1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
72 PP/01 Solaris build fix for Oracle's LDAP libraries.
73 Bugzilla 1109, patch from Stephen Usher.
75 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
77 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
80 TF/02 Fix a couple more cases where we did not log the error message
81 when unlink() failed. See also change 4.74-TF/03.
83 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
84 lock up or crash if it happened to be inside a call to libc when it
85 got a SIGUSR1 from exiwhat.
87 The SIGUSR1 handler appends the current process status to the process
88 log which is later printed by exiwhat. It used to use the general
89 purpose logging code to do this, but several functions it calls are
92 The new output code in the SIGUSR1 handler is specific to the process
93 log, and simple enough that it's easy to inspect for signal safety.
94 Removing some special cases also simplifies the general logging code.
95 Removing the spurious timestamps from the process log simplifies
98 TF/04 Improved ratelimit ACL condition.
100 The /noupdate option has been deprecated in favour of /readonly which
101 has clearer semantics. The /leaky, /strict, and /readonly update modes
102 are mutually exclusive. The update mode is no longer included in the
103 database key; it just determines when the database is updated. (This
104 means that when you upgrde Exim will forget old rate measurements.)
106 Exim now checks that the per_* options are used with an update mode that
107 makes sense for the current ACL. For example, when Exim is processing a
108 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
109 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
110 must specify per_mail/readonly. If you omit the update mode it defaults to
111 /leaky where that makes sense (as before) or /readonly where required.
113 The /noupdate option is now undocumented but still supported for
114 backwards compatibility. It is equivalent to /readonly except that in
115 ACLs where /readonly is required you may specify /leaky/noupdate or
116 /strict/noupdate which are treated the same as /readonly.
118 A useful new feature is the /count= option. This is a generalization
119 of the per_byte option, so that you can measure the throughput of other
120 aggregate values. For example, the per_byte option is now equivalent
121 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
123 The per_rcpt option has been generalized using the /count= mechanism
124 (though it's more complicated than the per_byte equivalence). When it is
125 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
126 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
127 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
128 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
129 Note that using per_rcpt with a non-readonly update mode in more than
130 one ACL will cause the recipients to be double-counted. (The per_mail
131 and per_byte options don't have this problem.)
133 The handling of very low rates has changed slightly. If the computed rate
134 is less than the event's count (usually one) then this event is the first
135 after a long gap. In this case the rate is set to the same as this event's
136 count, so that the first message of a spam run is counted properly.
138 The major new feature is a mechanism for counting the rate of unique
139 events. The new per_addr option counts the number of different
140 recipients that someone has sent messages to in the last time period. It
141 behaves like per_rcpt if all the recipient addresses are different, but
142 duplicate recipient addresses do not increase the measured rate. Like
143 the /count= option this is a general mechanism, so the per_addr option
144 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
145 example, measure the rate that a client uses different sender addresses
146 with the options per_mail/unique=$sender_address. There are further
147 details in the main documentation.
149 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
151 TF/06 Removed a few PCRE remnants.
153 TF/07 Automatically extract Exim's version number from tags in the git
154 repository when doing development or release builds.
156 PP/02 Raise smtp_cmd_buffer_size to 16kB.
157 Bugzilla 879. Patch from Paul Fisher.
159 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
160 Heavily based on revision 40f9a89a from Simon Arlott's tree.
163 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
165 PP/05 Variable $av_failed, true if the AV scanner deferred.
166 Bugzilla 1078. Patch from John Horne.
168 PP/06 Stop make process more reliably on build failure.
169 Bugzilla 1087. Patch from Heiko Schlittermann.
171 PP/07 Make maildir_use_size_file an _expandable_ boolean.
172 Bugzilla 1089. Patch from Heiko Schlittermann.
174 PP/08 Handle ${run} returning more data than OS pipe buffer size.
175 Bugzilla 1131. Patch from Holger Weiß.
177 PP/09 Handle IPv6 addresses with SPF.
178 Bugzilla 860. Patch from Wolfgang Breyha.
180 PP/10 GnuTLS: support TLS 1.2 & 1.1.
182 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
185 PP/11 match_* no longer expand right-hand-side by default.
186 New compile-time build option, EXPAND_LISTMATCH_RHS.
187 New expansion conditions, "inlist", "inlisti".
189 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
191 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
193 PP/14 fix log_write() format string regression from TF/03.
194 Bugzilla 1152. Patch from Dmitry Isaikin.
200 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
202 PP/02 Harmonised TLS library version reporting; only show if debugging.
203 Layout now matches that introduced for other libraries in 4.74 PP/03.
205 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
207 PP/04 New "dns_use_edns0" global option.
209 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
212 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
213 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
215 TK/01 Updated PolarSSL code to 0.14.2.
216 Bugzilla 1097. Patch from Andreas Metzler.
218 PP/07 Catch divide-by-zero in ${eval:...}.
221 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
224 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
225 format-string attack -- SECURITY: remote arbitrary code execution.
227 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
228 time unintentionally subject to list matching rules, letting the header
229 cause arbitrary Exim lookups (of items which can occur in lists, *not*
230 arbitrary string expansion). This allowed for information disclosure.
232 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
233 INT_MIN/-1 -- value coerced to INT_MAX.
239 NM/01 Workround for PCRE version dependency in version reporting
242 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
243 This fixes portability to compilers other than gcc, notably
244 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
246 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
247 makefiles for portability to HP-UX and POSIX correctness.
249 PP/01 Permit LOOKUP_foo enabling on the make command-line.
250 Also via indented variable definition in the Makefile.
251 (Debugging by Oliver Heesakkers).
253 PP/02 Restore caching of spamd results with expanded spamd_address.
254 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
256 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
257 Improves build reliability. Fix from: Frank Elsner
259 NM/02 Fix wide character breakage in the rfc2047 coding
260 Fixes bug 1064. Patch from Andrey N. Oktyabrski
262 NM/03 Allow underscore in dnslist lookups
263 Fixes bug 1026. Patch from Graeme Fowler
265 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
266 Code patches from Adam Ciarcinski of NetBSD.
268 NM/04 Fixed exiqgrep to cope with mailq missing size issue
271 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
272 is logged, to avoid truncation. Patch from John Horne.
274 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
275 Patch from Jakob Hirsch.
277 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
278 SQL string expansion failure details.
279 Patch from Andrey Oktyabrski.
281 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
282 Patch from Simon Arlott.
284 PP/09 New lookups functionality failed to compile on old gcc which rejects
285 extern declarations in function scope.
286 Patch from Oliver Fleischmann
288 PP/10 Use sig_atomic_t for flags set from signal handlers.
289 Check getgroups() return and improve debugging.
290 Fixed developed for diagnosis in bug 927 (which turned out to be
293 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
294 Patch from Mark Zealey.
296 PP/12 Bugzilla 1056: Improved spamd server selection.
297 Patch from Mark Zealey.
299 PP/13 Bugzilla 1086: Deal with maildir quota file races.
300 Based on patch from Heiko Schlittermann.
302 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
303 Patch from Uwe Doering, sign-off by Michael Haardt.
305 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
306 variable declaration deep within a block. Bug and patch from
309 PP/15 lookups-Makefile IRIX compatibilty coercion.
311 PP/16 Make DISABLE_DKIM build knob functional.
313 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
314 Patch by Simon Arlott
316 TF/03 Fix valgrind.h portability to C89 compilers that do not support
317 variable argument macros. Our copy now differs from upstream.
323 TF/01 Failure to get a lock on a hints database can have serious
324 consequences so log it to the panic log.
326 TF/02 Log LMTP confirmation messages in the same way as SMTP,
327 controlled using the smtp_confirmation log selector.
329 TF/03 Include the error message when we fail to unlink a spool file.
331 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
332 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
333 for maintaining out-of-tree patches for some time.
335 PP/01 Bugzilla 139: Documentation and portability issues.
336 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
337 Handle per-OS dynamic-module compilation flags.
339 PP/02 Let /dev/null have normal permissions.
340 The 4.73 fixes were a little too stringent and complained about the
341 permissions on /dev/null. Exempt it from some checks.
342 Reported by Andreas M. Kirchwitz.
344 PP/03 Report version information for many libraries, including
345 Exim version information for dynamically loaded libraries. Created
346 version.h, now support a version extension string for distributors
347 who patch heavily. Dynamic module ABI change.
349 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
350 privilege escalation vulnerability whereby the Exim run-time user
351 can cause root to append content of the attacker's choosing to
354 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
357 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
358 If dropping privileges for untrusted macros, we disabled normal logging
359 on the basis that it would fail; for the Exim run-time user, this is not
360 the case, and it resulted in successful deliveries going unlogged.
361 Fixed. Reported by Andreas Metzler.
367 PP/01 Date: & Message-Id: revert to normally being appended to a message,
368 only prepend for the Resent-* case. Fixes regression introduced in
369 Exim 4.70 by NM/22 for Bugzilla 607.
371 PP/02 Include check_rfc2047_length in configure.default because we're seeing
372 increasing numbers of administrators be bitten by this.
374 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
376 PP/03 Bugzilla 994: added openssl_options main configuration option.
378 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
380 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
382 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
384 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
385 without a peer certificate, leading to a segfault because of an
386 assumption that peers always have certificates. Be a little more
387 paranoid. Problem reported by Martin Tscholak.
389 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
390 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
391 NB: ClamAV planning to remove STREAM in "middle of 2010".
392 CL also introduces -bmalware, various -d+acl logging additions and
393 more caution in buffer sizes.
395 PP/09 Implemented reverse_ip expansion operator.
397 PP/10 Bugzilla 937: provide a "debug" ACL control.
399 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
401 PP/12 Bugzilla 973: Implement --version.
403 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
405 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
407 PP/15 Bugzilla 816: support multiple condition rules on Routers.
409 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
410 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
411 ignore trailing whitespace.
413 JJ/02 prevent non-panic DKIM error from being sent to paniclog
415 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
418 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
419 Notification from Dr Andrew Aitchison.
421 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
422 ExtendedDetectionInfo response format.
423 Notification from John Horne.
425 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
428 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
429 XSL and documented dependency on system catalogs, with examples of how
432 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
435 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
436 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
437 configuration file which is writeable by the Exim user or group.
439 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
440 of configuration files to cover files specified with the -C option if
441 they are going to be used with root privileges, not just the default
444 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
445 option (effectively making it always true).
447 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
448 files to be used while preserving root privileges.
450 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
451 that rogue child processes cannot use them.
453 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
454 run-time user, instead of root.
456 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
457 Exim run-time user without dropping privileges.
459 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
460 result string, instead of calling string_vformat() twice with the same
463 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
464 for other users. Others should always drop root privileges if they use
465 -C on the command line, even for a whitelisted configure file.
467 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
469 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
475 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
476 $data_path, and $header_path variables; fixed documentation bugs and
479 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
480 exipick to access non-standard spools, including the "frozen" queue
483 NM/01 Bugzilla 965: Support mysql stored procedures.
484 Patch from Alain Williams
486 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
488 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
489 Patch from Andreas Metzler
491 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
492 Patch from Kirill Miazine
494 NM/05 Bugzilla 671: Added umask to procmail example.
496 JJ/03 installed exipick 20100323.0, fixing doc bug
498 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
499 directory. Notification and patch from Dan Rosenberg.
501 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
503 TK/02 Improve log output when DKIM signing operation fails.
505 MH/01 Treat the transport option dkim_domain as a colon separated
506 list, not as a single string, and sign the message with each element,
507 omitting multiple occurences of the same signer.
509 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
510 Bugzilla 985, 986. Patch by Simon Arlott
512 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
513 Patch by Simon Arlott
515 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
516 MBX locking. Notification from Dan Rosenberg.
522 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
524 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
526 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
528 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
530 NM/04 Bugzilla 915: Flag broken perl installation during build.
536 TK/01 Added patch by Johannes Berg that expands the main option
537 "spamd_address" if it starts with a dollar sign.
539 TK/02 Write list of recipients to X-Envelope-Sender header when building
540 the mbox-format spool file for content scanning (suggested by Jakob
543 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
544 (http://www.dcc-servers.net/) support via dccifd. Activated by
545 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
547 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
548 by Mark Daniel Reidel <mr@df.eu>.
550 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
551 When building exim an external PCRE library is now needed -
552 PCRE is a system library on the majority of modern systems.
553 See entry on PCRE_LIBS in EDITME file.
555 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
556 conversation. Added nologin parameter to request.
557 Patch contributed by Kirill Miazine.
559 TF/01 Do not log submission mode rewrites if they do not change the address.
561 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
563 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
564 log files in place. Contributed by Roberto Lima.
566 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
568 TF/03 Bugzilla 615: When checking the local_parts router precondition
569 after a local_part_suffix or local_part_prefix option, Exim now
570 does not use the address's named list lookup cache, since this
571 contains cached lookups for the whole local part.
573 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
574 Robert Millan. Documentation is in experimental-spec.txt.
576 TF/04 Bugzilla 668: Fix parallel build (make -j).
578 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
580 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
581 Patch provided by Jan Srzednicki.
583 TF/05 Leading white space used to be stripped from $spam_report which
584 wrecked the formatting. Now it is preserved.
586 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
587 that they are available at delivery time.
589 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
591 TF/08 TLS error reporting now respects the incoming_interface and
592 incoming_port log selectors.
594 TF/09 Produce a more useful error message if an SMTP transport's hosts
595 setting expands to an empty string.
597 NM/06 Bugzilla 744: EXPN did not work under TLS.
598 Patch provided by Phil Pennock.
600 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
601 Patch provided by Richard Godbee.
603 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
604 acl_smtp_notquit, added index entry.
606 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
607 Patch provided by Eugene Bujak.
609 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
610 accept(). Patch provided by Maxim Dounin.
612 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
613 Patch provided by Phil Pennock.
615 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
617 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
618 Patch provided by Brad "anomie" Jorsch.
620 NM/14 Bugzilla 622: Added timeout setting to transport filter.
621 Patch provided by Dean Brooks.
623 TK/05 Add native DKIM support (does not depend on external libraries).
625 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
626 Patch provided by Graeme Fowler.
628 NM/16 Bugzilla 851: Documentation example syntax fix.
630 NM/17 Changed NOTICE file to remove references to embedded PCRE.
632 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
635 NM/19 Bugzilla 745: TLS version reporting.
636 Patch provided by Phil Pennock.
638 NM/20 Bugzilla 167: bool: condition support.
639 Patch provided by Phil Pennock.
641 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
642 clients. Patch provided by Phil Pennock.
644 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
645 Patch provided by Brad "anomie" Jorsch.
647 NM/23 Bugzilla 687: Fix misparses in eximstats.
648 Patch provided by Heiko Schlittermann.
650 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
651 Patch provided by Heiko Schlittermann.
653 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
654 plus update to original patch.
656 NM/26 Bugzilla 799: Documentation correction for ratelimit.
658 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
659 Patch provided by David Brownlee.
661 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
663 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
665 NM/30 Bugzilla 888: TLS documentation bugfixes.
667 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
669 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
670 Unlike the original bugzilla I have changed all shell scripts in src tree.
672 NM/33 Bugzilla 898: Transport filter timeout fix.
673 Patch by Todd Rinaldo.
675 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
676 Patch by Serge Demonchaux.
678 NM/35 Bugzilla 39: Base64 decode bug fixes.
679 Patch by Jakob Hirsch.
681 NM/36 Bugzilla 909: Correct connect() call in dcc code.
683 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
685 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
687 NM/39 Bugzilla 911: Fixed MakeLinks build script.
693 TK/01 Add preliminary DKIM support. Currently requires a forked version of
694 ALT-N's libdkim that I have put here:
695 http://duncanthrax.net/exim-experimental/
697 Note to Michael Haardt: I had to rename some vars in sieve.c. They
698 were called 'true' and it seems that C99 defines that as a reserved
699 keyword to be used with 'bool' variable types. That means you could
700 not include C99-style headers which use bools without triggering
701 build errors in sieve.c.
703 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
704 as mailq or other aliases. Changed the --help handling significantly
705 to do whats expected. exim_usage() emits usage/help information.
707 SC/01 Added the -bylocaldomain option to eximstats.
709 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
711 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
713 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
719 PH/01 Another patch from the Sieve maintainer.
721 PH/02 When an IPv6 address is converted to a string for single-key lookup
722 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
723 dots are used instead of colons so that keys in lsearch files need not
724 contain colons. This was done some time before quoting was made available
725 in lsearch files. However, iplsearch files do require colons in IPv6 keys
726 (notated using the quote facility) so as to distinguish them from IPv4
727 keys. This meant that lookups for IP addresses in host lists did not work
728 for iplsearch lookups.
730 This has been fixed by arranging for IPv6 addresses to be expressed with
731 colons if the lookup type is iplsearch. This is not incompatible, because
732 previously such lookups could never work.
734 The situation is now rather anomolous, since one *can* have colons in
735 ordinary lsearch keys. However, making the change in all cases is
736 incompatible and would probably break a number of configurations.
738 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
741 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
742 conversion specification without a maximum field width, thereby enabling
743 a rogue spamd server to cause a buffer overflow. While nobody in their
744 right mind would setup Exim to query an untrusted spamd server, an
745 attacker that gains access to a server running spamd could potentially
746 exploit this vulnerability to run arbitrary code as the Exim user.
748 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
749 $primary_hostname instead of what libspf2 thinks the hosts name is.
751 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
752 a directory entry by the name of the lookup key. Previously, if a
753 symlink pointed to a non-existing file or a file in a directory that
754 Exim lacked permissions to read, a lookup for a key matching that
755 symlink would fail. Now it is enough that a matching directory entry
756 exists, symlink or not. (Bugzilla 503.)
758 PH/03 The body_linecount and body_zerocount variables are now exported in the
761 PH/04 Added the $dnslist_matched variable.
763 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
764 This means they are set thereafter only if the connection becomes
767 PH/06 Added the client_condition to authenticators so that some can be skipped
768 by clients under certain conditions.
770 PH/07 The error message for a badly-placed control=no_multiline_responses left
771 "_responses" off the end of the name.
773 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
775 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
776 (without spaces) instead of just copying the configuration text.
778 PH/10 Added the /noupdate option to the ratelimit ACL condition.
780 PH/11 Added $max_received_linelength.
782 PH/12 Added +ignore_defer and +include_defer to host lists.
784 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
785 way in which PCRE > 7.0 is built.
787 PH/14 Implemented queue_only_load_latch.
789 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
790 MAIL command. The effect was to mangle the value on 64-bit systems.
792 PH/16 Another patch from the Sieve maintainer.
794 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
796 PH/18 If a system quota error occurred while trying to create the file for
797 a maildir delivery, the message "Mailbox is full" was not appended to the
798 bounce if the delivery eventually timed out. Change 4.67/27 below applied
799 only to a quota excession during the actual writing of the file.
801 PH/19 It seems that peer DN values may contain newlines (and other non-printing
802 characters?) which causes problems in log lines. The DN values are now
803 passed through string_printing() before being added to log lines.
805 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
806 and InterBase are left for another time.)
808 PH/21 Added message_body_newlines option.
810 PH/22 Guard against possible overflow in moan_check_errorcopy().
812 PH/23 POSIX allows open() to be a macro; guard against that.
814 PH/24 If the recipient of an error message contained an @ in the local part
815 (suitably quoted, of course), incorrect values were put in $domain and
816 $local_part during the evaluation of errors_copy.
822 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
823 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
826 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
827 issue a MAIL command.
829 PH/02 In an ACL statement such as
831 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
833 if a client was not listed at all, or was listed with a value other than
834 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
835 the condition was not true (as it should be), so access was not denied.
836 The bug was that the ! inversion was incorrectly passed on to the second
837 item. This has been fixed.
839 PH/03 Added additional dnslists conditions == and =& which are different from
840 = and & when the dns lookup returns more than one IP address.
842 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
843 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
845 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
846 FSYNC, which compiles an option called disable_fsync that allows for
847 bypassing fsync(). The documentation is heavily laced with warnings.
849 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
851 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
852 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
853 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
854 including adding "make clean"; (3) Added -fPIC when compiling the test
855 dynamically loaded module, to get rid of a warning.
857 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
858 message fails, move_frozen_messages = true and ignore_bounce_errors_after
859 = 0s. The bug is otherwise harmless.
861 PH/07 There was a bug in the dovecot authenticator such that the value of
862 $auth1 could be overwritten, and so not correctly preserved, after a
863 successful authentication. This usually meant that the value preserved by
864 the server_setid option was incorrect.
866 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
868 PH/09 Installed PCRE release 7.0.
870 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
871 run for batched SMTP input. It is now run at the start of every message
872 in the batch. While fixing this I discovered that the process information
873 (output by running exiwhat) was not always getting set for -bs and -bS
874 input. This is fixed, and it now also says "batched" for BSMTP.
876 PH/11 Added control=no_pipelining.
878 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
879 patch, slightly modified), and move the expansion of helo_data till after
880 the connection is made in the smtp transport (so it can use these
883 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
885 PH/14 Added log_selector = +pid.
887 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
889 PH/16 Add ${if forany and ${if forall.
891 PH/17 Added dsn_from option to vary the From: line in DSNs.
893 PH/18 Flush SMTP output before performing a callout, unless control =
894 no_callout_flush is set.
896 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
897 was true (the default) a successful delivery failed to delete the retry
898 item, thus causing premature timeout of the address. The bug is now
901 PH/20 Added hosts_avoid_pipelining to the smtp transport.
903 PH/21 Long custom messages for fakedefer and fakereject are now split up
904 into multiline reponses in the same way that messages for "deny" and
905 other ACL rejections are.
907 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
908 with slight modification.
910 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
911 draft, changing the syntax and factoring some duplicate code".
913 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
914 for deliveries of the second and subsequent messages over the same SMTP
917 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
918 ${reduce, with only minor "tidies".
920 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
922 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
923 expansion side effects.
925 PH/27 When a message times out after an over-quota error from an Exim-imposed
926 quota, the bounce message says "mailbox is full". This message was not
927 being given when it was a system quota that was exceeded. It now should
930 MH/03 Made $recipients available in local_scan(). local_scan() already has
931 better access to the recipient list through recipients_list[], but
932 $recipients can be useful in postmaster-provided expansion strings.
934 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
935 in the case of a MAIL command with additional options following the
936 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
937 were accidentally chopped off.
939 PH/29 SMTP synchronization checks are implemented when a command is read -
940 there is a check that no more input is waiting when there shouldn't be
941 any. However, for some commands, a delay in an ACL can mean that it is
942 some time before the response is written. In this time, more input might
943 arrive, invalidly. So now there are extra checks after an ACL has run for
944 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
945 pipelining has not been advertised.
947 PH/30 MH's patch to allow iscntrl() characters to be list separators.
949 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
950 returned in the SMTP response when smtp_return_error_details was false.
953 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
954 instead of the C I/O that was originally supplied, because problems were
957 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
958 Exim which did not show up earlier: it was assuming that a call to
959 SSL_CTX_set_info_callback() might give an error value. In fact, there is
960 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
961 was a macro that became an assignment, so it seemed to work. This has
962 changed to a proper function call with a void return, hence the compile
963 error. Exim's code has been fixed.
965 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
968 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
970 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
972 PH/37 If a message is not accepted after it has had an id assigned (e.g.
973 because it turns out to be too big or there is a timeout) there is no
974 "Completed" line in the log. When some messages of this type were
975 selected by exigrep, they were listed as "not completed". Others were
976 picked up by some special patterns. I have improved the selection
977 criteria to be more general.
979 PH/38 The host_find_failed option in the manualroute router can now be set
980 to "ignore", to completely ignore a host whose IP address cannot be
981 found. If all hosts are ignored, the behaviour is controlled by the new
982 host_all_ignored option.
984 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
985 homing or because of multiple MX records with /mx) generated more than
986 one IP address, and the following item turned out to be the local host,
987 all the secondary addresses of the first item were incorrectly removed
988 from the list, along with the local host and any following hosts (which
989 is what is supposed to happen).
991 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
992 whoever called Exim into the -H file. In the case of the daemon it was
993 behaving confusingly. When first started, it used values for whoever
994 started the daemon, but after a SIGHUP it used the Exim user (because it
995 calls itself on a restart). I have changed the code so that it now always
998 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
999 message are rejected with the same error (e.g. no authentication or bad
1000 sender address), and a DATA command is nevertheless sent (as can happen
1001 with PIPELINING or a stupid MUA), the error message that was given to the
1002 RCPT commands is included in the rejection of the DATA command. This is
1003 intended to be helpful for MUAs that show only the final error to their
1006 PH/42 Another patch from the Sieve maintainer.
1008 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1009 Eximstats - Fixed some broken HTML links and added missing column headers
1011 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1012 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1014 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1016 PH/43 Yet another patch from the Sieve maintainer.
1018 PH/44 I found a way to check for a TCP/IP connection going away before sending
1019 the response to the final '.' that terminates a message, but only in the
1020 case where the client has not sent further data following the '.'
1021 (unfortunately, this is allowed). However, in many cases there won't be
1022 any further data because there won't be any more messages to send. A call
1023 to select() can be used: if it shows that the input is "ready", there is
1024 either input waiting, or the socket has been closed. An attempt to read
1025 the next input character can distinguish the two cases. Previously, Exim
1026 would have sent an OK response which the client would never have see.
1027 This could lead to message repetition. This fix should cure that, at
1028 least in a lot of common cases.
1030 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1031 advertised in response to EHLO.
1037 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1038 fixed by 4.65/MH/01 (is this a record?) are fixed:
1040 (i) An empty string was always treated as zero by the numeric comparison
1041 operators. This behaviour has been restored.
1043 (ii) It is documented that the numeric comparison operators always treat
1044 their arguments as decimal numbers. This was broken in that numbers
1045 starting with 0 were being interpreted as octal.
1047 While fixing these problems I realized that there was another issue that
1048 hadn't been noticed. Values of message_size_limit (both the global option
1049 and the transport option) were treated as octal if they started with 0.
1050 The documentation was vague. These values are now always treated as
1051 decimal, and I will make that clear in the documentation.
1057 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1058 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1061 MH/01 Don't check that the operands of numeric comparison operators are
1062 integers when their expansion is in "skipping" mode (fixes bug
1063 introduced by 4.64-PH/07).
1065 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1066 child addresses, Exim now panics and dies. Previously, because the count
1067 is held in a short int, deliveries were likely to be lost. As such a
1068 large number of recipients for a single message is ridiculous
1069 (performance will be very, very poor), I have chosen to impose a limit
1070 rather than extend the field.
1076 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1077 leftover -K file (the existence of which was triggered by #402).
1078 While we were at it, introduced process PID as part of the -K
1079 filename. This should rule out race conditions when creating
1082 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1083 processing considerably. Previous code took too long for large mails,
1084 triggering a timeout which in turn triggers #401.
1086 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1087 in the DK code in transports.c. sendfile() is not really portable,
1088 hence the _LINUX specificness.
1090 TF/01 In the add_headers option to the mail command in an Exim filter,
1091 there was a bug that Exim would claim a syntax error in any
1092 header after the first one which had an odd number of characters
1095 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1096 callout verification, Exim cached a "reject" for the entire domain. This
1097 is correct for most verifications, but it is not correct for a recipient
1098 verification with use_sender or use_postmaster set, because in that case
1099 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1100 case of MAIL FROM:<> rejection from other early rejections (e.g.
1101 rejection of HELO). When verifying a recipient using a non-null MAIL
1102 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1103 Whatever the result of the callout, the value of the domain cache is
1104 left unchanged (for any other kind of callout, getting as far as trying
1105 RCPT means that the domain itself is ok).
1107 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1110 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1111 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1112 session, the daemon ignores ECONNECT errors and logs others; it now
1113 ignores EPIPE as well.
1115 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1116 (quoted-printable decoding).
1118 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1119 later the small subsequent patch to fix an introduced bug.
1121 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1123 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1125 PH/08 An error is now given if message_size_limit is specified negative.
1127 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1128 to be given (somewhat) arbitrary names.
1130 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1133 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1134 miscellaneous code fixes
1136 PH/10 Added the log_reject_target ACL modifier to specify where to log
1139 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1140 hostname. This is wrong, because it relates to the incoming message (and
1141 probably the interface on which it is arriving) and not to the outgoing
1142 callout (which could be using a different interface). This has been
1143 changed to use the value of the helo_data option from the smtp transport
1144 instead - this is what is used when a message is actually being sent. If
1145 there is no remote transport (possible with a router that sets up host
1146 addresses), $smtp_active_hostname is used.
1148 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1149 tweaks were necessary in order to get it to work (see also 21 below):
1150 (a) The code assumed that strncpy() returns a negative number on buffer
1151 overflow, which isn't the case. Replaced with Exim's string_format()
1153 (b) There were several signed/unsigned issues. I just did the minimum
1154 hacking in of casts. There is scope for a larger refactoring.
1155 (c) The code used strcasecmp() which is not a standard C function.
1156 Replaced with Exim's strcmpic() function.
1157 (d) The code set only $1; it now sets $auth1 as well.
1158 (e) A simple test gave the error "authentication client didn't specify
1159 service in request". It would seem that Dovecot has changed its
1160 interface. Fortunately there's a specification; I followed it and
1161 changed what the client sends and it appears to be working now.
1163 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1166 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1167 address A is aliased to B and C, where B exists and C does not. Without
1168 -v the output is "A verified" because verification stops after a
1169 successful redirection if more than one address is generated. However,
1170 with -v the child addresses are also verified. Exim was outputting "A
1171 failed to verify" and then showing the successful verification for C,
1172 with its parentage. It now outputs "B failed to verify", showing B's
1173 parentage before showing the successful verification of C.
1175 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1176 look up a TXT record in a specific list after matching in a combined
1179 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1180 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1181 they consult the DNS. I had assumed they would set it the way they
1182 wanted; and indeed my experiments on Linux seem to show that in some
1183 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1184 To be on the safe side, however, I have now made the interface to
1185 host_find_byname() similar to host_find_bydns(), with an argument
1186 containing the DNS resolver options. The host_find_byname() function now
1187 sets these options at its start, just as host_find_bydns() does. The smtp
1188 transport options dns_qualify_single and dns_search_parents are passed to
1189 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1190 of host_find_byname() use the default settings of RES_DEFNAMES
1191 (qualify_single) but not RES_DNSRCH (search_parents).
1193 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1194 spool_read_header() do less string testing, by means of a preliminary
1195 switch on the second character of optional "-foo" lines. (This is
1196 overdue, caused by the large number of possibilities that now exist.
1197 Originally there were few.) While I was there, I also converted the
1198 str(n)cmp tests so they don't re-test the leading "-" and the first
1199 character, in the hope this might squeeze out yet more improvement.
1201 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1202 flag allowing group syntax was set by the header_syntax check but not
1203 turned off, possible causing trouble later; (2) The flag was not being
1204 set at all for the header_verify test, causing "group"-style headers to
1205 be rejected. I have now set it in this case, and also caused header_
1206 verify to ignore an empty address taken from a group. While doing this, I
1207 came across some other cases where the code for allowing group syntax
1208 while scanning a header line wasn't quite right (mostly, not resetting
1209 the flag correctly in the right place). These bugs could have caused
1210 trouble for malformed header lines. I hope it is now all correct.
1212 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1213 with the "reply" argument non-NULL. The code, however (which originally
1214 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1215 but it didn't always do it. This confused somebody who was copying the
1216 code for some other use. I have removed all the tests.
1218 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1219 feature that was used to support insecure browsers during the U.S. crypto
1220 embargo. It requires special client support, and Exim is probably the
1221 only MTA that supported it -- and would never use it because real RSA is
1222 always available. This code has been removed, because it had the bad
1223 effect of slowing Exim down by computing (never used) parameters for the
1224 RSA_EXPORT functionality.
1226 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1227 authenticator to fail if there's a tab character in the incoming data
1228 (there should never be unless someone is messing about, as it's supposed
1229 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1230 if the connection is using TLS or if the remote IP is the same as the
1231 local IP, and the "valid-client-cert option" if a client certificate has
1234 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1235 authenticators. This can be used for authorization after authentication
1236 succeeds. (In the case of plaintext, it servers for both authentication
1239 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1240 if any retry times were supplied.
1242 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1243 connection, where there is no client IP address to check. In this
1244 situation, the verify now always succeeds.
1246 PH/25 Applied John Jetmore's -Mset patch.
1248 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1250 PH/27 In a string expansion for a processed (not raw) header when multiple
1251 headers of the same name were present, leading whitespace was being
1252 removed from all of them, but trailing whitespace was being removed only
1253 from the last one. Now trailing whitespace is removed from each header
1254 before concatenation. Completely empty headers in a concatenation (as
1255 before) are ignored.
1257 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1258 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1260 PH/29 [Removed. This was a change that I later backed out, and forgot to
1261 correct the ChangeLog entry (that I had efficiently created) before
1262 committing the later change.]
1264 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1265 address errors (4xx response to RCPT) over the same connection as other
1266 messages routed to the same hosts. Such deliveries are always "forced",
1267 so retry times are not inspected. This resulted in far too many retries
1268 for the affected addresses. The effect occurred only when there were more
1269 hosts than the hosts_max_try setting in the smtp transport when it had
1270 the 4xx errors. Those hosts that it had tried were not added to the list
1271 of hosts for which the message was waiting, so if all were tried, there
1272 was no problem. Two fixes have been applied:
1274 (i) If there are any address or message errors in an SMTP delivery, none
1275 of the hosts (tried or untried) are now added to the list of hosts
1276 for which the message is waiting, so the message should not be a
1277 candidate for sending over the same connection that was used for a
1278 successful delivery of some other message. This seems entirely
1279 reasonable: after all the message is NOT "waiting for some host".
1280 This is so "obvious" that I'm not sure why it wasn't done
1281 previously. Hope I haven't missed anything, but it can't do any
1282 harm, as the worst effect is to miss an optimization.
1284 (ii) If, despite (i), such a delivery is accidentally attempted, the
1285 routing retry time is respected, so at least it doesn't keep
1286 hammering the server.
1288 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1289 in ${readsocket because some servers need this prod.
1291 PH/32 Added some extra debug output when updating a wait-xxx database.
1293 PH/33 The hint "could be header name not terminated by colon", which has been
1294 given for certain expansion errors for a long time, was not being given
1295 for the ${if def:h_colon_omitted{... case.
1297 PH/34 The spec says: "With one important exception, whenever a domain list is
1298 being scanned, $domain contains the subject domain." There was at least
1299 one case where this was not true.
1301 PH/35 The error "getsockname() failed: connection reset by peer" was being
1302 written to the panic log as well as the main log, but it isn't really
1303 panic-worthy as it just means the connection died rather early on. I have
1304 removed the panic log writing for the ECONNRESET error when getsockname()
1307 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1308 runs only) independently of the message's sender address. This meant
1309 that, if the 4xx error was in fact related to the sender, a different
1310 message to the same recipient with a different sender could confuse
1311 things. In particualar, this can happen when sending to a greylisting
1312 server, but other circumstances could also provoke similar problems.
1313 I have changed the default so that the retry time for these errors is now
1314 based a combination of the sender and recipient addresses. This change
1315 can be overridden by setting address_retry_include_sender=false in the
1318 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1319 remote server are returned as part of bounce messages. This was not
1320 happening for LMTP over a pipe (the lmtp transport), but now it is the
1321 same for both kinds of LMTP.
1323 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1324 in header lines that were in fact CNAMEs. This is no longer the case.
1326 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1327 and queue runs started by the daemon processed all messages. This has
1328 been fixed so that -R and -S can now usefully be given with -q<time>.
1330 PH/40 Import PCRE release 6.7 (fixes some bugs).
1332 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1334 PH/42 Give an error if -q is specified more than once.
1336 PH/43 Renamed the variables $interface_address and $interface_port as
1337 $received_ip_address and $received_port, to make it clear that these
1338 values apply to message reception, and not to the outgoing interface when
1339 a message is delivered. (The old names remain recognized, of course.)
1341 PH/44 There was no timeout on the connect() call when using a Unix domain
1342 socket in the ${readsocket expansion. There now is.
1344 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1345 be meaningful with "accept".
1347 SC/01 Eximstats V1.43
1348 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1350 SC/02 Eximstats V1.44
1351 Use a glob alias rather than an array ref in the generated
1352 parser. This improves both readability and performance.
1354 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1355 Collect SpamAssassin and rejection statistics.
1356 Don't display local sender or destination tables unless
1357 there is data to show.
1358 Added average volumes into the top table text output.
1360 SC/04 Eximstats V1.46
1361 Collect data on the number of addresses (recipients)
1362 as well as the number of messages.
1364 SC/05 Eximstats V1.47
1365 Added 'Message too big' to the list of mail rejection
1366 reasons (thanks to Marco Gaiarin).
1368 SC/06 Eximstats V1.48
1369 Mainlog lines which have GMT offsets and are too short to
1370 have a flag are now skipped.
1372 SC/07 Eximstats V1.49 (Alain Williams)
1373 Added the -emptyok flag.
1375 SC/08 Eximstats V1.50
1376 Fixes for obtaining the IP address from reject messages.
1378 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1379 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1380 whitesspace changes from 4.64-PH/27
1382 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1385 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1386 are found, allow negative numbers in numeric criteria)
1388 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1390 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1391 to match changes made in 4.64-PH/43
1393 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1395 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1396 "domain" is actually an IP address, and force a failure. This locks out
1397 those revolvers/nameservers that support "A-for-A" lookups, in
1398 contravention of the specifications.
1400 PH/48 When a host name was looked up from an IP address, and the subsequent
1401 forward lookup of the name timed out, the host name was left in
1402 $sender_host_name, contrary to the specification.
1404 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1405 restricted to single-key lookups, Exim was not diagnosing an error if
1406 * or *@ was used with a query-style lookup.
1408 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1410 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1411 long ago, but noone interested enough thought of it. Let's just say that
1412 the "1.1" means that there are some new functions that weren't there at
1413 some point in the past.
1415 PH/51 Error processing for expansion failure of helo_data from an smtp
1416 transport during callout processing was broken.
1418 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1419 tested/used via the -bh/-bhc/-bs options.
1421 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1422 bug, fixed in subsequent PCRE releases).
1424 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1425 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1427 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1433 SC/01 Use a glob alias rather than an array ref in eximstats generated
1434 parser. This improves both readability and performance.
1436 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1437 Don't display local sender or destination tables in eximstats unless
1438 there is data to show.
1439 Added average volumes into the eximstats top table text output.
1441 SC/03 Collect data on the number of addresses (recipients) as well
1442 as the number of messages in eximstats.
1444 TF/01 Correct an error in the documentation for the redirect router. Exim
1445 does (usually) call initgroups() when daemonizing.
1447 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1448 with consistent privilege compared to when running as a daemon.
1450 TF/03 Note in the spec that $authenticated_id is not set for local
1451 submissions from trusted users.
1453 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1454 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1456 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1457 by adding some example configuration directives to the default
1458 configuration file. A little bit of work is required to uncomment the
1459 directives and define how usernames and passwords are checked, but
1460 there is now a framework to start from.
1462 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1463 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1464 without this. I don't know how relevant this is to other LDAP libraries.
1466 PH/02 Add the verb name to the "unknown ACL verb" error.
1468 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1470 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1472 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1473 directory not expanded when it should be if an expanded home directory
1474 was set for the address (which is overridden by the transport).
1476 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1479 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1480 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1481 because it is too late at that time, and has no effect.
1483 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1484 security issue with \' (bugzilla #107). I could not use the
1485 PQescapeStringConn() function, because it needs a PGconn value as one of
1488 PH/08 When testing addresses using -bt, indicate those final addresses that
1489 are duplicates that would not cause an additional delivery. At least one
1490 person was confused, thinking that -bt output corresponded to deliveries.
1491 (Suppressing duplicates isn't a good idea as you lose the information
1492 about possibly different redirections that led to the duplicates.)
1494 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1495 systems where poll() doesn't work, in particular OS X.
1497 PH/10 Added more information to debugging output for retry time not reached.
1499 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1500 operations in malware.c.
1502 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1505 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1506 syslog_duplication set false, log lines that would normally be written
1507 both the the main log and to the reject log were not written to syslog at
1510 PH/14 In the default configuration, change the use of "message" in ACL warn
1511 statements to "add_header".
1513 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1514 not followed by a command (e.g. "seen endif").
1516 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1517 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1520 PH/17 Added extra conditions to the default value of delay_warning_condition
1524 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1525 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1526 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1529 The Auto-Submitted: and various List- headers are standardised, whereas I
1530 don't think Precedence: ever was.
1532 PH/18 Refactored debugging code in route_finduser() to show more information,
1533 in particular, the error code if getpwnam() issues one.
1535 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1536 This is apparently needed in addition to the PH/07 change above to avoid
1537 any possible encoding problems.
1539 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1540 but not after initializing Perl.
1542 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1543 output them only if debugging. By default they are written stderr,
1544 apparently, which is not desirable.
1546 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1549 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1552 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1554 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1555 authenticated or an ident call has been made. Suppress the default
1556 values for $authenticated_id and $authenticated_sender (but permit -oMai
1557 and -oMas) when testing with -bh.
1559 PH/24 Re-jigged the order of the tests in the default configuration so that the
1560 tests for valid domains and recipients precede the DNS black list and CSA
1561 tests, on the grounds that those ones are more expensive.
1563 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1564 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1565 This bug exists in every version of Exim that I still have, right back to
1568 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1569 However, an attempt to turn on case-sensitivity in a regex key by
1570 including (?-i) didn't work because the subject string was already
1571 lowercased, and the effects were non-intuitive. It turns out that a
1572 one-line patch can be used to allow (?-i) to work as expected.
1578 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1579 other effects) broke the use of negated acl sub-conditions.
1581 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1584 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1585 "Deny" causes Exim to reject the incoming connection with a 554 error.
1586 Unfortunately, if there is a major crisis, such as a disk failure,
1587 tcp-wrappers gives "deny", whereas what one would like would be some
1588 kind of temporary error. A kludge has been added to help with this.
1589 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1590 554 error is used if errno is still zero or contains ENOENT (which occurs
1591 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1594 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1596 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1597 errors. Otherwise a message that provokes a temporary error (when other
1598 messages do not) can cause a whole host to time out.
1600 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1601 addresses were routed directly to files or pipes from a redirect router.
1602 File deliveries just didn't batch; pipe deliveries might have suffered
1605 PH/06 A failure to get a lock for a hints database would erroneously always say
1606 "Failed to get write lock", even when it was really a read lock.
1608 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1609 of 0600. This has been changed to use the value of the lockfile_mode
1610 option (which defaults to 0600).
1612 PH/08 Applied small patch from the Sieve maintainer.
1614 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1615 folder from quota calculations, a direct delivery into this folder messed
1616 up the contents of the maildirsize file. This was because the regex was
1617 used only to exclude .Trash (or whatever) when the size of the mailbox
1618 was calculated. There was no check that a delivery was happening into an
1619 excluded directory. This bug has been fixed by ignoring all quota
1620 processing for deliveries into excluded directories.
1622 PH/10 Added the maildirfolder_create_regex option to appendfile.
1628 PH/01 The code for finding all the local interface addresses on a FreeBSD
1629 system running IPv6 was broken. This may well have applied to all BSD
1630 systems, as well as to others that have similar system calls. The broken
1631 code found IPv4 interfaces correctly, but gave incorrect values for the
1632 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1633 that it would not match correctly against @[] and not recognize the IPv6
1636 PH/02 The ipliteral router was not recognizing addresses of the form user@
1637 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1639 PH/03 Added disable_ipv6.
1641 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1642 decoded form, because it is most often used to construct To: headers
1643 lines in autoreplies, and the decoded form may well be syntactically
1644 invalid. However, $reply_address has leading white space removed, and all
1645 newlines turned into spaces so that the autoreply transport does not
1648 PH/05 If group was specified without a user on a router, and no group or user
1649 was specified on a transport, the group from the router was ignored.
1651 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1652 for visible compile-time settings that can be used to change these
1653 numbers, for those that want even more. Backwards compatibility with old
1654 spool files has been maintained. However, going back to a previous Exim
1655 release will lost any variables that are in spool files.
1657 PH/07 Two small changes when running in the test harness: increase delay when
1658 passing a TCP/IP connection to a new process, in case the original
1659 process has to generate a bounce, and remove special handling of
1660 127.0.0.2 (sic), which is no longer necessary.
1662 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1663 be the same on different OS.
1665 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1668 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1669 whether --show-vars was specified or not
1671 JJ/02 exipick: Added support for new ACL variable spool format introduced
1674 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1675 syntactically invalid From: or Reply-to: line, and a filter used this to
1676 generate an autoreply, and therefore failed to obtain an address for the
1677 autoreply, Exim could try to deliver to a non-existent relative file
1678 name, causing unrelated and misleading errors. What now happens is that
1679 it logs this as a hard delivery error, but does not attempt to create a
1682 PH/11 The exinext utility has a -C option for testing purposes, but although
1683 the given file was scanned by exinext itself; it wasn't being passed on
1684 when Exim was called.
1686 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1687 an end-of-file indication when reading a command response.
1689 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1690 compiled. In many other places in Exim, IPv6 addresses are always
1691 recognized, so I have changed this. It also means that IPv4 domain
1692 literals of the form [IPV4:n.n.n.n] are now always recognized.
1694 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1695 used if the router is not running as root, for example, when verifying at
1696 ACL time, or when using -bh. The debugging output from this situation was
1697 non-existent - all you got was a failure to exec. I have made two
1700 (a) Failures to set uid/gid, the current directory, or a process leader
1701 in a subprocess such as that created by queryprogram now generate
1702 suitable debugging ouput when -d is set.
1704 (b) The queryprogram router detects when it is not running as root,
1705 outputs suitable debugging information if -d is set, and then runs
1706 the subprocess without attempting to change uid/gid.
1708 PH/15 Minor change to Makefile for building test_host (undocumented testing
1711 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1712 additional section of a DNS packet that returns MX or SRV records.
1713 Instead, it always explicitly searches for A/AAAA records. This avoids
1714 major problems that occur when a DNS server includes only records of one
1715 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1716 fixed another bug: if SRV records were looked up and the corresponding
1717 address records were *not* found in the additional section, the port
1718 values from the SRV records were lost.
1720 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1721 using the correct key (the original address) when searching the retry
1722 rules in order to find which one to use for generating the retry hint.
1724 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1725 adding the default one. Similarly, if it contains a Reply-To: header, the
1726 errors_reply_to option, if set, is not used.
1728 PH/19 When calculating a retry time, Exim used to measure the "time since
1729 failure" by looking at the "first failed" field in the retry record. Now
1730 it does not use this if it is later than than the arrival time of the
1731 message. Instead it uses the arrival time. This makes for better
1732 behaviour in cases where some deliveries succeed, thus re-setting the
1733 "first failed" field. An example is a quota failure for a huge message
1734 when small messages continue to be delivered. Without this change, the
1735 "time since failure" will always be short, possible causing more frequent
1736 delivery attempts for the huge message than are intended.
1737 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1739 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1740 $1, $2, $3) because the numerical variables can be reset during some
1741 expansion items (e.g. "match"), thereby losing the authentication data.
1743 PH/21 Make -bV show the size of off_t variables so that the test suite can
1744 decide whether to run tests for quotas > 2G.
1746 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1747 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1748 filecount value is greater than 2G or if a quota value is greater than 2G
1749 on a system where the size of off_t is not greater than 4, a panic error
1752 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1753 never match. The debug and -bh output now contains an explicit error
1754 message indicating a malformed IPv4 address or mask.
1756 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1757 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1758 PH/23 above applies.
1760 PH/25 Do not write to syslog when running in the test harness. The only
1761 occasion when this arises is a failure to open the main or panic logs
1762 (for which there is an explicit test).
1764 PH/26 Added the /no_tell option to "control=freeze".
1766 PH/27 If a host name lookup failed very early in a connection, for example, if
1767 the IP address matched host_lookup and the reverse lookup yielded a name
1768 that did not have a forward lookup, an error message of the form "no IP
1769 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1770 could be logged. Now it outputs the IP address instead of "NULL".
1772 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1773 allows the sender and the authenticated sender to be set when
1774 submitting a message from within Exim. Since child_open_exim() is
1775 documented for local_scan(), the new function should be too.
1777 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1778 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1779 results in an empty string is now treated as unset.
1781 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1783 PH/31 Added sender_verify_fail logging option.
1785 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1786 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1787 tidied the source and removed it altogether.
1789 PH/33 When a queue run was abandoned because the load average was too high, a
1790 log line was always written; now it is written only if the queue_run log
1791 selector is set. In addition, the log line for abandonment now contains
1792 information about the queue run such as the pid. This is always present
1793 in "start" and "stop" lines but was omitted from the "abandon" line.
1795 PH/34 Omit spaces between a header name and the colon in the error message that
1796 is given when verify = headers_syntax fails (if there are lots of them,
1797 the message gets confusing).
1799 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1800 names, as there are now some PTR records that contain slashes. This check
1801 is only to protect against broken name servers that fall over on strange
1802 characters, so the fact that it applies to all lookups doesn't matter.
1804 PH/36 Now that the new test suite is complete, we can remove some of the
1805 special code in Exim that was needed for the old test suite. For example,
1806 sorting DNS records because real resolvers return them in an arbitrary
1807 order. The new test suite's fake resolver always returns records in the
1810 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1811 bounces) except when queue_only is set, to avoid logging races between
1812 the different processes.
1814 PH/38 Panic-die if .include specifies a non-absolute path.
1816 PH/39 A tweak to the "H" retry rule from its user.
1818 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1819 a label. They prevented compilation on older perls.
1821 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1822 a warning to be raised on newish perls.
1824 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1825 on queue. Changes to match documented behaviour of showing count of
1826 messages matching specified criteria.
1828 PH/40 Changed the default ident timeout from 30s to 5s.
1830 PH/41 Added support for the use of login_cap features, on those BSD systems
1831 that have them, for controlling the resources used by pipe deliveries.
1833 PH/42 The content-scanning code uses fopen() to create files in which to put
1834 message data. Previously it was not paying any attention to the mode of
1835 the files. Exim runs with umask(0) because the rest of the code creates
1836 files with open(), and sets the required mode explicitly. Thus, these
1837 files were ending up world-writeable. This was not a big issue, because,
1838 being within the spool directory, they were not world-accessible. I have
1839 created a function called modefopen, which takes an additional mode
1840 argument. It sets umask(777), creates the file, chmods it to the required
1841 mode, then resets the umask. All the relevant calls to fopen() in the
1842 content scanning code have been changed to use this function.
1844 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1845 to 24 hours. This avoids potential overflow problems when processing G
1846 and H retry rules. I suspect nobody ever tinkers with this value.
1848 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1850 PH/45 When the plaintext authenticator is running as a client, the server's
1851 challenges are checked to ensure they are valid base64 strings. By
1852 default, the authentication attempt is cancelled if an invalid string is
1853 received. Setting client_ignore_invalid_base64 true ignores these errors.
1854 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1855 they are received. Thus, the responses can be made to depend on the
1856 challenges. If an invalid string is ignored, an empty string is placed in
1859 PH/46 Messages that are created by the autoreply transport now contains a
1860 References: header, in accordance with RFCs 2822 and 3834.
1862 PH/47 Added authenticated_sender_force to the smtp transport.
1864 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1866 PH/49 Installed latest patch from the Sieve maintainer.
1868 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1869 was also set, the appendfile transport was unnecessarily scanning a
1870 directory of message files (e.g. for maildir delivery) to find the count
1871 of files (along with the size), even though it did not need this
1872 information. It now does the scan only if it needs to find either the
1873 size of the count of files.
1875 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1877 PH/52 Two bugs concerned with error handling when the smtp transport is
1880 (i) Exim was not creating retry information for temporary errors given
1881 for individual recipients after the DATA command when the smtp transport
1882 was used in LMTP mode. This meant that they could be retried too
1883 frequently, and not timed out correctly.
1885 (ii) Exim was setting the flag that allows error details to be returned
1886 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1887 recipients that were returned after the DATA command.
1889 PH/53 This is related to PH/52, but is more general: for any failing address,
1890 when detailed error information was permitted to be returned to the
1891 sender, but the error was temporary, then after the final timeout, only
1892 "retry timeout exceeded" was returned. Now it returns the full error as
1893 well as "retry timeout exceeded".
1895 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1896 do this, and (what is worse) MTAs that accept it.
1898 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1899 will now be deprecated.
1901 PH/56 New os.c-cygwin from the Cygwin maintainer.
1903 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1904 formats (previously only available in exim formats via -bpr, -bpru,
1905 and -bpra. Now also available in native and exiqgrep formats)
1907 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1908 with very large, slow to parse queues
1910 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1912 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1914 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1915 responses to authentication challenges, though it was showing the
1916 challenges; (ii) I've removed the CR characters from the debug output for
1919 PH/58 Allow for the insertion of a newline as well as a space when a string
1920 is turned into more than one encoded-word during RFC 2047 encoding. The
1921 Sieve code now uses this.
1923 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1924 data_4xx, lost_connection, tls_required.
1926 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1927 message was being sent as an SMTP response.
1929 PH/61 Add -l and -k options to exicyclog.
1931 PH/62 When verifying, if an address was redirected to one new address, so that
1932 verification continued, and the new address failed or deferred after
1933 having set something in $address_data, the value of $address_data was not
1934 passed back to the ACL. This was different to the case when no
1935 redirection occurred. The value is now passed back in both cases.
1937 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1938 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1939 use login_cap.h, so on its own it isn't the distinguishing feature. The
1940 new name refers directly to the setclassresources() function.
1942 PH/65 Added configuration files for NetBSD3.
1944 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1946 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1947 is preferred over IPv4.
1949 PH/68 The bounce_return_message and bounce_return_body options were not being
1950 honoured for bounces generated during the reception of non-SMTP messages.
1951 In particular, this applied to messages rejected by the ACL. This bug has
1952 been fixed. However, if bounce_return_message is true and bounce_return_
1953 body is false, the headers that are returned for a non-SMTP message
1954 include only those that have been read before the error was detected.
1955 (In the case of an ACL rejection, they have all been read.)
1957 PH/69 The HTML version of the specification is now built in a directory called
1958 spec_html instead of spec.html, because the latter looks like a path with
1959 a MIME-type, and this confuses some software.
1961 PH/70 Catch two compiler warnings in sieve.c.
1963 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1964 function verify_get_ident() calls ip_connect() to connect a socket, but
1965 if the "connect()" function timed out, ip_connect() used to close the
1966 socket. However, verify_get_ident() also closes the socket later, and in
1967 between Exim writes to the log, which may get opened at this point. When
1968 the socket was closed in ip_connect(), the log could get the same file
1969 descriptor number as the socket. This naturally causes chaos. The fix is
1970 not to close the socket in ip_connect(); the socket should be closed by
1971 the function that creates it. There was only one place in the code where
1972 this was missing, in the iplookup router, which I don't think anybody now
1973 uses, but I've fixed it anyway.
1975 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1976 well as to direct DNS lookups. Otherwise the handling of names in host
1977 lists is inconsistent and therefore confusing.
1983 PH/01 Two changes to the default runtime configuration:
1985 (1) Move the checks for relay_from_hosts and authenticated clients from
1986 after to before the (commented out) DNS black list checks.
1988 (2) Add control=submission to the relay_from_hosts and authenticated
1989 clients checks, on the grounds that messages accepted by these
1990 statements are most likely to be submissions.
1992 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1994 (1) Generate an error if the third argument for the ${prvs expansion is
1997 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2000 (3) Reset the variables that are obtained from the first argument of
2001 ${prvscheck and used in the second argument before leaving the code,
2002 because their memory is reclaimed, so using them afterwards may do
2005 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2006 one (it's much easier than Tom thought :-).
2008 (5) Because of (4), we can now allow for the use of $prvscheck_result
2009 inside the third argument.
2011 PH/03 For some reason, the default setting of PATH when running a command from
2012 a pipe transport was just "/usr/bin". I have changed it to
2015 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2016 anything to be listed in the output from -bV.
2018 PH/05 When a filter generated an autoreply, the entire To: header line was
2019 quoted in the delivery log line, like this:
2021 => >A.N.Other <ano@some.domain> <original@ddress> ...
2023 This has been changed so that it extracts the operative address. There
2024 may be more than one such address. If so, they are comma-separated, like
2027 => >ano@some.domain,ona@other.domain <original@ddress> ...
2029 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2030 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2031 not being looked up in the rDNS to get a host name, Exim was showing the
2032 IP address twice in Received: lines, even though the IP addresses were
2033 identical. For example:
2035 Received: from [1.2.3.4] (helo=[1.2.3.4])
2037 However, if the real host name was known, it was omitting the HELO data
2038 if it matched the actual IP address. This has been tidied up so that it
2039 doesn't show the same IP address twice.
2041 PH/07 When both +timestamp and +memory debugging was on, the value given by
2042 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2043 called by the debug printing, thereby overwriting the timestamp buffer.
2044 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2046 PH/08 When the original message was included in an autoreply transport, it
2047 always said "this is a copy of the message, including all the headers",
2048 even if body_only or headers_only was set. It now gives an appropriate
2051 PH/09 Applied a patch from the Sieve maintainer which:
2053 o fixes some comments
2054 o adds the (disabled) notify extension core
2055 o adds some debug output for the result of if/elsif tests
2056 o points to the current vacation draft in the documentation
2057 and documents the missing references header update
2061 o fixes a bug in processing the envelope test (when testing
2062 multiple envelope elements, the last element determinted the
2065 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2066 Electronic Mail") by including:
2068 Auto-submitted: auto-generated
2070 in the messages that it generates (bounce messages and others, such as
2071 warnings). In the case of bounce messages for non-SMTP mesages, there was
2072 also a typo: it was using "Auto_submitted" (underscore instead of
2073 hyphen). Since every message generated by Exim is necessarily in response
2074 to another message, thes have all been changed to:
2076 Auto-Submitted: auto-replied
2078 in accordance with these statements in the RFC:
2080 The auto-replied keyword:
2082 - SHOULD be used on messages sent in direct response to another
2083 message by an automatic process,
2085 - MUST NOT be used on manually-generated messages,
2087 - MAY be used on Delivery Status Notifications (DSNs) and Message
2088 Disposition Notifications (MDNs),
2090 - MUST NOT be used on messages generated by automatic or periodic
2091 processes, except for messages which are automatic responses to
2094 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2095 to the default Received: header definition.
2097 PH/12 Added log selector acl_warn_skipped (default on).
2099 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2100 variables because (a) they are in the wrong storage pool and (b) even if
2101 they were copied, it wouldn't work properly because of the caching.
2103 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2104 checking when decoding. Apparently there are clients that generate
2105 overlong encoded strings. Why am I not surprised?
2107 PH/15 If the first argument of "${if match_address" was not empty, but did not
2108 contain an "@" character, Exim crashed. Now it writes a panic log message
2109 and treats the condition as false.
2111 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2113 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2114 extension "envelope-auth". The code is finished and in agreement with
2115 other implementations, but there is no documentation so far and in fact,
2116 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2117 not changing the active code.
2119 Print executed "if" and "elsif" statements when debugging is used. This
2120 helps a great deal to understand what a filter does.
2122 Document more things not specified clearly in RFC3028. I had all this
2123 sorted out, when out of a sudden new issues came to my mind. Oops."
2125 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2128 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2129 canonical form (as documented). However, after a host name lookup from
2130 the IP address, check_host() was doing a simple string comparison with
2131 addresses acquired from the DNS when checking that the found name did
2132 have the original IP as one of its addresses. Since any found IPv6
2133 addresses are likely to be in abbreviated form, the comparison could
2134 fail. Luckily, there already exists a function for doing the comparison
2135 by converting both addresses to binary, so now that is used instead of
2136 the text comparison.
2138 PH/20 There was another similar case to PH/19, when a complete host name was
2139 given in a host list; looking up its IP address could give an abbreviated
2140 form, whereas the current host's name might or might not be abbreviated.
2141 The same fix has been applied.
2147 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2148 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2151 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2152 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2154 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2156 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2157 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2158 Also, for "Auto-Submitted", treat anything other than "no" as
2159 non-personal, in accordance with RFC 3834. (Previously it treated
2160 anything starting "auto-" as non-personal.)
2162 TF/01 The control=submission/name=... option had a problem with syntax
2163 errors if the name included a slash character. The /name= option
2164 now slurps the rest of the string, so it can include any characters
2165 but it must come last in the list of options (after /sender_retain
2168 PH/05 Some modifications to the interface to the fake nameserver for the new
2176 TK/01 Added the "success_on_redirect" address verification option. See
2177 NewStuff for rationale and an example.
2179 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2181 PH/02 Patch to exigrep to allow it to work on syslog lines.
2183 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2184 fread() to skip over the body file's header line, because in Cygwin the
2185 header line is locked and is inaccessible.
2187 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2188 co-exist for some time) to make it clear that it is the Exim ID that is
2189 referenced, not the Message-ID: header line.
2191 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2192 string_format() function, because snprintf() does not exist on all
2195 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2196 ${stat: expansion item.
2198 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2199 protocol synchronization error", to keep the pedants happy.
2201 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2202 well as for IRIX systems, when gcc is being used. See the host.c source
2205 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2207 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2210 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2211 transport and to the smtp transport in LMTP mode.
2213 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2215 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2216 run a filter in a subprocess. This could lead to confusion in subsequent
2217 lookups in the parent process. There should also be a search_tidyup() at
2218 the end of the subprocess.
2220 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2221 only if the host matched helo_try_verify_hosts, which caused the
2222 verification to occur when the EHLO/HELO command was issued. The ACL just
2223 tested the remembered result. Now, if a previous verification attempt has
2224 not happened, "verify = helo" does it there and then.
2226 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2228 TK/03 Fix log output including CR from clamd.
2230 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2231 exist provoked a memory error which could cause a segfault.
2233 PH/15 Installed PCRE 6.2
2235 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2237 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2238 of the problem. Specifically, suggested +O2 rather than +O1 for the
2241 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2243 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2244 record, the port was not correctly logged when the outgoing_port log
2245 selector was set (it logged the transort's default port).
2247 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2248 fallback_hosts, and "hosts" in the smtp transport.
2250 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2251 host errors such as "Connection refused".
2253 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2254 authentication with radiusclient 0.4.9:
2256 - Error returned from rc_read_config was caught wrongly
2257 - Username/password not passed on to radius server due to wrong length.
2259 The presumption is that some radiusclient API changes for 4.51/PH/17
2260 were not taken care of correctly. The code is still untested by me (my
2261 Linux distribution still has 0.3.2 of radiusclient), but it was
2262 contributed by a Radius user.
2264 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2265 expanding the "port" option of the smtp transport.
2267 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2268 while reading a MIME header. Thanks to Tom Hughes for a patch.
2270 PH/24 Include config.h inside local_scan.h so that configuration settings are
2273 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2274 that in an ACL for RCPT (for example), you can examine exactly what was
2277 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2278 commands, but it was not correctly comparing the address with the actual
2279 client host address. Thus, it would show the EHLO address in Received:
2280 header lines when this was not necessary.
2282 PH/27 Added the % operator to ${eval:}.
2284 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2285 it should be ignoring failures (because with -C, for example, it has lost
2286 privilege). It wasn't ignoring creation failures other than "already
2289 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2292 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2293 because an input line was too long, either on its own, or by virtue of
2294 too many continuations, the temporary file was not being removed, and the
2295 return code was incorrect.
2297 PH/31 Missing "BOOL" in function definition in filtertest.c.
2299 PH/32 Applied Sieve patches from the maintainer.
2301 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2303 PH/33 Added "verify = not_blind".
2305 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2306 Local/Makefile (with some defaults set). These are used in built scripts
2307 such as exicyclog, but they have never been used in the exim_install
2308 script (though there are many overriding facilities there). I have
2309 arranged that the exim_install script now takes note of these two
2312 PH/35 Installed configuration files for Dragonfly.
2314 PH/36 When a locally submitted message by a trusted user did not contain a
2315 From: header, and the sender address was obtained from -f or from an SMTP
2316 MAIL command, and the trusted user did not use -F to supply a sender
2317 name, $originator_name was incorrectly used when constructing a From:
2318 header. Furthermore, $originator_name was used for submission mode
2319 messages from external hosts without From: headers in a similar way,
2320 which is clearly wrong.
2322 PH/37 Added control=suppress_local_fixups.
2324 PH/38 When log_selector = +received_sender was set, and the addition of the
2325 sender made the log line's construction buffer exactly full, or one byte
2326 less than full, an overflow happened when the terminating "\n" was
2329 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2330 when the result of a list match is failure because a DNS lookup failed.
2332 PH/40 RM_COMMAND is now used in the building process.
2334 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2335 the "build-* directories that it finds.
2337 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2338 address in a domain literal was a prefix of an interface address.
2340 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2341 when verifying a sender address, unless rewrite_headers is false.
2343 PH/44 Wrote a long comment about why errors_to addresses are verified as
2344 recipients, not senders.
2346 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2347 the ratelimit ACL was added.
2349 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2351 PH/46 Added extra information about PostgreSQL errors to the error string.
2353 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2354 avoiding the need to install special zones in a real server. This is
2355 backwards compatible; if it can't find the fake resolver, it drops back.
2356 Thus, both old and new test suites can be run.
2358 TF/02 Added util/ratelimit.pl
2360 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2361 clock is set back in time.
2363 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2364 Candler <B.Candler@pobox.com>.
2366 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2367 OK for addresses that are the result of redirections.
2369 PH/48 A number of further additions for the benefit of the new test suite,
2370 including a fake gethostbyname() that interfaces to the fake DNS resolver
2373 TF/06 The fix for widen_domains has also been applied to qualify_single and
2374 search_parents which are the other dnslookup options that can cause
2377 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2380 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2382 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2383 service" if it encountered an MX record with an empty target hostname.
2384 The message is now "an MX or SRV record indicated no SMTP service".
2386 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2387 if the DNS of the sending site is misconfigured. This is quite a
2388 common situation. This change restores the behaviour of treating a
2389 helo verification defer as a failure.
2391 PH/51 If self=fail was set on a router, the bounce message did not include the
2392 actual error message.
2398 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2400 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2401 command itself ended in error, the underlying message about the transport
2402 filter timeout was being overwritten with the pipe command error. Now the
2403 underlying error message should be appended to the second error message.
2405 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2407 PH/02 Reduce the amount of output that "make" produces by default. Full output
2408 can still be requested.
2410 PH/03 The warning log line about a condition test deferring for a "warn" verb
2411 was being output only once per connection, rather than after each
2412 occurrence (because it was using the same function as for successful
2413 "warn" verbs). This seems wrong, so I have changed it.
2415 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2416 it should not have, which might have caused a crash in the right
2417 circumstances, but probably never did.
2419 PH/04 Installed a modified version of Tony Finch's patch to make submission
2420 mode fix the return path as well as the Sender: header line, and to
2421 add a /name= option so that you can make the user's friendly name appear
2424 TF/03 Added the control = fakedefer ACL modifier.
2426 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2427 Mark Lowes for thorough testing.
2429 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2431 TK/03 Merged latest SRS patch from Miles Wilton.
2433 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2434 with the definition in sysexits.h (which is #included earlier).
2435 Fortunately, Exim does not actually use EX_OK. The code used to try to
2436 preserve the sysexits.h value, by assumimg that macro definitions were
2437 scanned for macro replacements. I have been disabused of this notion,
2438 so now the code just undefines EX_OK before #including unistd.h.
2440 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2441 in the smtp transport. When a block could not be written in a single
2442 write() function, the timeout was being re-applied to each part-write.
2443 This seems wrong - if the receiver was accepting one byte at a time it
2444 would take for ever. The timeout is now adjusted when this happens. It
2445 doesn't have to be particularly precise.
2447 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2448 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2450 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2451 without a domain if the check to <postmaster@domain> fails.
2453 SC/01 Eximstats: added -xls and the ability to specify output files
2454 (patch written by Frank Heydlauf).
2456 SC/02 Eximstats: use FileHandles for outputing results.
2458 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2460 SC/04 Eximstats: fixed display of large numbers with -nvr option
2462 SC/05 Eximstats: fixed merging of reports with empty tables.
2464 SC/06 Eximstats: added the -include_original_destination flag
2466 SC/07 Eximstats: removed tabs and trailing whitespace.
2468 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2470 TK/06 MBOX spool code: Add real "From " MBOX separator line
2471 so the .eml file is really in mbox format (even though
2472 most programs do not really care). Patch from Alex Miller.
2474 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2475 The latter is generated from $received_to and is only set if the
2476 message has one envelope recipient. SA can use these headers,
2477 obviously out-of-the-box. Patch from Alex Miller.
2479 PH/08 The ${def test on a variable was returning false if the variable's
2480 value was "0", contrary to what the specification has always said!
2481 The result should be true unless the variable is empty.
2483 PH/09 The syntax error of a character other than { following "${if
2484 def:variable_name" (after optional whitespace) was not being diagnosed.
2485 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2486 accidental colon was present, for example, could give incorrect results.
2488 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2489 result is used (not including appendfile, where other changes are about
2492 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2493 This involved changing a lot of size variables from int to off_t. It
2494 should work with maildirs and everything.
2496 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2497 spamd dying while we are connected to it.
2499 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2502 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2503 function for BDB 4.3.
2505 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2507 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2508 the sources for PCRE have been re-arranged and more files are now
2511 PH/15 The code I had for printing potentially long long variables in PH/11
2512 above was not the best (it lost precision). The length of off_t variables
2513 is now inspected at build time, and an appropriate printing format (%ld
2514 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2515 to be "long long int" or "long int". This is needed for the internal
2516 formatting function string_vformat().
2518 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2519 the configuration file to be ":syslog", then the script "guesses" where
2520 the logs files are, rather than using the compiled in default. In our
2521 case the guess is not the same as the compiled default, so the script
2522 suddenly stopped working when I started to use syslog. The patch checks
2523 to see if log_file_path is "". If so, it attempts to read it from exim
2524 with no configuration file to get the compiled in version, before it
2525 falls back to the previous guessing code."
2527 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2528 implementing BATV in an Exim configuration. See NewStuff for the gory
2531 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2532 Makefile that are specific to HP-UX.
2534 PH/18 If the "use_postmaster" option was set for a recipient callout together
2535 with the "random" option, the postmaster address was used as the MAIL
2536 FROM address for the random test, but not for the subsequent recipient
2537 test. It is now used for both.
2539 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2540 patch removes a few documentation additions to RFC 3028, because the
2541 latest draft now contains them. It adds the new en;ascii-case comparator
2542 and a new error check for 8bit text in MIME parts. Comparator and
2543 require names are now matched exactly. I enabled the subaddress
2544 extension, but it is not well tested yet (read: it works for me)."
2546 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2547 rework some of the code of TK/09 above to avoid the hardwired use of
2548 "%lld" and "long long". Replaced the call to snprintf() with a call to
2551 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2552 records point to non-existent hosts", "retry timeout exceeded", and
2553 "retry time not reached for any host after a long failure period".
2555 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2556 experimental DomainKeys support:
2558 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2559 (2) On an error such as an illegally used "control", the wrong name for
2560 the control was given.
2562 These problems did NOT occur unless DomainKeys support was compiled.
2564 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2566 PH/24 Added ${if match_ip condition.
2568 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2569 databases so that it will be absolutely obvious if a crash occurs in the
2570 DB library. This is a regular occurrence (often caused by mis-matched
2573 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2574 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2575 picked up on a user's system that detects such things. There doesn't seem
2576 to be a gcc warning option for this - only an attribute that has to be
2577 put on the function's prototype. It seems that in Fedora Core 4 they have
2578 set this on a number of new functions. No doubt there will be more in due
2581 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2582 specify a transport. However, if an address that was verified by such a
2583 router was the subject of a callout, Exim crashed because it tried to
2584 read the rcpt_include_affixes from the non-existent transport. Now it
2585 just assumes that the setting of that option is false. This bug was
2586 introduced by 4.51/PH/31.
2588 PH/28 Changed -d+all to exclude +memory, because that information is very
2589 rarely of interest, but it makes the output a lot bigger. People tend to
2590 do -d+all out of habit.
2592 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2593 code in os-type was giving problems when libc.so lives in lib64, like on
2596 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2597 aren't the modern standard, and it seems that some systems' include files
2598 don't always have them. Exim was already checking for some of the newer
2599 ones like T_AAAA, and defining it itself. I've added checks for all the
2600 record types that Exim uses.
2602 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2603 not automatically generating a new one, as it is supposed to. This
2604 prevented TLS from working. If the file did exist, but contained invalid
2605 data, a new version was generated, as expected. It was only the case of a
2606 non-existent file that was broken.
2608 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2609 with a change in libdomainkeys > 0.64.
2611 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2612 from DNS. If the selector record carries the flag, it now has
2613 precedence over the domain-wide flag.
2615 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2617 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2618 the use of a port name that isn't defined in /etc/services) occurred, the
2619 message was deferred as in a normal delivery, and thus remained on the
2620 spool, instead of being failed because of the mua_wrapper setting. This
2621 is now fixed, and I tidied up some of the mua_wrapper messages at the
2624 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2625 the messages in a hash of arrays rather than using individual hashes.
2626 This is a bit cleaner and results in dramatic memory savings, albeit
2627 at a slight CPU cost.
2629 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2630 as requested by Marc Sherman.
2632 SC/10 Eximstats: added histograms for user specified patterns as requested
2635 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2637 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2638 fopen() in the content-scanning modules that did not already have it.
2644 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2645 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2647 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2649 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2651 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2652 to test Sieve filters that use "vacation".
2654 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2655 that changes the way the GnuTLS parameters are stored in the cache file.
2656 The new format can be generated externally. For backward compatibility,
2657 if the data in the cache doesn't make sense, Exim assumes it has read an
2658 old-format file, and it generates new data and writes a new file. This
2659 means that you can't go back to an older release without removing the
2662 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2663 work if there are any delivery delays because "one_time" forces the
2664 parent to be marked "delivered", so its unseen clone is never tried
2665 again. For this reason, Exim now forbids the simultaneous setting of
2668 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2669 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2670 entry if you want to know the details. The fix, however, neglected to
2671 consider the case when local delivery batching is involved. The test for
2672 "previously delivered" was not happening when checking to see if an
2673 address could be batched with a previous (undelivered) one; under
2674 certain circumstances this could lead to multiple deliveries to the same
2677 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2678 in its include files, and this causes problems building Exim.
2680 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2681 header_syntax) but Exim was just ignoring anything given after a slash.
2682 In particular, this caused confusion with an attempt to use "verify =
2683 reverse_host_lookup/defer_ok". An error is now given when options are
2684 supplied for verify items that do not have them. (Maybe reverse_host_
2685 lookup should have a defer_ok option, but that's a different point.)
2687 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2688 defined by RFC 821) to 2048, because there were problems with some AUTH
2689 commands, and RFC 1869 says the size should be increased for extended
2690 SMTP commands that take arguments.
2692 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2695 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2696 "unknown" error; now it says that the functionality isn't in the binary.
2698 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2699 an address' error message when a string expansion fails (syntax or
2700 whatever). Otherwise the password may appear in the log. Following change
2701 PH/42 below, there is no longer a chance of it appearing in a bounce
2704 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2706 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2707 of its addresses was ever tried. (Bugzilla bug #2.)
2709 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2710 the result incorrectly in the debug output. (It correctly added a newline
2711 to what was transported.)
2713 TF/01 Added $received_time.
2715 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2716 commented out examples of how to interface to a virus scanner and to
2717 SpamAssassin. Also added commented examples of av_scanner and
2718 spamd_address settings.
2720 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2721 and controls are allowed in which ACLs. There were a couple of minor
2722 errors. Some of the entries in the conditions table (which is a table of
2723 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2724 negation of where the condition IS allowed.
2726 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2728 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2729 header file does not have a version number, so I've had to invent a new
2730 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2731 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2732 radiusclient), but it was contributed by a Radius user.
2734 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2735 files or directories, for OpenSSL.
2737 PH/19 When an Exim process that is running as root has to create an Exim log
2738 file, it does so in a subprocess that runs as exim:exim so as to get the
2739 ownership right at creation (otherwise, other Exim processes might see
2740 the file with the wrong ownership). There was no test for failure of this
2741 fork() call, which would lead to the process getting stuck as it waited
2742 for a non-existent subprocess. Forks do occasionally fail when resources
2743 run out. I reviewed all the other calls to fork(); they all seem to check
2746 PH/20 When checking for unexpected SMTP input at connect time (before writing
2747 the banner), Exim was not dealing correctly with a non-positive return
2748 from the read() function. If the client had disconnected by this time,
2749 the result was a log entry for a synchronization error with an empty
2750 string after "input=" when read() returned zero. If read() returned -1
2751 (an event I could not check), uninitialized data bytes were printed.
2752 There were reports of junk text (parts of files, etc) appearing after
2755 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2757 PH/22 Added support for macro redefinition, and (re)definition in between
2758 driver and ACL definitions.
2760 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2761 forgetting to use the resulting value; it was using the unexpanded value.
2763 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2764 hadn't been configured. The fix is from Juergen Kreileder, who
2765 understands it better than I do:
2767 "Here's what I see happening with three configured cyrus_sasl
2768 authenticators configured (plain, login, cram-md5):
2770 On startup auth_cyrus_sasl_init() gets called for each of these.
2771 This means three calls to sasl_listmech() without a specified mech_list.
2772 => SASL tests which mechs of all available mechs actually work
2773 => three warnings about OTP not working
2774 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2776 With the patch, sasl_listmech() also gets called three times. But now
2777 SASL's mech_list option is set to the server_mech specified in the the
2778 authenticator. Or in other words, the answer from sasl_listmech()
2779 gets limited to just the mech you're testing for (which is different
2781 => the return list contains just 'plain' or 'login', 'cram-md5' or
2782 nothing depending on the value of ob->server_mech.
2784 I've just tested the patch: Authentication still works fine,
2785 unavailable mechs specified in the exim configuration are still
2786 caught, and the auth.log warnings about OTP are gone."
2788 PH/25 When debugging is enabled, the contents of the command line are added
2789 to the debugging output, even when log_selector=+arguments is not
2792 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2793 answer is "GNU", and only if the return is "GNU/something" is the answer
2796 PH/27 $acl_verify_message is now set immediately after the failure of a
2797 verification in an ACL, and so is available in subsequent modifiers. In
2798 particular, the message can be preserved by coding like this:
2800 warn !verify = sender
2801 set acl_m0 = $acl_verify_message
2803 Previously, $acl_verify_message was set only while expanding "message"
2804 and "log_message" when a very denied access.
2806 PH/28 Modified OS/os.c-Linux with
2808 -#ifndef OS_LOAD_AVERAGE
2809 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2811 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2812 nomenclature these days.)
2814 PH/29 Installed patch from the Sieve maintainer that adds the options
2815 sieve_useraddress and sieve_subaddress to the redirect router.
2817 PH/30 In these circumstances:
2818 . Two addresses routed to the same list of hosts;
2819 . First host does not offer TLS;
2820 . First host accepts first address;
2821 . First host gives temporary error to second address;
2822 . Second host offers TLS and a TLS session is established;
2823 . Second host accepts second address.
2824 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2825 and peerdn, if requested) that were in fact used only for the second
2828 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2829 attention to any local part prefix or suffix that was matched by the
2830 router that accepted the address. It now behaves in the same way as it
2831 does for delivery: the affixes are removed from the local part unless
2832 rcpt_include_affixes is set on the transport.
2834 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2835 timeout during the DATA phase of an incoming message.
2837 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2838 applied a patch sanctioned by the Sieve maintainer.
2840 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2841 the uid or gid is negative. A case of a negative gid caused this to be
2842 noticed. The fix allows for either to be negative.
2844 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2845 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2848 PH/36 The change PH/12 above was broken. Fixed it.
2850 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2851 the grounds that routing the same address twice would always produce the
2852 same answer. This might have been true once, but it is certainly no
2853 longer true now. Routing a child address may depend on the previous
2854 routing that produced that child. Some complicated redirection strategies
2855 went wrong when messages had multiple recipients, and made Exim's
2856 behaviour dependent on the order in which the addresses were given.
2858 I have moved the duplicate checking until after the routing is complete.
2859 Exim scans the addresses that are assigned to local and remote
2860 transports, and removes any duplicates. This means that more work will be
2861 done, as duplicates will always all be routed, but duplicates are
2862 presumably rare, so I don't expect this is of any significance.
2864 For deliveries to pipes, files, and autoreplies, the duplicate checking
2865 still happens during the routing process, since they are not going to be
2868 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2869 It corrects a timeout issue with spamd. This is Ian's comment: "The
2870 background is that sometimes spamd either never reads data from a
2871 connection it has accepted, or it never writes response data. The exiscan
2872 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2873 blindly assumes that writes won't block so it may never time out."
2875 PH/39 Allow G after quota size as well as K and M.
2877 PH/40 The value set for $authenticated_id in an authenticator may not contain
2878 binary zeroes or newlines because the value is written to log lines and
2879 to spool files. There was no check on this. Now the value is run through
2880 the string_printing() function so that such characters are converted to
2881 printable escape sequences.
2883 PH/41 $message_linecount is a new variable that contains the total number of
2884 lines in the message. Compare $body_linecount, which is the count for the
2887 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2888 bounce and delay warning messages, except in certain special cases, which
2891 (a) An SMTP error message from a remote host;
2892 (b) A message specified in a :fail: redirection;
2893 (c) A message specified in a "fail" command in a system filter;
2894 (d) A message specified in a FAIL return from the queryprogram router;
2895 (e) A message specified by the cannot_route_message router option.
2897 In these cases only, Exim does include the error details in bounce and
2898 warning messages. There are also a few cases where bland messages such
2899 as "unrouteable address" or "local delivery error" are given.
2901 PH/43 $value is now also set for the "else" part of a ${run expansion.
2903 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2904 being worked on, but at least Exim now implements the latest version to
2907 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2908 process to complete was treated as a delivery failure, a timeout while
2909 writing the message to the pipe was logged, but erroneously treated as a
2910 successful delivery. Such timeouts include transport filter timeouts. For
2911 consistency with the overall process timeout, these timeouts are now
2912 treated as errors, giving rise to delivery failures by default. However,
2913 there is now a new Boolean option for the pipe transport called
2914 timeout_defer, which, if set TRUE, converts the failures into defers for
2915 both kinds of timeout. A transport filter timeout is now identified in
2918 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2919 systems where "make" and "gmake" are different, calling "gmake" at top
2920 level broke things. I've arranged for the value of $(MAKE) to be passed
2921 from the Makefile to this script so that it can call the same version of
2925 A note about Exim versions 4.44 and 4.50
2926 ----------------------------------------
2928 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2929 changes of various kinds. As a consequence, a big documentation update was
2930 needed. This delayed the release for rather longer than seemed good, especially
2931 in the light of a couple of (minor) security issues. Therefore, the changes
2932 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2933 release. So 4.44 and 4.50 are in effect two different branches that both start
2936 I have left the 4.50 change log unchanged; it contains all the changes since
2937 4.43. The change log for 4.44 is below; many of its items are identical to
2938 those for 4.50. This seems to be the most sensible way to preserve the
2939 historical information.
2945 1. Minor wording change to the doc/README.SIEVE file.
2947 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2948 computation of the current number of files was incorrect.
2950 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2951 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2954 4. Give more explanation in the error message when the command for a transport
2955 filter fails to execute.
2957 5. There are several places where Exim runs a non-Exim command in a
2958 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2959 was being done only for the command run by the queryprogram router. It is
2960 now done for all such subprocesses. The other cases are: ${run, transport
2961 filters, and the commands run by the lmtp and pipe transports.
2963 6. Added CONFIGURE_GROUP build-time option.
2965 7. Some older OS have a limit of 256 on the maximum number of file
2966 descriptors. Exim was using setrlimit() to set 1000 as a large value
2967 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2968 systems. I've change it so that if it can't get 1000, it tries for 256.
2970 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2971 was an oversight, and furthermore, ever since the addition of extra
2972 controls (e.g. 4.43/32), the checks on when to allow different forms of
2973 "control" were broken. There should now be diagnostics for all cases when a
2974 control that does not make sense is encountered.
2976 9. Added the /retain_sender option to "control=submission".
2978 10. $recipients is now available in the predata ACL (oversight).
2980 11. Tidy the search cache before the fork to do a delivery from a message
2981 received from the command line. Otherwise the child will trigger a lookup
2982 failure and thereby defer the delivery if it tries to use (for example) a
2983 cached ldap connection that the parent has called unbind on.
2985 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2986 of $address_data from the recipient verification was clobbered by the
2987 sender verification.
2989 13. The value of address_data from a sender verification is now available in
2990 $sender_address_data in subsequent conditions in the ACL statement.
2992 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2994 15. Added a new option "connect=<time>" to callout options, to set a different
2997 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2998 was its contents. (It was OK if the option was not defined at all.)
3000 17. A "Completed" log line is now written for messages that are removed from
3001 the spool by the -Mrm option.
3003 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3004 information about exactly what failed.
3006 19. Added -dd to debug only the daemon process.
3008 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3009 handles timeouts, both on the server side and network timeouts. Renamed the
3010 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3012 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3013 It is now set to "smtps".
3015 22. $host_address is now set to the target address during the checking of
3016 ignore_target_hosts.
3018 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3019 being passed; this would have caused $sender_host_name to have been used if
3020 matching the list had actually called for a host name (not very likely,
3021 since this list is usually IP addresses). A host name is now passed as
3024 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3025 code that specifies a non-restarting handler (typically sigaction() in
3026 modern systems) in an attempt to fix a rare and obscure crash bug.
3028 25. Narrowed the window for a race in the daemon that could cause it to ignore
3029 SIGCHLD signals. This is not a major problem, because they are used only to
3030 wake it up if nothing else does.
3032 26. A malformed maildirsize file could cause Exim to calculate negative values
3033 for the mailbox size or file count. Odd effects could occur as a result.
3034 The maildirsize information is now recalculated if the size or filecount
3037 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3038 support for a long time. Removed HAVE_SYS_VFS_H.
3040 28. Installed the latest version of exipick from John Jetmore.
3042 29. In an address list, if the pattern was not a regular expression, an empty
3043 subject address (from a bounce message) matched only if the pattern was an
3044 empty string. Non-empty patterns were not even tested. This was the wrong
3045 because it is perfectly reasonable to use an empty address as part of a
3046 database query. An empty address is now tested by patterns that are
3047 lookups. However, all the other forms of pattern expect the subject to
3048 contain a local part and a domain, and therefore, for them, an empty
3049 address still always fails if the pattern is not itself empty.
3051 30. Exim went into a mad DNS loop when attempting to do a callout where the
3052 host was specified on an smtp transport, and looking it up yielded more
3053 than one IP address.
3055 31. Re-factored the code for checking spool and log partition space into a
3056 function that finds that data and another that does the check. The former
3057 is then used to implement four new variables: $spool_space, $log_space,
3058 $spool_inodes, and $log_inodes.
3060 32. The RFC2047 encoding function was originally intended for short strings
3061 such as real names; it was not keeping to the 75-character limit for
3062 encoded words that the RFC imposes. It now respects the limit, and
3063 generates multiple encoded words if necessary. To be on the safe side, I
3064 have increased the buffer size for the ${rfc2047: expansion operator from
3067 33. It is now permitted to omit both strings after an "if" condition; if the
3068 condition is true, the result is "true". As before, when the second string
3069 is omitted, a false condition yields an empty string. This makes it less
3070 cumbersome to write custom ACL and router conditions.
3072 34. Failure to deliver a bounce message always caused it to be frozen, even if
3073 there was an errors_to setting on the router. The errors_to setting is now
3076 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3077 canonical form (fully expanded) before being placed in
3078 $sender_host_address.
3080 36. The table in the code that translates DNS record types into text (T_A to
3081 "A" for instance) was missing entries for NS and CNAME. It is just possible
3082 that this could have caused confusion if both these types were looked up
3083 for the same domain, because the text type is used as part of Exim's
3084 per-process caching. But the chance of anyone hitting this buglet seems
3087 37. The dnsdb lookup has been extended in a number of ways.
3089 (1) There is a new type, "zns", which walks up the domain tree until it
3090 finds some nameserver records. It should be used with care.
3092 (2) There is a new type, "mxh", which is like "mx" except that it returns
3093 just the host names, not the priorities.
3095 (3) It is now possible to give a list of domains (or IP addresses) to be
3096 looked up. The behaviour when one of the lookups defers can be
3097 controlled by a keyword.
3099 (4) It is now possible to specify the separator character for use when
3100 multiple records are returned.
3102 38. The dnslists ACL condition has been extended: it is now possible to supply
3103 a list of IP addresses and/or domains to be looked up in a particular DNS
3106 39. Added log_selector=+queue_time_overall.
3108 40. When running the queue in the test harness, wait just a tad after forking a
3109 delivery process, to get repeatability of debugging output.
3111 41. Include certificate and key file names in error message when GnuTLS fails
3112 to set them up, because the GnuTLS error message doesn't include the name
3113 of the failing file when there is a problem reading it.
3115 42. Allow both -bf and -bF in the same test run.
3117 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3119 44. The "Exiscan patch" is now merged into the mainline Exim source.
3121 45. Sometimes the final signoff response after QUIT could fail to get
3122 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3123 < 0 before doing a fflush(). This bug looks as though it goes back to the
3124 introduction of TLS in release 3.20, but "sometimes" must have been rare
3125 because the tests only now provoked it.
3127 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3128 (this can affect the format of dates).
3130 47. exim_tidydb, when checking for the continued existence of a message for
3131 which it has found a message-specific retry record, was not finding
3132 messages that were in split spool directories. Consequently, it was
3133 deleting retry records that should have stayed in existence.
3135 48. Steve fixed some bugs in eximstats.
3137 49. The SPA authentication driver was not abandoning authentication and moving
3138 on to the next authenticator when an expansion was forced to fail,
3139 contradicting the general specification for all authenticators. Instead it
3140 was generating a temporary error. It now behaves as specified.
3142 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3143 (the order specifies the preference for clients). The order is now AES256,
3144 AES128, 3DES, ARCFOUR128.
3146 51. Small patch to Sieve code - explicitly set From: when generating an
3149 52. Exim crashed if a remote delivery caused a very long error message to be
3150 recorded - for instance if somebody sent an entire SpamAssassin report back
3151 as a large number of 550 error lines. This bug was coincidentally fixed by
3152 increasing the size of one of Exim's internal buffers (big_buffer) that
3153 happened as part of the Exiscan merge. However, to be on the safe side, I
3154 have made the code more robust (and fixed the comments that describe what
3157 53. Now that there can be additional text after "Completed" in log lines (if
3158 the queue_time_overall log selector is set), a one-byte patch to exigrep
3159 was needed to allow it to recognize "Completed" as not the last thing in
3162 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3163 patch that reportedly fixes this has been added. I am not expert enough to
3164 create a test for it. This is what the patch creator wrote:
3166 "I found a little strange behaviour of ldap code when working with
3167 Windows 2003 AD Domain, where users was placed in more than one
3168 Organization Units. When I tried to give exim partial DN, the exit code
3169 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3170 But simultaneously result of request was absolutely normal ldap result,
3171 so I produce this patch..."
3173 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3174 REFERENCE, so I have modified the code to exclude the patch when that macro
3177 55. Some experimental protocols are using DNS PTR records for new purposes. The
3178 keys for these records are domain names, not reversed IP addresses. The
3179 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3180 leaves it alone. Component reversal etc. now happens only for IP addresses.
3183 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3185 57. Double the size of the debug message buffer (to 2048) so that more of very
3186 long debug lines gets shown.
3188 58. The exicyclog utility now does better if the number of log files to keep
3189 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3191 59. Two changes related to the smtp_active_hostname option:
3193 (1) $smtp_active_hostname is now available as a variable.
3194 (2) The default for smtp_banner uses $smtp_active_hostname instead
3195 of $primary_hostname.
3197 60. The host_aton() function is supposed to be passed a string that is known
3198 to be a valid IP address. However, in the case of IPv6 addresses, it was
3199 not checking this. This is a hostage to fortune. Exim now panics and dies
3200 if the condition is not met. A case was found where this could be provoked
3201 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3202 components; fortuitously, this particular loophole had already been fixed
3203 by change 4.50/55 above.
3205 If there are any other similar loopholes, the new check in host_aton()
3206 itself should stop them being exploited. The report I received stated that
3207 data on the command line could provoke the exploit when Exim was running as
3208 exim, but did not say which command line option was involved. All I could
3209 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3210 running as the user.
3213 61. There was a buffer overflow vulnerability in the SPA authentication code
3214 (which came originally from the Samba project). I have added a test to the
3215 spa_base64_to_bits() function which I hope fixes it.
3218 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3219 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3221 63. The daemon start-up calls getloadavg() while still root for those OS that
3222 need the first call to be done as root, but it missed one case: when
3223 deliver_queue_load_max is set with deliver_drop_privilege. This is
3224 necessary for the benefit of the queue runner, because there is no re-exec
3225 when deliver_drop_privilege is set.
3227 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3228 This has been fixed.
3230 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3231 was in use, was not putting the data itself into the right store pool;
3232 consequently, it could be overwritten for a subsequent message in the same
3233 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3236 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3238 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3239 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3240 the function were treating the return as a boolean value, which happened to
3241 work because 0=false and not-0=true, but is not correct code.
3243 68. The host_aton() function was not handling scoped IPv6 addresses (those
3244 with, for example, "%eth0" on the end) correctly.
3246 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3247 negated items (that is, ~something) in unsigned ints. Some compilers
3248 apparently mutter when there is no cast.
3250 70. If an address verification called from an ACL failed, and did not produce a
3251 user-specific message (i.e. there was only a "system" message), nothing was
3252 put in $acl_verify_message. In this situation, it now puts the system
3255 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3256 session; change 4.31/43 added the unwanted input to the log line - except
3257 that it did not do this in the start of session case. It now does.
3259 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3260 This is wrong and can cause the other end to generate a synchronization
3261 error if it is another Exim or anything else that does the synchronization
3262 check. A QUIT command is no longer sent after a timeout.
3264 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3265 during host lookups.
3267 74. The defer_ok option of callout verification was not working if it was used
3268 when verifying addresses in header lines, that is, for this case:
3270 verify = header_sender/callout=defer_ok
3272 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3273 those file descriptors could be used for SMTP connections. If anything
3274 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3275 could be sent to the SMTP client, causing chaos. The daemon now opens
3276 stdin, stdout, and stderr to /dev/null when it puts itself into the
3279 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3280 log by default. The user can override this with suitable Perl magic.
3282 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3283 the log message when discard triggers, was not working for the DATA ACL or
3284 for the non-SMTP ACL.
3286 78. Error message wording change in sieve.c.
3288 79. If smtp_accept_max_per_host was set, the number of connections could be
3289 restricted to fewer than expected, because the daemon was trying to set up
3290 a new connection before checking whether the processes handling previous
3291 connections had finished. The check for completed processes is now done
3292 earlier. On busy systems, this bug wouldn't be noticed because something
3293 else would have woken the daemon, and it would have reaped the completed
3296 80. If a message was submitted locally by a user whose login name contained one
3297 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3298 It caused a spool format error. I have fixed the spool reading code. A
3299 related problem was that the "from" clause in the Received: line became
3300 illegal because of the space(s). It is now covered by ${quote_local_part.
3302 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3305 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3307 83. Patch from Sieve maintainer to fix unterminated string problem in
3308 "vacation" handling.
3310 84. Some minor changes to the Linux configuration files to help with other
3311 OS variants using glibc.
3313 85. One more patch for Sieve to update vacation handling to latest spec.
3316 ----------------------------------------------------
3317 See the note above about the 4.44 and 4.50 releases.
3318 ----------------------------------------------------
3324 1. Change 4.43/35 introduced a bug that caused file counts to be
3325 incorrectly computed when quota_filecount was set in an appendfile
3328 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3329 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3332 3. Give more explanation in the error message when the command for a transport
3333 filter fails to execute.
3335 4. There are several places where Exim runs a non-Exim command in a
3336 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3337 was being done only for the command run by the queryprogram router. It is
3338 now done for all such subprocesses. The other cases are: ${run, transport
3339 filters, and the commands run by the lmtp and pipe transports.
3341 5. Some older OS have a limit of 256 on the maximum number of file
3342 descriptors. Exim was using setrlimit() to set 1000 as a large value
3343 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3344 systems. I've change it so that if it can't get 1000, it tries for 256.
3346 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3347 was an oversight, and furthermore, ever since the addition of extra
3348 controls (e.g. 4.43/32), the checks on when to allow different forms of
3349 "control" were broken. There should now be diagnostics for all cases when a
3350 control that does not make sense is encountered.
3352 7. $recipients is now available in the predata ACL (oversight).
3354 8. Tidy the search cache before the fork to do a delivery from a message
3355 received from the command line. Otherwise the child will trigger a lookup
3356 failure and thereby defer the delivery if it tries to use (for example) a
3357 cached ldap connection that the parent has called unbind on.
3359 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3360 of $address_data from the recipient verification was clobbered by the
3361 sender verification.
3363 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3364 was its contents. (It was OK if the option was not defined at all.)
3366 11. A "Completed" log line is now written for messages that are removed from
3367 the spool by the -Mrm option.
3369 12. $host_address is now set to the target address during the checking of
3370 ignore_target_hosts.
3372 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3373 being passed; this would have caused $sender_host_name to have been used if
3374 matching the list had actually called for a host name (not very likely,
3375 since this list is usually IP addresses). A host name is now passed as
3378 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3379 code that specifies a non-restarting handler (typically sigaction() in
3380 modern systems) in an attempt to fix a rare and obscure crash bug.
3382 15. Narrowed the window for a race in the daemon that could cause it to ignore
3383 SIGCHLD signals. This is not a major problem, because they are used only to
3384 wake it up if nothing else does.
3386 16. A malformed maildirsize file could cause Exim to calculate negative values
3387 for the mailbox size or file count. Odd effects could occur as a result.
3388 The maildirsize information is now recalculated if the size or filecount
3391 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3392 support for a long time. Removed HAVE_SYS_VFS_H.
3394 18. Updated exipick to current release from John Jetmore.
3396 19. Allow an empty sender to be matched against a lookup in an address list.
3397 Previously the only cases considered were a regular expression, or an
3400 20. Exim went into a mad DNS lookup loop when doing a callout where the
3401 host was specified on the transport, if the DNS lookup yielded more than
3404 21. The RFC2047 encoding function was originally intended for short strings
3405 such as real names; it was not keeping to the 75-character limit for
3406 encoded words that the RFC imposes. It now respects the limit, and
3407 generates multiple encoded words if necessary. To be on the safe side, I
3408 have increased the buffer size for the ${rfc2047: expansion operator from
3411 22. Failure to deliver a bounce message always caused it to be frozen, even if
3412 there was an errors_to setting on the router. The errors_to setting is now
3415 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3416 canonical form (fully expanded) before being placed in
3417 $sender_host_address.
3419 24. Updated eximstats to version 1.33
3421 25. Include certificate and key file names in error message when GnuTLS fails
3422 to set them up, because the GnuTLS error message doesn't include the name
3423 of the failing file when there is a problem reading it.
3425 26. Expand error message when OpenSSL has problems setting up cert/key files.
3428 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3429 (this can affect the format of dates).
3431 28. exim_tidydb, when checking for the continued existence of a message for
3432 which it has found a message-specific retry record, was not finding
3433 messages that were in split spool directories. Consequently, it was
3434 deleting retry records that should have stayed in existence.
3436 29. eximstats updated to version 1.35
3437 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3438 1.35 - bugfix such that pie charts by volume are generated correctly
3440 30. The SPA authentication driver was not abandoning authentication and moving
3441 on to the next authenticator when an expansion was forced to fail,
3442 contradicting the general specification for all authenticators. Instead it
3443 was generating a temporary error. It now behaves as specified.
3445 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3446 (the order specifies the preference for clients). The order is now AES256,
3447 AES128, 3DES, ARCFOUR128.
3449 31. Small patch to Sieve code - explicitly set From: when generating an
3452 32. Exim crashed if a remote delivery caused a very long error message to be
3453 recorded - for instance if somebody sent an entire SpamAssassin report back
3454 as a large number of 550 error lines. This bug was coincidentally fixed by
3455 increasing the size of one of Exim's internal buffers (big_buffer) that
3456 happened as part of the Exiscan merge. However, to be on the safe side, I
3457 have made the code more robust (and fixed the comments that describe what
3460 33. Some experimental protocols are using DNS PTR records for new purposes. The
3461 keys for these records are domain names, not reversed IP addresses. The
3462 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3463 leaves it alone. Component reversal etc. now happens only for IP addresses.
3466 34. The host_aton() function is supposed to be passed a string that is known
3467 to be a valid IP address. However, in the case of IPv6 addresses, it was
3468 not checking this. This is a hostage to fortune. Exim now panics and dies
3469 if the condition is not met. A case was found where this could be provoked
3470 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3471 components; fortuitously, this particular loophole had already been fixed
3472 by change 4.50/55 or 4.44/33 above.
3474 If there are any other similar loopholes, the new check in host_aton()
3475 itself should stop them being exploited. The report I received stated that
3476 data on the command line could provoke the exploit when Exim was running as
3477 exim, but did not say which command line option was involved. All I could
3478 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3479 running as the user.
3482 35. There was a buffer overflow vulnerability in the SPA authentication code
3483 (which came originally from the Samba project). I have added a test to the
3484 spa_base64_to_bits() function which I hope fixes it.
3487 36. The daemon start-up calls getloadavg() while still root for those OS that
3488 need the first call to be done as root, but it missed one case: when
3489 deliver_queue_load_max is set with deliver_drop_privilege. This is
3490 necessary for the benefit of the queue runner, because there is no re-exec
3491 when deliver_drop_privilege is set.
3493 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3494 was in use, was not putting the data itself into the right store pool;
3495 consequently, it could be overwritten for a subsequent message in the same
3496 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3499 38. Sometimes the final signoff response after QUIT could fail to get
3500 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3501 < 0 before doing a fflush(). This bug looks as though it goes back to the
3502 introduction of TLS in release 3.20, but "sometimes" must have been rare
3503 because the tests only now provoked it.
3509 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3510 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3511 Now it returns an integer. A number of calls were still expecting a T/F
3512 return. Fortuitously, in all cases, the tests worked in OK situations,
3513 which is the norm. However, things would have gone wrong on any write
3514 failures on the smtp file descriptor. This function is used when sending
3515 messages over SMTP and also when doing verify callouts.
3517 2. When Exim is called to do synchronous delivery of a locally submitted
3518 message (the -odf or -odi options), it no longer closes stderr before doing
3521 3. Implemented the mua_wrapper option.
3523 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3525 5. Implemented the functions header_remove(), header_testname(),
3526 header_add_at_position(), and receive_remove_recipient(), and exported them
3529 6. If an ACL "warn" statement specified the addition of headers, Exim already
3530 inserted X-ACL-Warn: at the start if there was no header name. However, it
3531 was not making this test for the second and subsequent header lines if
3532 there were newlines in the string. This meant that an invalid header could
3533 be inserted if Exim was badly configured.
3535 7. Allow an ACL "warn" statement to add header lines at the start or after all
3536 the Received: headers, as well as at the end.
3538 8. Added the rcpt_4xx retry error code.
3540 9. Added postmaster_mailfrom=xxx to callout verification option.
3542 10. Added mailfrom=xxxx to the callout verification option, for verify=
3545 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3546 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3548 12. Inserted some casts to stop certain compilers warning when using pointer
3549 differences as field lengths or precisions in printf-type calls (mostly
3550 affecting debugging statements).
3552 13. Added optional readline() support for -be (dynamically loaded).
3554 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3555 same clock tick as a message's arrival, so that its received time was the
3556 same as the "first fail" time on the retry record, and that message
3557 remained on the queue past the ultimate address timeout, every queue runner
3558 would try a delivery (because it was past the ultimate address timeout) but
3559 after another failure, the ultimate address timeout, which should have then
3560 bounced the address, did not kick in. This was a "< instead of <=" error;
3561 in most cases the first failure would have been in the next clock tick
3562 after the received time, and all would be well.
3564 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3565 being recognized when the domain list was tested by the match_domain
3566 condition in an expansion string.
3568 16. Added the ${str2b64: operator.
3570 17. Exim was always calling setrlimit() to set a large limit for the number of
3571 processes, without checking whether the existing limit was already
3572 adequate. (It did check for the limit on file descriptors.) Furthermore,
3573 errors from getrlimit() and setrlimit() were being ignored. Now they are
3574 logged to the main and panic logs, but Exim does carry on, to try to do its
3575 job under whatever limits there are.
3577 18. Imported PCRE 5.0.
3579 19. Trivial typo in log message " temporarily refused connection" (the leading
3582 20. If the log selector return_path_on_delivery was set and an address was
3583 redirected to /dev/null, the delivery process crashed because it assumed
3584 that a return path would always be set for a "successful" delivery. In this
3585 case, the whole delivery is bypassed as an optimization, and therefore no
3588 21. Internal re-arrangement: the function for sending a challenge and reading
3589 a response while authentication was assuming a zero-terminated challenge
3590 string. It's now changed to take a pointer and a length, to allow for
3591 binary data in such strings.
3593 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3595 23. Exim was not respecting finduser_retries when seeking the login of the
3596 uid under which it was called; it was always trying 10 times. (The default
3597 setting of finduser_retries is zero.) Also, it was sleeping after the final
3598 failure, which is pointless.
3600 24. Implemented tls_on_connect_ports.
3602 25. Implemented acl_smtp_predata.
3604 26. If the domain in control=submission is set empty, Exim assumes that the
3605 authenticated id is a complete email address when it generates From: or
3606 Sender: header lines.
3608 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3609 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3610 chown and chgrp in /bin and hostname in /usr/bin.
3612 28. Exim was keeping the "process log" file open after each use, just as it
3613 does for the main log. This opens the possibility of it remaining open for
3614 long periods when the USR1 signal hits a daemon. Occasional processlog
3615 errors were reported, that could have been caused by this. Anyway, it seems
3616 much more sensible not to leave this file open at all, so that is what now
3619 29. The long-running daemon process does not normally write to the log once it
3620 has entered its main loop, and it closes the log before doing so. This is
3621 so that log files can straightforwardly be renamed and moved. However,
3622 there are a couple of unusual error situations where the daemon does write
3623 log entries, and I had neglected to close the log afterwards.
3625 30. The text of an SMTP error response that was received during a remote
3626 delivery was being truncated at 512 bytes. This is too short for some of
3627 the long messages that one sometimes sees. I've increased the limit to
3630 31. It is now possible to make retry rules that apply only when a message has a
3631 specific sender, in particular, an empty sender.
3633 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3634 it possible to be selective about when SMTP synchronization is enforced.
3636 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3638 32. Implemented hosts_connection_nolog.
3640 33. Added an ACL for QUIT.
3642 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3645 35. Added mailbox_size and mailbox_filecount to appendfile.
3647 36. Added control = no_multiline_responses to ACLs.
3649 37. There was a bug in the logic of the code that waits for the clock to tick
3650 in the case where the clock went backwards by a substantial amount such
3651 that the microsecond fraction of "now" was more than the microsecond
3652 fraction of "then" (but the whole seconds number was less).
3654 38. Added support for the libradius Radius client library this is found on
3655 FreeBSD (previously only the radiusclient library was supported).
3661 1. When certain lookups returned multiple values in the form name=value, the
3662 quoting of the values was not always being done properly. Specifically:
3663 (a) If the value started with a double quote, but contained no whitespace,
3665 (b) If the value contained whitespace other than a space character (i.e.
3666 tabs or newlines or carriage returns) it was not quoted.
3667 This fix has been applied to the mysql and pgsql lookups by writing a
3668 separate quoting function and calling it from the lookup code. The fix
3669 should probably also be applied to nisplus, ibase and oracle lookups, but
3670 since I cannot test any of those, I have not disturbed their existing code.
3672 2. A hit in the callout cache for a specific address caused a log line with no
3673 reason for rejecting RCPT. Now it says "Previous (cached) callout
3674 verification failure".
3676 3. There was an off-by-one bug in the queryprogram router. An over-long
3677 return line was truncated at 256 instead of 255 characters, thereby
3678 overflowing its buffer with the terminating zero. As well as fixing this, I
3679 have increased the buffer size to 1024 (and made a note to document this).
3681 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3682 when Exim is waiting for an SMTP response from a remote server, Exim
3683 restarts its select() call on the socket, thereby resetting its timeout.
3684 This is not a problem when such interrupts are rare. Somebody set up a cron
3685 job to run exiwhat every 2 minutes, which is less than the normal select()
3686 timeout (5 or 10 minutes). This meant that the select() timeout never
3687 kicked in because it was always reset. I have fixed this by comparing the
3688 time when an interrupt arrives with the time at the start of the first call
3689 to select(). If more time than the timeout has elapsed, the interrupt is
3690 treated as a timeout.
3692 5. Some internal re-factoring in preparation for the addition of Sieve
3693 extensions (by MH). In particular, the "personal" test is moved to a
3694 separate function, and given an option for scanning Cc: and Bcc: (which is
3695 not set for Exim filters).
3697 6. When Exim created an email address using the login of the caller as the
3698 local part (e.g. when creating a From: or Sender: header line), it was not
3699 quoting the local part when it contained special characters such as @.
3701 7. Installed new OpenBSD configuration files.
3703 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3704 try to make them clearer.
3706 9. Callout options, other than the timeout value, were being ignored when
3707 verifying sender addresses in header lines. For example, when using
3709 verify = header_sender/callout=no_cache
3711 the cache was (incorrectly) being used.
3713 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3714 only the Cygwin environment.
3716 11. When return_path_on_delivery was set as a log selector, if different remote
3717 addresses in the same message used different return paths and parallel
3718 remote delivery occurred, the wrong values would sometimes be logged.
3719 (Whenever a remote delivery process finished, the return path value from
3720 the most recently started remote delivery process was logged.)
3722 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3723 lines when AUTH and/or TLS are in use. This is the "received protocol"
3724 field. Exim used to use "asmtp" for authenticated SMTP, without any
3725 indication (in the protocol name) for TLS use. Now it follows the RFC and
3726 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3727 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3728 names appear in log lines as well as in Received: header lines.
3730 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3731 extensions, and comparison tests, and to fix some bugs.
3733 14. Changes to the "personal" filter test:
3735 (1) The test was buggy in that it was just doing the equivalent of
3736 "contains" tests on header lines. For example, if a user's address was
3737 anne@some.where, the "personal" test would incorrectly be true for
3739 To: susanne@some.where
3741 This test is now done by extracting each address from the header in turn,
3742 and checking the entire address. Other tests that are part of "personal"
3743 are now done using regular expressions (for example, to check local parts
3744 of addresses in From: header lines).
3746 (2) The list of non-personal local parts in From: addresses has been
3747 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3748 taken from the Sieve specification recommendations.
3750 (3) If the message contains any header line starting with "List-" it is
3751 treated as non-personal.
3753 (4) The test for "circular" in the Subject: header line has been removed
3754 because it now seems ill-conceived.
3756 15. Minor typos in src/EDITME comments corrected.
3758 16. Installed latest exipick from John Jetmore.
3760 17. If headers_add on a router specified a text string that was too long for
3761 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3762 of string_sprintf() is now avoided.
3764 18. $message_body_size was not set (it was always zero) when running the DATA
3765 ACL and the local_scan() function.
3767 19. For the "mail" command in an Exim filter, no default was being set for
3768 the once_repeat time, causing a random time value to be used if "once" was
3769 specified. (If the value happened to be <= 0, no repeat happened.) The
3770 default is now 0s, meaning "never repeat". The "vacation" command was OK
3771 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3772 (I found it when inspecting the code).
3774 20. There is now an overall timeout for performing a callout verification. It
3775 defaults to 4 times the callout timeout, which applies to individual SMTP
3776 commands during the callout. The overall timeout applies when there is more
3777 than one host that can be tried. The timeout is checked before trying the
3778 next host. This prevents very long delays if there are a large number of
3779 hosts and all are timing out (e.g. when the network connections are timing
3780 out). The value of the overall timeout can be changed by specifying an
3781 additional sub-option for "callout", called "maxwait". For example:
3783 verify = sender/callout=5s,maxwait=20s
3785 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3786 to the end before writing, but this should make it even safer).
3788 22. Exim was forgetting that it had advertised PIPELINING for the second and
3789 subsequent messages on an SMTP connection. It was also not resetting its
3790 memory on STARTTLS and an internal HELO.
3792 23. When Exim logs an SMTP synchronization error within a session, it now
3793 records whether PIPELINING has been advertised or not.
3795 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3796 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3797 rather than long int.
3799 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3801 26. Added the never_mail option to autoreply.
3807 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3808 crash if the getsockname() call failed; this can happen if a connection is
3809 closed very soon after it is established. The problem was simply in the
3810 order in which certain operations were done, causing Exim to try to write
3811 to the SMTP stream before it had set up the file descriptor. The bug has
3812 been fixed by making things happen in the correct order.
3818 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3819 before the connection was closed, thus losing the rejection response.
3821 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3822 some early Solaris releases, but causes trouble in current releases where
3823 socklen_t is defined.
3825 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3828 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3831 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3832 that releasing the top of it at the end releases what was used for sub-
3833 expansions (unless the block got too big). However, discard this block if
3834 the first thing is a variable or header, so that we can use its block when
3835 it is dynamic (useful for very large $message_headers, for example).
3837 6. Lookups now cache *every* query, not just the most recent. A new, separate
3838 store pool is used for this. It can be recovered when all lookup caches are
3839 flushed. Lookups now release memory at the end of their result strings.
3840 This has involved some general refactoring of the lookup sources.
3842 7. Some code has been added to the store_xxx() functions to reduce the amount
3843 of flapping under certain conditions.
3845 8. log_incoming_interface used to affect only the <= reception log lines. Now
3846 it causes the local interface and port to be added to several more SMTP log
3847 lines, for example "SMTP connection from", and rejection lines.
3849 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3851 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3853 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3854 could be overwritten at the end of the current message (or the start of a
3855 new message if it was set in a HELO ACL). The value is now preserved for
3856 the duration of the SMTP connection.
3858 12. If a transport had a headers_rewrite setting, and a matching header line
3859 contained an unqualified address, that address was qualified, even if it
3860 did not match any rewriting rules. The underlying bug was that the values
3861 of the flags that permit the existence of unqualified sender and recipient
3862 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3863 non-local messages, and by -bnq for local messages) were not being
3864 preserved with the message after it was received.
3866 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3867 "next input=" as part of the text comprising the host identity instead of
3868 the correct text. The code was using the same buffer for two different
3869 strings. However, depending on which order the printing function evaluated
3870 its arguments, the bug did not always show up. Under Linux, for example, my
3871 test suite worked just fine.
3873 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3874 doesn't work with some older versions of Perl. It has been changed to "my",
3875 which in any case is probably the better facility to use.
3877 15. A really picky compiler found some instances of statements for creating
3878 error messages that either had too many or two few arguments for the format
3881 16. The size of the buffer for calls to the DNS resolver has been increased
3882 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3883 for addresses that have a lot of PTR records. This alleviates a problem; it
3884 does not fully solve it.
3886 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3887 buffer now truncates the list and logs the incident, which is the same
3888 action as happens when Exim is looking up a host name and its aliases.
3889 Previously in this situation something unpredictable would happen;
3890 sometimes it was "internal error: store_reset failed".
3892 18. If a server dropped the connection unexpectedly when an Exim client was
3893 using GnuTLS and trying to read a response, the client delivery process
3894 crashed while trying to generate an error log message.
3896 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3897 string, for example:
3899 warn message = H1: something\nH2: something
3901 the text was added as a single header line from Exim's point of view
3902 though it ended up OK in the delivered message. However, searching for the
3903 second and subsequent header lines using $h_h2: did not work. This has been
3904 fixed. Similarly, if a system filter added multiple headers in this way,
3905 the routers could not see them.
3907 20. Expanded the error message when iplsearch is called with an invalid key to
3908 suggest using net-iplsearch in a host list.
3910 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3911 ACLs are no longer actually imposed (and a message to that effect is
3914 22. If a "gecos" field in a passwd entry contained escaped characters, in
3915 particular, if it contained a \" sequence, Exim got it wrong when building
3916 a From: or a Sender: header from that name. A second bug also caused
3917 incorrect handling when an unquoted " was present following a character
3918 that needed quoting.
3920 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3921 was not being matched caselessly.
3923 24. Arranged for all hyphens in the exim.8 source to be escaped with
3926 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3927 properly. Recipient callout cache records were still being keyed to include
3928 the sender, even when use_sender was set false. This led to far more
3929 callouts that were necessary. The sender is no longer included in the key
3930 when use_sender is false.
3932 26. Added "control = submission" modifier to ACLs.
3934 27. Added the ${base62d: operator to decode base 62 numbers.
3936 28. dnsdb lookups can now access SRV records.
3938 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3939 the configuration file.
3941 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3942 (-v) mode. This makes the output for a verbose queue run more intelligible.
3944 31. Added a use_postmaster feature to recipient callouts.
3946 32. Added the $body_zerocount variable, containing the number of binary zero
3947 bytes in the message body.
3949 33. The time of last modification of the "new" subdirectory is now used as the
3950 "mailbox time last read" when there is a quota error for a maildir
3953 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3955 35. Added +ignore_unknown as a special item in host lists.
3957 36. Code for decoding IPv6 addresses in host lists is now included, even if
3958 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3959 address was recognized as an IP address, but was then not correctly decoded
3960 into binary, causing unexpected and incorrect effects when compared with
3967 1. Very minor rewording of debugging text in manualroute to say "list of
3968 hosts" instead of "hostlist".
3970 2. If verify=header_syntax was set, and a header line with an unqualified
3971 address (no domain) and a large number of spaces between the end of the
3972 name and the colon was received, the reception process suffered a buffer
3973 overflow, and (when I tested it) crashed. This was caused by some obsolete
3974 code that should have been removed. The fix is to remove it!
3976 3. When running in the test harness, delay a bit after writing a bounce
3977 message to get a bit more predictability in the log output.
3979 4. Added a call to search_tidyup() just before forking a reception process. In
3980 theory, someone could use a lookup in the expansion of smtp_accept_max_
3981 per_host which, without the tidyup, could leave open a database connection.
3983 5. Added the variables $recipient_data and $sender_data which get set from a
3984 lookup success in an ACL "recipients" or "senders" condition, or a router
3985 "senders" option, similar to $domain_data and $local_part_data.
3987 6. Moved the writing of debug_print from before to after the "senders" test
3990 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3991 problems for message scanning, either using a data ACL, or using
3992 local_scan() because the Received: header was not generated till after they
3993 were called (in order to set the time as the time of reception completion).
3994 I have revised the way this works. The header is now generated after the
3995 body is received, but before the ACL or local_scan() are called. After they
3996 are run, the timestamp in the header is updated.
4002 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4003 before starting a queue runner without re-exec. This happened only when
4004 deliver_drop_privilege was set or when the Exim user was set to root. The
4005 effect of the bug was that timeouts during subsequent deliveries caused
4006 crashes instead of being properly handled. The handler is now left at its
4007 default (and expected) setting.
4009 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4010 message, again when deliver_drop_privilege is set or Exim is run as root.
4011 The bug described in (1) was not present in this case, but the tidying up
4012 of the other signals was missing. I have made the two cases consistent.
4014 3. The ignore_target_hosts setting on a manualroute router was being ignored
4015 for hosts that were looked up using the /MX notation.
4017 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4020 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4021 operated on the sender address. After changing the $sender_address to <>
4022 for the sender address verify, Exim was re-instated it as the original
4023 (before rewriting) address, but remembering that it had rewritten it, so it
4024 wasn't rewriting it again. This bug also had the effect of breaking the
4025 sender address verification caching when the sender address was rewritten.
4027 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4028 This has been changed so that if the ip literal address matches
4029 ignore_target_hosts, the router declines.
4031 7. Added expansion conditions match_domain, match_address, and match_local_
4032 part (NOT match_host).
4034 8. The placeholder for the Received: header didn't have a length field set.
4036 9. Added code to Exim itself and to exim_lock to test for a specific race
4037 condition that could lead to file corruption when using MBX delivery. The
4038 issue is with the lockfile that is created in /tmp. If this file is removed
4039 after a process has opened it but before that process has acquired a lock,
4040 there is the potential for a second process to recreate the file and also
4041 acquire a lock. This could lead to two Exim processes writing to the file
4042 at the same time. The added code performs the same test as UW imapd; it
4043 checks after acquiring the lock that its file descriptor still refers to
4044 the same named file.
4046 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4047 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4050 11. Added the smtp_active_hostname option. If used, this will typically be made
4051 to depend on the incoming interface address. Because $interface_address is
4052 not set up until the daemon has forked a reception process, error responses
4053 that can happen earlier (such as "too many connections") no longer contain
4056 12. If an expansion in a condition on a "warn" statement fails because a lookup
4057 defers, the "warn" statement is abandoned, and the next ACL statement is
4058 processed. Previously this caused the whole ACL to be aborted.
4060 13. Added the iplsearch lookup type.
4062 14. Added ident_timeout as a log selector.
4064 15. Added tls_certificate_verified as a log selector.
4066 16. Added a global option tls_require_ciphers (compare the smtp transport
4067 option of the same name). This controls incoming TLS connections.
4069 17. I finally figured out how to make tls_require_ciphers do a similar thing
4070 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4071 before starting the TLS session.
4073 18. Tabs are now shown as \t in -bP output.
4075 19. If the log selector return_path_on_delivery was set, Exim crashed when
4076 bouncing a message because it had too many Received: header lines.
4078 20. If two routers both had headers_remove settings, and the first one included
4079 a superfluous trailing colon, the final name in the first list and the
4080 first name in the second list were incorrectly joined into one item (with a
4081 colon in the middle).
4087 1. Added -C and -D options to the exinext utility, mainly to make it easier
4088 to include in the automated testing, but these could be helpful when
4089 multiple configurations are in use.
4091 2. The exinext utility was not formatting the output nicely when there was
4092 an alternate port involved in the retry record key, nor when there was a
4093 message id as well (for retries that were specific to a specific message
4094 and a specific host). It was also confused by IPv6 addresses, because of
4095 the additional colons they contain. I have fixed the IPv4 problem, and
4096 patched it up to do a reasonable job for IPv6.
4098 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4099 delivery, the log line now contains "pipelined" if PIPELINING was used.
4101 4. An SMTP transport process used to panic and die if the bind() call to set
4102 an explicit outgoing interface failed. This has been changed; it is now
4103 treated in the same way as a connect() failure.
4105 5. A reference to $sender_host_name in the part of a conditional expansion
4106 that was being skipped was still causing a DNS lookup. This no longer
4109 6. The def: expansion condition was not recognizing references to header lines
4110 that used bh_ and bheader_.
4112 7. Added the _cache feature to named lists.
4114 8. The code for checking quota_filecount in the appendfile transport was
4115 allowing one more file than it should have been.
4117 9. For compatibility with Sendmail, the command line option
4125 and sets the incoming protocol and host name (for trusted callers). The
4126 host name and its colon can be omitted when only the protocol is to be set.
4127 Note the Exim already has two private options, -pd and -ps, that refer to
4128 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4129 "s", but I don't think that's a major issue.
4131 10. A number of refactoring changes to the code, none of which should affect
4134 (a) The number of logging options was getting close to filling up the
4135 32-bit word that was used as a bit map. I have split them into two classes:
4136 those that are passed in the argument to log_write(), and those that are
4137 only ever tested independently outside of that function. These are now in
4138 separate 32-bit words, so there is plenty of room for expansion again.
4139 There is no change in the user interface or the logging behaviour.
4141 (b) When building, for example, log lines, the code previously used a
4142 macro that called string_cat() twice, in order to add two strings. This is
4143 not really sufficiently general. Furthermore, there was one instance where
4144 it was actually wrong because one of the argument was used twice, and in
4145 one call a function was used. (As it happened, calling the function twice
4146 did not affect the overall behaviour.) The macro has been replaced by a
4147 function that can join an arbitrary number of extra strings onto a growing
4150 (c) The code for expansion conditions now uses a table and a binary chop
4151 instead of a serial search (which was left over from when there were very
4152 few conditions). Also, it now recognizes conditions like "pam" even when
4153 the relevant support is not compiled in: a suitably worded error message is
4154 given if an attempt is made to use such a condition.
4156 11. Added ${time_interval:xxxxx}.
4158 12. A bug was causing one of the ddress fields not to be passed back correctly
4159 from remote delivery subprocesses. The field in question was not being
4160 subsequently used, so this caused to problems in practice.
4162 13. Added new log selectors queue_time and deliver_time.
4164 14. Might have fixed a bug in maildirsizefile handling that threw up
4165 "unexpected character" debug warnings, and recalculated the data
4166 unnecessarily. In any case, I expanded the warning message to give more
4169 15. Added the message "Restricted characters in address" to the statements in
4170 the default ACL that block characters like @ and % in local parts.
4172 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4173 Three changes have been made:
4175 (a) There was a serious bug; a negative response to MAIL caused the whole
4176 recipient domain to be cached as invalid, thereby blocking all messages
4177 to all local parts at the same domain, from all senders. This bug has
4178 been fixed. The domain is no longer cached after a negative response to
4179 MAIL if the sender used is not empty.
4181 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4184 (c) A new callout option, "use_sender" has been added for people who want
4185 the modified behaviour.
4191 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4194 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4195 indeed breaks things for older releases.
4197 3. Added additional logging to the case where there is a problem reading data
4198 from a filter that is running in a subprocess using a pipe, in order to
4199 try to track down a specific problem.
4201 4. Testing facility fudge: when running in the test harness and attempting
4202 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4203 getting "No route to host". Convert this to a timeout.
4205 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4208 6. Some OS don't have socklen_t but use size_t instead. This affects the
4209 fifth argument of getsockopt() amongst other things. This is now
4210 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4211 can be set for individual OS. I have set it for SunOS5, OSF1, and
4212 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4213 some earlier ones do not.
4215 7. Change 4.30/15 was not doing the test caselessly.
4217 8. The standard form for an IPv6 address literal was being rejected by address
4218 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4219 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4220 this, as well as the form without the "IPv6" on the front (but only when
4221 address literals are enabled, of course).
4223 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4225 10. Exim crashed if a message with an empty sender address specified by -f
4226 encountered a router with an errors_to setting. This could be provoked only
4227 by a command such as
4231 where an empty string was supplied; "<>" did not hit this bug.
4233 11. Installed PCRE release 4.5.
4235 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4236 remained set. It is now erased.
4238 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4239 times from message ids (which are base 36 rather than the normal 62).
4241 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4242 were being counted as actual protocol errors, and logged if the log
4243 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4244 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4245 and DATA following a set of rejected RCPTs do not count as protocol errors.
4246 In other words, Exim assumes they were pipelined, though this may not
4247 actually be the case. Of course, in all cases the client gets an
4248 appropriate error code.
4250 15. If a lookup fails in an ACL condition, a message about the failure may
4251 be available; it is used if testing the ACL cannot continue, because most
4252 such messages specify what the cause of the deferral is. However, some
4253 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4254 that caused an old message to be retained and used if a later statement
4255 caused a defer, replacing the real cause of the deferral.
4257 16. If an IP address had so many PTR records that the DNS lookup buffer
4258 was not large enough to hold them, Exim could crash while trying to process
4259 the truncated data. It now detects and logs this case.
4261 17. Further to 4.21/58, another change has been made: if (and only if) the
4262 first line of a message (the first header line) ends with CRLF, a bare LF
4263 in a subsequent header line has a space inserted after it, so as not to
4264 terminate the header.
4266 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4267 unnecessarily, used atoi() instead of strtol(), and didn't check the
4268 termination when getting file sizes from file names by regex.
4270 19. Completely re-implemented the support for maildirsize files, in the light
4271 of a number of problems with the previous contributed implementation
4272 (4.30/29). In particular:
4274 . If the quota is zero, the maildirsize file is maintained, but no quota is
4277 . If the maildir directory does not exist, it is created before any attempt
4278 to write a maildirsize file.
4280 . The quota value in the file is just a cache; if the quota is changed in
4281 the transport, the new value overrides.
4283 . A regular expression is available for excluding directories from the
4286 20. The autoreply transport checks the characters in options that define the
4287 message's headers; it allows continued headers, but it was checking with
4288 isspace() after an embedded newline instead of explicitly looking for a
4291 21. If all the "regular" hosts to which an address was routed had passed their
4292 expiry times, and had not reached their retry times, the address was
4293 bounced, even if fallback hosts were defined. Now Exim should go on to try
4296 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4297 equivalent code in the SMTP transport. Some hosts send humungous responses
4298 to HELO/EHLO, more than 1024 it seems.
4300 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4301 gives compiler warnings in some environments. I've now done it "properly",
4304 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4305 (because of problems with the built-in one) was declared to return uschar *
4306 instead of char *, causing compiler failure.
4308 25. Fixed a file descriptor leak when processing alias/forward files.
4310 26. Fixed a minor format string issue in dbfn.c.
4312 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4314 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4315 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4318 29. When a sender address is verified, it is cached, to save repeating the test
4319 when there is more than one recipient in a message. However, when the
4320 verification involves a callout, it is possible for different callout
4321 options to be set for different recipients. It is too complicated to keep
4322 track of this in the cache, so now Exim always runs a verification when a
4323 callout is required, relying on the callout cache for the optimization.
4324 The overhead is duplication of the address routing, but this should not be
4327 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4328 to be verified with callout=postmaster, and the main callout worked but the
4329 postmaster check failed, the verification correctly failed. However, if a
4330 subsequent RCPT command asked for sender verification *without* the
4331 postmaster check, incorrect caching caused this verification also to fail,
4334 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4335 it was not caching the DNS options (qualify_single, search_parents) that
4336 were used when the lookup failed. A subsequent lookup with different
4337 options therefore always gave the same answer, though there were cases
4338 where it should not have. (Example: a "domains = !$mx_any" option on a
4339 dnslookup router: the "domains" option is always processed without any
4340 widening, but the router might have qualify_single set.) Now Exim uses the
4341 cached value only when the same options are set.
4343 32. Added John Jetmore's "exipick" utility to the distribution.
4345 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4346 than a timeout (e.g. a certificate is required, and is not provided), an
4347 Exim server now closes the connection immediately. Previously it waited for
4348 the client to close - but if the client is SSL, it seems that they each
4349 wait for each other, leading to a delay before one of them times out.
4351 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4352 maintained 0.8.x compatibility because I don't think many are using it, and
4353 it is clearly obsolete.
4355 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4358 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4359 client certificate was expired. A simple patch fixes this, though I don't
4360 understand the full logic of why the verify callback is called multiple
4363 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4364 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4365 which causes problems with some clients (such as the Certicom SSL Plus
4366 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4367 disables the coutermeasure allowing Eudora to connect."
4369 38. Exim was not checking that a write() to a log file succeeded. This could
4370 lead to Bad Things if a log got too big, in particular if it hit a file
4371 size limit. Exim now panics and dies if it cannot write to a log file, just
4372 as it does if it cannot open a log file.
4374 39. Modified OS/Makefile-Linux so that it now contains
4376 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4378 The two -D definitions ensure that Exim is compiled with large file
4379 support, which makes it possible to handle log files that are bigger than
4382 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4383 instance) a domain was checked against a named list that involved a lookup,
4384 causing $domain_data to be set, then another domain was checked against the
4385 same list, then the first domain was re-checked, the value of $domain_data
4386 after the final check could be wrong. In particular, if the second check
4387 failed, it could be set empty. This bug probably also applied to
4390 41. The strip_trailing_dot option was not being applied to the address given
4391 with the -f command-line option.
4393 42. The code for reading a message's header from the spool was incrementing
4394 $received_count, but never initializing it. This meant that the value was
4395 incorrect (doubled) while delivering a message in the same process in which
4396 it was received. In the most common configuration of Exim, this never
4397 happens - a fresh exec is done - but it can happen when
4398 deliver_drop_privilege is set.
4400 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4401 it now includes up to 150 characters of the unexpected data in the log
4404 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4405 and building data strings. The size of both of these buffers was 10 000
4406 bytes - far larger than anybody would *ever* want, thought I. Needless to
4407 say, somebody hit the limit. I have increased the maximum line length to
4408 20 000 and the maximum data length of concatenated lines to 100 000. I have
4409 also fixed two bugs, because there was no checking on these buffers. Tsk,
4410 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4411 buffer is too small.
4413 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4414 lsearch lookups. Now it does.
4416 46. When parsing a route_list item in a manualroute router, a fixed-length
4417 buffer was used for the list of hosts. I made this 1024 bytes long,
4418 thinking that nobody would ever have a list of hosts that long. Wrong.
4419 Somebody had a whole pile of complicated expansion conditions, and the
4420 string was silently truncated, leading to an expansion error. It turns out
4421 that it is easier to change to an unlimited length (owing to other changes
4422 that have happened since this code was originally written) than to build
4423 structure for giving a limitation error. The length of the item that
4424 expands into the list of hosts is now unlimited.
4426 47. The lsearch lookup could not handle data where the length of text line was
4427 more than 4095 characters. Such lines were truncated, leading to shortened
4428 data being returned. It should now handle lines of any length.
4430 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4431 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4434 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4435 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4436 no longer happens in comment lines. A list of replacements is now placed
4437 at the head of all of the source files, except those whose only change is
4438 to replace PERL_COMMAND in the very first #! line.
4440 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4441 messages on the queue, with a bottom-up merge sort, using code contributed
4442 by Michael Haardt. This should make operations like -bp somewhat faster on
4443 large queues. It won't affect queue runners, except when queue_run_in_order
4446 51. Installed eximstats 1.31 in the distribution.
4448 52. Added support for SRV lookups to the dnslookup router.
4450 53. If an ACL referred to $message_body or $message_body_end, the value was not
4451 reset for any messages that followed in the same SMTP session.
4453 54. The store-handling optimization for building very long strings was not
4454 differentiating between the different store pools. I don't think this
4455 actually made any difference in practice, but I've tidied it.
4457 55. While running the routers to verify a sender address, $sender_address
4458 was still set to the sender address. This is wrong, because when routing to
4459 send a bounce to the sender, it would be empty. Therefore, I have changed
4460 it so that, while verifying a sender address, $sender_address is set to <>.
4461 (There is no change to what happens when verifying a recipient address.)
4463 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4464 target A or AAAA records (if not already returned) without resetting the
4465 qualify_single or search_parents options of the DNS resolver. These are
4466 inappropriate in this case because the targets of MX and SRV records must
4467 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4468 target that, when qualified, matched something in the local domain. These
4469 two options are now turned off when doing these lookups.
4471 57. It seems that at least some releases of Reiserfs (which does not have the
4472 concept of a fixed number of inodes) returns zero and not -1 for the
4473 number of available inodes. This interacted badly with check_spool_inodes,
4474 which assumed that -1 was the "no such thing" setting. What I have done is
4475 to check that the total number of inodes is greater than zero before doing
4476 the test of how many are available.
4478 58. When a "warn" ACL statement has a log_message modifier, the message is
4479 remembered, and not repeated. This is to avoid a lot of repetition when a
4480 message has many recipients that cause the same warning to be written.
4481 Howewer, Exim was preserving the list of already written lines for an
4482 entire SMTP session, which doesn't seem right. The memory is now reset if a
4483 new message is started.
4485 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4486 debugging output unless log_rewrite was also set.
4488 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4489 of (int)(handle) when we know that handle contains (void *)(-1).
4491 61. The Exim daemon panic-logs an error return when it closes the incoming
4492 connection. However "connection reset by peer" seems to be common, and
4493 isn't really an error worthy of noting specially, so that particular error
4496 62. When Exim is trying to find all the local interfaces, it used to panic and
4497 die if the ioctl to get the interface flags failed. However, it seems that
4498 on at least one OS (Solaris 9) it is possible to have an interface that is
4499 included in the list of interfaces, but for which you get a failure error
4500 for this call. This happens when the interface is not "plumbed" into a
4501 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4502 failure of the "get flags" call assumes that the interface is down.
4504 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4505 makes life easier for people who are doing arithmetic on fields extracted
4506 from dates, where you often get leading zeros that should not be
4507 interpreted as octal.
4509 64. Added qualify_domain to the redirect router, to override the global
4512 65. If a pathologically long header line contained very many addresses (the
4513 report of this problem mentioned 10 000) and each of them was rewritten,
4514 Exim could use up a very large amount of memory. (It kept on making new
4515 copies of the header line as it rewrote, and never released the old ones.)
4516 At the expense of a bit more processing, the header rewriting function has
4517 been changed so that it no longer eats memory in this way.
4519 66. The generation of the Received: header has been moved from the time that a
4520 message starts to be received, to the time that it finishes. The timestamp
4521 in the Received: header should now be very close to that of the <= log
4522 line. There are two side-effects of this change:
4524 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4525 logged header lines no longer include the local Received: line, because
4526 it has not yet been created. The same applies to a copy of the message
4527 that is returned to a non-SMTP sender when a message is rejected.
4529 (b) When a filter file is tested using -bf, no additional Received: header
4530 is added to the test message. After some thought, I decided that this
4533 This change does not affect the value of $received_for. It is still set
4534 after address rewriting, but before local_scan() is called.
4536 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4538 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4539 gave an unhelpful panic error message, and a defer error. I have managed to
4540 change this behaviour so that it now rejects any supplied certificate,
4541 which seems right, as the list of acceptable certificates is empty.
4543 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4544 gave an unhelpful defer error. I have not managed to make this reject any
4545 supplied certificates, but the error message it gives is "no certificate
4546 supplied", which is not helpful.
4548 70. exigrep's output now also includes lines that are not associated with any
4549 message, but which match the given pattern. Implemented by a patch from
4550 Martin Sluka, which also tidied up the Perl a bit.
4552 71. Recipient callout verification, like sender verification, was using <> in
4553 the MAIL FROM command. This isn't really the right thing, since the actual
4554 sender may affect whether the remote host accepts the recipient or not. I
4555 have changed it to use the actual sender in the callout; this means that
4556 the cache record is now keyed on a recipient/sender pair, not just the
4557 recipient address. There doesn't seem to be a real danger of callout loops,
4558 since a callout by the remote host to check the sender would use <>.
4559 [SEE ABOVE: changed after hitting problems.]
4561 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4562 temporary errors. However, in the case of such a code being given after
4563 the end of a data transmission (i.e. after ".") Exim was failing to write
4564 a retry record for the message. (Yes, there was some broken host that was
4565 actually sending 8xx at this point.)
4567 73. An unknown lookup type in a host list could cause Exim to panic-die when
4568 the list was checked. (An example that provoked this was putting <; in the
4569 middle of a list instead of at the start.) If this happened during a DATA
4570 ACL check, a -D file could be left lying around. This kind of configuration
4571 error no longer causes Exim to die; instead it causes a defer errror. The
4572 incident is still logged to the main and panic logs.
4574 74. Buglet left over from Exim 3 conversion. The message "too many messages
4575 in one connection" was written to the rejectlog but not the mainlog, except
4576 when address rewriting (yes!) was being logged.
4578 75. Added write_rejectlog option.
4580 76. When a system filter was run not as root (that is, when system_filter_user
4581 was set), the values of the $n variables were not being returned to the
4582 main process; thus, they were not subsequently available in the $sn
4585 77. Added +return_path_on_delivery log selector.
4587 78. A connection timeout was being treated differently from recipients deferred
4588 when testing hosts_max_try with a message that was older than the host's
4589 retry timeout. (The host should not be counted, thus allowing all hosts to
4590 be tried at least once before bouncing.) This may have been the cause of an
4591 occasionally reported bug whereby a message would remain on the queue
4592 longer than the retry timeout, but would be bounced if a delivery was
4593 forced. I say "may" because I never totally pinned down the problem;
4594 setting up timeout/retry tests is difficult. See also the next item.
4596 79. The ultimate address timeout was not being applied to errors that involved
4597 a combination of host plus message (for example, a timeout on a MAIL
4598 command). When an address resolved to a number of possible hosts, and they
4599 were not all tried for each delivery (e.g. because of hosts_max_try), a
4600 message could remain on the queue longer than the retry timeout.
4602 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4605 81. Fixed an obscure SMTP outgoing bug which required at least the following
4606 conditions: (a) there was another message waiting for the same server;
4607 (b) the server returned 5xx to all RCPT commands in the first message so
4608 that the message was not completed; (c) the server dropped the connection
4609 or gave a negative response to the RSET that Exim sends to abort the
4610 transaction. The observed case was a dropped connection after DATA that had
4611 been sent in pipelining mode. That is, the server had advertised PIPELINING
4612 but was not implementing it correctly. The effect of the bug was incorrect
4613 behaviour, such as trying another host, and this could lead to a crash.
4619 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4620 and daemon.c were passed as pointers to ints; they should have been
4621 pointers to socklen_t variables (which are typically unsigned ints).
4623 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4626 3. Fixed a really odd bug that affected only the testing scheme; patching a
4627 certain fixed string in the binary changed the value of another string that
4628 happened to be identical to the end of the original first string.
4630 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4631 name", it returns that address as the IP address. On some operating
4632 systems (e.g. Solaris), it also passes back the IP address string as the
4633 "host name". However, on others (e.g. Linux), it passes back an empty
4634 string. Exim wasn't checking for this, and was changing the host name to an
4635 empty string, assuming it had been canonicized.
4637 5. Although rare, it is permitted to have more than one PTR record for a given
4638 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4639 all the names associated with an address, because they do in Solaris.
4640 However, it seems that they do not in Linux for data that comes from the
4641 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4642 I found this out when I moved to a new Linux workstation and tried to run
4643 the Exim test suite.
4645 To get round this problem I have changed the code so that it now does its
4646 own call to the DNS to look up PTR records when searching for a host name.
4647 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4648 addresses that are only in /etc/hosts are still found.
4650 This behaviour is, however, controlled by an option called host_lookup_
4651 order, which defaults to "bydns:byaddr". If people want to use the other
4652 order, or indeed, just use one or the other means of lookup, they can
4653 specify it in this variable.
4655 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4656 some operating systems, this comes back from gethostbyaddr() as an empty
4657 string, and this is what Exim used to test for. However, it seems that in
4658 other systems, "." is yielded. Exim now tests for this case too.
4660 7. The values of check_spool_space and check_log_space are now held internally
4661 as a number of kilobytes instead of an absolute number of bytes. If a
4662 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4663 kilobyte. This means that much larger values can be stored.
4665 8. Exim monitor: an attempt to get the action menu when not actually pointing
4666 at a message produces an empty menu entitled "No message selected". This
4667 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4668 no entries in it ("Shell widget menu has zero width and/or height"). So I
4669 have added a single, blank menu entry in this case.
4671 9. Added ${quote_local_part.
4673 10. MIME decoding is now applied to the contents of Subject: header lines when
4676 11. Now that a reference to $sender_host_address automatically causes a reverse
4677 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4678 host lookup before query-style lookups in lists that might use this
4679 variable. This has therefore been abolished, and the "net-" prefix is no
4680 longer necessary for query-style lookups.
4682 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4683 have been a typo for LFLAGS, so it has been changed.
4685 13. The install script calls Exim with "-C /dev/null" in order to find the
4686 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4687 to be output. Howeve, since Exim outputs its version number before the
4688 error, it didn't break the script. It just looked ugly. I fixed this by
4689 always allowing "-C /dev/null" if the caller is root.
4691 14. Ignore overlarge ACL variable number when reading spool file - insurance
4692 against a later release with more variables having written the file.
4694 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4695 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4696 this, as well as the form without the "IPv6" on the front.
4698 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4699 OS/Makefile-Darwin file.
4701 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4702 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4704 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4705 variable. However, some OS use ints for inodes. I've added cast to long int
4706 to get rid of the compiler warning.
4708 19. I had forgotten to lock out "/../" in configuration file names when
4709 ALT_CONFIG_PREFIX was set.
4711 20. Routers used for verification do not need to specify transports. However,
4712 if such a router generated a host list, and callout was configured, Exim
4713 crashed, because it could not find a port number from the (non-existent)
4714 transport. It now assumes port 25 in this circumstance.
4716 21. Added the -t option to exigrep.
4718 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4719 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4720 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4721 them is set, LOOKUP_LSEARCH is forced.
4723 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4725 24. Added sender and host information to the "rejected by local_scan()" log
4726 line; previously there was no indication of these.
4728 25. Added .include_if_exists.
4730 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4731 Linux. It turns out that not all file systems support this. Apparently some
4732 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4733 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4734 not supported on the file descriptor, is now ignored when Exim is trying to
4735 sync a directory. This applies only to Linux.
4737 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4739 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4740 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4743 29. Added support for maildirsize files from supplied patch (modified a bit).
4745 30. The use of :fail: followed by an empty string could lead Exim to respond to
4746 sender verification failures with (e.g.):
4748 550 Verification failed for <xxx>
4749 550 Sender verify failed
4751 where the first response line was missing the '-' that indicates it is not
4752 the final line of the response.
4754 31. The loop for finding the name of the user that called Exim had a hardwired
4755 limit of 10; it now uses the value of finduser_retries, which is used for
4756 all other user lookups.
4758 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4761 33. Exim was neglecting to zero errno before one call of strtol() when
4762 expanding a string and expecting an integer value. On some systems this
4763 resulted in spurious "integer overflow" errors. Also, it was casting the
4764 result into an int without checking.
4766 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4767 did not work. The code looks as if it has *never* worked, though it appears
4768 to have been documented since at least releast 1.62. I have made it work.
4770 35. The "timeout_DNS" error in retry rules, also documented since at least
4771 1.62, also never worked. As it isn't clear exactly what this means, and
4772 clearly it isn't a major issue, I have abolished the feature by treating it
4773 as "timeout", and writing a warning to the main and panic logs.
4775 36. The display of retry rules for -brt wasn't always showing the error code
4778 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4779 timeout_connect_A, timeout_connect_MX.
4781 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4782 to the empty sender.
4784 39. The daemon was not analysing the content of -oX till after it had closed
4785 stderr and disconnected from the controlling terminal. This meant that any
4786 syntax errors were only noted on the panic log, and the return code from
4787 the command was 0. By re-arranging the code a little, I've made the
4788 decoding happen first, so such errors now appear on stderr, and the return
4789 code is 1. However, the actual setting up of the sockets still happens in
4790 the disconnected process, so errors there are still only recorded on the
4793 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4794 connections (as happens on some IP stacks) was logged at start up time as
4795 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4796 it from "IPv6 and IPv4", which means that two separate sockets are being
4799 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4800 says whether AF_INET or AF_INET6 was passed as an argument.
4802 42. Exiwhat output was messed up when time zones were included in log
4805 43. Exiwhat now gives more information about the daemon's listening ports,
4806 and whether -tls-on-connect was used.
4808 44. The "port" option of the smtp transport is now expanded.
4810 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4811 silently ignored. Now an error message is written to the main and panic
4814 46. There's a new ACL modifier called "logwrite" which writes to a log file
4815 as soon as it is encountered.
4817 47. Added $local_user_uid and $local_user_gid at routing time.
4819 48. Exim crashed when trying to verify a sender address that was being
4822 49. Exim was recognizing only a space character after ".include". It now also
4823 recognizes a tab character.
4825 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4826 extracting the relevant information from the specification. The man page no
4827 longer contains scrambled data for the -d option, and I've added a section
4828 at the front about calling Exim under different names.
4830 51. Added "extra_headers" argument to the "mail" command in filter files.
4832 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4835 53. Installed eximstats 1.29.
4837 54. Added transport_filter_timeout as a generic transport option.
4839 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4840 Cc: header lines. This was required by RFC 822, but it not required by RFC
4843 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4844 incoming messages that did not have them. Now it does so only if the
4845 message originates locally, that is, if there is no associated remote host
4846 address. When Resent- header lines are present, this applies to the Resent-
4847 lines rather than the non-Resent- lines.
4849 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4850 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4852 58. Messages for configuration errors now include the name of the main
4853 configuration file - useful now that there may be more than one file in a
4854 list (.included file names were always shown).
4856 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4857 for those rare installations that do not start the daemon as root or run it
4858 setuid root. I've cut out the call to initgroups() if the daemon is not
4861 60. The Exim user and group can now be bound into the binary as text strings
4862 that are looked up at the start of Exim's processing.
4864 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4866 62. Added $mailstore_basename variable.
4868 63. Installed patch to sieve.c from Michael Haardt.
4870 64. When Exim failed to open the panic log after failing to open the main log,
4871 the original message it was trying to log was written to stderr and debug
4872 output, but if they were not available (the usual case in production), it
4873 was lost. Now it is written to syslog before the two lines that record the
4874 failures to open the logs.
4876 65. Users' Exim filters run in subprocesses under the user's uid. It is
4877 possible for a "deliver" command or an alias in a "personal" command to
4878 provoke an address rewrite. If logging of address rewriting is configured,
4879 this fails because the process is not running as root or exim. There may be
4880 a better way of dealing with this, but for the moment (because 4.30 needs
4881 to be released), I have disabled address rewrite logging when running a
4882 filter in a non-root, non-exim process.
4888 1. The buildconfig auxiliary program wasn't quoting the value set for
4889 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4890 not defined. This bug was masked in 4.22 by the effect that was fixed in
4893 2. Some messages that were rejected after a message id was allocated were
4894 shown as "incomplete" by exigrep. It no longer does this for messages that
4895 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4897 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4898 have allow_domain_literals set, the ID did not get logged in the <= line.
4899 Domain literals are now always recognized in Message-ID: header lines.
4901 4. The first argument for a ${extract expansion item is the key name or field
4902 number. Leading and trailing spaces in this item were not being ignored,
4903 causing some misleading effects.
4905 5. When deliver_drop_privilege was set, single queue runner processes started
4906 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4907 same command in the process it spins off) were not dropping privilege.
4909 6. When the daemon running as "exim" started a queue runner, it always
4910 re-executed Exim in the spun-off process. This is a waste of effort when
4911 deliver_drop_privilege is set. The new process now just calls the
4912 queue-runner function directly.
4918 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4921 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4922 ignored. Though the use of -oP was forcing the writing of a pid file, it
4923 was always written to the default place.
4925 3. If the message "no IP address found for host xxxx" is generated during
4926 incoming verification, it is now followed by identification of the incoming
4927 connection (so you can more easily find what provoked it).
4929 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4931 5. Added some features to "harden" Exim a bit more against certain attacks:
4933 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4934 be put in Local/Makefile. This is like the never_users runtime option,
4935 but it cannot be overridden. The default setting is "root".
4937 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4938 prefix string with which any file named in a -C command line option
4941 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4942 is retained for -C and -D only if the caller of Exim is root. Without
4943 it, the exim user may also use -C and -D and retain privilege.
4945 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4946 command line option is disabled.
4948 6. Macro names set by the -D option must start with an upper case letter, just
4949 like macro names defined in the configuration file.
4951 7. Added "dereference=" facility to LDAP.
4953 8. Two instances of the typo "uknown" in the source files are fixed.
4955 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4956 the Configure-Makefile script screwed up while processing it.
4958 10. Incorporated PCRE 4.4.
4960 11. The SMTP synchronization check was not operating right at the start of an
4961 SMTP session. For example, it could not catch a HELO sent before the client
4962 waited for the greeting. There is now a check for outstanding input at the
4963 point when the greeting is written. Because of the duplex, asynchronous
4964 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4965 way, but not yet received, when the check is performed.
4967 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4968 on TCP/IP sockets, because this apparently causes some broken clients to
4971 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4972 unchanged) from the Cygwin maintainer.
4974 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4975 was supported instead of testing for mbx. Effectively a typo.
4977 15. The spa authenticator server code was not checking that the input it
4978 received was valid base64.
4980 16. The debug output line for the "set" modifier in ACLs was not showing the
4981 name of the variable that was being set.
4983 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4985 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4986 lookup on its own. Something else was needed to trigger the lookup. For
4987 example, a match in host_lookup or the need for a host name in a host list.
4988 Now, if $sender_host_name is referenced and the host name has not yet been
4989 looked up, a lookup is performed. If the lookup fails, the variable remains
4990 empty, and $host_lookup_failed is set to "1".
4992 19. Added "eqi" as a case-independent comparison operator.
4994 20. The saslauthd authentication condition could segfault if neither service
4995 nor realm was specified.
4997 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4998 error message that was logged was misleading, and incoming SMTP
4999 connections were dropped. The message is now more accurate, and temporary
5000 errors are given to SMTP connections.
5002 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5003 (except RSET) with a 421 error, until QUIT is received. However, it was
5004 failing to send a response to QUIT.
5006 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5007 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5008 ACL is now run after the helo_try_verify_hosts code.
5010 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5011 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5012 case-independent, so other case variants are also recognized). Apparently
5013 some systems use these upper case variants.
5015 25. If more than two messages were waiting for the same host, and a transport
5016 filter was specified for the transport, Exim sent two messages over the
5017 same TCP/IP connection, and then failed with "socket operation on non-
5018 socket" when it tried to send the third.
5020 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5022 27. The extern definition of crypt16() in expand.c was not being excluded when
5023 the OS had its own crypt16() function.
5025 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5026 as a preferred synonym for return_size_limit, both as an option and as an
5029 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5031 30. Changed the default configuration ACL to relax the local part checking rule
5032 for addresses that are not in any local domains. For these addresses,
5033 slashes and pipe symbols are allowed within local parts, but the sequence
5034 /../ is explicitly forbidden.
5036 31. SPA server authentication was not clearing the challenge buffer before
5039 32. log_message in a "warn" ACL statement was writing to the reject log as
5040 well as to the main log, which contradicts the documentation and doesn't
5041 seem right (because no rejection is happening). So I have stopped it.
5043 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5044 However, I am unable to do any testing of this.
5046 34. Fixed an infelicity in the appendfile transport. When checking directories
5047 for a mailbox, to see if any needed to be created, it was accidentally
5048 using path names with one or more superfluous leading slashes; tracing
5049 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5051 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5052 contents are added to the log line that is written for every discarded
5053 recipient. (Previously a log_message setting was ignored.)
5055 36. The ${quote: operator now quotes the string if it is empty.
5057 37. The install script runs exim in order to find its version number. If for
5058 some reason other than non-existence or emptiness, which it checks, it
5059 could not run './exim', it was installing it with an empty version number,
5060 i.e. as "exim-". This error state is now caught, and the installation is
5063 38. An argument was missing from the function that creates an error message
5064 when Exim fails to connect to the socket for saslauthd authentication.
5065 This could cause Exim to crash, or give a corrupted message.
5067 39. Added isip, isip4, and isip6 to ${if conditions.
5069 40. The ACL variables $acl_xx are now saved with the message, and can be
5070 accessed later in routers, transports, and filters.
5072 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5073 strings in the file are not string-expanded.
5075 42. If a MAIL command specified a SIZE value that was too large to fit into an
5076 int variable, the check against message_size_limit failed. Such values are
5077 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5078 day this will have to be increased, but I don't think I want to be around
5079 when emails are that large.
5086 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5087 iconv() is not standard in FreeBSD.
5089 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5090 IPv6 enabled. The observed symptom was a segmentation fault on return from
5091 the function os_common_find_running_interfaces() in src/os.c.
5093 3. In the check_special_case() function in daemon.c I had used "errno" as an
5094 argument name, which causes warnings on some systems. This was basically a
5095 typo, since it was named "eno" in the comments!
5097 4. The code that waits for the clock to tick (at a resolution of some fraction
5098 of a second) so as to ensure message-id uniqueness was always waiting for
5099 at least one whole tick, when it could have waited for less. [This is
5100 almost certainly not relevant at current processor speeds, where it is
5101 unlikely to ever wait at all. But we try to future-proof.]
5103 5. The function that sleeps for a time interval that includes fractions of a
5104 second contained a race. It did not block SIGALRM between setting the
5105 timer, and suspending (a couple of lines later). If the interval was short
5106 and the sigsuspend() was delayed until after it had expired, the suspension
5107 never ended. On busy systems this could lead to processes getting stuck for
5110 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5111 process, before it forks any delivery processes. The open lookup caching
5112 mechanism meant that the open file or database connection was passed into
5113 the delivery process. The problem was that delivery processes always tidy
5114 up cached lookup data. This could cause a problem for the next delivery
5115 process started by the queue runner, because the external queue runner
5116 process does not know about the closure. So the next delivery process
5117 still has data in the lookup cache. In the case of a file lookup, there was
5118 no problem because closing a file descriptor in a subprocess doesn't affect
5119 the parent. However, if the lookup was caching a connection to a database,
5120 the connection was closed, and the second delivery process was likely to
5121 see errors such as "PGSQL: query failed: server closed the connection
5122 unexpectedly". The problem has been fixed by closing all cached lookups
5123 in a queue runner before running a delivery process.
5125 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5126 seem to have the "const" qualifier which it has on other OS. I've
5129 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5130 *of the same type* (client or server) with the same public name that an
5131 error should be diagnosed.
5133 9. When Exim looked up a host name for an IP address, but failed to find the
5134 original IP address when looking up the host name (a safety check), it
5135 output the message "<ip address> does not match any IP for NULL", which was
5136 confusing, to say the least. The bug was that the host name should have
5137 appeared instead of "NULL".
5139 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5140 user that is built into the binary, and the -C or -D options is used, root
5141 privilege is dropped before the configuration file is read. In addition,
5142 logging is switched to stderr instead of the normal log files. If the
5143 configuration then re-defines the Exim user, the unprivileged environment
5144 is probably not what is expected, so Exim logs a panic warning message (but
5147 However, if deliver_drop_privilege is set, the unprivileged state may well
5148 be exactly what is intended, so the warning has been cut out in that case,
5149 and Exim is allowed to try to write to its normal log files.
5155 1. smtp_return_error_details was not giving details for temporary sender
5156 or receiver verification errors.
5158 2. Diagnose a configuration error if two authenticators have the same public
5161 3. Exim used not to create the message log file for a message until the first
5162 delivery attempt. This could be confusing when incoming messages were held
5163 for policy or load reasons. The message log file is now created at the time
5164 the message is received, and an initial "Received" line is written to it.
5166 4. The automatically generated man page for command line options had a minor
5167 bug that caused no ill effects; however, a more serious problem was that
5168 the procedure for building the man page automatically didn't always
5169 operate. Consequently, release 4.20 contains an out-of-date version. This
5170 shouldn't happen again.
5172 5. When building Exim with embedded Perl support, the script that builds the
5173 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5174 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5176 6. The freeze_tell option was not being used for messages that were frozen on
5177 arrival, either by an ACL or by local_scan().
5179 7. Added the smtp_incomplete_transaction log selector.
5181 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5182 was accepting AUTH without a new EHLO.
5184 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5185 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5188 10. Logging of TCP/IP connections (when configured) now happens in the main
5189 daemon process instead of the child process, so that the TCP/IP connection
5190 count is more accurate (but it can never be perfect).
5192 11. The use of "drop" in a nested ACL was not being handled correctly in the
5193 outer ACL. Now, if condition failure induced by the nested "drop" causes
5194 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5195 or "require"), the connection is dropped.
5197 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5198 that yield "discard" can now be used with an "accept" or a "discard" verb,
5199 but an error is generated for any others (because I can't see a useful way
5200 to define what should happen).
5202 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5203 are now processed in the same way as lines in the Exim configuration file.
5204 In particular, continuation lines are supported.
5206 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5208 15. Added -ti meaning -t -i.
5210 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5211 domains, and warn by logging if others are found.
5213 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5214 returned from GIFCONF when Exim is trying to find the list of interfaces on
5215 a host. The code in os.c has been modified to copy each ifreq to an aligned
5216 structure in all cases.
5218 Also, in some cases, the returned ifreq's were being copied to a 'struct
5219 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5220 means the last couple of bytes of an IPv6 address could be chopped if the
5221 ifreq contained only a normal sockaddr (14 bytes storage).
5223 18. Named domain lists were not supported in the hosts_treat_as_local option.
5224 An entry such as +xxxx was not recognized, and was treated as a literal
5227 19. Ensure that header lines added by a DATA ACL are included in the reject log
5228 if the ACL subsequently rejects the message.
5230 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5231 MD5 (which is deprecated).
5233 21. When testing a filter file using -bf, Exim was writing a message when it
5234 took the sender from a "From " line in the message, but it was not doing so
5235 when it took $return_path from a Return-Path: header line. It now does.
5237 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5238 with a valid header line field name (a series of printing characters
5239 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5241 23. Changed "disc" in the source to "disk" to conform to the documentation and
5242 the book and for uniformity.
5244 24. Ignore Sendmail's -Ooption=value command line item.
5246 25. When execve() failed while trying to run a command in a pipe transport,
5247 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5248 could be confused with a return value of 69 from the command itself. This
5249 has been changed to 127, the value the shell returns if it is asked to run
5250 a non-existent command. The wording for the related log line suggests a
5251 non-existent command as the problem.
5253 26. If received_header_text expands to an empty string, do not add a Received:
5254 header line to the message. (Well, it adds a token one on the spool, but
5255 marks it "old" so that it doesn't get used or transmitted.)
5257 27. Installed eximstats 1.28 (addition of -nt option).
5259 28. There was no check for failure on the call to getsockname() in the daemon
5260 code. This can fail if there is a shortage of resources on the system, with
5261 ENOMEM, for example. A temporary error is now given on failure.
5263 29. Contrary to the C standard, it seems that in some environments, the
5264 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5265 program. Exim now does this explicitly; it affects the formatting of
5266 timestamps using strftime().
5268 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5269 complaints. I've now initialized all the variables, to avoid this.
5271 32. Header lines added by a system filter were not being "seen" during
5272 transport-time rewrites.
5274 33. The info_callback() function passed to OpenSSL is set up with type void
5275 (*)(SSL *, int, int), as described somewhere. However, when calling the
5276 function (actually a macro) that sets it up, the type void(*)() is
5277 expected. I've put in a cast to prevent warnings from picky compilers.
5279 34. If a DNS black list lookup found a CNAME record, but there were no A
5280 records associated with the domain it pointed at, Exim crashed.
5282 35. If a DNS black list lookup returned more than one A record, Exim ignored
5283 all but the first. It now scans all returned addresses if a particular IP
5284 value is being sought. In this situation, the contents of the
5285 $dnslist_value variable are a list of all the addresses, separated by a
5288 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5289 to accept a host name and all its aliases if the forward lookup for any of
5290 them yielded the IP address of the incoming connection. Now it accepts only
5291 those names whose forward lookup yields the correct IP address. Any other
5292 names are discarded. This closes a loophole whereby a rogue DNS
5293 administrator could create reverse DNS records to break through a
5294 wildcarded host restriction in an ACL.
5296 37. If a user filter or a system filter that ran in a subprocess used any of
5297 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5298 the wrong values were passed to the pipe command ($thisaddress had the
5299 value of $0, $0 had the value of $1, etc). This bug was introduced by
5300 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5302 38. Improved the line breaking for long SMTP error messages from ACLs.
5303 Previously, if there was no break point between 40 and 75 characters, Exim
5304 left the rest of the message alone. Two changes have been made: (a) I've
5305 reduced the minimum length to 35 characters; (b) if it can't find a break
5306 point between 35 and 75 characters, it looks ahead and uses the first one
5307 that it finds. This may give the occasional overlong line, but at least the
5308 remaining text gets split now.
5310 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5311 file descriptors might be low, and that setting 1000 would always raise it.
5312 It turns out that in some environments, the limit is already over 1000 and
5313 that lowering it causes trouble. So now Exim takes care not to decrease it.
5315 40. When delivering a message, the value of $return_path is set to $sender_
5316 address at the start of routing (routers may change the value). By an
5317 oversight, this default was not being set up when an address was tested by
5318 -bt or -bv, which affected the outcome if any router or filter referred to
5321 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5322 log only when "message" or "log_message" are set. However, if one of the
5323 conditions was an address verification, or a call to a nested ACL, the
5324 messages generated by the underlying test were being passed through. This
5325 no longer happens. The underlying message is available in $acl_verify_
5326 message for both "message" and "log_message" expansions, so it can be
5327 passed through if needed.
5329 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5330 new expansion $bh_ to give the encoded byte string without charset
5331 translation. Translation happens only if iconv() is available; HAVE_ICONV
5332 indicates this at build time. HEADERS_CHARSET gives the charset to
5333 translate to; headers_charset can change it in the configuration, and
5334 "headers charset" can change it in an individual filter file.
5336 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5337 that creates RFC 2047 encoded "words" labels them as that charset instead
5338 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5339 expansion operator; (ii) when Exim creates a From: line for a local
5340 message; (iii) when a header line is rewritten to include a "phrase" part.
5342 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5343 buggy, causing it to skip the first lines of messages whose message ID
5344 ended in 'D'. This would not have bitten before Exim release 4.14, because
5345 message IDs were unlikely to end in 'D' before then. The effect was to have
5346 incorrect size information for certain domains.
5348 45. #include "config.h" was missing at the start of the crypt16.c module. This
5349 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5352 46. If there was a timeout during a "random" callout check, Exim treated it as
5353 a failure of the random address, and carried on sending RSET and the real
5354 address. If the delay was just some slowness somewhere, the response to the
5355 original RCPT would be taken as a response to RSET and so on, causing
5356 mayhem of various kinds.
5358 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5359 when I implemented it. It didn't allow for the fact that some option values
5360 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5361 the right test for positive values.
5363 48. Domain names in DNS records are case-independent. Exim always looks them up
5364 in lower case. Some resolvers return domain names in exactly the case they
5365 appear in the zone file, that is, they may contain uppercase letters. Not
5366 all resolvers do this - some return always lower case. Exim was treating a
5367 change of case by a resolver as a change of domain, similar to a widening
5368 of a domain abbreviation. This triggered its re-routing code and so it was
5369 trying to route what was effectively the same domain again. This normally
5370 caused routing to fail (because the router wouldn't handle the domain
5371 twice). Now Exim checks for this case specially, and just changes the
5372 casing of the domain that it ultimately uses when it transmits the message
5375 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5378 50. If a filter generated a file delivery with a non-absolute name (possible if
5379 no home directory exists for the router), the forbid_file option was not
5382 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5383 the existing equality matching.
5385 52. Exim was using ints instead of ino_t variables in some places where it was
5386 dealing with inode numbers.
5388 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5389 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5390 called TMPDIR, and if it finds it is different, it changes its value.
5392 54. The smtp_printf() function is now made available to local_scan() so
5393 additional output lines can be written before returning. There is also an
5394 smtp_fflush() function to enable the detection of a dropped connection.
5395 The variables smtp_input and smtp_batched_input are exported to
5398 55. Changed the default runtime configuration: the message "Unknown user"
5399 has been removed from the ACL, and instead placed on the localuser router,
5400 using the cannot_route_message feature. This means that any verification
5401 failures that generate their own messages won't get overridden. Similarly,
5402 the "Unrouteable address" message that was in the ACL for unverifiable
5403 relay addresses has also been removed.
5405 56. Added hosts_avoid_esmtp to the smtp transport.
5407 57. The exicyclog script was not checking for the esoteric option
5408 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5409 will work only if exicyclog is run under the appropriate euid.
5411 58. Following a discussion on the list, the rules by which Exim recognises line
5412 endings on incoming messages have been changed. The -dropcr and drop_cr
5413 options are now no-ops, retained only for backwards compatibility. The
5414 following line terminators are recognized: LF CRLF CR. However, special
5415 processing applies to CR:
5417 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5418 nor a local message in the state where . is a terminator.
5420 (ii) If a bare CR is encountered in a header line, an extra space is added
5421 after the line terminator so as not to end the header. The reasoning
5422 behind this is that bare CRs in header lines are most likely either
5423 to be mistakes, or people trying to play silly games.
5425 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5426 was being incorrectly given as 18 bytes larger than it should have been.
5427 This is a VOB (very old bug).
5429 60. This may never have affected anything current, but just in case it has:
5430 When the local host is found other than at the start of a list of hosts,
5431 the local host, those with the same MX, and any that follow, are discarded.
5432 When the list in question was part of a longer list of hosts, the following
5433 hosts (not currently being processed) were also being discarded. This no
5434 longer happens. I'm not sure if this situation could ever has previously
5437 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5440 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5441 header. This is something that is recommended in a new Internet Draft, and
5442 is something that is documented as being done by Sendmail. There are two
5443 possible values. For messages generated by the autoreply transport, Exim
5446 Auto-Submitted: auto-replied
5448 whereas for all other generated messages (e.g. bounces) it adds
5450 Auto-Submitted: auto-generated
5452 63. The "personal" condition in filters now includes a test for the
5453 Auto-Submitted: header. If it contains the string "auto-" the message it
5454 not considered personal.
5456 64. Added rcpt_include_affixes as a generic transport option.
5458 65. Added queue_only_override (default true).
5460 66. Added the syslog_duplication option.
5462 67. If what should have been the first header line of a message consisted of
5463 a space followed by a colon, Exim was mis-interpreting it as a header line.
5464 It isn't of course - it is syntactically invalid and should therefore be
5465 treated as the start of the message body. The misbehaviour could have
5466 caused a number of strange effects, including loss of data in subsequent
5467 header lines, and spool format errors.
5469 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5470 client host had authenticated. This control can now be exercised by an ACL
5471 for more flexibility.
5473 69. By default, callouts do not happen when testing with -bh. There is now a
5474 variant, -bhc, which does actually run the callout code, including
5475 consulting and updating the callout cache.
5477 70. Added support for saslauthd authentication, courtesy of Alexander
5480 71. If statvfs() failed on the spool or log directories while checking their
5481 size for availability, Exim confusingly gave the error "space shortage".
5482 Furthermore, in debugging mode it crashed with a floating point exception.
5483 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5484 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5485 serious problem, Exim now writes to the main and panic logs when this
5486 happens, with details of the failure. It then refuses to accept the
5487 incoming message, giving the message "spool directory problem" or "log
5488 directory problem" with a 421 code for SMTP messages.
5490 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5491 0, 1, and 2 exist, because some OS complain for execs without them (see
5492 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5493 descriptors. However, the code omitted to check that the open succeeded,
5494 causing mysterious errors if for some reason the permissions on /dev/null
5495 got screwed. Now Exim writes a message to the main and panic logs, and
5496 bombs out if it can't open /dev/null.
5498 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5499 interact so that it is all more flexible. It is supposed to remain
5500 backwards compatible. Also added extra_local_interfaces.
5502 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5503 to bomb out with an assertion failure - to the client this appears as a
5504 connection drop. This problem occurs in the part of the code that was taken
5505 from the Samba project. Fortunately, the assertion is in a very simple
5506 function, so I have fixed this by reproducing the function inline in the
5507 one place where it is called, and arranging for authentication to fail
5508 instead of killing the process with assert().
5510 75. The SPA client code was not working when the server requested OEM rather
5511 than Unicode encoding.
5513 76. Added code to make require_files with a specific uid setting more usable in
5514 the case where statting the file as root fails - usually a non-root-mounted
5515 NFS file system. When this happens and the failure is EACCES, Exim now
5516 forks a subprocess and does the per-uid checking as the relevant uid.
5518 77. Added process_log_path.
5520 78. If log_file_path was not explicitly set, a setting of check_log_space or
5521 check_log_inodes was ignored.
5523 79. If a space check for the spool or log partitions fails, the incident is now
5524 logged. Of course, in the latter case the data may get lost...
5526 80. Added the %p formatting code to string_format() so that it can be used to
5527 print addresses in debug_print(). Adjusted all the address printing in the
5528 debugging in store.c to use %p rather than %d.
5530 81. There was a concern that a line of code in smtp_in.c could overflow a
5531 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5532 initially expressed, the concern was not well-founded, because trailing
5533 spaces are removed early. However, if the trailing spaces were followed by
5534 a NULL, they did not get removed, so the overflow was possible. Two fixes
5537 (a) I re-wrote the offending code in a cleaner fashion.
5538 (b) If an incoming SMTP command contains a NULL character, it is rejected
5541 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5542 runs initgroups(), so that if the Exim user is in any additional groups,
5543 they will be used during message reception.
5549 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob