 commands, and RFC 1869 says the size should be increased for extended
2653 SMTP commands that take arguments.
2655 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2658 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2659 "unknown" error; now it says that the functionality isn't in the binary.
2661 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2662 an address' error message when a string expansion fails (syntax or
2663 whatever). Otherwise the password may appear in the log. Following change
2664 PH/42 below, there is no longer a chance of it appearing in a bounce
2667 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2669 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2670 of its addresses was ever tried. (Bugzilla bug #2.)
2672 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2673 the result incorrectly in the debug output. (It correctly added a newline
2674 to what was transported.)
2676 TF/01 Added $received_time.
2678 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2679 commented out examples of how to interface to a virus scanner and to
2680 SpamAssassin. Also added commented examples of av_scanner and
2681 spamd_address settings.
2683 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2684 and controls are allowed in which ACLs. There were a couple of minor
2685 errors. Some of the entries in the conditions table (which is a table of
2686 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2687 negation of where the condition IS allowed.
2689 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2691 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2692 header file does not have a version number, so I've had to invent a new
2693 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2694 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2695 radiusclient), but it was contributed by a Radius user.
2697 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2698 files or directories, for OpenSSL.
2700 PH/19 When an Exim process that is running as root has to create an Exim log
2701 file, it does so in a subprocess that runs as exim:exim so as to get the
2702 ownership right at creation (otherwise, other Exim processes might see
2703 the file with the wrong ownership). There was no test for failure of this
2704 fork() call, which would lead to the process getting stuck as it waited
2705 for a non-existent subprocess. Forks do occasionally fail when resources
2706 run out. I reviewed all the other calls to fork(); they all seem to check
2709 PH/20 When checking for unexpected SMTP input at connect time (before writing
2710 the banner), Exim was not dealing correctly with a non-positive return
2711 from the read() function. If the client had disconnected by this time,
2712 the result was a log entry for a synchronization error with an empty
2713 string after "input=" when read() returned zero. If read() returned -1
2714 (an event I could not check), uninitialized data bytes were printed.
2715 There were reports of junk text (parts of files, etc) appearing after
2718 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2720 PH/22 Added support for macro redefinition, and (re)definition in between
2721 driver and ACL definitions.
2723 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2724 forgetting to use the resulting value; it was using the unexpanded value.
2726 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2727 hadn't been configured. The fix is from Juergen Kreileder, who
2728 understands it better than I do:
2730 "Here's what I see happening with three configured cyrus_sasl
2731 authenticators configured (plain, login, cram-md5):
2733 On startup auth_cyrus_sasl_init() gets called for each of these.
2734 This means three calls to sasl_listmech() without a specified mech_list.
2735 => SASL tests which mechs of all available mechs actually work
2736 => three warnings about OTP not working
2737 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2739 With the patch, sasl_listmech() also gets called three times. But now
2740 SASL's mech_list option is set to the server_mech specified in the the
2741 authenticator. Or in other words, the answer from sasl_listmech()
2742 gets limited to just the mech you're testing for (which is different
2744 => the return list contains just 'plain' or 'login', 'cram-md5' or
2745 nothing depending on the value of ob->server_mech.
2747 I've just tested the patch: Authentication still works fine,
2748 unavailable mechs specified in the exim configuration are still
2749 caught, and the auth.log warnings about OTP are gone."
2751 PH/25 When debugging is enabled, the contents of the command line are added
2752 to the debugging output, even when log_selector=+arguments is not
2755 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2756 answer is "GNU", and only if the return is "GNU/something" is the answer
2759 PH/27 $acl_verify_message is now set immediately after the failure of a
2760 verification in an ACL, and so is available in subsequent modifiers. In
2761 particular, the message can be preserved by coding like this:
2763 warn !verify = sender
2764 set acl_m0 = $acl_verify_message
2766 Previously, $acl_verify_message was set only while expanding "message"
2767 and "log_message" when a very denied access.
2769 PH/28 Modified OS/os.c-Linux with
2771 -#ifndef OS_LOAD_AVERAGE
2772 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2774 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2775 nomenclature these days.)
2777 PH/29 Installed patch from the Sieve maintainer that adds the options
2778 sieve_useraddress and sieve_subaddress to the redirect router.
2780 PH/30 In these circumstances:
2781 . Two addresses routed to the same list of hosts;
2782 . First host does not offer TLS;
2783 . First host accepts first address;
2784 . First host gives temporary error to second address;
2785 . Second host offers TLS and a TLS session is established;
2786 . Second host accepts second address.
2787 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2788 and peerdn, if requested) that were in fact used only for the second
2791 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2792 attention to any local part prefix or suffix that was matched by the
2793 router that accepted the address. It now behaves in the same way as it
2794 does for delivery: the affixes are removed from the local part unless
2795 rcpt_include_affixes is set on the transport.
2797 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2798 timeout during the DATA phase of an incoming message.
2800 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2801 applied a patch sanctioned by the Sieve maintainer.
2803 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2804 the uid or gid is negative. A case of a negative gid caused this to be
2805 noticed. The fix allows for either to be negative.
2807 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2808 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2811 PH/36 The change PH/12 above was broken. Fixed it.
2813 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2814 the grounds that routing the same address twice would always produce the
2815 same answer. This might have been true once, but it is certainly no
2816 longer true now. Routing a child address may depend on the previous
2817 routing that produced that child. Some complicated redirection strategies
2818 went wrong when messages had multiple recipients, and made Exim's
2819 behaviour dependent on the order in which the addresses were given.
2821 I have moved the duplicate checking until after the routing is complete.
2822 Exim scans the addresses that are assigned to local and remote
2823 transports, and removes any duplicates. This means that more work will be
2824 done, as duplicates will always all be routed, but duplicates are
2825 presumably rare, so I don't expect this is of any significance.
2827 For deliveries to pipes, files, and autoreplies, the duplicate checking
2828 still happens during the routing process, since they are not going to be
2831 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2832 It corrects a timeout issue with spamd. This is Ian's comment: "The
2833 background is that sometimes spamd either never reads data from a
2834 connection it has accepted, or it never writes response data. The exiscan
2835 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2836 blindly assumes that writes won't block so it may never time out."
2838 PH/39 Allow G after quota size as well as K and M.
2840 PH/40 The value set for $authenticated_id in an authenticator may not contain
2841 binary zeroes or newlines because the value is written to log lines and
2842 to spool files. There was no check on this. Now the value is run through
2843 the string_printing() function so that such characters are converted to
2844 printable escape sequences.
2846 PH/41 $message_linecount is a new variable that contains the total number of
2847 lines in the message. Compare $body_linecount, which is the count for the
2850 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2851 bounce and delay warning messages, except in certain special cases, which
2854 (a) An SMTP error message from a remote host;
2855 (b) A message specified in a :fail: redirection;
2856 (c) A message specified in a "fail" command in a system filter;
2857 (d) A message specified in a FAIL return from the queryprogram router;
2858 (e) A message specified by the cannot_route_message router option.
2860 In these cases only, Exim does include the error details in bounce and
2861 warning messages. There are also a few cases where bland messages such
2862 as "unrouteable address" or "local delivery error" are given.
2864 PH/43 $value is now also set for the "else" part of a ${run expansion.
2866 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2867 being worked on, but at least Exim now implements the latest version to
2870 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2871 process to complete was treated as a delivery failure, a timeout while
2872 writing the message to the pipe was logged, but erroneously treated as a
2873 successful delivery. Such timeouts include transport filter timeouts. For
2874 consistency with the overall process timeout, these timeouts are now
2875 treated as errors, giving rise to delivery failures by default. However,
2876 there is now a new Boolean option for the pipe transport called
2877 timeout_defer, which, if set TRUE, converts the failures into defers for
2878 both kinds of timeout. A transport filter timeout is now identified in
2881 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2882 systems where "make" and "gmake" are different, calling "gmake" at top
2883 level broke things. I've arranged for the value of $(MAKE) to be passed
2884 from the Makefile to this script so that it can call the same version of
2888 A note about Exim versions 4.44 and 4.50
2889 ----------------------------------------
2891 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2892 changes of various kinds. As a consequence, a big documentation update was
2893 needed. This delayed the release for rather longer than seemed good, especially
2894 in the light of a couple of (minor) security issues. Therefore, the changes
2895 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2896 release. So 4.44 and 4.50 are in effect two different branches that both start
2899 I have left the 4.50 change log unchanged; it contains all the changes since
2900 4.43. The change log for 4.44 is below; many of its items are identical to
2901 those for 4.50. This seems to be the most sensible way to preserve the
2902 historical information.
2908 1. Minor wording change to the doc/README.SIEVE file.
2910 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2911 computation of the current number of files was incorrect.
2913 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2914 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2917 4. Give more explanation in the error message when the command for a transport
2918 filter fails to execute.
2920 5. There are several places where Exim runs a non-Exim command in a
2921 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2922 was being done only for the command run by the queryprogram router. It is
2923 now done for all such subprocesses. The other cases are: ${run, transport
2924 filters, and the commands run by the lmtp and pipe transports.
2926 6. Added CONFIGURE_GROUP build-time option.
2928 7. Some older OS have a limit of 256 on the maximum number of file
2929 descriptors. Exim was using setrlimit() to set 1000 as a large value
2930 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2931 systems. I've change it so that if it can't get 1000, it tries for 256.
2933 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2934 was an oversight, and furthermore, ever since the addition of extra
2935 controls (e.g. 4.43/32), the checks on when to allow different forms of
2936 "control" were broken. There should now be diagnostics for all cases when a
2937 control that does not make sense is encountered.
2939 9. Added the /retain_sender option to "control=submission".
2941 10. $recipients is now available in the predata ACL (oversight).
2943 11. Tidy the search cache before the fork to do a delivery from a message
2944 received from the command line. Otherwise the child will trigger a lookup
2945 failure and thereby defer the delivery if it tries to use (for example) a
2946 cached ldap connection that the parent has called unbind on.
2948 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2949 of $address_data from the recipient verification was clobbered by the
2950 sender verification.
2952 13. The value of address_data from a sender verification is now available in
2953 $sender_address_data in subsequent conditions in the ACL statement.
2955 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2957 15. Added a new option "connect=<time>" to callout options, to set a different
2960 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2961 was its contents. (It was OK if the option was not defined at all.)
2963 17. A "Completed" log line is now written for messages that are removed from
2964 the spool by the -Mrm option.
2966 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2967 information about exactly what failed.
2969 19. Added -dd to debug only the daemon process.
2971 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2972 handles timeouts, both on the server side and network timeouts. Renamed the
2973 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2975 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2976 It is now set to "smtps".
2978 22. $host_address is now set to the target address during the checking of
2979 ignore_target_hosts.
2981 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2982 being passed; this would have caused $sender_host_name to have been used if
2983 matching the list had actually called for a host name (not very likely,
2984 since this list is usually IP addresses). A host name is now passed as
2987 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2988 code that specifies a non-restarting handler (typically sigaction() in
2989 modern systems) in an attempt to fix a rare and obscure crash bug.
2991 25. Narrowed the window for a race in the daemon that could cause it to ignore
2992 SIGCHLD signals. This is not a major problem, because they are used only to
2993 wake it up if nothing else does.
2995 26. A malformed maildirsize file could cause Exim to calculate negative values
2996 for the mailbox size or file count. Odd effects could occur as a result.
2997 The maildirsize information is now recalculated if the size or filecount
3000 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3001 support for a long time. Removed HAVE_SYS_VFS_H.
3003 28. Installed the latest version of exipick from John Jetmore.
3005 29. In an address list, if the pattern was not a regular expression, an empty
3006 subject address (from a bounce message) matched only if the pattern was an
3007 empty string. Non-empty patterns were not even tested. This was the wrong
3008 because it is perfectly reasonable to use an empty address as part of a
3009 database query. An empty address is now tested by patterns that are
3010 lookups. However, all the other forms of pattern expect the subject to
3011 contain a local part and a domain, and therefore, for them, an empty
3012 address still always fails if the pattern is not itself empty.
3014 30. Exim went into a mad DNS loop when attempting to do a callout where the
3015 host was specified on an smtp transport, and looking it up yielded more
3016 than one IP address.
3018 31. Re-factored the code for checking spool and log partition space into a
3019 function that finds that data and another that does the check. The former
3020 is then used to implement four new variables: $spool_space, $log_space,
3021 $spool_inodes, and $log_inodes.
3023 32. The RFC2047 encoding function was originally intended for short strings
3024 such as real names; it was not keeping to the 75-character limit for
3025 encoded words that the RFC imposes. It now respects the limit, and
3026 generates multiple encoded words if necessary. To be on the safe side, I
3027 have increased the buffer size for the ${rfc2047: expansion operator from
3030 33. It is now permitted to omit both strings after an "if" condition; if the
3031 condition is true, the result is "true". As before, when the second string
3032 is omitted, a false condition yields an empty string. This makes it less
3033 cumbersome to write custom ACL and router conditions.
3035 34. Failure to deliver a bounce message always caused it to be frozen, even if
3036 there was an errors_to setting on the router. The errors_to setting is now
3039 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3040 canonical form (fully expanded) before being placed in
3041 $sender_host_address.
3043 36. The table in the code that translates DNS record types into text (T_A to
3044 "A" for instance) was missing entries for NS and CNAME. It is just possible
3045 that this could have caused confusion if both these types were looked up
3046 for the same domain, because the text type is used as part of Exim's
3047 per-process caching. But the chance of anyone hitting this buglet seems
3050 37. The dnsdb lookup has been extended in a number of ways.
3052 (1) There is a new type, "zns", which walks up the domain tree until it
3053 finds some nameserver records. It should be used with care.
3055 (2) There is a new type, "mxh", which is like "mx" except that it returns
3056 just the host names, not the priorities.
3058 (3) It is now possible to give a list of domains (or IP addresses) to be
3059 looked up. The behaviour when one of the lookups defers can be
3060 controlled by a keyword.
3062 (4) It is now possible to specify the separator character for use when
3063 multiple records are returned.
3065 38. The dnslists ACL condition has been extended: it is now possible to supply
3066 a list of IP addresses and/or domains to be looked up in a particular DNS
3069 39. Added log_selector=+queue_time_overall.
3071 40. When running the queue in the test harness, wait just a tad after forking a
3072 delivery process, to get repeatability of debugging output.
3074 41. Include certificate and key file names in error message when GnuTLS fails
3075 to set them up, because the GnuTLS error message doesn't include the name
3076 of the failing file when there is a problem reading it.
3078 42. Allow both -bf and -bF in the same test run.
3080 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3082 44. The "Exiscan patch" is now merged into the mainline Exim source.
3084 45. Sometimes the final signoff response after QUIT could fail to get
3085 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3086 < 0 before doing a fflush(). This bug looks as though it goes back to the
3087 introduction of TLS in release 3.20, but "sometimes" must have been rare
3088 because the tests only now provoked it.
3090 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3091 (this can affect the format of dates).
3093 47. exim_tidydb, when checking for the continued existence of a message for
3094 which it has found a message-specific retry record, was not finding
3095 messages that were in split spool directories. Consequently, it was
3096 deleting retry records that should have stayed in existence.
3098 48. Steve fixed some bugs in eximstats.
3100 49. The SPA authentication driver was not abandoning authentication and moving
3101 on to the next authenticator when an expansion was forced to fail,
3102 contradicting the general specification for all authenticators. Instead it
3103 was generating a temporary error. It now behaves as specified.
3105 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3106 (the order specifies the preference for clients). The order is now AES256,
3107 AES128, 3DES, ARCFOUR128.
3109 51. Small patch to Sieve code - explicitly set From: when generating an
3112 52. Exim crashed if a remote delivery caused a very long error message to be
3113 recorded - for instance if somebody sent an entire SpamAssassin report back
3114 as a large number of 550 error lines. This bug was coincidentally fixed by
3115 increasing the size of one of Exim's internal buffers (big_buffer) that
3116 happened as part of the Exiscan merge. However, to be on the safe side, I
3117 have made the code more robust (and fixed the comments that describe what
3120 53. Now that there can be additional text after "Completed" in log lines (if
3121 the queue_time_overall log selector is set), a one-byte patch to exigrep
3122 was needed to allow it to recognize "Completed" as not the last thing in
3125 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3126 patch that reportedly fixes this has been added. I am not expert enough to
3127 create a test for it. This is what the patch creator wrote:
3129 "I found a little strange behaviour of ldap code when working with
3130 Windows 2003 AD Domain, where users was placed in more than one
3131 Organization Units. When I tried to give exim partial DN, the exit code
3132 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3133 But simultaneously result of request was absolutely normal ldap result,
3134 so I produce this patch..."
3136 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3137 REFERENCE, so I have modified the code to exclude the patch when that macro
3140 55. Some experimental protocols are using DNS PTR records for new purposes. The
3141 keys for these records are domain names, not reversed IP addresses. The
3142 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3143 leaves it alone. Component reversal etc. now happens only for IP addresses.
3146 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3148 57. Double the size of the debug message buffer (to 2048) so that more of very
3149 long debug lines gets shown.
3151 58. The exicyclog utility now does better if the number of log files to keep
3152 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3154 59. Two changes related to the smtp_active_hostname option:
3156 (1) $smtp_active_hostname is now available as a variable.
3157 (2) The default for smtp_banner uses $smtp_active_hostname instead
3158 of $primary_hostname.
3160 60. The host_aton() function is supposed to be passed a string that is known
3161 to be a valid IP address. However, in the case of IPv6 addresses, it was
3162 not checking this. This is a hostage to fortune. Exim now panics and dies
3163 if the condition is not met. A case was found where this could be provoked
3164 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3165 components; fortuitously, this particular loophole had already been fixed
3166 by change 4.50/55 above.
3168 If there are any other similar loopholes, the new check in host_aton()
3169 itself should stop them being exploited. The report I received stated that
3170 data on the command line could provoke the exploit when Exim was running as
3171 exim, but did not say which command line option was involved. All I could
3172 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3173 running as the user.
3176 61. There was a buffer overflow vulnerability in the SPA authentication code
3177 (which came originally from the Samba project). I have added a test to the
3178 spa_base64_to_bits() function which I hope fixes it.
3181 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3182 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3184 63. The daemon start-up calls getloadavg() while still root for those OS that
3185 need the first call to be done as root, but it missed one case: when
3186 deliver_queue_load_max is set with deliver_drop_privilege. This is
3187 necessary for the benefit of the queue runner, because there is no re-exec
3188 when deliver_drop_privilege is set.
3190 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3191 This has been fixed.
3193 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3194 was in use, was not putting the data itself into the right store pool;
3195 consequently, it could be overwritten for a subsequent message in the same
3196 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3199 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3201 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3202 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3203 the function were treating the return as a boolean value, which happened to
3204 work because 0=false and not-0=true, but is not correct code.
3206 68. The host_aton() function was not handling scoped IPv6 addresses (those
3207 with, for example, "%eth0" on the end) correctly.
3209 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3210 negated items (that is, ~something) in unsigned ints. Some compilers
3211 apparently mutter when there is no cast.
3213 70. If an address verification called from an ACL failed, and did not produce a
3214 user-specific message (i.e. there was only a "system" message), nothing was
3215 put in $acl_verify_message. In this situation, it now puts the system
3218 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3219 session; change 4.31/43 added the unwanted input to the log line - except
3220 that it did not do this in the start of session case. It now does.
3222 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3223 This is wrong and can cause the other end to generate a synchronization
3224 error if it is another Exim or anything else that does the synchronization
3225 check. A QUIT command is no longer sent after a timeout.
3227 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3228 during host lookups.
3230 74. The defer_ok option of callout verification was not working if it was used
3231 when verifying addresses in header lines, that is, for this case:
3233 verify = header_sender/callout=defer_ok
3235 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3236 those file descriptors could be used for SMTP connections. If anything
3237 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3238 could be sent to the SMTP client, causing chaos. The daemon now opens
3239 stdin, stdout, and stderr to /dev/null when it puts itself into the
3242 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3243 log by default. The user can override this with suitable Perl magic.
3245 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3246 the log message when discard triggers, was not working for the DATA ACL or
3247 for the non-SMTP ACL.
3249 78. Error message wording change in sieve.c.
3251 79. If smtp_accept_max_per_host was set, the number of connections could be
3252 restricted to fewer than expected, because the daemon was trying to set up
3253 a new connection before checking whether the processes handling previous
3254 connections had finished. The check for completed processes is now done
3255 earlier. On busy systems, this bug wouldn't be noticed because something
3256 else would have woken the daemon, and it would have reaped the completed
3259 80. If a message was submitted locally by a user whose login name contained one
3260 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3261 It caused a spool format error. I have fixed the spool reading code. A
3262 related problem was that the "from" clause in the Received: line became
3263 illegal because of the space(s). It is now covered by ${quote_local_part.
3265 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3268 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3270 83. Patch from Sieve maintainer to fix unterminated string problem in
3271 "vacation" handling.
3273 84. Some minor changes to the Linux configuration files to help with other
3274 OS variants using glibc.
3276 85. One more patch for Sieve to update vacation handling to latest spec.
3279 ----------------------------------------------------
3280 See the note above about the 4.44 and 4.50 releases.
3281 ----------------------------------------------------
3287 1. Change 4.43/35 introduced a bug that caused file counts to be
3288 incorrectly computed when quota_filecount was set in an appendfile
3291 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3292 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3295 3. Give more explanation in the error message when the command for a transport
3296 filter fails to execute.
3298 4. There are several places where Exim runs a non-Exim command in a
3299 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3300 was being done only for the command run by the queryprogram router. It is
3301 now done for all such subprocesses. The other cases are: ${run, transport
3302 filters, and the commands run by the lmtp and pipe transports.
3304 5. Some older OS have a limit of 256 on the maximum number of file
3305 descriptors. Exim was using setrlimit() to set 1000 as a large value
3306 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3307 systems. I've change it so that if it can't get 1000, it tries for 256.
3309 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3310 was an oversight, and furthermore, ever since the addition of extra
3311 controls (e.g. 4.43/32), the checks on when to allow different forms of
3312 "control" were broken. There should now be diagnostics for all cases when a
3313 control that does not make sense is encountered.
3315 7. $recipients is now available in the predata ACL (oversight).
3317 8. Tidy the search cache before the fork to do a delivery from a message
3318 received from the command line. Otherwise the child will trigger a lookup
3319 failure and thereby defer the delivery if it tries to use (for example) a
3320 cached ldap connection that the parent has called unbind on.
3322 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3323 of $address_data from the recipient verification was clobbered by the
3324 sender verification.
3326 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3327 was its contents. (It was OK if the option was not defined at all.)
3329 11. A "Completed" log line is now written for messages that are removed from
3330 the spool by the -Mrm option.
3332 12. $host_address is now set to the target address during the checking of
3333 ignore_target_hosts.
3335 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3336 being passed; this would have caused $sender_host_name to have been used if
3337 matching the list had actually called for a host name (not very likely,
3338 since this list is usually IP addresses). A host name is now passed as
3341 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3342 code that specifies a non-restarting handler (typically sigaction() in
3343 modern systems) in an attempt to fix a rare and obscure crash bug.
3345 15. Narrowed the window for a race in the daemon that could cause it to ignore
3346 SIGCHLD signals. This is not a major problem, because they are used only to
3347 wake it up if nothing else does.
3349 16. A malformed maildirsize file could cause Exim to calculate negative values
3350 for the mailbox size or file count. Odd effects could occur as a result.
3351 The maildirsize information is now recalculated if the size or filecount
3354 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3355 support for a long time. Removed HAVE_SYS_VFS_H.
3357 18. Updated exipick to current release from John Jetmore.
3359 19. Allow an empty sender to be matched against a lookup in an address list.
3360 Previously the only cases considered were a regular expression, or an
3363 20. Exim went into a mad DNS lookup loop when doing a callout where the
3364 host was specified on the transport, if the DNS lookup yielded more than
3367 21. The RFC2047 encoding function was originally intended for short strings
3368 such as real names; it was not keeping to the 75-character limit for
3369 encoded words that the RFC imposes. It now respects the limit, and
3370 generates multiple encoded words if necessary. To be on the safe side, I
3371 have increased the buffer size for the ${rfc2047: expansion operator from
3374 22. Failure to deliver a bounce message always caused it to be frozen, even if
3375 there was an errors_to setting on the router. The errors_to setting is now
3378 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3379 canonical form (fully expanded) before being placed in
3380 $sender_host_address.
3382 24. Updated eximstats to version 1.33
3384 25. Include certificate and key file names in error message when GnuTLS fails
3385 to set them up, because the GnuTLS error message doesn't include the name
3386 of the failing file when there is a problem reading it.
3388 26. Expand error message when OpenSSL has problems setting up cert/key files.
3391 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3392 (this can affect the format of dates).
3394 28. exim_tidydb, when checking for the continued existence of a message for
3395 which it has found a message-specific retry record, was not finding
3396 messages that were in split spool directories. Consequently, it was
3397 deleting retry records that should have stayed in existence.
3399 29. eximstats updated to version 1.35
3400 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3401 1.35 - bugfix such that pie charts by volume are generated correctly
3403 30. The SPA authentication driver was not abandoning authentication and moving
3404 on to the next authenticator when an expansion was forced to fail,
3405 contradicting the general specification for all authenticators. Instead it
3406 was generating a temporary error. It now behaves as specified.
3408 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3409 (the order specifies the preference for clients). The order is now AES256,
3410 AES128, 3DES, ARCFOUR128.
3412 31. Small patch to Sieve code - explicitly set From: when generating an
3415 32. Exim crashed if a remote delivery caused a very long error message to be
3416 recorded - for instance if somebody sent an entire SpamAssassin report back
3417 as a large number of 550 error lines. This bug was coincidentally fixed by
3418 increasing the size of one of Exim's internal buffers (big_buffer) that
3419 happened as part of the Exiscan merge. However, to be on the safe side, I
3420 have made the code more robust (and fixed the comments that describe what
3423 33. Some experimental protocols are using DNS PTR records for new purposes. The
3424 keys for these records are domain names, not reversed IP addresses. The
3425 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3426 leaves it alone. Component reversal etc. now happens only for IP addresses.
3429 34. The host_aton() function is supposed to be passed a string that is known
3430 to be a valid IP address. However, in the case of IPv6 addresses, it was
3431 not checking this. This is a hostage to fortune. Exim now panics and dies
3432 if the condition is not met. A case was found where this could be provoked
3433 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3434 components; fortuitously, this particular loophole had already been fixed
3435 by change 4.50/55 or 4.44/33 above.
3437 If there are any other similar loopholes, the new check in host_aton()
3438 itself should stop them being exploited. The report I received stated that
3439 data on the command line could provoke the exploit when Exim was running as
3440 exim, but did not say which command line option was involved. All I could
3441 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3442 running as the user.
3445 35. There was a buffer overflow vulnerability in the SPA authentication code
3446 (which came originally from the Samba project). I have added a test to the
3447 spa_base64_to_bits() function which I hope fixes it.
3450 36. The daemon start-up calls getloadavg() while still root for those OS that
3451 need the first call to be done as root, but it missed one case: when
3452 deliver_queue_load_max is set with deliver_drop_privilege. This is
3453 necessary for the benefit of the queue runner, because there is no re-exec
3454 when deliver_drop_privilege is set.
3456 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3457 was in use, was not putting the data itself into the right store pool;
3458 consequently, it could be overwritten for a subsequent message in the same
3459 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3462 38. Sometimes the final signoff response after QUIT could fail to get
3463 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3464 < 0 before doing a fflush(). This bug looks as though it goes back to the
3465 introduction of TLS in release 3.20, but "sometimes" must have been rare
3466 because the tests only now provoked it.
3472 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3473 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3474 Now it returns an integer. A number of calls were still expecting a T/F
3475 return. Fortuitously, in all cases, the tests worked in OK situations,
3476 which is the norm. However, things would have gone wrong on any write
3477 failures on the smtp file descriptor. This function is used when sending
3478 messages over SMTP and also when doing verify callouts.
3480 2. When Exim is called to do synchronous delivery of a locally submitted
3481 message (the -odf or -odi options), it no longer closes stderr before doing
3484 3. Implemented the mua_wrapper option.
3486 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3488 5. Implemented the functions header_remove(), header_testname(),
3489 header_add_at_position(), and receive_remove_recipient(), and exported them
3492 6. If an ACL "warn" statement specified the addition of headers, Exim already
3493 inserted X-ACL-Warn: at the start if there was no header name. However, it
3494 was not making this test for the second and subsequent header lines if
3495 there were newlines in the string. This meant that an invalid header could
3496 be inserted if Exim was badly configured.
3498 7. Allow an ACL "warn" statement to add header lines at the start or after all
3499 the Received: headers, as well as at the end.
3501 8. Added the rcpt_4xx retry error code.
3503 9. Added postmaster_mailfrom=xxx to callout verification option.
3505 10. Added mailfrom=xxxx to the callout verification option, for verify=
3508 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3509 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3511 12. Inserted some casts to stop certain compilers warning when using pointer
3512 differences as field lengths or precisions in printf-type calls (mostly
3513 affecting debugging statements).
3515 13. Added optional readline() support for -be (dynamically loaded).
3517 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3518 same clock tick as a message's arrival, so that its received time was the
3519 same as the "first fail" time on the retry record, and that message
3520 remained on the queue past the ultimate address timeout, every queue runner
3521 would try a delivery (because it was past the ultimate address timeout) but
3522 after another failure, the ultimate address timeout, which should have then
3523 bounced the address, did not kick in. This was a "< instead of <=" error;
3524 in most cases the first failure would have been in the next clock tick
3525 after the received time, and all would be well.
3527 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3528 being recognized when the domain list was tested by the match_domain
3529 condition in an expansion string.
3531 16. Added the ${str2b64: operator.
3533 17. Exim was always calling setrlimit() to set a large limit for the number of
3534 processes, without checking whether the existing limit was already
3535 adequate. (It did check for the limit on file descriptors.) Furthermore,
3536 errors from getrlimit() and setrlimit() were being ignored. Now they are
3537 logged to the main and panic logs, but Exim does carry on, to try to do its
3538 job under whatever limits there are.
3540 18. Imported PCRE 5.0.
3542 19. Trivial typo in log message " temporarily refused connection" (the leading
3545 20. If the log selector return_path_on_delivery was set and an address was
3546 redirected to /dev/null, the delivery process crashed because it assumed
3547 that a return path would always be set for a "successful" delivery. In this
3548 case, the whole delivery is bypassed as an optimization, and therefore no
3551 21. Internal re-arrangement: the function for sending a challenge and reading
3552 a response while authentication was assuming a zero-terminated challenge
3553 string. It's now changed to take a pointer and a length, to allow for
3554 binary data in such strings.
3556 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3558 23. Exim was not respecting finduser_retries when seeking the login of the
3559 uid under which it was called; it was always trying 10 times. (The default
3560 setting of finduser_retries is zero.) Also, it was sleeping after the final
3561 failure, which is pointless.
3563 24. Implemented tls_on_connect_ports.
3565 25. Implemented acl_smtp_predata.
3567 26. If the domain in control=submission is set empty, Exim assumes that the
3568 authenticated id is a complete email address when it generates From: or
3569 Sender: header lines.
3571 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3572 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3573 chown and chgrp in /bin and hostname in /usr/bin.
3575 28. Exim was keeping the "process log" file open after each use, just as it
3576 does for the main log. This opens the possibility of it remaining open for
3577 long periods when the USR1 signal hits a daemon. Occasional processlog
3578 errors were reported, that could have been caused by this. Anyway, it seems
3579 much more sensible not to leave this file open at all, so that is what now
3582 29. The long-running daemon process does not normally write to the log once it
3583 has entered its main loop, and it closes the log before doing so. This is
3584 so that log files can straightforwardly be renamed and moved. However,
3585 there are a couple of unusual error situations where the daemon does write
3586 log entries, and I had neglected to close the log afterwards.
3588 30. The text of an SMTP error response that was received during a remote
3589 delivery was being truncated at 512 bytes. This is too short for some of
3590 the long messages that one sometimes sees. I've increased the limit to
3593 31. It is now possible to make retry rules that apply only when a message has a
3594 specific sender, in particular, an empty sender.
3596 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3597 it possible to be selective about when SMTP synchronization is enforced.
3599 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3601 32. Implemented hosts_connection_nolog.
3603 33. Added an ACL for QUIT.
3605 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3608 35. Added mailbox_size and mailbox_filecount to appendfile.
3610 36. Added control = no_multiline_responses to ACLs.
3612 37. There was a bug in the logic of the code that waits for the clock to tick
3613 in the case where the clock went backwards by a substantial amount such
3614 that the microsecond fraction of "now" was more than the microsecond
3615 fraction of "then" (but the whole seconds number was less).
3617 38. Added support for the libradius Radius client library this is found on
3618 FreeBSD (previously only the radiusclient library was supported).
3624 1. When certain lookups returned multiple values in the form name=value, the
3625 quoting of the values was not always being done properly. Specifically:
3626 (a) If the value started with a double quote, but contained no whitespace,
3628 (b) If the value contained whitespace other than a space character (i.e.
3629 tabs or newlines or carriage returns) it was not quoted.
3630 This fix has been applied to the mysql and pgsql lookups by writing a
3631 separate quoting function and calling it from the lookup code. The fix
3632 should probably also be applied to nisplus, ibase and oracle lookups, but
3633 since I cannot test any of those, I have not disturbed their existing code.
3635 2. A hit in the callout cache for a specific address caused a log line with no
3636 reason for rejecting RCPT. Now it says "Previous (cached) callout
3637 verification failure".
3639 3. There was an off-by-one bug in the queryprogram router. An over-long
3640 return line was truncated at 256 instead of 255 characters, thereby
3641 overflowing its buffer with the terminating zero. As well as fixing this, I
3642 have increased the buffer size to 1024 (and made a note to document this).
3644 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3645 when Exim is waiting for an SMTP response from a remote server, Exim
3646 restarts its select() call on the socket, thereby resetting its timeout.
3647 This is not a problem when such interrupts are rare. Somebody set up a cron
3648 job to run exiwhat every 2 minutes, which is less than the normal select()
3649 timeout (5 or 10 minutes). This meant that the select() timeout never
3650 kicked in because it was always reset. I have fixed this by comparing the
3651 time when an interrupt arrives with the time at the start of the first call
3652 to select(). If more time than the timeout has elapsed, the interrupt is
3653 treated as a timeout.
3655 5. Some internal re-factoring in preparation for the addition of Sieve
3656 extensions (by MH). In particular, the "personal" test is moved to a
3657 separate function, and given an option for scanning Cc: and Bcc: (which is
3658 not set for Exim filters).
3660 6. When Exim created an email address using the login of the caller as the
3661 local part (e.g. when creating a From: or Sender: header line), it was not
3662 quoting the local part when it contained special characters such as @.
3664 7. Installed new OpenBSD configuration files.
3666 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3667 try to make them clearer.
3669 9. Callout options, other than the timeout value, were being ignored when
3670 verifying sender addresses in header lines. For example, when using
3672 verify = header_sender/callout=no_cache
3674 the cache was (incorrectly) being used.
3676 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3677 only the Cygwin environment.
3679 11. When return_path_on_delivery was set as a log selector, if different remote
3680 addresses in the same message used different return paths and parallel
3681 remote delivery occurred, the wrong values would sometimes be logged.
3682 (Whenever a remote delivery process finished, the return path value from
3683 the most recently started remote delivery process was logged.)
3685 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3686 lines when AUTH and/or TLS are in use. This is the "received protocol"
3687 field. Exim used to use "asmtp" for authenticated SMTP, without any
3688 indication (in the protocol name) for TLS use. Now it follows the RFC and
3689 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3690 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3691 names appear in log lines as well as in Received: header lines.
3693 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3694 extensions, and comparison tests, and to fix some bugs.
3696 14. Changes to the "personal" filter test:
3698 (1) The test was buggy in that it was just doing the equivalent of
3699 "contains" tests on header lines. For example, if a user's address was
3700 anne@some.where, the "personal" test would incorrectly be true for
3702 To: susanne@some.where
3704 This test is now done by extracting each address from the header in turn,
3705 and checking the entire address. Other tests that are part of "personal"
3706 are now done using regular expressions (for example, to check local parts
3707 of addresses in From: header lines).
3709 (2) The list of non-personal local parts in From: addresses has been
3710 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3711 taken from the Sieve specification recommendations.
3713 (3) If the message contains any header line starting with "List-" it is
3714 treated as non-personal.
3716 (4) The test for "circular" in the Subject: header line has been removed
3717 because it now seems ill-conceived.
3719 15. Minor typos in src/EDITME comments corrected.
3721 16. Installed latest exipick from John Jetmore.
3723 17. If headers_add on a router specified a text string that was too long for
3724 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3725 of string_sprintf() is now avoided.
3727 18. $message_body_size was not set (it was always zero) when running the DATA
3728 ACL and the local_scan() function.
3730 19. For the "mail" command in an Exim filter, no default was being set for
3731 the once_repeat time, causing a random time value to be used if "once" was
3732 specified. (If the value happened to be <= 0, no repeat happened.) The
3733 default is now 0s, meaning "never repeat". The "vacation" command was OK
3734 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3735 (I found it when inspecting the code).
3737 20. There is now an overall timeout for performing a callout verification. It
3738 defaults to 4 times the callout timeout, which applies to individual SMTP
3739 commands during the callout. The overall timeout applies when there is more
3740 than one host that can be tried. The timeout is checked before trying the
3741 next host. This prevents very long delays if there are a large number of
3742 hosts and all are timing out (e.g. when the network connections are timing
3743 out). The value of the overall timeout can be changed by specifying an
3744 additional sub-option for "callout", called "maxwait". For example:
3746 verify = sender/callout=5s,maxwait=20s
3748 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3749 to the end before writing, but this should make it even safer).
3751 22. Exim was forgetting that it had advertised PIPELINING for the second and
3752 subsequent messages on an SMTP connection. It was also not resetting its
3753 memory on STARTTLS and an internal HELO.
3755 23. When Exim logs an SMTP synchronization error within a session, it now
3756 records whether PIPELINING has been advertised or not.
3758 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3759 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3760 rather than long int.
3762 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3764 26. Added the never_mail option to autoreply.
3770 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3771 crash if the getsockname() call failed; this can happen if a connection is
3772 closed very soon after it is established. The problem was simply in the
3773 order in which certain operations were done, causing Exim to try to write
3774 to the SMTP stream before it had set up the file descriptor. The bug has
3775 been fixed by making things happen in the correct order.
3781 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3782 before the connection was closed, thus losing the rejection response.
3784 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3785 some early Solaris releases, but causes trouble in current releases where
3786 socklen_t is defined.
3788 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3791 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3794 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3795 that releasing the top of it at the end releases what was used for sub-
3796 expansions (unless the block got too big). However, discard this block if
3797 the first thing is a variable or header, so that we can use its block when
3798 it is dynamic (useful for very large $message_headers, for example).
3800 6. Lookups now cache *every* query, not just the most recent. A new, separate
3801 store pool is used for this. It can be recovered when all lookup caches are
3802 flushed. Lookups now release memory at the end of their result strings.
3803 This has involved some general refactoring of the lookup sources.
3805 7. Some code has been added to the store_xxx() functions to reduce the amount
3806 of flapping under certain conditions.
3808 8. log_incoming_interface used to affect only the <= reception log lines. Now
3809 it causes the local interface and port to be added to several more SMTP log
3810 lines, for example "SMTP connection from", and rejection lines.
3812 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3814 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3816 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3817 could be overwritten at the end of the current message (or the start of a
3818 new message if it was set in a HELO ACL). The value is now preserved for
3819 the duration of the SMTP connection.
3821 12. If a transport had a headers_rewrite setting, and a matching header line
3822 contained an unqualified address, that address was qualified, even if it
3823 did not match any rewriting rules. The underlying bug was that the values
3824 of the flags that permit the existence of unqualified sender and recipient
3825 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3826 non-local messages, and by -bnq for local messages) were not being
3827 preserved with the message after it was received.
3829 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3830 "next input=" as part of the text comprising the host identity instead of
3831 the correct text. The code was using the same buffer for two different
3832 strings. However, depending on which order the printing function evaluated
3833 its arguments, the bug did not always show up. Under Linux, for example, my
3834 test suite worked just fine.
3836 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3837 doesn't work with some older versions of Perl. It has been changed to "my",
3838 which in any case is probably the better facility to use.
3840 15. A really picky compiler found some instances of statements for creating
3841 error messages that either had too many or two few arguments for the format
3844 16. The size of the buffer for calls to the DNS resolver has been increased
3845 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3846 for addresses that have a lot of PTR records. This alleviates a problem; it
3847 does not fully solve it.
3849 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3850 buffer now truncates the list and logs the incident, which is the same
3851 action as happens when Exim is looking up a host name and its aliases.
3852 Previously in this situation something unpredictable would happen;
3853 sometimes it was "internal error: store_reset failed".
3855 18. If a server dropped the connection unexpectedly when an Exim client was
3856 using GnuTLS and trying to read a response, the client delivery process
3857 crashed while trying to generate an error log message.
3859 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3860 string, for example:
3862 warn message = H1: something\nH2: something
3864 the text was added as a single header line from Exim's point of view
3865 though it ended up OK in the delivered message. However, searching for the
3866 second and subsequent header lines using $h_h2: did not work. This has been
3867 fixed. Similarly, if a system filter added multiple headers in this way,
3868 the routers could not see them.
3870 20. Expanded the error message when iplsearch is called with an invalid key to
3871 suggest using net-iplsearch in a host list.
3873 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3874 ACLs are no longer actually imposed (and a message to that effect is
3877 22. If a "gecos" field in a passwd entry contained escaped characters, in
3878 particular, if it contained a \" sequence, Exim got it wrong when building
3879 a From: or a Sender: header from that name. A second bug also caused
3880 incorrect handling when an unquoted " was present following a character
3881 that needed quoting.
3883 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3884 was not being matched caselessly.
3886 24. Arranged for all hyphens in the exim.8 source to be escaped with
3889 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3890 properly. Recipient callout cache records were still being keyed to include
3891 the sender, even when use_sender was set false. This led to far more
3892 callouts that were necessary. The sender is no longer included in the key
3893 when use_sender is false.
3895 26. Added "control = submission" modifier to ACLs.
3897 27. Added the ${base62d: operator to decode base 62 numbers.
3899 28. dnsdb lookups can now access SRV records.
3901 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3902 the configuration file.
3904 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3905 (-v) mode. This makes the output for a verbose queue run more intelligible.
3907 31. Added a use_postmaster feature to recipient callouts.
3909 32. Added the $body_zerocount variable, containing the number of binary zero
3910 bytes in the message body.
3912 33. The time of last modification of the "new" subdirectory is now used as the
3913 "mailbox time last read" when there is a quota error for a maildir
3916 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3918 35. Added +ignore_unknown as a special item in host lists.
3920 36. Code for decoding IPv6 addresses in host lists is now included, even if
3921 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3922 address was recognized as an IP address, but was then not correctly decoded
3923 into binary, causing unexpected and incorrect effects when compared with
3930 1. Very minor rewording of debugging text in manualroute to say "list of
3931 hosts" instead of "hostlist".
3933 2. If verify=header_syntax was set, and a header line with an unqualified
3934 address (no domain) and a large number of spaces between the end of the
3935 name and the colon was received, the reception process suffered a buffer
3936 overflow, and (when I tested it) crashed. This was caused by some obsolete
3937 code that should have been removed. The fix is to remove it!
3939 3. When running in the test harness, delay a bit after writing a bounce
3940 message to get a bit more predictability in the log output.
3942 4. Added a call to search_tidyup() just before forking a reception process. In
3943 theory, someone could use a lookup in the expansion of smtp_accept_max_
3944 per_host which, without the tidyup, could leave open a database connection.
3946 5. Added the variables $recipient_data and $sender_data which get set from a
3947 lookup success in an ACL "recipients" or "senders" condition, or a router
3948 "senders" option, similar to $domain_data and $local_part_data.
3950 6. Moved the writing of debug_print from before to after the "senders" test
3953 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3954 problems for message scanning, either using a data ACL, or using
3955 local_scan() because the Received: header was not generated till after they
3956 were called (in order to set the time as the time of reception completion).
3957 I have revised the way this works. The header is now generated after the
3958 body is received, but before the ACL or local_scan() are called. After they
3959 are run, the timestamp in the header is updated.
3965 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3966 before starting a queue runner without re-exec. This happened only when
3967 deliver_drop_privilege was set or when the Exim user was set to root. The
3968 effect of the bug was that timeouts during subsequent deliveries caused
3969 crashes instead of being properly handled. The handler is now left at its
3970 default (and expected) setting.
3972 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3973 message, again when deliver_drop_privilege is set or Exim is run as root.
3974 The bug described in (1) was not present in this case, but the tidying up
3975 of the other signals was missing. I have made the two cases consistent.
3977 3. The ignore_target_hosts setting on a manualroute router was being ignored
3978 for hosts that were looked up using the /MX notation.
3980 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3983 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3984 operated on the sender address. After changing the $sender_address to <>
3985 for the sender address verify, Exim was re-instated it as the original
3986 (before rewriting) address, but remembering that it had rewritten it, so it
3987 wasn't rewriting it again. This bug also had the effect of breaking the
3988 sender address verification caching when the sender address was rewritten.
3990 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3991 This has been changed so that if the ip literal address matches
3992 ignore_target_hosts, the router declines.
3994 7. Added expansion conditions match_domain, match_address, and match_local_
3995 part (NOT match_host).
3997 8. The placeholder for the Received: header didn't have a length field set.
3999 9. Added code to Exim itself and to exim_lock to test for a specific race
4000 condition that could lead to file corruption when using MBX delivery. The
4001 issue is with the lockfile that is created in /tmp. If this file is removed
4002 after a process has opened it but before that process has acquired a lock,
4003 there is the potential for a second process to recreate the file and also
4004 acquire a lock. This could lead to two Exim processes writing to the file
4005 at the same time. The added code performs the same test as UW imapd; it
4006 checks after acquiring the lock that its file descriptor still refers to
4007 the same named file.
4009 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4010 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4013 11. Added the smtp_active_hostname option. If used, this will typically be made
4014 to depend on the incoming interface address. Because $interface_address is
4015 not set up until the daemon has forked a reception process, error responses
4016 that can happen earlier (such as "too many connections") no longer contain
4019 12. If an expansion in a condition on a "warn" statement fails because a lookup
4020 defers, the "warn" statement is abandoned, and the next ACL statement is
4021 processed. Previously this caused the whole ACL to be aborted.
4023 13. Added the iplsearch lookup type.
4025 14. Added ident_timeout as a log selector.
4027 15. Added tls_certificate_verified as a log selector.
4029 16. Added a global option tls_require_ciphers (compare the smtp transport
4030 option of the same name). This controls incoming TLS connections.
4032 17. I finally figured out how to make tls_require_ciphers do a similar thing
4033 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4034 before starting the TLS session.
4036 18. Tabs are now shown as \t in -bP output.
4038 19. If the log selector return_path_on_delivery was set, Exim crashed when
4039 bouncing a message because it had too many Received: header lines.
4041 20. If two routers both had headers_remove settings, and the first one included
4042 a superfluous trailing colon, the final name in the first list and the
4043 first name in the second list were incorrectly joined into one item (with a
4044 colon in the middle).
4050 1. Added -C and -D options to the exinext utility, mainly to make it easier
4051 to include in the automated testing, but these could be helpful when
4052 multiple configurations are in use.
4054 2. The exinext utility was not formatting the output nicely when there was
4055 an alternate port involved in the retry record key, nor when there was a
4056 message id as well (for retries that were specific to a specific message
4057 and a specific host). It was also confused by IPv6 addresses, because of
4058 the additional colons they contain. I have fixed the IPv4 problem, and
4059 patched it up to do a reasonable job for IPv6.
4061 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4062 delivery, the log line now contains "pipelined" if PIPELINING was used.
4064 4. An SMTP transport process used to panic and die if the bind() call to set
4065 an explicit outgoing interface failed. This has been changed; it is now
4066 treated in the same way as a connect() failure.
4068 5. A reference to $sender_host_name in the part of a conditional expansion
4069 that was being skipped was still causing a DNS lookup. This no longer
4072 6. The def: expansion condition was not recognizing references to header lines
4073 that used bh_ and bheader_.
4075 7. Added the _cache feature to named lists.
4077 8. The code for checking quota_filecount in the appendfile transport was
4078 allowing one more file than it should have been.
4080 9. For compatibility with Sendmail, the command line option
4088 and sets the incoming protocol and host name (for trusted callers). The
4089 host name and its colon can be omitted when only the protocol is to be set.
4090 Note the Exim already has two private options, -pd and -ps, that refer to
4091 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4092 "s", but I don't think that's a major issue.
4094 10. A number of refactoring changes to the code, none of which should affect
4097 (a) The number of logging options was getting close to filling up the
4098 32-bit word that was used as a bit map. I have split them into two classes:
4099 those that are passed in the argument to log_write(), and those that are
4100 only ever tested independently outside of that function. These are now in
4101 separate 32-bit words, so there is plenty of room for expansion again.
4102 There is no change in the user interface or the logging behaviour.
4104 (b) When building, for example, log lines, the code previously used a
4105 macro that called string_cat() twice, in order to add two strings. This is
4106 not really sufficiently general. Furthermore, there was one instance where
4107 it was actually wrong because one of the argument was used twice, and in
4108 one call a function was used. (As it happened, calling the function twice
4109 did not affect the overall behaviour.) The macro has been replaced by a
4110 function that can join an arbitrary number of extra strings onto a growing
4113 (c) The code for expansion conditions now uses a table and a binary chop
4114 instead of a serial search (which was left over from when there were very
4115 few conditions). Also, it now recognizes conditions like "pam" even when
4116 the relevant support is not compiled in: a suitably worded error message is
4117 given if an attempt is made to use such a condition.
4119 11. Added ${time_interval:xxxxx}.
4121 12. A bug was causing one of the ddress fields not to be passed back correctly
4122 from remote delivery subprocesses. The field in question was not being
4123 subsequently used, so this caused to problems in practice.
4125 13. Added new log selectors queue_time and deliver_time.
4127 14. Might have fixed a bug in maildirsizefile handling that threw up
4128 "unexpected character" debug warnings, and recalculated the data
4129 unnecessarily. In any case, I expanded the warning message to give more
4132 15. Added the message "Restricted characters in address" to the statements in
4133 the default ACL that block characters like @ and % in local parts.
4135 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4136 Three changes have been made:
4138 (a) There was a serious bug; a negative response to MAIL caused the whole
4139 recipient domain to be cached as invalid, thereby blocking all messages
4140 to all local parts at the same domain, from all senders. This bug has
4141 been fixed. The domain is no longer cached after a negative response to
4142 MAIL if the sender used is not empty.
4144 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4147 (c) A new callout option, "use_sender" has been added for people who want
4148 the modified behaviour.
4154 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4157 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4158 indeed breaks things for older releases.
4160 3. Added additional logging to the case where there is a problem reading data
4161 from a filter that is running in a subprocess using a pipe, in order to
4162 try to track down a specific problem.
4164 4. Testing facility fudge: when running in the test harness and attempting
4165 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4166 getting "No route to host". Convert this to a timeout.
4168 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4171 6. Some OS don't have socklen_t but use size_t instead. This affects the
4172 fifth argument of getsockopt() amongst other things. This is now
4173 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4174 can be set for individual OS. I have set it for SunOS5, OSF1, and
4175 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4176 some earlier ones do not.
4178 7. Change 4.30/15 was not doing the test caselessly.
4180 8. The standard form for an IPv6 address literal was being rejected by address
4181 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4182 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4183 this, as well as the form without the "IPv6" on the front (but only when
4184 address literals are enabled, of course).
4186 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4188 10. Exim crashed if a message with an empty sender address specified by -f
4189 encountered a router with an errors_to setting. This could be provoked only
4190 by a command such as
4194 where an empty string was supplied; "<>" did not hit this bug.
4196 11. Installed PCRE release 4.5.
4198 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4199 remained set. It is now erased.
4201 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4202 times from message ids (which are base 36 rather than the normal 62).
4204 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4205 were being counted as actual protocol errors, and logged if the log
4206 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4207 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4208 and DATA following a set of rejected RCPTs do not count as protocol errors.
4209 In other words, Exim assumes they were pipelined, though this may not
4210 actually be the case. Of course, in all cases the client gets an
4211 appropriate error code.
4213 15. If a lookup fails in an ACL condition, a message about the failure may
4214 be available; it is used if testing the ACL cannot continue, because most
4215 such messages specify what the cause of the deferral is. However, some
4216 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4217 that caused an old message to be retained and used if a later statement
4218 caused a defer, replacing the real cause of the deferral.
4220 16. If an IP address had so many PTR records that the DNS lookup buffer
4221 was not large enough to hold them, Exim could crash while trying to process
4222 the truncated data. It now detects and logs this case.
4224 17. Further to 4.21/58, another change has been made: if (and only if) the
4225 first line of a message (the first header line) ends with CRLF, a bare LF
4226 in a subsequent header line has a space inserted after it, so as not to
4227 terminate the header.
4229 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4230 unnecessarily, used atoi() instead of strtol(), and didn't check the
4231 termination when getting file sizes from file names by regex.
4233 19. Completely re-implemented the support for maildirsize files, in the light
4234 of a number of problems with the previous contributed implementation
4235 (4.30/29). In particular:
4237 . If the quota is zero, the maildirsize file is maintained, but no quota is
4240 . If the maildir directory does not exist, it is created before any attempt
4241 to write a maildirsize file.
4243 . The quota value in the file is just a cache; if the quota is changed in
4244 the transport, the new value overrides.
4246 . A regular expression is available for excluding directories from the
4249 20. The autoreply transport checks the characters in options that define the
4250 message's headers; it allows continued headers, but it was checking with
4251 isspace() after an embedded newline instead of explicitly looking for a
4254 21. If all the "regular" hosts to which an address was routed had passed their
4255 expiry times, and had not reached their retry times, the address was
4256 bounced, even if fallback hosts were defined. Now Exim should go on to try
4259 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4260 equivalent code in the SMTP transport. Some hosts send humungous responses
4261 to HELO/EHLO, more than 1024 it seems.
4263 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4264 gives compiler warnings in some environments. I've now done it "properly",
4267 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4268 (because of problems with the built-in one) was declared to return uschar *
4269 instead of char *, causing compiler failure.
4271 25. Fixed a file descriptor leak when processing alias/forward files.
4273 26. Fixed a minor format string issue in dbfn.c.
4275 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4277 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4278 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4281 29. When a sender address is verified, it is cached, to save repeating the test
4282 when there is more than one recipient in a message. However, when the
4283 verification involves a callout, it is possible for different callout
4284 options to be set for different recipients. It is too complicated to keep
4285 track of this in the cache, so now Exim always runs a verification when a
4286 callout is required, relying on the callout cache for the optimization.
4287 The overhead is duplication of the address routing, but this should not be
4290 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4291 to be verified with callout=postmaster, and the main callout worked but the
4292 postmaster check failed, the verification correctly failed. However, if a
4293 subsequent RCPT command asked for sender verification *without* the
4294 postmaster check, incorrect caching caused this verification also to fail,
4297 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4298 it was not caching the DNS options (qualify_single, search_parents) that
4299 were used when the lookup failed. A subsequent lookup with different
4300 options therefore always gave the same answer, though there were cases
4301 where it should not have. (Example: a "domains = !$mx_any" option on a
4302 dnslookup router: the "domains" option is always processed without any
4303 widening, but the router might have qualify_single set.) Now Exim uses the
4304 cached value only when the same options are set.
4306 32. Added John Jetmore's "exipick" utility to the distribution.
4308 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4309 than a timeout (e.g. a certificate is required, and is not provided), an
4310 Exim server now closes the connection immediately. Previously it waited for
4311 the client to close - but if the client is SSL, it seems that they each
4312 wait for each other, leading to a delay before one of them times out.
4314 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4315 maintained 0.8.x compatibility because I don't think many are using it, and
4316 it is clearly obsolete.
4318 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4321 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4322 client certificate was expired. A simple patch fixes this, though I don't
4323 understand the full logic of why the verify callback is called multiple
4326 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4327 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4328 which causes problems with some clients (such as the Certicom SSL Plus
4329 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4330 disables the coutermeasure allowing Eudora to connect."
4332 38. Exim was not checking that a write() to a log file succeeded. This could
4333 lead to Bad Things if a log got too big, in particular if it hit a file
4334 size limit. Exim now panics and dies if it cannot write to a log file, just
4335 as it does if it cannot open a log file.
4337 39. Modified OS/Makefile-Linux so that it now contains
4339 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4341 The two -D definitions ensure that Exim is compiled with large file
4342 support, which makes it possible to handle log files that are bigger than
4345 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4346 instance) a domain was checked against a named list that involved a lookup,
4347 causing $domain_data to be set, then another domain was checked against the
4348 same list, then the first domain was re-checked, the value of $domain_data
4349 after the final check could be wrong. In particular, if the second check
4350 failed, it could be set empty. This bug probably also applied to
4353 41. The strip_trailing_dot option was not being applied to the address given
4354 with the -f command-line option.
4356 42. The code for reading a message's header from the spool was incrementing
4357 $received_count, but never initializing it. This meant that the value was
4358 incorrect (doubled) while delivering a message in the same process in which
4359 it was received. In the most common configuration of Exim, this never
4360 happens - a fresh exec is done - but it can happen when
4361 deliver_drop_privilege is set.
4363 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4364 it now includes up to 150 characters of the unexpected data in the log
4367 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4368 and building data strings. The size of both of these buffers was 10 000
4369 bytes - far larger than anybody would *ever* want, thought I. Needless to
4370 say, somebody hit the limit. I have increased the maximum line length to
4371 20 000 and the maximum data length of concatenated lines to 100 000. I have
4372 also fixed two bugs, because there was no checking on these buffers. Tsk,
4373 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4374 buffer is too small.
4376 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4377 lsearch lookups. Now it does.
4379 46. When parsing a route_list item in a manualroute router, a fixed-length
4380 buffer was used for the list of hosts. I made this 1024 bytes long,
4381 thinking that nobody would ever have a list of hosts that long. Wrong.
4382 Somebody had a whole pile of complicated expansion conditions, and the
4383 string was silently truncated, leading to an expansion error. It turns out
4384 that it is easier to change to an unlimited length (owing to other changes
4385 that have happened since this code was originally written) than to build
4386 structure for giving a limitation error. The length of the item that
4387 expands into the list of hosts is now unlimited.
4389 47. The lsearch lookup could not handle data where the length of text line was
4390 more than 4095 characters. Such lines were truncated, leading to shortened
4391 data being returned. It should now handle lines of any length.
4393 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4394 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4397 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4398 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4399 no longer happens in comment lines. A list of replacements is now placed
4400 at the head of all of the source files, except those whose only change is
4401 to replace PERL_COMMAND in the very first #! line.
4403 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4404 messages on the queue, with a bottom-up merge sort, using code contributed
4405 by Michael Haardt. This should make operations like -bp somewhat faster on
4406 large queues. It won't affect queue runners, except when queue_run_in_order
4409 51. Installed eximstats 1.31 in the distribution.
4411 52. Added support for SRV lookups to the dnslookup router.
4413 53. If an ACL referred to $message_body or $message_body_end, the value was not
4414 reset for any messages that followed in the same SMTP session.
4416 54. The store-handling optimization for building very long strings was not
4417 differentiating between the different store pools. I don't think this
4418 actually made any difference in practice, but I've tidied it.
4420 55. While running the routers to verify a sender address, $sender_address
4421 was still set to the sender address. This is wrong, because when routing to
4422 send a bounce to the sender, it would be empty. Therefore, I have changed
4423 it so that, while verifying a sender address, $sender_address is set to <>.
4424 (There is no change to what happens when verifying a recipient address.)
4426 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4427 target A or AAAA records (if not already returned) without resetting the
4428 qualify_single or search_parents options of the DNS resolver. These are
4429 inappropriate in this case because the targets of MX and SRV records must
4430 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4431 target that, when qualified, matched something in the local domain. These
4432 two options are now turned off when doing these lookups.
4434 57. It seems that at least some releases of Reiserfs (which does not have the
4435 concept of a fixed number of inodes) returns zero and not -1 for the
4436 number of available inodes. This interacted badly with check_spool_inodes,
4437 which assumed that -1 was the "no such thing" setting. What I have done is
4438 to check that the total number of inodes is greater than zero before doing
4439 the test of how many are available.
4441 58. When a "warn" ACL statement has a log_message modifier, the message is
4442 remembered, and not repeated. This is to avoid a lot of repetition when a
4443 message has many recipients that cause the same warning to be written.
4444 Howewer, Exim was preserving the list of already written lines for an
4445 entire SMTP session, which doesn't seem right. The memory is now reset if a
4446 new message is started.
4448 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4449 debugging output unless log_rewrite was also set.
4451 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4452 of (int)(handle) when we know that handle contains (void *)(-1).
4454 61. The Exim daemon panic-logs an error return when it closes the incoming
4455 connection. However "connection reset by peer" seems to be common, and
4456 isn't really an error worthy of noting specially, so that particular error
4459 62. When Exim is trying to find all the local interfaces, it used to panic and
4460 die if the ioctl to get the interface flags failed. However, it seems that
4461 on at least one OS (Solaris 9) it is possible to have an interface that is
4462 included in the list of interfaces, but for which you get a failure error
4463 for this call. This happens when the interface is not "plumbed" into a
4464 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4465 failure of the "get flags" call assumes that the interface is down.
4467 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4468 makes life easier for people who are doing arithmetic on fields extracted
4469 from dates, where you often get leading zeros that should not be
4470 interpreted as octal.
4472 64. Added qualify_domain to the redirect router, to override the global
4475 65. If a pathologically long header line contained very many addresses (the
4476 report of this problem mentioned 10 000) and each of them was rewritten,
4477 Exim could use up a very large amount of memory. (It kept on making new
4478 copies of the header line as it rewrote, and never released the old ones.)
4479 At the expense of a bit more processing, the header rewriting function has
4480 been changed so that it no longer eats memory in this way.
4482 66. The generation of the Received: header has been moved from the time that a
4483 message starts to be received, to the time that it finishes. The timestamp
4484 in the Received: header should now be very close to that of the <= log
4485 line. There are two side-effects of this change:
4487 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4488 logged header lines no longer include the local Received: line, because
4489 it has not yet been created. The same applies to a copy of the message
4490 that is returned to a non-SMTP sender when a message is rejected.
4492 (b) When a filter file is tested using -bf, no additional Received: header
4493 is added to the test message. After some thought, I decided that this
4496 This change does not affect the value of $received_for. It is still set
4497 after address rewriting, but before local_scan() is called.
4499 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4501 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4502 gave an unhelpful panic error message, and a defer error. I have managed to
4503 change this behaviour so that it now rejects any supplied certificate,
4504 which seems right, as the list of acceptable certificates is empty.
4506 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4507 gave an unhelpful defer error. I have not managed to make this reject any
4508 supplied certificates, but the error message it gives is "no certificate
4509 supplied", which is not helpful.
4511 70. exigrep's output now also includes lines that are not associated with any
4512 message, but which match the given pattern. Implemented by a patch from
4513 Martin Sluka, which also tidied up the Perl a bit.
4515 71. Recipient callout verification, like sender verification, was using <> in
4516 the MAIL FROM command. This isn't really the right thing, since the actual
4517 sender may affect whether the remote host accepts the recipient or not. I
4518 have changed it to use the actual sender in the callout; this means that
4519 the cache record is now keyed on a recipient/sender pair, not just the
4520 recipient address. There doesn't seem to be a real danger of callout loops,
4521 since a callout by the remote host to check the sender would use <>.
4522 [SEE ABOVE: changed after hitting problems.]
4524 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4525 temporary errors. However, in the case of such a code being given after
4526 the end of a data transmission (i.e. after ".") Exim was failing to write
4527 a retry record for the message. (Yes, there was some broken host that was
4528 actually sending 8xx at this point.)
4530 73. An unknown lookup type in a host list could cause Exim to panic-die when
4531 the list was checked. (An example that provoked this was putting <; in the
4532 middle of a list instead of at the start.) If this happened during a DATA
4533 ACL check, a -D file could be left lying around. This kind of configuration
4534 error no longer causes Exim to die; instead it causes a defer errror. The
4535 incident is still logged to the main and panic logs.
4537 74. Buglet left over from Exim 3 conversion. The message "too many messages
4538 in one connection" was written to the rejectlog but not the mainlog, except
4539 when address rewriting (yes!) was being logged.
4541 75. Added write_rejectlog option.
4543 76. When a system filter was run not as root (that is, when system_filter_user
4544 was set), the values of the $n variables were not being returned to the
4545 main process; thus, they were not subsequently available in the $sn
4548 77. Added +return_path_on_delivery log selector.
4550 78. A connection timeout was being treated differently from recipients deferred
4551 when testing hosts_max_try with a message that was older than the host's
4552 retry timeout. (The host should not be counted, thus allowing all hosts to
4553 be tried at least once before bouncing.) This may have been the cause of an
4554 occasionally reported bug whereby a message would remain on the queue
4555 longer than the retry timeout, but would be bounced if a delivery was
4556 forced. I say "may" because I never totally pinned down the problem;
4557 setting up timeout/retry tests is difficult. See also the next item.
4559 79. The ultimate address timeout was not being applied to errors that involved
4560 a combination of host plus message (for example, a timeout on a MAIL
4561 command). When an address resolved to a number of possible hosts, and they
4562 were not all tried for each delivery (e.g. because of hosts_max_try), a
4563 message could remain on the queue longer than the retry timeout.
4565 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4568 81. Fixed an obscure SMTP outgoing bug which required at least the following
4569 conditions: (a) there was another message waiting for the same server;
4570 (b) the server returned 5xx to all RCPT commands in the first message so
4571 that the message was not completed; (c) the server dropped the connection
4572 or gave a negative response to the RSET that Exim sends to abort the
4573 transaction. The observed case was a dropped connection after DATA that had
4574 been sent in pipelining mode. That is, the server had advertised PIPELINING
4575 but was not implementing it correctly. The effect of the bug was incorrect
4576 behaviour, such as trying another host, and this could lead to a crash.
4582 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4583 and daemon.c were passed as pointers to ints; they should have been
4584 pointers to socklen_t variables (which are typically unsigned ints).
4586 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4589 3. Fixed a really odd bug that affected only the testing scheme; patching a
4590 certain fixed string in the binary changed the value of another string that
4591 happened to be identical to the end of the original first string.
4593 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4594 name", it returns that address as the IP address. On some operating
4595 systems (e.g. Solaris), it also passes back the IP address string as the
4596 "host name". However, on others (e.g. Linux), it passes back an empty
4597 string. Exim wasn't checking for this, and was changing the host name to an
4598 empty string, assuming it had been canonicized.
4600 5. Although rare, it is permitted to have more than one PTR record for a given
4601 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4602 all the names associated with an address, because they do in Solaris.
4603 However, it seems that they do not in Linux for data that comes from the
4604 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4605 I found this out when I moved to a new Linux workstation and tried to run
4606 the Exim test suite.
4608 To get round this problem I have changed the code so that it now does its
4609 own call to the DNS to look up PTR records when searching for a host name.
4610 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4611 addresses that are only in /etc/hosts are still found.
4613 This behaviour is, however, controlled by an option called host_lookup_
4614 order, which defaults to "bydns:byaddr". If people want to use the other
4615 order, or indeed, just use one or the other means of lookup, they can
4616 specify it in this variable.
4618 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4619 some operating systems, this comes back from gethostbyaddr() as an empty
4620 string, and this is what Exim used to test for. However, it seems that in
4621 other systems, "." is yielded. Exim now tests for this case too.
4623 7. The values of check_spool_space and check_log_space are now held internally
4624 as a number of kilobytes instead of an absolute number of bytes. If a
4625 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4626 kilobyte. This means that much larger values can be stored.
4628 8. Exim monitor: an attempt to get the action menu when not actually pointing
4629 at a message produces an empty menu entitled "No message selected". This
4630 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4631 no entries in it ("Shell widget menu has zero width and/or height"). So I
4632 have added a single, blank menu entry in this case.
4634 9. Added ${quote_local_part.
4636 10. MIME decoding is now applied to the contents of Subject: header lines when
4639 11. Now that a reference to $sender_host_address automatically causes a reverse
4640 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4641 host lookup before query-style lookups in lists that might use this
4642 variable. This has therefore been abolished, and the "net-" prefix is no
4643 longer necessary for query-style lookups.
4645 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4646 have been a typo for LFLAGS, so it has been changed.
4648 13. The install script calls Exim with "-C /dev/null" in order to find the
4649 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4650 to be output. Howeve, since Exim outputs its version number before the
4651 error, it didn't break the script. It just looked ugly. I fixed this by
4652 always allowing "-C /dev/null" if the caller is root.
4654 14. Ignore overlarge ACL variable number when reading spool file - insurance
4655 against a later release with more variables having written the file.
4657 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4658 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4659 this, as well as the form without the "IPv6" on the front.
4661 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4662 OS/Makefile-Darwin file.
4664 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4665 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4667 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4668 variable. However, some OS use ints for inodes. I've added cast to long int
4669 to get rid of the compiler warning.
4671 19. I had forgotten to lock out "/../" in configuration file names when
4672 ALT_CONFIG_PREFIX was set.
4674 20. Routers used for verification do not need to specify transports. However,
4675 if such a router generated a host list, and callout was configured, Exim
4676 crashed, because it could not find a port number from the (non-existent)
4677 transport. It now assumes port 25 in this circumstance.
4679 21. Added the -t option to exigrep.
4681 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4682 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4683 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4684 them is set, LOOKUP_LSEARCH is forced.
4686 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4688 24. Added sender and host information to the "rejected by local_scan()" log
4689 line; previously there was no indication of these.
4691 25. Added .include_if_exists.
4693 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4694 Linux. It turns out that not all file systems support this. Apparently some
4695 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4696 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4697 not supported on the file descriptor, is now ignored when Exim is trying to
4698 sync a directory. This applies only to Linux.
4700 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4702 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4703 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4706 29. Added support for maildirsize files from supplied patch (modified a bit).
4708 30. The use of :fail: followed by an empty string could lead Exim to respond to
4709 sender verification failures with (e.g.):
4711 550 Verification failed for <xxx>
4712 550 Sender verify failed
4714 where the first response line was missing the '-' that indicates it is not
4715 the final line of the response.
4717 31. The loop for finding the name of the user that called Exim had a hardwired
4718 limit of 10; it now uses the value of finduser_retries, which is used for
4719 all other user lookups.
4721 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4724 33. Exim was neglecting to zero errno before one call of strtol() when
4725 expanding a string and expecting an integer value. On some systems this
4726 resulted in spurious "integer overflow" errors. Also, it was casting the
4727 result into an int without checking.
4729 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4730 did not work. The code looks as if it has *never* worked, though it appears
4731 to have been documented since at least releast 1.62. I have made it work.
4733 35. The "timeout_DNS" error in retry rules, also documented since at least
4734 1.62, also never worked. As it isn't clear exactly what this means, and
4735 clearly it isn't a major issue, I have abolished the feature by treating it
4736 as "timeout", and writing a warning to the main and panic logs.
4738 36. The display of retry rules for -brt wasn't always showing the error code
4741 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4742 timeout_connect_A, timeout_connect_MX.
4744 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4745 to the empty sender.
4747 39. The daemon was not analysing the content of -oX till after it had closed
4748 stderr and disconnected from the controlling terminal. This meant that any
4749 syntax errors were only noted on the panic log, and the return code from
4750 the command was 0. By re-arranging the code a little, I've made the
4751 decoding happen first, so such errors now appear on stderr, and the return
4752 code is 1. However, the actual setting up of the sockets still happens in
4753 the disconnected process, so errors there are still only recorded on the
4756 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4757 connections (as happens on some IP stacks) was logged at start up time as
4758 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4759 it from "IPv6 and IPv4", which means that two separate sockets are being
4762 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4763 says whether AF_INET or AF_INET6 was passed as an argument.
4765 42. Exiwhat output was messed up when time zones were included in log
4768 43. Exiwhat now gives more information about the daemon's listening ports,
4769 and whether -tls-on-connect was used.
4771 44. The "port" option of the smtp transport is now expanded.
4773 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4774 silently ignored. Now an error message is written to the main and panic
4777 46. There's a new ACL modifier called "logwrite" which writes to a log file
4778 as soon as it is encountered.
4780 47. Added $local_user_uid and $local_user_gid at routing time.
4782 48. Exim crashed when trying to verify a sender address that was being
4785 49. Exim was recognizing only a space character after ".include". It now also
4786 recognizes a tab character.
4788 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4789 extracting the relevant information from the specification. The man page no
4790 longer contains scrambled data for the -d option, and I've added a section
4791 at the front about calling Exim under different names.
4793 51. Added "extra_headers" argument to the "mail" command in filter files.
4795 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4798 53. Installed eximstats 1.29.
4800 54. Added transport_filter_timeout as a generic transport option.
4802 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4803 Cc: header lines. This was required by RFC 822, but it not required by RFC
4806 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4807 incoming messages that did not have them. Now it does so only if the
4808 message originates locally, that is, if there is no associated remote host
4809 address. When Resent- header lines are present, this applies to the Resent-
4810 lines rather than the non-Resent- lines.
4812 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4813 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4815 58. Messages for configuration errors now include the name of the main
4816 configuration file - useful now that there may be more than one file in a
4817 list (.included file names were always shown).
4819 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4820 for those rare installations that do not start the daemon as root or run it
4821 setuid root. I've cut out the call to initgroups() if the daemon is not
4824 60. The Exim user and group can now be bound into the binary as text strings
4825 that are looked up at the start of Exim's processing.
4827 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4829 62. Added $mailstore_basename variable.
4831 63. Installed patch to sieve.c from Michael Haardt.
4833 64. When Exim failed to open the panic log after failing to open the main log,
4834 the original message it was trying to log was written to stderr and debug
4835 output, but if they were not available (the usual case in production), it
4836 was lost. Now it is written to syslog before the two lines that record the
4837 failures to open the logs.
4839 65. Users' Exim filters run in subprocesses under the user's uid. It is
4840 possible for a "deliver" command or an alias in a "personal" command to
4841 provoke an address rewrite. If logging of address rewriting is configured,
4842 this fails because the process is not running as root or exim. There may be
4843 a better way of dealing with this, but for the moment (because 4.30 needs
4844 to be released), I have disabled address rewrite logging when running a
4845 filter in a non-root, non-exim process.
4851 1. The buildconfig auxiliary program wasn't quoting the value set for
4852 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4853 not defined. This bug was masked in 4.22 by the effect that was fixed in
4856 2. Some messages that were rejected after a message id was allocated were
4857 shown as "incomplete" by exigrep. It no longer does this for messages that
4858 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4860 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4861 have allow_domain_literals set, the ID did not get logged in the <= line.
4862 Domain literals are now always recognized in Message-ID: header lines.
4864 4. The first argument for a ${extract expansion item is the key name or field
4865 number. Leading and trailing spaces in this item were not being ignored,
4866 causing some misleading effects.
4868 5. When deliver_drop_privilege was set, single queue runner processes started
4869 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4870 same command in the process it spins off) were not dropping privilege.
4872 6. When the daemon running as "exim" started a queue runner, it always
4873 re-executed Exim in the spun-off process. This is a waste of effort when
4874 deliver_drop_privilege is set. The new process now just calls the
4875 queue-runner function directly.
4881 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4884 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4885 ignored. Though the use of -oP was forcing the writing of a pid file, it
4886 was always written to the default place.
4888 3. If the message "no IP address found for host xxxx" is generated during
4889 incoming verification, it is now followed by identification of the incoming
4890 connection (so you can more easily find what provoked it).
4892 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4894 5. Added some features to "harden" Exim a bit more against certain attacks:
4896 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4897 be put in Local/Makefile. This is like the never_users runtime option,
4898 but it cannot be overridden. The default setting is "root".
4900 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4901 prefix string with which any file named in a -C command line option
4904 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4905 is retained for -C and -D only if the caller of Exim is root. Without
4906 it, the exim user may also use -C and -D and retain privilege.
4908 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4909 command line option is disabled.
4911 6. Macro names set by the -D option must start with an upper case letter, just
4912 like macro names defined in the configuration file.
4914 7. Added "dereference=" facility to LDAP.
4916 8. Two instances of the typo "uknown" in the source files are fixed.
4918 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4919 the Configure-Makefile script screwed up while processing it.
4921 10. Incorporated PCRE 4.4.
4923 11. The SMTP synchronization check was not operating right at the start of an
4924 SMTP session. For example, it could not catch a HELO sent before the client
4925 waited for the greeting. There is now a check for outstanding input at the
4926 point when the greeting is written. Because of the duplex, asynchronous
4927 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4928 way, but not yet received, when the check is performed.
4930 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4931 on TCP/IP sockets, because this apparently causes some broken clients to
4934 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4935 unchanged) from the Cygwin maintainer.
4937 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4938 was supported instead of testing for mbx. Effectively a typo.
4940 15. The spa authenticator server code was not checking that the input it
4941 received was valid base64.
4943 16. The debug output line for the "set" modifier in ACLs was not showing the
4944 name of the variable that was being set.
4946 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4948 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4949 lookup on its own. Something else was needed to trigger the lookup. For
4950 example, a match in host_lookup or the need for a host name in a host list.
4951 Now, if $sender_host_name is referenced and the host name has not yet been
4952 looked up, a lookup is performed. If the lookup fails, the variable remains
4953 empty, and $host_lookup_failed is set to "1".
4955 19. Added "eqi" as a case-independent comparison operator.
4957 20. The saslauthd authentication condition could segfault if neither service
4958 nor realm was specified.
4960 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4961 error message that was logged was misleading, and incoming SMTP
4962 connections were dropped. The message is now more accurate, and temporary
4963 errors are given to SMTP connections.
4965 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4966 (except RSET) with a 421 error, until QUIT is received. However, it was
4967 failing to send a response to QUIT.
4969 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4970 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4971 ACL is now run after the helo_try_verify_hosts code.
4973 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4974 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4975 case-independent, so other case variants are also recognized). Apparently
4976 some systems use these upper case variants.
4978 25. If more than two messages were waiting for the same host, and a transport
4979 filter was specified for the transport, Exim sent two messages over the
4980 same TCP/IP connection, and then failed with "socket operation on non-
4981 socket" when it tried to send the third.
4983 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4985 27. The extern definition of crypt16() in expand.c was not being excluded when
4986 the OS had its own crypt16() function.
4988 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4989 as a preferred synonym for return_size_limit, both as an option and as an
4992 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4994 30. Changed the default configuration ACL to relax the local part checking rule
4995 for addresses that are not in any local domains. For these addresses,
4996 slashes and pipe symbols are allowed within local parts, but the sequence
4997 /../ is explicitly forbidden.
4999 31. SPA server authentication was not clearing the challenge buffer before
5002 32. log_message in a "warn" ACL statement was writing to the reject log as
5003 well as to the main log, which contradicts the documentation and doesn't
5004 seem right (because no rejection is happening). So I have stopped it.
5006 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5007 However, I am unable to do any testing of this.
5009 34. Fixed an infelicity in the appendfile transport. When checking directories
5010 for a mailbox, to see if any needed to be created, it was accidentally
5011 using path names with one or more superfluous leading slashes; tracing
5012 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5014 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5015 contents are added to the log line that is written for every discarded
5016 recipient. (Previously a log_message setting was ignored.)
5018 36. The ${quote: operator now quotes the string if it is empty.
5020 37. The install script runs exim in order to find its version number. If for
5021 some reason other than non-existence or emptiness, which it checks, it
5022 could not run './exim', it was installing it with an empty version number,
5023 i.e. as "exim-". This error state is now caught, and the installation is
5026 38. An argument was missing from the function that creates an error message
5027 when Exim fails to connect to the socket for saslauthd authentication.
5028 This could cause Exim to crash, or give a corrupted message.
5030 39. Added isip, isip4, and isip6 to ${if conditions.
5032 40. The ACL variables $acl_xx are now saved with the message, and can be
5033 accessed later in routers, transports, and filters.
5035 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5036 strings in the file are not string-expanded.
5038 42. If a MAIL command specified a SIZE value that was too large to fit into an
5039 int variable, the check against message_size_limit failed. Such values are
5040 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5041 day this will have to be increased, but I don't think I want to be around
5042 when emails are that large.
5049 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5050 iconv() is not standard in FreeBSD.
5052 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5053 IPv6 enabled. The observed symptom was a segmentation fault on return from
5054 the function os_common_find_running_interfaces() in src/os.c.
5056 3. In the check_special_case() function in daemon.c I had used "errno" as an
5057 argument name, which causes warnings on some systems. This was basically a
5058 typo, since it was named "eno" in the comments!
5060 4. The code that waits for the clock to tick (at a resolution of some fraction
5061 of a second) so as to ensure message-id uniqueness was always waiting for
5062 at least one whole tick, when it could have waited for less. [This is
5063 almost certainly not relevant at current processor speeds, where it is
5064 unlikely to ever wait at all. But we try to future-proof.]
5066 5. The function that sleeps for a time interval that includes fractions of a
5067 second contained a race. It did not block SIGALRM between setting the
5068 timer, and suspending (a couple of lines later). If the interval was short
5069 and the sigsuspend() was delayed until after it had expired, the suspension
5070 never ended. On busy systems this could lead to processes getting stuck for
5073 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5074 process, before it forks any delivery processes. The open lookup caching
5075 mechanism meant that the open file or database connection was passed into
5076 the delivery process. The problem was that delivery processes always tidy
5077 up cached lookup data. This could cause a problem for the next delivery
5078 process started by the queue runner, because the external queue runner
5079 process does not know about the closure. So the next delivery process
5080 still has data in the lookup cache. In the case of a file lookup, there was
5081 no problem because closing a file descriptor in a subprocess doesn't affect
5082 the parent. However, if the lookup was caching a connection to a database,
5083 the connection was closed, and the second delivery process was likely to
5084 see errors such as "PGSQL: query failed: server closed the connection
5085 unexpectedly". The problem has been fixed by closing all cached lookups
5086 in a queue runner before running a delivery process.
5088 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5089 seem to have the "const" qualifier which it has on other OS. I've
5092 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5093 *of the same type* (client or server) with the same public name that an
5094 error should be diagnosed.
5096 9. When Exim looked up a host name for an IP address, but failed to find the
5097 original IP address when looking up the host name (a safety check), it
5098 output the message "<ip address> does not match any IP for NULL", which was
5099 confusing, to say the least. The bug was that the host name should have
5100 appeared instead of "NULL".
5102 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5103 user that is built into the binary, and the -C or -D options is used, root
5104 privilege is dropped before the configuration file is read. In addition,
5105 logging is switched to stderr instead of the normal log files. If the
5106 configuration then re-defines the Exim user, the unprivileged environment
5107 is probably not what is expected, so Exim logs a panic warning message (but
5110 However, if deliver_drop_privilege is set, the unprivileged state may well
5111 be exactly what is intended, so the warning has been cut out in that case,
5112 and Exim is allowed to try to write to its normal log files.
5118 1. smtp_return_error_details was not giving details for temporary sender
5119 or receiver verification errors.
5121 2. Diagnose a configuration error if two authenticators have the same public
5124 3. Exim used not to create the message log file for a message until the first
5125 delivery attempt. This could be confusing when incoming messages were held
5126 for policy or load reasons. The message log file is now created at the time
5127 the message is received, and an initial "Received" line is written to it.
5129 4. The automatically generated man page for command line options had a minor
5130 bug that caused no ill effects; however, a more serious problem was that
5131 the procedure for building the man page automatically didn't always
5132 operate. Consequently, release 4.20 contains an out-of-date version. This
5133 shouldn't happen again.
5135 5. When building Exim with embedded Perl support, the script that builds the
5136 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5137 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5139 6. The freeze_tell option was not being used for messages that were frozen on
5140 arrival, either by an ACL or by local_scan().
5142 7. Added the smtp_incomplete_transaction log selector.
5144 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5145 was accepting AUTH without a new EHLO.
5147 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5148 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5151 10. Logging of TCP/IP connections (when configured) now happens in the main
5152 daemon process instead of the child process, so that the TCP/IP connection
5153 count is more accurate (but it can never be perfect).
5155 11. The use of "drop" in a nested ACL was not being handled correctly in the
5156 outer ACL. Now, if condition failure induced by the nested "drop" causes
5157 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5158 or "require"), the connection is dropped.
5160 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5161 that yield "discard" can now be used with an "accept" or a "discard" verb,
5162 but an error is generated for any others (because I can't see a useful way
5163 to define what should happen).
5165 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5166 are now processed in the same way as lines in the Exim configuration file.
5167 In particular, continuation lines are supported.
5169 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5171 15. Added -ti meaning -t -i.
5173 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5174 domains, and warn by logging if others are found.
5176 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5177 returned from GIFCONF when Exim is trying to find the list of interfaces on
5178 a host. The code in os.c has been modified to copy each ifreq to an aligned
5179 structure in all cases.
5181 Also, in some cases, the returned ifreq's were being copied to a 'struct
5182 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5183 means the last couple of bytes of an IPv6 address could be chopped if the
5184 ifreq contained only a normal sockaddr (14 bytes storage).
5186 18. Named domain lists were not supported in the hosts_treat_as_local option.
5187 An entry such as +xxxx was not recognized, and was treated as a literal
5190 19. Ensure that header lines added by a DATA ACL are included in the reject log
5191 if the ACL subsequently rejects the message.
5193 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5194 MD5 (which is deprecated).
5196 21. When testing a filter file using -bf, Exim was writing a message when it
5197 took the sender from a "From " line in the message, but it was not doing so
5198 when it took $return_path from a Return-Path: header line. It now does.
5200 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5201 with a valid header line field name (a series of printing characters
5202 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5204 23. Changed "disc" in the source to "disk" to conform to the documentation and
5205 the book and for uniformity.
5207 24. Ignore Sendmail's -Ooption=value command line item.
5209 25. When execve() failed while trying to run a command in a pipe transport,
5210 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5211 could be confused with a return value of 69 from the command itself. This
5212 has been changed to 127, the value the shell returns if it is asked to run
5213 a non-existent command. The wording for the related log line suggests a
5214 non-existent command as the problem.
5216 26. If received_header_text expands to an empty string, do not add a Received:
5217 header line to the message. (Well, it adds a token one on the spool, but
5218 marks it "old" so that it doesn't get used or transmitted.)
5220 27. Installed eximstats 1.28 (addition of -nt option).
5222 28. There was no check for failure on the call to getsockname() in the daemon
5223 code. This can fail if there is a shortage of resources on the system, with
5224 ENOMEM, for example. A temporary error is now given on failure.
5226 29. Contrary to the C standard, it seems that in some environments, the
5227 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5228 program. Exim now does this explicitly; it affects the formatting of
5229 timestamps using strftime().
5231 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5232 complaints. I've now initialized all the variables, to avoid this.
5234 32. Header lines added by a system filter were not being "seen" during
5235 transport-time rewrites.
5237 33. The info_callback() function passed to OpenSSL is set up with type void
5238 (*)(SSL *, int, int), as described somewhere. However, when calling the
5239 function (actually a macro) that sets it up, the type void(*)() is
5240 expected. I've put in a cast to prevent warnings from picky compilers.
5242 34. If a DNS black list lookup found a CNAME record, but there were no A
5243 records associated with the domain it pointed at, Exim crashed.
5245 35. If a DNS black list lookup returned more than one A record, Exim ignored
5246 all but the first. It now scans all returned addresses if a particular IP
5247 value is being sought. In this situation, the contents of the
5248 $dnslist_value variable are a list of all the addresses, separated by a
5251 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5252 to accept a host name and all its aliases if the forward lookup for any of
5253 them yielded the IP address of the incoming connection. Now it accepts only
5254 those names whose forward lookup yields the correct IP address. Any other
5255 names are discarded. This closes a loophole whereby a rogue DNS
5256 administrator could create reverse DNS records to break through a
5257 wildcarded host restriction in an ACL.
5259 37. If a user filter or a system filter that ran in a subprocess used any of
5260 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5261 the wrong values were passed to the pipe command ($thisaddress had the
5262 value of $0, $0 had the value of $1, etc). This bug was introduced by
5263 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5265 38. Improved the line breaking for long SMTP error messages from ACLs.
5266 Previously, if there was no break point between 40 and 75 characters, Exim
5267 left the rest of the message alone. Two changes have been made: (a) I've
5268 reduced the minimum length to 35 characters; (b) if it can't find a break
5269 point between 35 and 75 characters, it looks ahead and uses the first one
5270 that it finds. This may give the occasional overlong line, but at least the
5271 remaining text gets split now.
5273 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5274 file descriptors might be low, and that setting 1000 would always raise it.
5275 It turns out that in some environments, the limit is already over 1000 and
5276 that lowering it causes trouble. So now Exim takes care not to decrease it.
5278 40. When delivering a message, the value of $return_path is set to $sender_
5279 address at the start of routing (routers may change the value). By an
5280 oversight, this default was not being set up when an address was tested by
5281 -bt or -bv, which affected the outcome if any router or filter referred to
5284 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5285 log only when "message" or "log_message" are set. However, if one of the
5286 conditions was an address verification, or a call to a nested ACL, the
5287 messages generated by the underlying test were being passed through. This
5288 no longer happens. The underlying message is available in $acl_verify_
5289 message for both "message" and "log_message" expansions, so it can be
5290 passed through if needed.
5292 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5293 new expansion $bh_ to give the encoded byte string without charset
5294 translation. Translation happens only if iconv() is available; HAVE_ICONV
5295 indicates this at build time. HEADERS_CHARSET gives the charset to
5296 translate to; headers_charset can change it in the configuration, and
5297 "headers charset" can change it in an individual filter file.
5299 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5300 that creates RFC 2047 encoded "words" labels them as that charset instead
5301 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5302 expansion operator; (ii) when Exim creates a From: line for a local
5303 message; (iii) when a header line is rewritten to include a "phrase" part.
5305 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5306 buggy, causing it to skip the first lines of messages whose message ID
5307 ended in 'D'. This would not have bitten before Exim release 4.14, because
5308 message IDs were unlikely to end in 'D' before then. The effect was to have
5309 incorrect size information for certain domains.
5311 45. #include "config.h" was missing at the start of the crypt16.c module. This
5312 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5315 46. If there was a timeout during a "random" callout check, Exim treated it as
5316 a failure of the random address, and carried on sending RSET and the real
5317 address. If the delay was just some slowness somewhere, the response to the
5318 original RCPT would be taken as a response to RSET and so on, causing
5319 mayhem of various kinds.
5321 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5322 when I implemented it. It didn't allow for the fact that some option values
5323 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5324 the right test for positive values.
5326 48. Domain names in DNS records are case-independent. Exim always looks them up
5327 in lower case. Some resolvers return domain names in exactly the case they
5328 appear in the zone file, that is, they may contain uppercase letters. Not
5329 all resolvers do this - some return always lower case. Exim was treating a
5330 change of case by a resolver as a change of domain, similar to a widening
5331 of a domain abbreviation. This triggered its re-routing code and so it was
5332 trying to route what was effectively the same domain again. This normally
5333 caused routing to fail (because the router wouldn't handle the domain
5334 twice). Now Exim checks for this case specially, and just changes the
5335 casing of the domain that it ultimately uses when it transmits the message
5338 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5341 50. If a filter generated a file delivery with a non-absolute name (possible if
5342 no home directory exists for the router), the forbid_file option was not
5345 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5346 the existing equality matching.
5348 52. Exim was using ints instead of ino_t variables in some places where it was
5349 dealing with inode numbers.
5351 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5352 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5353 called TMPDIR, and if it finds it is different, it changes its value.
5355 54. The smtp_printf() function is now made available to local_scan() so
5356 additional output lines can be written before returning. There is also an
5357 smtp_fflush() function to enable the detection of a dropped connection.
5358 The variables smtp_input and smtp_batched_input are exported to
5361 55. Changed the default runtime configuration: the message "Unknown user"
5362 has been removed from the ACL, and instead placed on the localuser router,
5363 using the cannot_route_message feature. This means that any verification
5364 failures that generate their own messages won't get overridden. Similarly,
5365 the "Unrouteable address" message that was in the ACL for unverifiable
5366 relay addresses has also been removed.
5368 56. Added hosts_avoid_esmtp to the smtp transport.
5370 57. The exicyclog script was not checking for the esoteric option
5371 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5372 will work only if exicyclog is run under the appropriate euid.
5374 58. Following a discussion on the list, the rules by which Exim recognises line
5375 endings on incoming messages have been changed. The -dropcr and drop_cr
5376 options are now no-ops, retained only for backwards compatibility. The
5377 following line terminators are recognized: LF CRLF CR. However, special
5378 processing applies to CR:
5380 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5381 nor a local message in the state where . is a terminator.
5383 (ii) If a bare CR is encountered in a header line, an extra space is added
5384 after the line terminator so as not to end the header. The reasoning
5385 behind this is that bare CRs in header lines are most likely either
5386 to be mistakes, or people trying to play silly games.
5388 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5389 was being incorrectly given as 18 bytes larger than it should have been.
5390 This is a VOB (very old bug).
5392 60. This may never have affected anything current, but just in case it has:
5393 When the local host is found other than at the start of a list of hosts,
5394 the local host, those with the same MX, and any that follow, are discarded.
5395 When the list in question was part of a longer list of hosts, the following
5396 hosts (not currently being processed) were also being discarded. This no
5397 longer happens. I'm not sure if this situation could ever has previously
5400 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5403 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5404 header. This is something that is recommended in a new Internet Draft, and
5405 is something that is documented as being done by Sendmail. There are two
5406 possible values. For messages generated by the autoreply transport, Exim
5409 Auto-Submitted: auto-replied
5411 whereas for all other generated messages (e.g. bounces) it adds
5413 Auto-Submitted: auto-generated
5415 63. The "personal" condition in filters now includes a test for the
5416 Auto-Submitted: header. If it contains the string "auto-" the message it
5417 not considered personal.
5419 64. Added rcpt_include_affixes as a generic transport option.
5421 65. Added queue_only_override (default true).
5423 66. Added the syslog_duplication option.
5425 67. If what should have been the first header line of a message consisted of
5426 a space followed by a colon, Exim was mis-interpreting it as a header line.
5427 It isn't of course - it is syntactically invalid and should therefore be
5428 treated as the start of the message body. The misbehaviour could have
5429 caused a number of strange effects, including loss of data in subsequent
5430 header lines, and spool format errors.
5432 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5433 client host had authenticated. This control can now be exercised by an ACL
5434 for more flexibility.
5436 69. By default, callouts do not happen when testing with -bh. There is now a
5437 variant, -bhc, which does actually run the callout code, including
5438 consulting and updating the callout cache.
5440 70. Added support for saslauthd authentication, courtesy of Alexander
5443 71. If statvfs() failed on the spool or log directories while checking their
5444 size for availability, Exim confusingly gave the error "space shortage".
5445 Furthermore, in debugging mode it crashed with a floating point exception.
5446 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5447 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5448 serious problem, Exim now writes to the main and panic logs when this
5449 happens, with details of the failure. It then refuses to accept the
5450 incoming message, giving the message "spool directory problem" or "log
5451 directory problem" with a 421 code for SMTP messages.
5453 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5454 0, 1, and 2 exist, because some OS complain for execs without them (see
5455 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5456 descriptors. However, the code omitted to check that the open succeeded,
5457 causing mysterious errors if for some reason the permissions on /dev/null
5458 got screwed. Now Exim writes a message to the main and panic logs, and
5459 bombs out if it can't open /dev/null.
5461 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5462 interact so that it is all more flexible. It is supposed to remain
5463 backwards compatible. Also added extra_local_interfaces.
5465 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5466 to bomb out with an assertion failure - to the client this appears as a
5467 connection drop. This problem occurs in the part of the code that was taken
5468 from the Samba project. Fortunately, the assertion is in a very simple
5469 function, so I have fixed this by reproducing the function inline in the
5470 one place where it is called, and arranging for authentication to fail
5471 instead of killing the process with assert().
5473 75. The SPA client code was not working when the server requested OEM rather
5474 than Unicode encoding.
5476 76. Added code to make require_files with a specific uid setting more usable in
5477 the case where statting the file as root fails - usually a non-root-mounted
5478 NFS file system. When this happens and the failure is EACCES, Exim now
5479 forks a subprocess and does the per-uid checking as the relevant uid.
5481 77. Added process_log_path.
5483 78. If log_file_path was not explicitly set, a setting of check_log_space or
5484 check_log_inodes was ignored.
5486 79. If a space check for the spool or log partitions fails, the incident is now
5487 logged. Of course, in the latter case the data may get lost...
5489 80. Added the %p formatting code to string_format() so that it can be used to
5490 print addresses in debug_print(). Adjusted all the address printing in the
5491 debugging in store.c to use %p rather than %d.
5493 81. There was a concern that a line of code in smtp_in.c could overflow a
5494 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5495 initially expressed, the concern was not well-founded, because trailing
5496 spaces are removed early. However, if the trailing spaces were followed by
5497 a NULL, they did not get removed, so the overflow was possible. Two fixes
5500 (a) I re-wrote the offending code in a cleaner fashion.
5501 (b) If an incoming SMTP command contains a NULL character, it is rejected
5504 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5505 runs initgroups(), so that if the Exim user is in any additional groups,
5506 they will be used during message reception.
5512 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob