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 not 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.
68 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
70 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
72 Report and patch from Dmitry Banschikov.
74 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
75 Removed SSL_clear() after SSL_new() which led to protocol negotiation
76 failures. We appear to now support TLS1.1+ with Exim.
78 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
79 lets Exim select keys and certificates based upon TLS SNI from client.
80 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
81 before an outbound SMTP session. New log_selector, +tls_sni.
83 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
84 NULL dereference. Report and patch from Alun Jones.
86 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
87 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
88 Not seeing resolver debug output on NetBSD, but suspect this is a
89 resolver implementation change.
91 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
92 Left warnings. Added "eximon gdb" invocation mode.
94 PP/21 Defaulting "accept_8bitmime" to true, not false.
96 PP/22 Added -bw for inetd wait mode support.
98 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
99 locate the relevant includes and libraries. Made this the default.
101 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
102 Bugzilla 1246, report and most of solution from Tomasz Kusy.
104 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
105 This may cause build issues on older platforms.
107 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
108 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
109 gnutls_require_mac & gnutls_require_protocols (no longer supported).
110 Added SNI support via GnuTLS too.
111 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
113 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
115 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
116 Applied second patch from Janne, implementing suggestion to default
117 multiple-strings-in-record handling to match SPF spec.
119 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
125 PP/01 Solaris build fix for Oracle's LDAP libraries.
126 Bugzilla 1109, patch from Stephen Usher.
128 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
130 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
133 TF/02 Fix a couple more cases where we did not log the error message
134 when unlink() failed. See also change 4.74-TF/03.
136 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
137 lock up or crash if it happened to be inside a call to libc when it
138 got a SIGUSR1 from exiwhat.
140 The SIGUSR1 handler appends the current process status to the process
141 log which is later printed by exiwhat. It used to use the general
142 purpose logging code to do this, but several functions it calls are
143 not safe for signals.
145 The new output code in the SIGUSR1 handler is specific to the process
146 log, and simple enough that it's easy to inspect for signal safety.
147 Removing some special cases also simplifies the general logging code.
148 Removing the spurious timestamps from the process log simplifies
151 TF/04 Improved ratelimit ACL condition.
153 The /noupdate option has been deprecated in favour of /readonly which
154 has clearer semantics. The /leaky, /strict, and /readonly update modes
155 are mutually exclusive. The update mode is no longer included in the
156 database key; it just determines when the database is updated. (This
157 means that when you upgrde Exim will forget old rate measurements.)
159 Exim now checks that the per_* options are used with an update mode that
160 makes sense for the current ACL. For example, when Exim is processing a
161 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
162 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
163 must specify per_mail/readonly. If you omit the update mode it defaults to
164 /leaky where that makes sense (as before) or /readonly where required.
166 The /noupdate option is now undocumented but still supported for
167 backwards compatibility. It is equivalent to /readonly except that in
168 ACLs where /readonly is required you may specify /leaky/noupdate or
169 /strict/noupdate which are treated the same as /readonly.
171 A useful new feature is the /count= option. This is a generalization
172 of the per_byte option, so that you can measure the throughput of other
173 aggregate values. For example, the per_byte option is now equivalent
174 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
176 The per_rcpt option has been generalized using the /count= mechanism
177 (though it's more complicated than the per_byte equivalence). When it is
178 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
179 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
180 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
181 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
182 Note that using per_rcpt with a non-readonly update mode in more than
183 one ACL will cause the recipients to be double-counted. (The per_mail
184 and per_byte options don't have this problem.)
186 The handling of very low rates has changed slightly. If the computed rate
187 is less than the event's count (usually one) then this event is the first
188 after a long gap. In this case the rate is set to the same as this event's
189 count, so that the first message of a spam run is counted properly.
191 The major new feature is a mechanism for counting the rate of unique
192 events. The new per_addr option counts the number of different
193 recipients that someone has sent messages to in the last time period. It
194 behaves like per_rcpt if all the recipient addresses are different, but
195 duplicate recipient addresses do not increase the measured rate. Like
196 the /count= option this is a general mechanism, so the per_addr option
197 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
198 example, measure the rate that a client uses different sender addresses
199 with the options per_mail/unique=$sender_address. There are further
200 details in the main documentation.
202 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
204 TF/06 Removed a few PCRE remnants.
206 TF/07 Automatically extract Exim's version number from tags in the git
207 repository when doing development or release builds.
209 PP/02 Raise smtp_cmd_buffer_size to 16kB.
210 Bugzilla 879. Patch from Paul Fisher.
212 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
213 Heavily based on revision 40f9a89a from Simon Arlott's tree.
216 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
218 PP/05 Variable $av_failed, true if the AV scanner deferred.
219 Bugzilla 1078. Patch from John Horne.
221 PP/06 Stop make process more reliably on build failure.
222 Bugzilla 1087. Patch from Heiko Schlittermann.
224 PP/07 Make maildir_use_size_file an _expandable_ boolean.
225 Bugzilla 1089. Patch from Heiko Schlittermann.
227 PP/08 Handle ${run} returning more data than OS pipe buffer size.
228 Bugzilla 1131. Patch from Holger Weiß.
230 PP/09 Handle IPv6 addresses with SPF.
231 Bugzilla 860. Patch from Wolfgang Breyha.
233 PP/10 GnuTLS: support TLS 1.2 & 1.1.
235 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
238 PP/11 match_* no longer expand right-hand-side by default.
239 New compile-time build option, EXPAND_LISTMATCH_RHS.
240 New expansion conditions, "inlist", "inlisti".
242 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
244 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
246 PP/14 fix log_write() format string regression from TF/03.
247 Bugzilla 1152. Patch from Dmitry Isaikin.
253 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
255 PP/02 Harmonised TLS library version reporting; only show if debugging.
256 Layout now matches that introduced for other libraries in 4.74 PP/03.
258 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
260 PP/04 New "dns_use_edns0" global option.
262 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
265 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
266 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
268 TK/01 Updated PolarSSL code to 0.14.2.
269 Bugzilla 1097. Patch from Andreas Metzler.
271 PP/07 Catch divide-by-zero in ${eval:...}.
274 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
277 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
278 format-string attack -- SECURITY: remote arbitrary code execution.
280 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
281 time unintentionally subject to list matching rules, letting the header
282 cause arbitrary Exim lookups (of items which can occur in lists, *not*
283 arbitrary string expansion). This allowed for information disclosure.
285 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
286 INT_MIN/-1 -- value coerced to INT_MAX.
292 NM/01 Workround for PCRE version dependency in version reporting
295 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
296 This fixes portability to compilers other than gcc, notably
297 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
299 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
300 makefiles for portability to HP-UX and POSIX correctness.
302 PP/01 Permit LOOKUP_foo enabling on the make command-line.
303 Also via indented variable definition in the Makefile.
304 (Debugging by Oliver Heesakkers).
306 PP/02 Restore caching of spamd results with expanded spamd_address.
307 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
309 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
310 Improves build reliability. Fix from: Frank Elsner
312 NM/02 Fix wide character breakage in the rfc2047 coding
313 Fixes bug 1064. Patch from Andrey N. Oktyabrski
315 NM/03 Allow underscore in dnslist lookups
316 Fixes bug 1026. Patch from Graeme Fowler
318 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
319 Code patches from Adam Ciarcinski of NetBSD.
321 NM/04 Fixed exiqgrep to cope with mailq missing size issue
324 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
325 is logged, to avoid truncation. Patch from John Horne.
327 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
328 Patch from Jakob Hirsch.
330 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
331 SQL string expansion failure details.
332 Patch from Andrey Oktyabrski.
334 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
335 Patch from Simon Arlott.
337 PP/09 New lookups functionality failed to compile on old gcc which rejects
338 extern declarations in function scope.
339 Patch from Oliver Fleischmann
341 PP/10 Use sig_atomic_t for flags set from signal handlers.
342 Check getgroups() return and improve debugging.
343 Fixed developed for diagnosis in bug 927 (which turned out to be
346 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
347 Patch from Mark Zealey.
349 PP/12 Bugzilla 1056: Improved spamd server selection.
350 Patch from Mark Zealey.
352 PP/13 Bugzilla 1086: Deal with maildir quota file races.
353 Based on patch from Heiko Schlittermann.
355 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
356 Patch from Uwe Doering, sign-off by Michael Haardt.
358 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
359 variable declaration deep within a block. Bug and patch from
362 PP/15 lookups-Makefile IRIX compatibilty coercion.
364 PP/16 Make DISABLE_DKIM build knob functional.
366 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
367 Patch by Simon Arlott
369 TF/03 Fix valgrind.h portability to C89 compilers that do not support
370 variable argument macros. Our copy now differs from upstream.
376 TF/01 Failure to get a lock on a hints database can have serious
377 consequences so log it to the panic log.
379 TF/02 Log LMTP confirmation messages in the same way as SMTP,
380 controlled using the smtp_confirmation log selector.
382 TF/03 Include the error message when we fail to unlink a spool file.
384 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
385 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
386 for maintaining out-of-tree patches for some time.
388 PP/01 Bugzilla 139: Documentation and portability issues.
389 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
390 Handle per-OS dynamic-module compilation flags.
392 PP/02 Let /dev/null have normal permissions.
393 The 4.73 fixes were a little too stringent and complained about the
394 permissions on /dev/null. Exempt it from some checks.
395 Reported by Andreas M. Kirchwitz.
397 PP/03 Report version information for many libraries, including
398 Exim version information for dynamically loaded libraries. Created
399 version.h, now support a version extension string for distributors
400 who patch heavily. Dynamic module ABI change.
402 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
403 privilege escalation vulnerability whereby the Exim run-time user
404 can cause root to append content of the attacker's choosing to
407 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
410 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
411 If dropping privileges for untrusted macros, we disabled normal logging
412 on the basis that it would fail; for the Exim run-time user, this is not
413 the case, and it resulted in successful deliveries going unlogged.
414 Fixed. Reported by Andreas Metzler.
420 PP/01 Date: & Message-Id: revert to normally being appended to a message,
421 only prepend for the Resent-* case. Fixes regression introduced in
422 Exim 4.70 by NM/22 for Bugzilla 607.
424 PP/02 Include check_rfc2047_length in configure.default because we're seeing
425 increasing numbers of administrators be bitten by this.
427 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
429 PP/03 Bugzilla 994: added openssl_options main configuration option.
431 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
433 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
435 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
437 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
438 without a peer certificate, leading to a segfault because of an
439 assumption that peers always have certificates. Be a little more
440 paranoid. Problem reported by Martin Tscholak.
442 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
443 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
444 NB: ClamAV planning to remove STREAM in "middle of 2010".
445 CL also introduces -bmalware, various -d+acl logging additions and
446 more caution in buffer sizes.
448 PP/09 Implemented reverse_ip expansion operator.
450 PP/10 Bugzilla 937: provide a "debug" ACL control.
452 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
454 PP/12 Bugzilla 973: Implement --version.
456 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
458 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
460 PP/15 Bugzilla 816: support multiple condition rules on Routers.
462 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
463 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
464 ignore trailing whitespace.
466 JJ/02 prevent non-panic DKIM error from being sent to paniclog
468 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
471 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
472 Notification from Dr Andrew Aitchison.
474 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
475 ExtendedDetectionInfo response format.
476 Notification from John Horne.
478 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
481 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
482 XSL and documented dependency on system catalogs, with examples of how
485 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
488 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
489 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
490 configuration file which is writeable by the Exim user or group.
492 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
493 of configuration files to cover files specified with the -C option if
494 they are going to be used with root privileges, not just the default
497 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
498 option (effectively making it always true).
500 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
501 files to be used while preserving root privileges.
503 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
504 that rogue child processes cannot use them.
506 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
507 run-time user, instead of root.
509 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
510 Exim run-time user without dropping privileges.
512 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
513 result string, instead of calling string_vformat() twice with the same
516 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
517 for other users. Others should always drop root privileges if they use
518 -C on the command line, even for a whitelisted configure file.
520 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
522 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
528 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
529 $data_path, and $header_path variables; fixed documentation bugs and
532 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
533 exipick to access non-standard spools, including the "frozen" queue
536 NM/01 Bugzilla 965: Support mysql stored procedures.
537 Patch from Alain Williams
539 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
541 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
542 Patch from Andreas Metzler
544 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
545 Patch from Kirill Miazine
547 NM/05 Bugzilla 671: Added umask to procmail example.
549 JJ/03 installed exipick 20100323.0, fixing doc bug
551 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
552 directory. Notification and patch from Dan Rosenberg.
554 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
556 TK/02 Improve log output when DKIM signing operation fails.
558 MH/01 Treat the transport option dkim_domain as a colon separated
559 list, not as a single string, and sign the message with each element,
560 omitting multiple occurences of the same signer.
562 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
563 Bugzilla 985, 986. Patch by Simon Arlott
565 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
566 Patch by Simon Arlott
568 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
569 MBX locking. Notification from Dan Rosenberg.
575 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
577 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
579 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
581 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
583 NM/04 Bugzilla 915: Flag broken perl installation during build.
589 TK/01 Added patch by Johannes Berg that expands the main option
590 "spamd_address" if it starts with a dollar sign.
592 TK/02 Write list of recipients to X-Envelope-Sender header when building
593 the mbox-format spool file for content scanning (suggested by Jakob
596 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
597 (http://www.dcc-servers.net/) support via dccifd. Activated by
598 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
600 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
601 by Mark Daniel Reidel <mr@df.eu>.
603 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
604 When building exim an external PCRE library is now needed -
605 PCRE is a system library on the majority of modern systems.
606 See entry on PCRE_LIBS in EDITME file.
608 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
609 conversation. Added nologin parameter to request.
610 Patch contributed by Kirill Miazine.
612 TF/01 Do not log submission mode rewrites if they do not change the address.
614 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
616 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
617 log files in place. Contributed by Roberto Lima.
619 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
621 TF/03 Bugzilla 615: When checking the local_parts router precondition
622 after a local_part_suffix or local_part_prefix option, Exim now
623 does not use the address's named list lookup cache, since this
624 contains cached lookups for the whole local part.
626 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
627 Robert Millan. Documentation is in experimental-spec.txt.
629 TF/04 Bugzilla 668: Fix parallel build (make -j).
631 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
633 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
634 Patch provided by Jan Srzednicki.
636 TF/05 Leading white space used to be stripped from $spam_report which
637 wrecked the formatting. Now it is preserved.
639 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
640 that they are available at delivery time.
642 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
644 TF/08 TLS error reporting now respects the incoming_interface and
645 incoming_port log selectors.
647 TF/09 Produce a more useful error message if an SMTP transport's hosts
648 setting expands to an empty string.
650 NM/06 Bugzilla 744: EXPN did not work under TLS.
651 Patch provided by Phil Pennock.
653 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
654 Patch provided by Richard Godbee.
656 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
657 acl_smtp_notquit, added index entry.
659 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
660 Patch provided by Eugene Bujak.
662 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
663 accept(). Patch provided by Maxim Dounin.
665 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
666 Patch provided by Phil Pennock.
668 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
670 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
671 Patch provided by Brad "anomie" Jorsch.
673 NM/14 Bugzilla 622: Added timeout setting to transport filter.
674 Patch provided by Dean Brooks.
676 TK/05 Add native DKIM support (does not depend on external libraries).
678 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
679 Patch provided by Graeme Fowler.
681 NM/16 Bugzilla 851: Documentation example syntax fix.
683 NM/17 Changed NOTICE file to remove references to embedded PCRE.
685 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
688 NM/19 Bugzilla 745: TLS version reporting.
689 Patch provided by Phil Pennock.
691 NM/20 Bugzilla 167: bool: condition support.
692 Patch provided by Phil Pennock.
694 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
695 clients. Patch provided by Phil Pennock.
697 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
698 Patch provided by Brad "anomie" Jorsch.
700 NM/23 Bugzilla 687: Fix misparses in eximstats.
701 Patch provided by Heiko Schlittermann.
703 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
704 Patch provided by Heiko Schlittermann.
706 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
707 plus update to original patch.
709 NM/26 Bugzilla 799: Documentation correction for ratelimit.
711 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
712 Patch provided by David Brownlee.
714 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
716 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
718 NM/30 Bugzilla 888: TLS documentation bugfixes.
720 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
722 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
723 Unlike the original bugzilla I have changed all shell scripts in src tree.
725 NM/33 Bugzilla 898: Transport filter timeout fix.
726 Patch by Todd Rinaldo.
728 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
729 Patch by Serge Demonchaux.
731 NM/35 Bugzilla 39: Base64 decode bug fixes.
732 Patch by Jakob Hirsch.
734 NM/36 Bugzilla 909: Correct connect() call in dcc code.
736 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
738 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
740 NM/39 Bugzilla 911: Fixed MakeLinks build script.
746 TK/01 Add preliminary DKIM support. Currently requires a forked version of
747 ALT-N's libdkim that I have put here:
748 http://duncanthrax.net/exim-experimental/
750 Note to Michael Haardt: I had to rename some vars in sieve.c. They
751 were called 'true' and it seems that C99 defines that as a reserved
752 keyword to be used with 'bool' variable types. That means you could
753 not include C99-style headers which use bools without triggering
754 build errors in sieve.c.
756 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
757 as mailq or other aliases. Changed the --help handling significantly
758 to do whats expected. exim_usage() emits usage/help information.
760 SC/01 Added the -bylocaldomain option to eximstats.
762 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
764 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
766 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
772 PH/01 Another patch from the Sieve maintainer.
774 PH/02 When an IPv6 address is converted to a string for single-key lookup
775 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
776 dots are used instead of colons so that keys in lsearch files need not
777 contain colons. This was done some time before quoting was made available
778 in lsearch files. However, iplsearch files do require colons in IPv6 keys
779 (notated using the quote facility) so as to distinguish them from IPv4
780 keys. This meant that lookups for IP addresses in host lists did not work
781 for iplsearch lookups.
783 This has been fixed by arranging for IPv6 addresses to be expressed with
784 colons if the lookup type is iplsearch. This is not incompatible, because
785 previously such lookups could never work.
787 The situation is now rather anomolous, since one *can* have colons in
788 ordinary lsearch keys. However, making the change in all cases is
789 incompatible and would probably break a number of configurations.
791 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
794 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
795 conversion specification without a maximum field width, thereby enabling
796 a rogue spamd server to cause a buffer overflow. While nobody in their
797 right mind would setup Exim to query an untrusted spamd server, an
798 attacker that gains access to a server running spamd could potentially
799 exploit this vulnerability to run arbitrary code as the Exim user.
801 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
802 $primary_hostname instead of what libspf2 thinks the hosts name is.
804 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
805 a directory entry by the name of the lookup key. Previously, if a
806 symlink pointed to a non-existing file or a file in a directory that
807 Exim lacked permissions to read, a lookup for a key matching that
808 symlink would fail. Now it is enough that a matching directory entry
809 exists, symlink or not. (Bugzilla 503.)
811 PH/03 The body_linecount and body_zerocount variables are now exported in the
814 PH/04 Added the $dnslist_matched variable.
816 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
817 This means they are set thereafter only if the connection becomes
820 PH/06 Added the client_condition to authenticators so that some can be skipped
821 by clients under certain conditions.
823 PH/07 The error message for a badly-placed control=no_multiline_responses left
824 "_responses" off the end of the name.
826 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
828 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
829 (without spaces) instead of just copying the configuration text.
831 PH/10 Added the /noupdate option to the ratelimit ACL condition.
833 PH/11 Added $max_received_linelength.
835 PH/12 Added +ignore_defer and +include_defer to host lists.
837 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
838 way in which PCRE > 7.0 is built.
840 PH/14 Implemented queue_only_load_latch.
842 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
843 MAIL command. The effect was to mangle the value on 64-bit systems.
845 PH/16 Another patch from the Sieve maintainer.
847 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
849 PH/18 If a system quota error occurred while trying to create the file for
850 a maildir delivery, the message "Mailbox is full" was not appended to the
851 bounce if the delivery eventually timed out. Change 4.67/27 below applied
852 only to a quota excession during the actual writing of the file.
854 PH/19 It seems that peer DN values may contain newlines (and other non-printing
855 characters?) which causes problems in log lines. The DN values are now
856 passed through string_printing() before being added to log lines.
858 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
859 and InterBase are left for another time.)
861 PH/21 Added message_body_newlines option.
863 PH/22 Guard against possible overflow in moan_check_errorcopy().
865 PH/23 POSIX allows open() to be a macro; guard against that.
867 PH/24 If the recipient of an error message contained an @ in the local part
868 (suitably quoted, of course), incorrect values were put in $domain and
869 $local_part during the evaluation of errors_copy.
875 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
876 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
879 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
880 issue a MAIL command.
882 PH/02 In an ACL statement such as
884 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
886 if a client was not listed at all, or was listed with a value other than
887 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
888 the condition was not true (as it should be), so access was not denied.
889 The bug was that the ! inversion was incorrectly passed on to the second
890 item. This has been fixed.
892 PH/03 Added additional dnslists conditions == and =& which are different from
893 = and & when the dns lookup returns more than one IP address.
895 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
896 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
898 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
899 FSYNC, which compiles an option called disable_fsync that allows for
900 bypassing fsync(). The documentation is heavily laced with warnings.
902 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
904 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
905 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
906 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
907 including adding "make clean"; (3) Added -fPIC when compiling the test
908 dynamically loaded module, to get rid of a warning.
910 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
911 message fails, move_frozen_messages = true and ignore_bounce_errors_after
912 = 0s. The bug is otherwise harmless.
914 PH/07 There was a bug in the dovecot authenticator such that the value of
915 $auth1 could be overwritten, and so not correctly preserved, after a
916 successful authentication. This usually meant that the value preserved by
917 the server_setid option was incorrect.
919 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
921 PH/09 Installed PCRE release 7.0.
923 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
924 run for batched SMTP input. It is now run at the start of every message
925 in the batch. While fixing this I discovered that the process information
926 (output by running exiwhat) was not always getting set for -bs and -bS
927 input. This is fixed, and it now also says "batched" for BSMTP.
929 PH/11 Added control=no_pipelining.
931 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
932 patch, slightly modified), and move the expansion of helo_data till after
933 the connection is made in the smtp transport (so it can use these
936 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
938 PH/14 Added log_selector = +pid.
940 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
942 PH/16 Add ${if forany and ${if forall.
944 PH/17 Added dsn_from option to vary the From: line in DSNs.
946 PH/18 Flush SMTP output before performing a callout, unless control =
947 no_callout_flush is set.
949 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
950 was true (the default) a successful delivery failed to delete the retry
951 item, thus causing premature timeout of the address. The bug is now
954 PH/20 Added hosts_avoid_pipelining to the smtp transport.
956 PH/21 Long custom messages for fakedefer and fakereject are now split up
957 into multiline reponses in the same way that messages for "deny" and
958 other ACL rejections are.
960 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
961 with slight modification.
963 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
964 draft, changing the syntax and factoring some duplicate code".
966 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
967 for deliveries of the second and subsequent messages over the same SMTP
970 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
971 ${reduce, with only minor "tidies".
973 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
975 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
976 expansion side effects.
978 PH/27 When a message times out after an over-quota error from an Exim-imposed
979 quota, the bounce message says "mailbox is full". This message was not
980 being given when it was a system quota that was exceeded. It now should
983 MH/03 Made $recipients available in local_scan(). local_scan() already has
984 better access to the recipient list through recipients_list[], but
985 $recipients can be useful in postmaster-provided expansion strings.
987 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
988 in the case of a MAIL command with additional options following the
989 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
990 were accidentally chopped off.
992 PH/29 SMTP synchronization checks are implemented when a command is read -
993 there is a check that no more input is waiting when there shouldn't be
994 any. However, for some commands, a delay in an ACL can mean that it is
995 some time before the response is written. In this time, more input might
996 arrive, invalidly. So now there are extra checks after an ACL has run for
997 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
998 pipelining has not been advertised.
1000 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1002 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1003 returned in the SMTP response when smtp_return_error_details was false.
1004 This has been fixed.
1006 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1007 instead of the C I/O that was originally supplied, because problems were
1008 reported on Solaris.
1010 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1011 Exim which did not show up earlier: it was assuming that a call to
1012 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1013 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1014 was a macro that became an assignment, so it seemed to work. This has
1015 changed to a proper function call with a void return, hence the compile
1016 error. Exim's code has been fixed.
1018 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1021 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1023 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1025 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1026 because it turns out to be too big or there is a timeout) there is no
1027 "Completed" line in the log. When some messages of this type were
1028 selected by exigrep, they were listed as "not completed". Others were
1029 picked up by some special patterns. I have improved the selection
1030 criteria to be more general.
1032 PH/38 The host_find_failed option in the manualroute router can now be set
1033 to "ignore", to completely ignore a host whose IP address cannot be
1034 found. If all hosts are ignored, the behaviour is controlled by the new
1035 host_all_ignored option.
1037 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1038 homing or because of multiple MX records with /mx) generated more than
1039 one IP address, and the following item turned out to be the local host,
1040 all the secondary addresses of the first item were incorrectly removed
1041 from the list, along with the local host and any following hosts (which
1042 is what is supposed to happen).
1044 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1045 whoever called Exim into the -H file. In the case of the daemon it was
1046 behaving confusingly. When first started, it used values for whoever
1047 started the daemon, but after a SIGHUP it used the Exim user (because it
1048 calls itself on a restart). I have changed the code so that it now always
1051 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1052 message are rejected with the same error (e.g. no authentication or bad
1053 sender address), and a DATA command is nevertheless sent (as can happen
1054 with PIPELINING or a stupid MUA), the error message that was given to the
1055 RCPT commands is included in the rejection of the DATA command. This is
1056 intended to be helpful for MUAs that show only the final error to their
1059 PH/42 Another patch from the Sieve maintainer.
1061 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1062 Eximstats - Fixed some broken HTML links and added missing column headers
1064 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1065 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1067 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1069 PH/43 Yet another patch from the Sieve maintainer.
1071 PH/44 I found a way to check for a TCP/IP connection going away before sending
1072 the response to the final '.' that terminates a message, but only in the
1073 case where the client has not sent further data following the '.'
1074 (unfortunately, this is allowed). However, in many cases there won't be
1075 any further data because there won't be any more messages to send. A call
1076 to select() can be used: if it shows that the input is "ready", there is
1077 either input waiting, or the socket has been closed. An attempt to read
1078 the next input character can distinguish the two cases. Previously, Exim
1079 would have sent an OK response which the client would never have see.
1080 This could lead to message repetition. This fix should cure that, at
1081 least in a lot of common cases.
1083 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1084 advertised in response to EHLO.
1090 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1091 fixed by 4.65/MH/01 (is this a record?) are fixed:
1093 (i) An empty string was always treated as zero by the numeric comparison
1094 operators. This behaviour has been restored.
1096 (ii) It is documented that the numeric comparison operators always treat
1097 their arguments as decimal numbers. This was broken in that numbers
1098 starting with 0 were being interpreted as octal.
1100 While fixing these problems I realized that there was another issue that
1101 hadn't been noticed. Values of message_size_limit (both the global option
1102 and the transport option) were treated as octal if they started with 0.
1103 The documentation was vague. These values are now always treated as
1104 decimal, and I will make that clear in the documentation.
1110 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1111 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1114 MH/01 Don't check that the operands of numeric comparison operators are
1115 integers when their expansion is in "skipping" mode (fixes bug
1116 introduced by 4.64-PH/07).
1118 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1119 child addresses, Exim now panics and dies. Previously, because the count
1120 is held in a short int, deliveries were likely to be lost. As such a
1121 large number of recipients for a single message is ridiculous
1122 (performance will be very, very poor), I have chosen to impose a limit
1123 rather than extend the field.
1129 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1130 leftover -K file (the existence of which was triggered by #402).
1131 While we were at it, introduced process PID as part of the -K
1132 filename. This should rule out race conditions when creating
1135 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1136 processing considerably. Previous code took too long for large mails,
1137 triggering a timeout which in turn triggers #401.
1139 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1140 in the DK code in transports.c. sendfile() is not really portable,
1141 hence the _LINUX specificness.
1143 TF/01 In the add_headers option to the mail command in an Exim filter,
1144 there was a bug that Exim would claim a syntax error in any
1145 header after the first one which had an odd number of characters
1148 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1149 callout verification, Exim cached a "reject" for the entire domain. This
1150 is correct for most verifications, but it is not correct for a recipient
1151 verification with use_sender or use_postmaster set, because in that case
1152 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1153 case of MAIL FROM:<> rejection from other early rejections (e.g.
1154 rejection of HELO). When verifying a recipient using a non-null MAIL
1155 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1156 Whatever the result of the callout, the value of the domain cache is
1157 left unchanged (for any other kind of callout, getting as far as trying
1158 RCPT means that the domain itself is ok).
1160 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1163 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1164 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1165 session, the daemon ignores ECONNECT errors and logs others; it now
1166 ignores EPIPE as well.
1168 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1169 (quoted-printable decoding).
1171 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1172 later the small subsequent patch to fix an introduced bug.
1174 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1176 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1178 PH/08 An error is now given if message_size_limit is specified negative.
1180 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1181 to be given (somewhat) arbitrary names.
1183 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1186 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1187 miscellaneous code fixes
1189 PH/10 Added the log_reject_target ACL modifier to specify where to log
1192 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1193 hostname. This is wrong, because it relates to the incoming message (and
1194 probably the interface on which it is arriving) and not to the outgoing
1195 callout (which could be using a different interface). This has been
1196 changed to use the value of the helo_data option from the smtp transport
1197 instead - this is what is used when a message is actually being sent. If
1198 there is no remote transport (possible with a router that sets up host
1199 addresses), $smtp_active_hostname is used.
1201 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1202 tweaks were necessary in order to get it to work (see also 21 below):
1203 (a) The code assumed that strncpy() returns a negative number on buffer
1204 overflow, which isn't the case. Replaced with Exim's string_format()
1206 (b) There were several signed/unsigned issues. I just did the minimum
1207 hacking in of casts. There is scope for a larger refactoring.
1208 (c) The code used strcasecmp() which is not a standard C function.
1209 Replaced with Exim's strcmpic() function.
1210 (d) The code set only $1; it now sets $auth1 as well.
1211 (e) A simple test gave the error "authentication client didn't specify
1212 service in request". It would seem that Dovecot has changed its
1213 interface. Fortunately there's a specification; I followed it and
1214 changed what the client sends and it appears to be working now.
1216 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1219 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1220 address A is aliased to B and C, where B exists and C does not. Without
1221 -v the output is "A verified" because verification stops after a
1222 successful redirection if more than one address is generated. However,
1223 with -v the child addresses are also verified. Exim was outputting "A
1224 failed to verify" and then showing the successful verification for C,
1225 with its parentage. It now outputs "B failed to verify", showing B's
1226 parentage before showing the successful verification of C.
1228 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1229 look up a TXT record in a specific list after matching in a combined
1232 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1233 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1234 they consult the DNS. I had assumed they would set it the way they
1235 wanted; and indeed my experiments on Linux seem to show that in some
1236 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1237 To be on the safe side, however, I have now made the interface to
1238 host_find_byname() similar to host_find_bydns(), with an argument
1239 containing the DNS resolver options. The host_find_byname() function now
1240 sets these options at its start, just as host_find_bydns() does. The smtp
1241 transport options dns_qualify_single and dns_search_parents are passed to
1242 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1243 of host_find_byname() use the default settings of RES_DEFNAMES
1244 (qualify_single) but not RES_DNSRCH (search_parents).
1246 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1247 spool_read_header() do less string testing, by means of a preliminary
1248 switch on the second character of optional "-foo" lines. (This is
1249 overdue, caused by the large number of possibilities that now exist.
1250 Originally there were few.) While I was there, I also converted the
1251 str(n)cmp tests so they don't re-test the leading "-" and the first
1252 character, in the hope this might squeeze out yet more improvement.
1254 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1255 flag allowing group syntax was set by the header_syntax check but not
1256 turned off, possible causing trouble later; (2) The flag was not being
1257 set at all for the header_verify test, causing "group"-style headers to
1258 be rejected. I have now set it in this case, and also caused header_
1259 verify to ignore an empty address taken from a group. While doing this, I
1260 came across some other cases where the code for allowing group syntax
1261 while scanning a header line wasn't quite right (mostly, not resetting
1262 the flag correctly in the right place). These bugs could have caused
1263 trouble for malformed header lines. I hope it is now all correct.
1265 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1266 with the "reply" argument non-NULL. The code, however (which originally
1267 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1268 but it didn't always do it. This confused somebody who was copying the
1269 code for some other use. I have removed all the tests.
1271 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1272 feature that was used to support insecure browsers during the U.S. crypto
1273 embargo. It requires special client support, and Exim is probably the
1274 only MTA that supported it -- and would never use it because real RSA is
1275 always available. This code has been removed, because it had the bad
1276 effect of slowing Exim down by computing (never used) parameters for the
1277 RSA_EXPORT functionality.
1279 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1280 authenticator to fail if there's a tab character in the incoming data
1281 (there should never be unless someone is messing about, as it's supposed
1282 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1283 if the connection is using TLS or if the remote IP is the same as the
1284 local IP, and the "valid-client-cert option" if a client certificate has
1287 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1288 authenticators. This can be used for authorization after authentication
1289 succeeds. (In the case of plaintext, it servers for both authentication
1292 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1293 if any retry times were supplied.
1295 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1296 connection, where there is no client IP address to check. In this
1297 situation, the verify now always succeeds.
1299 PH/25 Applied John Jetmore's -Mset patch.
1301 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1303 PH/27 In a string expansion for a processed (not raw) header when multiple
1304 headers of the same name were present, leading whitespace was being
1305 removed from all of them, but trailing whitespace was being removed only
1306 from the last one. Now trailing whitespace is removed from each header
1307 before concatenation. Completely empty headers in a concatenation (as
1308 before) are ignored.
1310 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1311 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1313 PH/29 [Removed. This was a change that I later backed out, and forgot to
1314 correct the ChangeLog entry (that I had efficiently created) before
1315 committing the later change.]
1317 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1318 address errors (4xx response to RCPT) over the same connection as other
1319 messages routed to the same hosts. Such deliveries are always "forced",
1320 so retry times are not inspected. This resulted in far too many retries
1321 for the affected addresses. The effect occurred only when there were more
1322 hosts than the hosts_max_try setting in the smtp transport when it had
1323 the 4xx errors. Those hosts that it had tried were not added to the list
1324 of hosts for which the message was waiting, so if all were tried, there
1325 was no problem. Two fixes have been applied:
1327 (i) If there are any address or message errors in an SMTP delivery, none
1328 of the hosts (tried or untried) are now added to the list of hosts
1329 for which the message is waiting, so the message should not be a
1330 candidate for sending over the same connection that was used for a
1331 successful delivery of some other message. This seems entirely
1332 reasonable: after all the message is NOT "waiting for some host".
1333 This is so "obvious" that I'm not sure why it wasn't done
1334 previously. Hope I haven't missed anything, but it can't do any
1335 harm, as the worst effect is to miss an optimization.
1337 (ii) If, despite (i), such a delivery is accidentally attempted, the
1338 routing retry time is respected, so at least it doesn't keep
1339 hammering the server.
1341 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1342 in ${readsocket because some servers need this prod.
1344 PH/32 Added some extra debug output when updating a wait-xxx database.
1346 PH/33 The hint "could be header name not terminated by colon", which has been
1347 given for certain expansion errors for a long time, was not being given
1348 for the ${if def:h_colon_omitted{... case.
1350 PH/34 The spec says: "With one important exception, whenever a domain list is
1351 being scanned, $domain contains the subject domain." There was at least
1352 one case where this was not true.
1354 PH/35 The error "getsockname() failed: connection reset by peer" was being
1355 written to the panic log as well as the main log, but it isn't really
1356 panic-worthy as it just means the connection died rather early on. I have
1357 removed the panic log writing for the ECONNRESET error when getsockname()
1360 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1361 runs only) independently of the message's sender address. This meant
1362 that, if the 4xx error was in fact related to the sender, a different
1363 message to the same recipient with a different sender could confuse
1364 things. In particualar, this can happen when sending to a greylisting
1365 server, but other circumstances could also provoke similar problems.
1366 I have changed the default so that the retry time for these errors is now
1367 based a combination of the sender and recipient addresses. This change
1368 can be overridden by setting address_retry_include_sender=false in the
1371 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1372 remote server are returned as part of bounce messages. This was not
1373 happening for LMTP over a pipe (the lmtp transport), but now it is the
1374 same for both kinds of LMTP.
1376 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1377 in header lines that were in fact CNAMEs. This is no longer the case.
1379 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1380 and queue runs started by the daemon processed all messages. This has
1381 been fixed so that -R and -S can now usefully be given with -q<time>.
1383 PH/40 Import PCRE release 6.7 (fixes some bugs).
1385 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1387 PH/42 Give an error if -q is specified more than once.
1389 PH/43 Renamed the variables $interface_address and $interface_port as
1390 $received_ip_address and $received_port, to make it clear that these
1391 values apply to message reception, and not to the outgoing interface when
1392 a message is delivered. (The old names remain recognized, of course.)
1394 PH/44 There was no timeout on the connect() call when using a Unix domain
1395 socket in the ${readsocket expansion. There now is.
1397 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1398 be meaningful with "accept".
1400 SC/01 Eximstats V1.43
1401 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1403 SC/02 Eximstats V1.44
1404 Use a glob alias rather than an array ref in the generated
1405 parser. This improves both readability and performance.
1407 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1408 Collect SpamAssassin and rejection statistics.
1409 Don't display local sender or destination tables unless
1410 there is data to show.
1411 Added average volumes into the top table text output.
1413 SC/04 Eximstats V1.46
1414 Collect data on the number of addresses (recipients)
1415 as well as the number of messages.
1417 SC/05 Eximstats V1.47
1418 Added 'Message too big' to the list of mail rejection
1419 reasons (thanks to Marco Gaiarin).
1421 SC/06 Eximstats V1.48
1422 Mainlog lines which have GMT offsets and are too short to
1423 have a flag are now skipped.
1425 SC/07 Eximstats V1.49 (Alain Williams)
1426 Added the -emptyok flag.
1428 SC/08 Eximstats V1.50
1429 Fixes for obtaining the IP address from reject messages.
1431 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1432 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1433 whitesspace changes from 4.64-PH/27
1435 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1438 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1439 are found, allow negative numbers in numeric criteria)
1441 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1443 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1444 to match changes made in 4.64-PH/43
1446 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1448 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1449 "domain" is actually an IP address, and force a failure. This locks out
1450 those revolvers/nameservers that support "A-for-A" lookups, in
1451 contravention of the specifications.
1453 PH/48 When a host name was looked up from an IP address, and the subsequent
1454 forward lookup of the name timed out, the host name was left in
1455 $sender_host_name, contrary to the specification.
1457 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1458 restricted to single-key lookups, Exim was not diagnosing an error if
1459 * or *@ was used with a query-style lookup.
1461 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1463 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1464 long ago, but noone interested enough thought of it. Let's just say that
1465 the "1.1" means that there are some new functions that weren't there at
1466 some point in the past.
1468 PH/51 Error processing for expansion failure of helo_data from an smtp
1469 transport during callout processing was broken.
1471 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1472 tested/used via the -bh/-bhc/-bs options.
1474 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1475 bug, fixed in subsequent PCRE releases).
1477 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1478 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1480 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1486 SC/01 Use a glob alias rather than an array ref in eximstats generated
1487 parser. This improves both readability and performance.
1489 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1490 Don't display local sender or destination tables in eximstats unless
1491 there is data to show.
1492 Added average volumes into the eximstats top table text output.
1494 SC/03 Collect data on the number of addresses (recipients) as well
1495 as the number of messages in eximstats.
1497 TF/01 Correct an error in the documentation for the redirect router. Exim
1498 does (usually) call initgroups() when daemonizing.
1500 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1501 with consistent privilege compared to when running as a daemon.
1503 TF/03 Note in the spec that $authenticated_id is not set for local
1504 submissions from trusted users.
1506 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1507 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1509 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1510 by adding some example configuration directives to the default
1511 configuration file. A little bit of work is required to uncomment the
1512 directives and define how usernames and passwords are checked, but
1513 there is now a framework to start from.
1515 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1516 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1517 without this. I don't know how relevant this is to other LDAP libraries.
1519 PH/02 Add the verb name to the "unknown ACL verb" error.
1521 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1523 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1525 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1526 directory not expanded when it should be if an expanded home directory
1527 was set for the address (which is overridden by the transport).
1529 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1532 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1533 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1534 because it is too late at that time, and has no effect.
1536 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1537 security issue with \' (bugzilla #107). I could not use the
1538 PQescapeStringConn() function, because it needs a PGconn value as one of
1541 PH/08 When testing addresses using -bt, indicate those final addresses that
1542 are duplicates that would not cause an additional delivery. At least one
1543 person was confused, thinking that -bt output corresponded to deliveries.
1544 (Suppressing duplicates isn't a good idea as you lose the information
1545 about possibly different redirections that led to the duplicates.)
1547 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1548 systems where poll() doesn't work, in particular OS X.
1550 PH/10 Added more information to debugging output for retry time not reached.
1552 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1553 operations in malware.c.
1555 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1558 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1559 syslog_duplication set false, log lines that would normally be written
1560 both the the main log and to the reject log were not written to syslog at
1563 PH/14 In the default configuration, change the use of "message" in ACL warn
1564 statements to "add_header".
1566 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1567 not followed by a command (e.g. "seen endif").
1569 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1570 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1573 PH/17 Added extra conditions to the default value of delay_warning_condition
1577 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1578 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1579 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1582 The Auto-Submitted: and various List- headers are standardised, whereas I
1583 don't think Precedence: ever was.
1585 PH/18 Refactored debugging code in route_finduser() to show more information,
1586 in particular, the error code if getpwnam() issues one.
1588 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1589 This is apparently needed in addition to the PH/07 change above to avoid
1590 any possible encoding problems.
1592 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1593 but not after initializing Perl.
1595 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1596 output them only if debugging. By default they are written stderr,
1597 apparently, which is not desirable.
1599 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1602 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1605 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1607 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1608 authenticated or an ident call has been made. Suppress the default
1609 values for $authenticated_id and $authenticated_sender (but permit -oMai
1610 and -oMas) when testing with -bh.
1612 PH/24 Re-jigged the order of the tests in the default configuration so that the
1613 tests for valid domains and recipients precede the DNS black list and CSA
1614 tests, on the grounds that those ones are more expensive.
1616 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1617 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1618 This bug exists in every version of Exim that I still have, right back to
1621 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1622 However, an attempt to turn on case-sensitivity in a regex key by
1623 including (?-i) didn't work because the subject string was already
1624 lowercased, and the effects were non-intuitive. It turns out that a
1625 one-line patch can be used to allow (?-i) to work as expected.
1631 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1632 other effects) broke the use of negated acl sub-conditions.
1634 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1637 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1638 "Deny" causes Exim to reject the incoming connection with a 554 error.
1639 Unfortunately, if there is a major crisis, such as a disk failure,
1640 tcp-wrappers gives "deny", whereas what one would like would be some
1641 kind of temporary error. A kludge has been added to help with this.
1642 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1643 554 error is used if errno is still zero or contains ENOENT (which occurs
1644 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1647 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1649 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1650 errors. Otherwise a message that provokes a temporary error (when other
1651 messages do not) can cause a whole host to time out.
1653 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1654 addresses were routed directly to files or pipes from a redirect router.
1655 File deliveries just didn't batch; pipe deliveries might have suffered
1658 PH/06 A failure to get a lock for a hints database would erroneously always say
1659 "Failed to get write lock", even when it was really a read lock.
1661 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1662 of 0600. This has been changed to use the value of the lockfile_mode
1663 option (which defaults to 0600).
1665 PH/08 Applied small patch from the Sieve maintainer.
1667 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1668 folder from quota calculations, a direct delivery into this folder messed
1669 up the contents of the maildirsize file. This was because the regex was
1670 used only to exclude .Trash (or whatever) when the size of the mailbox
1671 was calculated. There was no check that a delivery was happening into an
1672 excluded directory. This bug has been fixed by ignoring all quota
1673 processing for deliveries into excluded directories.
1675 PH/10 Added the maildirfolder_create_regex option to appendfile.
1681 PH/01 The code for finding all the local interface addresses on a FreeBSD
1682 system running IPv6 was broken. This may well have applied to all BSD
1683 systems, as well as to others that have similar system calls. The broken
1684 code found IPv4 interfaces correctly, but gave incorrect values for the
1685 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1686 that it would not match correctly against @[] and not recognize the IPv6
1689 PH/02 The ipliteral router was not recognizing addresses of the form user@
1690 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1692 PH/03 Added disable_ipv6.
1694 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1695 decoded form, because it is most often used to construct To: headers
1696 lines in autoreplies, and the decoded form may well be syntactically
1697 invalid. However, $reply_address has leading white space removed, and all
1698 newlines turned into spaces so that the autoreply transport does not
1701 PH/05 If group was specified without a user on a router, and no group or user
1702 was specified on a transport, the group from the router was ignored.
1704 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1705 for visible compile-time settings that can be used to change these
1706 numbers, for those that want even more. Backwards compatibility with old
1707 spool files has been maintained. However, going back to a previous Exim
1708 release will lost any variables that are in spool files.
1710 PH/07 Two small changes when running in the test harness: increase delay when
1711 passing a TCP/IP connection to a new process, in case the original
1712 process has to generate a bounce, and remove special handling of
1713 127.0.0.2 (sic), which is no longer necessary.
1715 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1716 be the same on different OS.
1718 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1721 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1722 whether --show-vars was specified or not
1724 JJ/02 exipick: Added support for new ACL variable spool format introduced
1727 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1728 syntactically invalid From: or Reply-to: line, and a filter used this to
1729 generate an autoreply, and therefore failed to obtain an address for the
1730 autoreply, Exim could try to deliver to a non-existent relative file
1731 name, causing unrelated and misleading errors. What now happens is that
1732 it logs this as a hard delivery error, but does not attempt to create a
1735 PH/11 The exinext utility has a -C option for testing purposes, but although
1736 the given file was scanned by exinext itself; it wasn't being passed on
1737 when Exim was called.
1739 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1740 an end-of-file indication when reading a command response.
1742 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1743 compiled. In many other places in Exim, IPv6 addresses are always
1744 recognized, so I have changed this. It also means that IPv4 domain
1745 literals of the form [IPV4:n.n.n.n] are now always recognized.
1747 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1748 used if the router is not running as root, for example, when verifying at
1749 ACL time, or when using -bh. The debugging output from this situation was
1750 non-existent - all you got was a failure to exec. I have made two
1753 (a) Failures to set uid/gid, the current directory, or a process leader
1754 in a subprocess such as that created by queryprogram now generate
1755 suitable debugging ouput when -d is set.
1757 (b) The queryprogram router detects when it is not running as root,
1758 outputs suitable debugging information if -d is set, and then runs
1759 the subprocess without attempting to change uid/gid.
1761 PH/15 Minor change to Makefile for building test_host (undocumented testing
1764 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1765 additional section of a DNS packet that returns MX or SRV records.
1766 Instead, it always explicitly searches for A/AAAA records. This avoids
1767 major problems that occur when a DNS server includes only records of one
1768 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1769 fixed another bug: if SRV records were looked up and the corresponding
1770 address records were *not* found in the additional section, the port
1771 values from the SRV records were lost.
1773 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1774 using the correct key (the original address) when searching the retry
1775 rules in order to find which one to use for generating the retry hint.
1777 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1778 adding the default one. Similarly, if it contains a Reply-To: header, the
1779 errors_reply_to option, if set, is not used.
1781 PH/19 When calculating a retry time, Exim used to measure the "time since
1782 failure" by looking at the "first failed" field in the retry record. Now
1783 it does not use this if it is later than than the arrival time of the
1784 message. Instead it uses the arrival time. This makes for better
1785 behaviour in cases where some deliveries succeed, thus re-setting the
1786 "first failed" field. An example is a quota failure for a huge message
1787 when small messages continue to be delivered. Without this change, the
1788 "time since failure" will always be short, possible causing more frequent
1789 delivery attempts for the huge message than are intended.
1790 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1792 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1793 $1, $2, $3) because the numerical variables can be reset during some
1794 expansion items (e.g. "match"), thereby losing the authentication data.
1796 PH/21 Make -bV show the size of off_t variables so that the test suite can
1797 decide whether to run tests for quotas > 2G.
1799 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1800 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1801 filecount value is greater than 2G or if a quota value is greater than 2G
1802 on a system where the size of off_t is not greater than 4, a panic error
1805 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1806 never match. The debug and -bh output now contains an explicit error
1807 message indicating a malformed IPv4 address or mask.
1809 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1810 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1811 PH/23 above applies.
1813 PH/25 Do not write to syslog when running in the test harness. The only
1814 occasion when this arises is a failure to open the main or panic logs
1815 (for which there is an explicit test).
1817 PH/26 Added the /no_tell option to "control=freeze".
1819 PH/27 If a host name lookup failed very early in a connection, for example, if
1820 the IP address matched host_lookup and the reverse lookup yielded a name
1821 that did not have a forward lookup, an error message of the form "no IP
1822 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1823 could be logged. Now it outputs the IP address instead of "NULL".
1825 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1826 allows the sender and the authenticated sender to be set when
1827 submitting a message from within Exim. Since child_open_exim() is
1828 documented for local_scan(), the new function should be too.
1830 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1831 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1832 results in an empty string is now treated as unset.
1834 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1836 PH/31 Added sender_verify_fail logging option.
1838 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1839 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1840 tidied the source and removed it altogether.
1842 PH/33 When a queue run was abandoned because the load average was too high, a
1843 log line was always written; now it is written only if the queue_run log
1844 selector is set. In addition, the log line for abandonment now contains
1845 information about the queue run such as the pid. This is always present
1846 in "start" and "stop" lines but was omitted from the "abandon" line.
1848 PH/34 Omit spaces between a header name and the colon in the error message that
1849 is given when verify = headers_syntax fails (if there are lots of them,
1850 the message gets confusing).
1852 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1853 names, as there are now some PTR records that contain slashes. This check
1854 is only to protect against broken name servers that fall over on strange
1855 characters, so the fact that it applies to all lookups doesn't matter.
1857 PH/36 Now that the new test suite is complete, we can remove some of the
1858 special code in Exim that was needed for the old test suite. For example,
1859 sorting DNS records because real resolvers return them in an arbitrary
1860 order. The new test suite's fake resolver always returns records in the
1863 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1864 bounces) except when queue_only is set, to avoid logging races between
1865 the different processes.
1867 PH/38 Panic-die if .include specifies a non-absolute path.
1869 PH/39 A tweak to the "H" retry rule from its user.
1871 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1872 a label. They prevented compilation on older perls.
1874 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1875 a warning to be raised on newish perls.
1877 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1878 on queue. Changes to match documented behaviour of showing count of
1879 messages matching specified criteria.
1881 PH/40 Changed the default ident timeout from 30s to 5s.
1883 PH/41 Added support for the use of login_cap features, on those BSD systems
1884 that have them, for controlling the resources used by pipe deliveries.
1886 PH/42 The content-scanning code uses fopen() to create files in which to put
1887 message data. Previously it was not paying any attention to the mode of
1888 the files. Exim runs with umask(0) because the rest of the code creates
1889 files with open(), and sets the required mode explicitly. Thus, these
1890 files were ending up world-writeable. This was not a big issue, because,
1891 being within the spool directory, they were not world-accessible. I have
1892 created a function called modefopen, which takes an additional mode
1893 argument. It sets umask(777), creates the file, chmods it to the required
1894 mode, then resets the umask. All the relevant calls to fopen() in the
1895 content scanning code have been changed to use this function.
1897 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1898 to 24 hours. This avoids potential overflow problems when processing G
1899 and H retry rules. I suspect nobody ever tinkers with this value.
1901 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1903 PH/45 When the plaintext authenticator is running as a client, the server's
1904 challenges are checked to ensure they are valid base64 strings. By
1905 default, the authentication attempt is cancelled if an invalid string is
1906 received. Setting client_ignore_invalid_base64 true ignores these errors.
1907 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1908 they are received. Thus, the responses can be made to depend on the
1909 challenges. If an invalid string is ignored, an empty string is placed in
1912 PH/46 Messages that are created by the autoreply transport now contains a
1913 References: header, in accordance with RFCs 2822 and 3834.
1915 PH/47 Added authenticated_sender_force to the smtp transport.
1917 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1919 PH/49 Installed latest patch from the Sieve maintainer.
1921 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1922 was also set, the appendfile transport was unnecessarily scanning a
1923 directory of message files (e.g. for maildir delivery) to find the count
1924 of files (along with the size), even though it did not need this
1925 information. It now does the scan only if it needs to find either the
1926 size of the count of files.
1928 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1930 PH/52 Two bugs concerned with error handling when the smtp transport is
1933 (i) Exim was not creating retry information for temporary errors given
1934 for individual recipients after the DATA command when the smtp transport
1935 was used in LMTP mode. This meant that they could be retried too
1936 frequently, and not timed out correctly.
1938 (ii) Exim was setting the flag that allows error details to be returned
1939 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1940 recipients that were returned after the DATA command.
1942 PH/53 This is related to PH/52, but is more general: for any failing address,
1943 when detailed error information was permitted to be returned to the
1944 sender, but the error was temporary, then after the final timeout, only
1945 "retry timeout exceeded" was returned. Now it returns the full error as
1946 well as "retry timeout exceeded".
1948 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1949 do this, and (what is worse) MTAs that accept it.
1951 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1952 will now be deprecated.
1954 PH/56 New os.c-cygwin from the Cygwin maintainer.
1956 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1957 formats (previously only available in exim formats via -bpr, -bpru,
1958 and -bpra. Now also available in native and exiqgrep formats)
1960 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1961 with very large, slow to parse queues
1963 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1965 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1967 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1968 responses to authentication challenges, though it was showing the
1969 challenges; (ii) I've removed the CR characters from the debug output for
1972 PH/58 Allow for the insertion of a newline as well as a space when a string
1973 is turned into more than one encoded-word during RFC 2047 encoding. The
1974 Sieve code now uses this.
1976 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1977 data_4xx, lost_connection, tls_required.
1979 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1980 message was being sent as an SMTP response.
1982 PH/61 Add -l and -k options to exicyclog.
1984 PH/62 When verifying, if an address was redirected to one new address, so that
1985 verification continued, and the new address failed or deferred after
1986 having set something in $address_data, the value of $address_data was not
1987 passed back to the ACL. This was different to the case when no
1988 redirection occurred. The value is now passed back in both cases.
1990 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1991 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1992 use login_cap.h, so on its own it isn't the distinguishing feature. The
1993 new name refers directly to the setclassresources() function.
1995 PH/65 Added configuration files for NetBSD3.
1997 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1999 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2000 is preferred over IPv4.
2002 PH/68 The bounce_return_message and bounce_return_body options were not being
2003 honoured for bounces generated during the reception of non-SMTP messages.
2004 In particular, this applied to messages rejected by the ACL. This bug has
2005 been fixed. However, if bounce_return_message is true and bounce_return_
2006 body is false, the headers that are returned for a non-SMTP message
2007 include only those that have been read before the error was detected.
2008 (In the case of an ACL rejection, they have all been read.)
2010 PH/69 The HTML version of the specification is now built in a directory called
2011 spec_html instead of spec.html, because the latter looks like a path with
2012 a MIME-type, and this confuses some software.
2014 PH/70 Catch two compiler warnings in sieve.c.
2016 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2017 function verify_get_ident() calls ip_connect() to connect a socket, but
2018 if the "connect()" function timed out, ip_connect() used to close the
2019 socket. However, verify_get_ident() also closes the socket later, and in
2020 between Exim writes to the log, which may get opened at this point. When
2021 the socket was closed in ip_connect(), the log could get the same file
2022 descriptor number as the socket. This naturally causes chaos. The fix is
2023 not to close the socket in ip_connect(); the socket should be closed by
2024 the function that creates it. There was only one place in the code where
2025 this was missing, in the iplookup router, which I don't think anybody now
2026 uses, but I've fixed it anyway.
2028 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2029 well as to direct DNS lookups. Otherwise the handling of names in host
2030 lists is inconsistent and therefore confusing.
2036 PH/01 Two changes to the default runtime configuration:
2038 (1) Move the checks for relay_from_hosts and authenticated clients from
2039 after to before the (commented out) DNS black list checks.
2041 (2) Add control=submission to the relay_from_hosts and authenticated
2042 clients checks, on the grounds that messages accepted by these
2043 statements are most likely to be submissions.
2045 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2047 (1) Generate an error if the third argument for the ${prvs expansion is
2050 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2053 (3) Reset the variables that are obtained from the first argument of
2054 ${prvscheck and used in the second argument before leaving the code,
2055 because their memory is reclaimed, so using them afterwards may do
2058 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2059 one (it's much easier than Tom thought :-).
2061 (5) Because of (4), we can now allow for the use of $prvscheck_result
2062 inside the third argument.
2064 PH/03 For some reason, the default setting of PATH when running a command from
2065 a pipe transport was just "/usr/bin". I have changed it to
2068 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2069 anything to be listed in the output from -bV.
2071 PH/05 When a filter generated an autoreply, the entire To: header line was
2072 quoted in the delivery log line, like this:
2074 => >A.N.Other <ano@some.domain> <original@ddress> ...
2076 This has been changed so that it extracts the operative address. There
2077 may be more than one such address. If so, they are comma-separated, like
2080 => >ano@some.domain,ona@other.domain <original@ddress> ...
2082 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2083 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2084 not being looked up in the rDNS to get a host name, Exim was showing the
2085 IP address twice in Received: lines, even though the IP addresses were
2086 identical. For example:
2088 Received: from [1.2.3.4] (helo=[1.2.3.4])
2090 However, if the real host name was known, it was omitting the HELO data
2091 if it matched the actual IP address. This has been tidied up so that it
2092 doesn't show the same IP address twice.
2094 PH/07 When both +timestamp and +memory debugging was on, the value given by
2095 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2096 called by the debug printing, thereby overwriting the timestamp buffer.
2097 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2099 PH/08 When the original message was included in an autoreply transport, it
2100 always said "this is a copy of the message, including all the headers",
2101 even if body_only or headers_only was set. It now gives an appropriate
2104 PH/09 Applied a patch from the Sieve maintainer which:
2106 o fixes some comments
2107 o adds the (disabled) notify extension core
2108 o adds some debug output for the result of if/elsif tests
2109 o points to the current vacation draft in the documentation
2110 and documents the missing references header update
2114 o fixes a bug in processing the envelope test (when testing
2115 multiple envelope elements, the last element determinted the
2118 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2119 Electronic Mail") by including:
2121 Auto-submitted: auto-generated
2123 in the messages that it generates (bounce messages and others, such as
2124 warnings). In the case of bounce messages for non-SMTP mesages, there was
2125 also a typo: it was using "Auto_submitted" (underscore instead of
2126 hyphen). Since every message generated by Exim is necessarily in response
2127 to another message, thes have all been changed to:
2129 Auto-Submitted: auto-replied
2131 in accordance with these statements in the RFC:
2133 The auto-replied keyword:
2135 - SHOULD be used on messages sent in direct response to another
2136 message by an automatic process,
2138 - MUST NOT be used on manually-generated messages,
2140 - MAY be used on Delivery Status Notifications (DSNs) and Message
2141 Disposition Notifications (MDNs),
2143 - MUST NOT be used on messages generated by automatic or periodic
2144 processes, except for messages which are automatic responses to
2147 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2148 to the default Received: header definition.
2150 PH/12 Added log selector acl_warn_skipped (default on).
2152 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2153 variables because (a) they are in the wrong storage pool and (b) even if
2154 they were copied, it wouldn't work properly because of the caching.
2156 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2157 checking when decoding. Apparently there are clients that generate
2158 overlong encoded strings. Why am I not surprised?
2160 PH/15 If the first argument of "${if match_address" was not empty, but did not
2161 contain an "@" character, Exim crashed. Now it writes a panic log message
2162 and treats the condition as false.
2164 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2166 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2167 extension "envelope-auth". The code is finished and in agreement with
2168 other implementations, but there is no documentation so far and in fact,
2169 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2170 not changing the active code.
2172 Print executed "if" and "elsif" statements when debugging is used. This
2173 helps a great deal to understand what a filter does.
2175 Document more things not specified clearly in RFC3028. I had all this
2176 sorted out, when out of a sudden new issues came to my mind. Oops."
2178 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2181 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2182 canonical form (as documented). However, after a host name lookup from
2183 the IP address, check_host() was doing a simple string comparison with
2184 addresses acquired from the DNS when checking that the found name did
2185 have the original IP as one of its addresses. Since any found IPv6
2186 addresses are likely to be in abbreviated form, the comparison could
2187 fail. Luckily, there already exists a function for doing the comparison
2188 by converting both addresses to binary, so now that is used instead of
2189 the text comparison.
2191 PH/20 There was another similar case to PH/19, when a complete host name was
2192 given in a host list; looking up its IP address could give an abbreviated
2193 form, whereas the current host's name might or might not be abbreviated.
2194 The same fix has been applied.
2200 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2201 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2204 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2205 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2207 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2209 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2210 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2211 Also, for "Auto-Submitted", treat anything other than "no" as
2212 non-personal, in accordance with RFC 3834. (Previously it treated
2213 anything starting "auto-" as non-personal.)
2215 TF/01 The control=submission/name=... option had a problem with syntax
2216 errors if the name included a slash character. The /name= option
2217 now slurps the rest of the string, so it can include any characters
2218 but it must come last in the list of options (after /sender_retain
2221 PH/05 Some modifications to the interface to the fake nameserver for the new
2229 TK/01 Added the "success_on_redirect" address verification option. See
2230 NewStuff for rationale and an example.
2232 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2234 PH/02 Patch to exigrep to allow it to work on syslog lines.
2236 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2237 fread() to skip over the body file's header line, because in Cygwin the
2238 header line is locked and is inaccessible.
2240 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2241 co-exist for some time) to make it clear that it is the Exim ID that is
2242 referenced, not the Message-ID: header line.
2244 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2245 string_format() function, because snprintf() does not exist on all
2248 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2249 ${stat: expansion item.
2251 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2252 protocol synchronization error", to keep the pedants happy.
2254 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2255 well as for IRIX systems, when gcc is being used. See the host.c source
2258 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2260 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2263 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2264 transport and to the smtp transport in LMTP mode.
2266 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2268 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2269 run a filter in a subprocess. This could lead to confusion in subsequent
2270 lookups in the parent process. There should also be a search_tidyup() at
2271 the end of the subprocess.
2273 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2274 only if the host matched helo_try_verify_hosts, which caused the
2275 verification to occur when the EHLO/HELO command was issued. The ACL just
2276 tested the remembered result. Now, if a previous verification attempt has
2277 not happened, "verify = helo" does it there and then.
2279 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2281 TK/03 Fix log output including CR from clamd.
2283 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2284 exist provoked a memory error which could cause a segfault.
2286 PH/15 Installed PCRE 6.2
2288 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2290 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2291 of the problem. Specifically, suggested +O2 rather than +O1 for the
2294 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2296 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2297 record, the port was not correctly logged when the outgoing_port log
2298 selector was set (it logged the transort's default port).
2300 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2301 fallback_hosts, and "hosts" in the smtp transport.
2303 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2304 host errors such as "Connection refused".
2306 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2307 authentication with radiusclient 0.4.9:
2309 - Error returned from rc_read_config was caught wrongly
2310 - Username/password not passed on to radius server due to wrong length.
2312 The presumption is that some radiusclient API changes for 4.51/PH/17
2313 were not taken care of correctly. The code is still untested by me (my
2314 Linux distribution still has 0.3.2 of radiusclient), but it was
2315 contributed by a Radius user.
2317 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2318 expanding the "port" option of the smtp transport.
2320 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2321 while reading a MIME header. Thanks to Tom Hughes for a patch.
2323 PH/24 Include config.h inside local_scan.h so that configuration settings are
2326 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2327 that in an ACL for RCPT (for example), you can examine exactly what was
2330 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2331 commands, but it was not correctly comparing the address with the actual
2332 client host address. Thus, it would show the EHLO address in Received:
2333 header lines when this was not necessary.
2335 PH/27 Added the % operator to ${eval:}.
2337 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2338 it should be ignoring failures (because with -C, for example, it has lost
2339 privilege). It wasn't ignoring creation failures other than "already
2342 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2345 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2346 because an input line was too long, either on its own, or by virtue of
2347 too many continuations, the temporary file was not being removed, and the
2348 return code was incorrect.
2350 PH/31 Missing "BOOL" in function definition in filtertest.c.
2352 PH/32 Applied Sieve patches from the maintainer.
2354 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2356 PH/33 Added "verify = not_blind".
2358 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2359 Local/Makefile (with some defaults set). These are used in built scripts
2360 such as exicyclog, but they have never been used in the exim_install
2361 script (though there are many overriding facilities there). I have
2362 arranged that the exim_install script now takes note of these two
2365 PH/35 Installed configuration files for Dragonfly.
2367 PH/36 When a locally submitted message by a trusted user did not contain a
2368 From: header, and the sender address was obtained from -f or from an SMTP
2369 MAIL command, and the trusted user did not use -F to supply a sender
2370 name, $originator_name was incorrectly used when constructing a From:
2371 header. Furthermore, $originator_name was used for submission mode
2372 messages from external hosts without From: headers in a similar way,
2373 which is clearly wrong.
2375 PH/37 Added control=suppress_local_fixups.
2377 PH/38 When log_selector = +received_sender was set, and the addition of the
2378 sender made the log line's construction buffer exactly full, or one byte
2379 less than full, an overflow happened when the terminating "\n" was
2382 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2383 when the result of a list match is failure because a DNS lookup failed.
2385 PH/40 RM_COMMAND is now used in the building process.
2387 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2388 the "build-* directories that it finds.
2390 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2391 address in a domain literal was a prefix of an interface address.
2393 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2394 when verifying a sender address, unless rewrite_headers is false.
2396 PH/44 Wrote a long comment about why errors_to addresses are verified as
2397 recipients, not senders.
2399 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2400 the ratelimit ACL was added.
2402 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2404 PH/46 Added extra information about PostgreSQL errors to the error string.
2406 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2407 avoiding the need to install special zones in a real server. This is
2408 backwards compatible; if it can't find the fake resolver, it drops back.
2409 Thus, both old and new test suites can be run.
2411 TF/02 Added util/ratelimit.pl
2413 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2414 clock is set back in time.
2416 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2417 Candler <B.Candler@pobox.com>.
2419 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2420 OK for addresses that are the result of redirections.
2422 PH/48 A number of further additions for the benefit of the new test suite,
2423 including a fake gethostbyname() that interfaces to the fake DNS resolver
2426 TF/06 The fix for widen_domains has also been applied to qualify_single and
2427 search_parents which are the other dnslookup options that can cause
2430 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2433 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2435 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2436 service" if it encountered an MX record with an empty target hostname.
2437 The message is now "an MX or SRV record indicated no SMTP service".
2439 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2440 if the DNS of the sending site is misconfigured. This is quite a
2441 common situation. This change restores the behaviour of treating a
2442 helo verification defer as a failure.
2444 PH/51 If self=fail was set on a router, the bounce message did not include the
2445 actual error message.
2451 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2453 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2454 command itself ended in error, the underlying message about the transport
2455 filter timeout was being overwritten with the pipe command error. Now the
2456 underlying error message should be appended to the second error message.
2458 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2460 PH/02 Reduce the amount of output that "make" produces by default. Full output
2461 can still be requested.
2463 PH/03 The warning log line about a condition test deferring for a "warn" verb
2464 was being output only once per connection, rather than after each
2465 occurrence (because it was using the same function as for successful
2466 "warn" verbs). This seems wrong, so I have changed it.
2468 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2469 it should not have, which might have caused a crash in the right
2470 circumstances, but probably never did.
2472 PH/04 Installed a modified version of Tony Finch's patch to make submission
2473 mode fix the return path as well as the Sender: header line, and to
2474 add a /name= option so that you can make the user's friendly name appear
2477 TF/03 Added the control = fakedefer ACL modifier.
2479 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2480 Mark Lowes for thorough testing.
2482 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2484 TK/03 Merged latest SRS patch from Miles Wilton.
2486 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2487 with the definition in sysexits.h (which is #included earlier).
2488 Fortunately, Exim does not actually use EX_OK. The code used to try to
2489 preserve the sysexits.h value, by assumimg that macro definitions were
2490 scanned for macro replacements. I have been disabused of this notion,
2491 so now the code just undefines EX_OK before #including unistd.h.
2493 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2494 in the smtp transport. When a block could not be written in a single
2495 write() function, the timeout was being re-applied to each part-write.
2496 This seems wrong - if the receiver was accepting one byte at a time it
2497 would take for ever. The timeout is now adjusted when this happens. It
2498 doesn't have to be particularly precise.
2500 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2501 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2503 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2504 without a domain if the check to <postmaster@domain> fails.
2506 SC/01 Eximstats: added -xls and the ability to specify output files
2507 (patch written by Frank Heydlauf).
2509 SC/02 Eximstats: use FileHandles for outputing results.
2511 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2513 SC/04 Eximstats: fixed display of large numbers with -nvr option
2515 SC/05 Eximstats: fixed merging of reports with empty tables.
2517 SC/06 Eximstats: added the -include_original_destination flag
2519 SC/07 Eximstats: removed tabs and trailing whitespace.
2521 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2523 TK/06 MBOX spool code: Add real "From " MBOX separator line
2524 so the .eml file is really in mbox format (even though
2525 most programs do not really care). Patch from Alex Miller.
2527 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2528 The latter is generated from $received_to and is only set if the
2529 message has one envelope recipient. SA can use these headers,
2530 obviously out-of-the-box. Patch from Alex Miller.
2532 PH/08 The ${def test on a variable was returning false if the variable's
2533 value was "0", contrary to what the specification has always said!
2534 The result should be true unless the variable is empty.
2536 PH/09 The syntax error of a character other than { following "${if
2537 def:variable_name" (after optional whitespace) was not being diagnosed.
2538 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2539 accidental colon was present, for example, could give incorrect results.
2541 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2542 result is used (not including appendfile, where other changes are about
2545 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2546 This involved changing a lot of size variables from int to off_t. It
2547 should work with maildirs and everything.
2549 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2550 spamd dying while we are connected to it.
2552 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2555 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2556 function for BDB 4.3.
2558 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2560 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2561 the sources for PCRE have been re-arranged and more files are now
2564 PH/15 The code I had for printing potentially long long variables in PH/11
2565 above was not the best (it lost precision). The length of off_t variables
2566 is now inspected at build time, and an appropriate printing format (%ld
2567 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2568 to be "long long int" or "long int". This is needed for the internal
2569 formatting function string_vformat().
2571 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2572 the configuration file to be ":syslog", then the script "guesses" where
2573 the logs files are, rather than using the compiled in default. In our
2574 case the guess is not the same as the compiled default, so the script
2575 suddenly stopped working when I started to use syslog. The patch checks
2576 to see if log_file_path is "". If so, it attempts to read it from exim
2577 with no configuration file to get the compiled in version, before it
2578 falls back to the previous guessing code."
2580 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2581 implementing BATV in an Exim configuration. See NewStuff for the gory
2584 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2585 Makefile that are specific to HP-UX.
2587 PH/18 If the "use_postmaster" option was set for a recipient callout together
2588 with the "random" option, the postmaster address was used as the MAIL
2589 FROM address for the random test, but not for the subsequent recipient
2590 test. It is now used for both.
2592 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2593 patch removes a few documentation additions to RFC 3028, because the
2594 latest draft now contains them. It adds the new en;ascii-case comparator
2595 and a new error check for 8bit text in MIME parts. Comparator and
2596 require names are now matched exactly. I enabled the subaddress
2597 extension, but it is not well tested yet (read: it works for me)."
2599 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2600 rework some of the code of TK/09 above to avoid the hardwired use of
2601 "%lld" and "long long". Replaced the call to snprintf() with a call to
2604 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2605 records point to non-existent hosts", "retry timeout exceeded", and
2606 "retry time not reached for any host after a long failure period".
2608 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2609 experimental DomainKeys support:
2611 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2612 (2) On an error such as an illegally used "control", the wrong name for
2613 the control was given.
2615 These problems did NOT occur unless DomainKeys support was compiled.
2617 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2619 PH/24 Added ${if match_ip condition.
2621 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2622 databases so that it will be absolutely obvious if a crash occurs in the
2623 DB library. This is a regular occurrence (often caused by mis-matched
2626 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2627 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2628 picked up on a user's system that detects such things. There doesn't seem
2629 to be a gcc warning option for this - only an attribute that has to be
2630 put on the function's prototype. It seems that in Fedora Core 4 they have
2631 set this on a number of new functions. No doubt there will be more in due
2634 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2635 specify a transport. However, if an address that was verified by such a
2636 router was the subject of a callout, Exim crashed because it tried to
2637 read the rcpt_include_affixes from the non-existent transport. Now it
2638 just assumes that the setting of that option is false. This bug was
2639 introduced by 4.51/PH/31.
2641 PH/28 Changed -d+all to exclude +memory, because that information is very
2642 rarely of interest, but it makes the output a lot bigger. People tend to
2643 do -d+all out of habit.
2645 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2646 code in os-type was giving problems when libc.so lives in lib64, like on
2649 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2650 aren't the modern standard, and it seems that some systems' include files
2651 don't always have them. Exim was already checking for some of the newer
2652 ones like T_AAAA, and defining it itself. I've added checks for all the
2653 record types that Exim uses.
2655 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2656 not automatically generating a new one, as it is supposed to. This
2657 prevented TLS from working. If the file did exist, but contained invalid
2658 data, a new version was generated, as expected. It was only the case of a
2659 non-existent file that was broken.
2661 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2662 with a change in libdomainkeys > 0.64.
2664 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2665 from DNS. If the selector record carries the flag, it now has
2666 precedence over the domain-wide flag.
2668 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2670 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2671 the use of a port name that isn't defined in /etc/services) occurred, the
2672 message was deferred as in a normal delivery, and thus remained on the
2673 spool, instead of being failed because of the mua_wrapper setting. This
2674 is now fixed, and I tidied up some of the mua_wrapper messages at the
2677 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2678 the messages in a hash of arrays rather than using individual hashes.
2679 This is a bit cleaner and results in dramatic memory savings, albeit
2680 at a slight CPU cost.
2682 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2683 as requested by Marc Sherman.
2685 SC/10 Eximstats: added histograms for user specified patterns as requested
2688 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2690 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2691 fopen() in the content-scanning modules that did not already have it.
2697 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2698 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2700 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2702 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2704 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2705 to test Sieve filters that use "vacation".
2707 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2708 that changes the way the GnuTLS parameters are stored in the cache file.
2709 The new format can be generated externally. For backward compatibility,
2710 if the data in the cache doesn't make sense, Exim assumes it has read an
2711 old-format file, and it generates new data and writes a new file. This
2712 means that you can't go back to an older release without removing the
2715 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2716 work if there are any delivery delays because "one_time" forces the
2717 parent to be marked "delivered", so its unseen clone is never tried
2718 again. For this reason, Exim now forbids the simultaneous setting of
2721 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2722 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2723 entry if you want to know the details. The fix, however, neglected to
2724 consider the case when local delivery batching is involved. The test for
2725 "previously delivered" was not happening when checking to see if an
2726 address could be batched with a previous (undelivered) one; under
2727 certain circumstances this could lead to multiple deliveries to the same
2730 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2731 in its include files, and this causes problems building Exim.
2733 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2734 header_syntax) but Exim was just ignoring anything given after a slash.
2735 In particular, this caused confusion with an attempt to use "verify =
2736 reverse_host_lookup/defer_ok". An error is now given when options are
2737 supplied for verify items that do not have them. (Maybe reverse_host_
2738 lookup should have a defer_ok option, but that's a different point.)
2740 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2741 defined by RFC 821) to 2048, because there were problems with some AUTH
2742 commands, and RFC 1869 says the size should be increased for extended
2743 SMTP commands that take arguments.
2745 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2748 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2749 "unknown" error; now it says that the functionality isn't in the binary.
2751 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2752 an address' error message when a string expansion fails (syntax or
2753 whatever). Otherwise the password may appear in the log. Following change
2754 PH/42 below, there is no longer a chance of it appearing in a bounce
2757 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2759 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2760 of its addresses was ever tried. (Bugzilla bug #2.)
2762 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2763 the result incorrectly in the debug output. (It correctly added a newline
2764 to what was transported.)
2766 TF/01 Added $received_time.
2768 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2769 commented out examples of how to interface to a virus scanner and to
2770 SpamAssassin. Also added commented examples of av_scanner and
2771 spamd_address settings.
2773 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2774 and controls are allowed in which ACLs. There were a couple of minor
2775 errors. Some of the entries in the conditions table (which is a table of
2776 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2777 negation of where the condition IS allowed.
2779 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2781 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2782 header file does not have a version number, so I've had to invent a new
2783 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2784 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2785 radiusclient), but it was contributed by a Radius user.
2787 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2788 files or directories, for OpenSSL.
2790 PH/19 When an Exim process that is running as root has to create an Exim log
2791 file, it does so in a subprocess that runs as exim:exim so as to get the
2792 ownership right at creation (otherwise, other Exim processes might see
2793 the file with the wrong ownership). There was no test for failure of this
2794 fork() call, which would lead to the process getting stuck as it waited
2795 for a non-existent subprocess. Forks do occasionally fail when resources
2796 run out. I reviewed all the other calls to fork(); they all seem to check
2799 PH/20 When checking for unexpected SMTP input at connect time (before writing
2800 the banner), Exim was not dealing correctly with a non-positive return
2801 from the read() function. If the client had disconnected by this time,
2802 the result was a log entry for a synchronization error with an empty
2803 string after "input=" when read() returned zero. If read() returned -1
2804 (an event I could not check), uninitialized data bytes were printed.
2805 There were reports of junk text (parts of files, etc) appearing after
2808 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2810 PH/22 Added support for macro redefinition, and (re)definition in between
2811 driver and ACL definitions.
2813 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2814 forgetting to use the resulting value; it was using the unexpanded value.
2816 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2817 hadn't been configured. The fix is from Juergen Kreileder, who
2818 understands it better than I do:
2820 "Here's what I see happening with three configured cyrus_sasl
2821 authenticators configured (plain, login, cram-md5):
2823 On startup auth_cyrus_sasl_init() gets called for each of these.
2824 This means three calls to sasl_listmech() without a specified mech_list.
2825 => SASL tests which mechs of all available mechs actually work
2826 => three warnings about OTP not working
2827 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2829 With the patch, sasl_listmech() also gets called three times. But now
2830 SASL's mech_list option is set to the server_mech specified in the the
2831 authenticator. Or in other words, the answer from sasl_listmech()
2832 gets limited to just the mech you're testing for (which is different
2834 => the return list contains just 'plain' or 'login', 'cram-md5' or
2835 nothing depending on the value of ob->server_mech.
2837 I've just tested the patch: Authentication still works fine,
2838 unavailable mechs specified in the exim configuration are still
2839 caught, and the auth.log warnings about OTP are gone."
2841 PH/25 When debugging is enabled, the contents of the command line are added
2842 to the debugging output, even when log_selector=+arguments is not
2845 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2846 answer is "GNU", and only if the return is "GNU/something" is the answer
2849 PH/27 $acl_verify_message is now set immediately after the failure of a
2850 verification in an ACL, and so is available in subsequent modifiers. In
2851 particular, the message can be preserved by coding like this:
2853 warn !verify = sender
2854 set acl_m0 = $acl_verify_message
2856 Previously, $acl_verify_message was set only while expanding "message"
2857 and "log_message" when a very denied access.
2859 PH/28 Modified OS/os.c-Linux with
2861 -#ifndef OS_LOAD_AVERAGE
2862 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2864 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2865 nomenclature these days.)
2867 PH/29 Installed patch from the Sieve maintainer that adds the options
2868 sieve_useraddress and sieve_subaddress to the redirect router.
2870 PH/30 In these circumstances:
2871 . Two addresses routed to the same list of hosts;
2872 . First host does not offer TLS;
2873 . First host accepts first address;
2874 . First host gives temporary error to second address;
2875 . Second host offers TLS and a TLS session is established;
2876 . Second host accepts second address.
2877 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2878 and peerdn, if requested) that were in fact used only for the second
2881 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2882 attention to any local part prefix or suffix that was matched by the
2883 router that accepted the address. It now behaves in the same way as it
2884 does for delivery: the affixes are removed from the local part unless
2885 rcpt_include_affixes is set on the transport.
2887 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2888 timeout during the DATA phase of an incoming message.
2890 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2891 applied a patch sanctioned by the Sieve maintainer.
2893 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2894 the uid or gid is negative. A case of a negative gid caused this to be
2895 noticed. The fix allows for either to be negative.
2897 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2898 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2901 PH/36 The change PH/12 above was broken. Fixed it.
2903 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2904 the grounds that routing the same address twice would always produce the
2905 same answer. This might have been true once, but it is certainly no
2906 longer true now. Routing a child address may depend on the previous
2907 routing that produced that child. Some complicated redirection strategies
2908 went wrong when messages had multiple recipients, and made Exim's
2909 behaviour dependent on the order in which the addresses were given.
2911 I have moved the duplicate checking until after the routing is complete.
2912 Exim scans the addresses that are assigned to local and remote
2913 transports, and removes any duplicates. This means that more work will be
2914 done, as duplicates will always all be routed, but duplicates are
2915 presumably rare, so I don't expect this is of any significance.
2917 For deliveries to pipes, files, and autoreplies, the duplicate checking
2918 still happens during the routing process, since they are not going to be
2921 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2922 It corrects a timeout issue with spamd. This is Ian's comment: "The
2923 background is that sometimes spamd either never reads data from a
2924 connection it has accepted, or it never writes response data. The exiscan
2925 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2926 blindly assumes that writes won't block so it may never time out."
2928 PH/39 Allow G after quota size as well as K and M.
2930 PH/40 The value set for $authenticated_id in an authenticator may not contain
2931 binary zeroes or newlines because the value is written to log lines and
2932 to spool files. There was no check on this. Now the value is run through
2933 the string_printing() function so that such characters are converted to
2934 printable escape sequences.
2936 PH/41 $message_linecount is a new variable that contains the total number of
2937 lines in the message. Compare $body_linecount, which is the count for the
2940 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2941 bounce and delay warning messages, except in certain special cases, which
2944 (a) An SMTP error message from a remote host;
2945 (b) A message specified in a :fail: redirection;
2946 (c) A message specified in a "fail" command in a system filter;
2947 (d) A message specified in a FAIL return from the queryprogram router;
2948 (e) A message specified by the cannot_route_message router option.
2950 In these cases only, Exim does include the error details in bounce and
2951 warning messages. There are also a few cases where bland messages such
2952 as "unrouteable address" or "local delivery error" are given.
2954 PH/43 $value is now also set for the "else" part of a ${run expansion.
2956 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2957 being worked on, but at least Exim now implements the latest version to
2960 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2961 process to complete was treated as a delivery failure, a timeout while
2962 writing the message to the pipe was logged, but erroneously treated as a
2963 successful delivery. Such timeouts include transport filter timeouts. For
2964 consistency with the overall process timeout, these timeouts are now
2965 treated as errors, giving rise to delivery failures by default. However,
2966 there is now a new Boolean option for the pipe transport called
2967 timeout_defer, which, if set TRUE, converts the failures into defers for
2968 both kinds of timeout. A transport filter timeout is now identified in
2971 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2972 systems where "make" and "gmake" are different, calling "gmake" at top
2973 level broke things. I've arranged for the value of $(MAKE) to be passed
2974 from the Makefile to this script so that it can call the same version of
2978 A note about Exim versions 4.44 and 4.50
2979 ----------------------------------------
2981 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2982 changes of various kinds. As a consequence, a big documentation update was
2983 needed. This delayed the release for rather longer than seemed good, especially
2984 in the light of a couple of (minor) security issues. Therefore, the changes
2985 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2986 release. So 4.44 and 4.50 are in effect two different branches that both start
2989 I have left the 4.50 change log unchanged; it contains all the changes since
2990 4.43. The change log for 4.44 is below; many of its items are identical to
2991 those for 4.50. This seems to be the most sensible way to preserve the
2992 historical information.
2998 1. Minor wording change to the doc/README.SIEVE file.
3000 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3001 computation of the current number of files was incorrect.
3003 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3004 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3007 4. Give more explanation in the error message when the command for a transport
3008 filter fails to execute.
3010 5. There are several places where Exim runs a non-Exim command in a
3011 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3012 was being done only for the command run by the queryprogram router. It is
3013 now done for all such subprocesses. The other cases are: ${run, transport
3014 filters, and the commands run by the lmtp and pipe transports.
3016 6. Added CONFIGURE_GROUP build-time option.
3018 7. Some older OS have a limit of 256 on the maximum number of file
3019 descriptors. Exim was using setrlimit() to set 1000 as a large value
3020 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3021 systems. I've change it so that if it can't get 1000, it tries for 256.
3023 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3024 was an oversight, and furthermore, ever since the addition of extra
3025 controls (e.g. 4.43/32), the checks on when to allow different forms of
3026 "control" were broken. There should now be diagnostics for all cases when a
3027 control that does not make sense is encountered.
3029 9. Added the /retain_sender option to "control=submission".
3031 10. $recipients is now available in the predata ACL (oversight).
3033 11. Tidy the search cache before the fork to do a delivery from a message
3034 received from the command line. Otherwise the child will trigger a lookup
3035 failure and thereby defer the delivery if it tries to use (for example) a
3036 cached ldap connection that the parent has called unbind on.
3038 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3039 of $address_data from the recipient verification was clobbered by the
3040 sender verification.
3042 13. The value of address_data from a sender verification is now available in
3043 $sender_address_data in subsequent conditions in the ACL statement.
3045 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3047 15. Added a new option "connect=<time>" to callout options, to set a different
3050 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3051 was its contents. (It was OK if the option was not defined at all.)
3053 17. A "Completed" log line is now written for messages that are removed from
3054 the spool by the -Mrm option.
3056 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3057 information about exactly what failed.
3059 19. Added -dd to debug only the daemon process.
3061 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3062 handles timeouts, both on the server side and network timeouts. Renamed the
3063 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3065 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3066 It is now set to "smtps".
3068 22. $host_address is now set to the target address during the checking of
3069 ignore_target_hosts.
3071 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3072 being passed; this would have caused $sender_host_name to have been used if
3073 matching the list had actually called for a host name (not very likely,
3074 since this list is usually IP addresses). A host name is now passed as
3077 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3078 code that specifies a non-restarting handler (typically sigaction() in
3079 modern systems) in an attempt to fix a rare and obscure crash bug.
3081 25. Narrowed the window for a race in the daemon that could cause it to ignore
3082 SIGCHLD signals. This is not a major problem, because they are used only to
3083 wake it up if nothing else does.
3085 26. A malformed maildirsize file could cause Exim to calculate negative values
3086 for the mailbox size or file count. Odd effects could occur as a result.
3087 The maildirsize information is now recalculated if the size or filecount
3090 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3091 support for a long time. Removed HAVE_SYS_VFS_H.
3093 28. Installed the latest version of exipick from John Jetmore.
3095 29. In an address list, if the pattern was not a regular expression, an empty
3096 subject address (from a bounce message) matched only if the pattern was an
3097 empty string. Non-empty patterns were not even tested. This was the wrong
3098 because it is perfectly reasonable to use an empty address as part of a
3099 database query. An empty address is now tested by patterns that are
3100 lookups. However, all the other forms of pattern expect the subject to
3101 contain a local part and a domain, and therefore, for them, an empty
3102 address still always fails if the pattern is not itself empty.
3104 30. Exim went into a mad DNS loop when attempting to do a callout where the
3105 host was specified on an smtp transport, and looking it up yielded more
3106 than one IP address.
3108 31. Re-factored the code for checking spool and log partition space into a
3109 function that finds that data and another that does the check. The former
3110 is then used to implement four new variables: $spool_space, $log_space,
3111 $spool_inodes, and $log_inodes.
3113 32. The RFC2047 encoding function was originally intended for short strings
3114 such as real names; it was not keeping to the 75-character limit for
3115 encoded words that the RFC imposes. It now respects the limit, and
3116 generates multiple encoded words if necessary. To be on the safe side, I
3117 have increased the buffer size for the ${rfc2047: expansion operator from
3120 33. It is now permitted to omit both strings after an "if" condition; if the
3121 condition is true, the result is "true". As before, when the second string
3122 is omitted, a false condition yields an empty string. This makes it less
3123 cumbersome to write custom ACL and router conditions.
3125 34. Failure to deliver a bounce message always caused it to be frozen, even if
3126 there was an errors_to setting on the router. The errors_to setting is now
3129 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3130 canonical form (fully expanded) before being placed in
3131 $sender_host_address.
3133 36. The table in the code that translates DNS record types into text (T_A to
3134 "A" for instance) was missing entries for NS and CNAME. It is just possible
3135 that this could have caused confusion if both these types were looked up
3136 for the same domain, because the text type is used as part of Exim's
3137 per-process caching. But the chance of anyone hitting this buglet seems
3140 37. The dnsdb lookup has been extended in a number of ways.
3142 (1) There is a new type, "zns", which walks up the domain tree until it
3143 finds some nameserver records. It should be used with care.
3145 (2) There is a new type, "mxh", which is like "mx" except that it returns
3146 just the host names, not the priorities.
3148 (3) It is now possible to give a list of domains (or IP addresses) to be
3149 looked up. The behaviour when one of the lookups defers can be
3150 controlled by a keyword.
3152 (4) It is now possible to specify the separator character for use when
3153 multiple records are returned.
3155 38. The dnslists ACL condition has been extended: it is now possible to supply
3156 a list of IP addresses and/or domains to be looked up in a particular DNS
3159 39. Added log_selector=+queue_time_overall.
3161 40. When running the queue in the test harness, wait just a tad after forking a
3162 delivery process, to get repeatability of debugging output.
3164 41. Include certificate and key file names in error message when GnuTLS fails
3165 to set them up, because the GnuTLS error message doesn't include the name
3166 of the failing file when there is a problem reading it.
3168 42. Allow both -bf and -bF in the same test run.
3170 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3172 44. The "Exiscan patch" is now merged into the mainline Exim source.
3174 45. Sometimes the final signoff response after QUIT could fail to get
3175 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3176 < 0 before doing a fflush(). This bug looks as though it goes back to the
3177 introduction of TLS in release 3.20, but "sometimes" must have been rare
3178 because the tests only now provoked it.
3180 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3181 (this can affect the format of dates).
3183 47. exim_tidydb, when checking for the continued existence of a message for
3184 which it has found a message-specific retry record, was not finding
3185 messages that were in split spool directories. Consequently, it was
3186 deleting retry records that should have stayed in existence.
3188 48. Steve fixed some bugs in eximstats.
3190 49. The SPA authentication driver was not abandoning authentication and moving
3191 on to the next authenticator when an expansion was forced to fail,
3192 contradicting the general specification for all authenticators. Instead it
3193 was generating a temporary error. It now behaves as specified.
3195 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3196 (the order specifies the preference for clients). The order is now AES256,
3197 AES128, 3DES, ARCFOUR128.
3199 51. Small patch to Sieve code - explicitly set From: when generating an
3202 52. Exim crashed if a remote delivery caused a very long error message to be
3203 recorded - for instance if somebody sent an entire SpamAssassin report back
3204 as a large number of 550 error lines. This bug was coincidentally fixed by
3205 increasing the size of one of Exim's internal buffers (big_buffer) that
3206 happened as part of the Exiscan merge. However, to be on the safe side, I
3207 have made the code more robust (and fixed the comments that describe what
3210 53. Now that there can be additional text after "Completed" in log lines (if
3211 the queue_time_overall log selector is set), a one-byte patch to exigrep
3212 was needed to allow it to recognize "Completed" as not the last thing in
3215 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3216 patch that reportedly fixes this has been added. I am not expert enough to
3217 create a test for it. This is what the patch creator wrote:
3219 "I found a little strange behaviour of ldap code when working with
3220 Windows 2003 AD Domain, where users was placed in more than one
3221 Organization Units. When I tried to give exim partial DN, the exit code
3222 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3223 But simultaneously result of request was absolutely normal ldap result,
3224 so I produce this patch..."
3226 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3227 REFERENCE, so I have modified the code to exclude the patch when that macro
3230 55. Some experimental protocols are using DNS PTR records for new purposes. The
3231 keys for these records are domain names, not reversed IP addresses. The
3232 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3233 leaves it alone. Component reversal etc. now happens only for IP addresses.
3236 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3238 57. Double the size of the debug message buffer (to 2048) so that more of very
3239 long debug lines gets shown.
3241 58. The exicyclog utility now does better if the number of log files to keep
3242 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3244 59. Two changes related to the smtp_active_hostname option:
3246 (1) $smtp_active_hostname is now available as a variable.
3247 (2) The default for smtp_banner uses $smtp_active_hostname instead
3248 of $primary_hostname.
3250 60. The host_aton() function is supposed to be passed a string that is known
3251 to be a valid IP address. However, in the case of IPv6 addresses, it was
3252 not checking this. This is a hostage to fortune. Exim now panics and dies
3253 if the condition is not met. A case was found where this could be provoked
3254 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3255 components; fortuitously, this particular loophole had already been fixed
3256 by change 4.50/55 above.
3258 If there are any other similar loopholes, the new check in host_aton()
3259 itself should stop them being exploited. The report I received stated that
3260 data on the command line could provoke the exploit when Exim was running as
3261 exim, but did not say which command line option was involved. All I could
3262 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3263 running as the user.
3266 61. There was a buffer overflow vulnerability in the SPA authentication code
3267 (which came originally from the Samba project). I have added a test to the
3268 spa_base64_to_bits() function which I hope fixes it.
3271 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3272 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3274 63. The daemon start-up calls getloadavg() while still root for those OS that
3275 need the first call to be done as root, but it missed one case: when
3276 deliver_queue_load_max is set with deliver_drop_privilege. This is
3277 necessary for the benefit of the queue runner, because there is no re-exec
3278 when deliver_drop_privilege is set.
3280 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3281 This has been fixed.
3283 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3284 was in use, was not putting the data itself into the right store pool;
3285 consequently, it could be overwritten for a subsequent message in the same
3286 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3289 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3291 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3292 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3293 the function were treating the return as a boolean value, which happened to
3294 work because 0=false and not-0=true, but is not correct code.
3296 68. The host_aton() function was not handling scoped IPv6 addresses (those
3297 with, for example, "%eth0" on the end) correctly.
3299 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3300 negated items (that is, ~something) in unsigned ints. Some compilers
3301 apparently mutter when there is no cast.
3303 70. If an address verification called from an ACL failed, and did not produce a
3304 user-specific message (i.e. there was only a "system" message), nothing was
3305 put in $acl_verify_message. In this situation, it now puts the system
3308 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3309 session; change 4.31/43 added the unwanted input to the log line - except
3310 that it did not do this in the start of session case. It now does.
3312 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3313 This is wrong and can cause the other end to generate a synchronization
3314 error if it is another Exim or anything else that does the synchronization
3315 check. A QUIT command is no longer sent after a timeout.
3317 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3318 during host lookups.
3320 74. The defer_ok option of callout verification was not working if it was used
3321 when verifying addresses in header lines, that is, for this case:
3323 verify = header_sender/callout=defer_ok
3325 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3326 those file descriptors could be used for SMTP connections. If anything
3327 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3328 could be sent to the SMTP client, causing chaos. The daemon now opens
3329 stdin, stdout, and stderr to /dev/null when it puts itself into the
3332 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3333 log by default. The user can override this with suitable Perl magic.
3335 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3336 the log message when discard triggers, was not working for the DATA ACL or
3337 for the non-SMTP ACL.
3339 78. Error message wording change in sieve.c.
3341 79. If smtp_accept_max_per_host was set, the number of connections could be
3342 restricted to fewer than expected, because the daemon was trying to set up
3343 a new connection before checking whether the processes handling previous
3344 connections had finished. The check for completed processes is now done
3345 earlier. On busy systems, this bug wouldn't be noticed because something
3346 else would have woken the daemon, and it would have reaped the completed
3349 80. If a message was submitted locally by a user whose login name contained one
3350 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3351 It caused a spool format error. I have fixed the spool reading code. A
3352 related problem was that the "from" clause in the Received: line became
3353 illegal because of the space(s). It is now covered by ${quote_local_part.
3355 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3358 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3360 83. Patch from Sieve maintainer to fix unterminated string problem in
3361 "vacation" handling.
3363 84. Some minor changes to the Linux configuration files to help with other
3364 OS variants using glibc.
3366 85. One more patch for Sieve to update vacation handling to latest spec.
3369 ----------------------------------------------------
3370 See the note above about the 4.44 and 4.50 releases.
3371 ----------------------------------------------------
3377 1. Change 4.43/35 introduced a bug that caused file counts to be
3378 incorrectly computed when quota_filecount was set in an appendfile
3381 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3382 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3385 3. Give more explanation in the error message when the command for a transport
3386 filter fails to execute.
3388 4. There are several places where Exim runs a non-Exim command in a
3389 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3390 was being done only for the command run by the queryprogram router. It is
3391 now done for all such subprocesses. The other cases are: ${run, transport
3392 filters, and the commands run by the lmtp and pipe transports.
3394 5. Some older OS have a limit of 256 on the maximum number of file
3395 descriptors. Exim was using setrlimit() to set 1000 as a large value
3396 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3397 systems. I've change it so that if it can't get 1000, it tries for 256.
3399 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3400 was an oversight, and furthermore, ever since the addition of extra
3401 controls (e.g. 4.43/32), the checks on when to allow different forms of
3402 "control" were broken. There should now be diagnostics for all cases when a
3403 control that does not make sense is encountered.
3405 7. $recipients is now available in the predata ACL (oversight).
3407 8. Tidy the search cache before the fork to do a delivery from a message
3408 received from the command line. Otherwise the child will trigger a lookup
3409 failure and thereby defer the delivery if it tries to use (for example) a
3410 cached ldap connection that the parent has called unbind on.
3412 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3413 of $address_data from the recipient verification was clobbered by the
3414 sender verification.
3416 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3417 was its contents. (It was OK if the option was not defined at all.)
3419 11. A "Completed" log line is now written for messages that are removed from
3420 the spool by the -Mrm option.
3422 12. $host_address is now set to the target address during the checking of
3423 ignore_target_hosts.
3425 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3426 being passed; this would have caused $sender_host_name to have been used if
3427 matching the list had actually called for a host name (not very likely,
3428 since this list is usually IP addresses). A host name is now passed as
3431 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3432 code that specifies a non-restarting handler (typically sigaction() in
3433 modern systems) in an attempt to fix a rare and obscure crash bug.
3435 15. Narrowed the window for a race in the daemon that could cause it to ignore
3436 SIGCHLD signals. This is not a major problem, because they are used only to
3437 wake it up if nothing else does.
3439 16. A malformed maildirsize file could cause Exim to calculate negative values
3440 for the mailbox size or file count. Odd effects could occur as a result.
3441 The maildirsize information is now recalculated if the size or filecount
3444 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3445 support for a long time. Removed HAVE_SYS_VFS_H.
3447 18. Updated exipick to current release from John Jetmore.
3449 19. Allow an empty sender to be matched against a lookup in an address list.
3450 Previously the only cases considered were a regular expression, or an
3453 20. Exim went into a mad DNS lookup loop when doing a callout where the
3454 host was specified on the transport, if the DNS lookup yielded more than
3457 21. The RFC2047 encoding function was originally intended for short strings
3458 such as real names; it was not keeping to the 75-character limit for
3459 encoded words that the RFC imposes. It now respects the limit, and
3460 generates multiple encoded words if necessary. To be on the safe side, I
3461 have increased the buffer size for the ${rfc2047: expansion operator from
3464 22. Failure to deliver a bounce message always caused it to be frozen, even if
3465 there was an errors_to setting on the router. The errors_to setting is now
3468 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3469 canonical form (fully expanded) before being placed in
3470 $sender_host_address.
3472 24. Updated eximstats to version 1.33
3474 25. Include certificate and key file names in error message when GnuTLS fails
3475 to set them up, because the GnuTLS error message doesn't include the name
3476 of the failing file when there is a problem reading it.
3478 26. Expand error message when OpenSSL has problems setting up cert/key files.
3481 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3482 (this can affect the format of dates).
3484 28. exim_tidydb, when checking for the continued existence of a message for
3485 which it has found a message-specific retry record, was not finding
3486 messages that were in split spool directories. Consequently, it was
3487 deleting retry records that should have stayed in existence.
3489 29. eximstats updated to version 1.35
3490 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3491 1.35 - bugfix such that pie charts by volume are generated correctly
3493 30. The SPA authentication driver was not abandoning authentication and moving
3494 on to the next authenticator when an expansion was forced to fail,
3495 contradicting the general specification for all authenticators. Instead it
3496 was generating a temporary error. It now behaves as specified.
3498 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3499 (the order specifies the preference for clients). The order is now AES256,
3500 AES128, 3DES, ARCFOUR128.
3502 31. Small patch to Sieve code - explicitly set From: when generating an
3505 32. Exim crashed if a remote delivery caused a very long error message to be
3506 recorded - for instance if somebody sent an entire SpamAssassin report back
3507 as a large number of 550 error lines. This bug was coincidentally fixed by
3508 increasing the size of one of Exim's internal buffers (big_buffer) that
3509 happened as part of the Exiscan merge. However, to be on the safe side, I
3510 have made the code more robust (and fixed the comments that describe what
3513 33. Some experimental protocols are using DNS PTR records for new purposes. The
3514 keys for these records are domain names, not reversed IP addresses. The
3515 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3516 leaves it alone. Component reversal etc. now happens only for IP addresses.
3519 34. The host_aton() function is supposed to be passed a string that is known
3520 to be a valid IP address. However, in the case of IPv6 addresses, it was
3521 not checking this. This is a hostage to fortune. Exim now panics and dies
3522 if the condition is not met. A case was found where this could be provoked
3523 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3524 components; fortuitously, this particular loophole had already been fixed
3525 by change 4.50/55 or 4.44/33 above.
3527 If there are any other similar loopholes, the new check in host_aton()
3528 itself should stop them being exploited. The report I received stated that
3529 data on the command line could provoke the exploit when Exim was running as
3530 exim, but did not say which command line option was involved. All I could
3531 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3532 running as the user.
3535 35. There was a buffer overflow vulnerability in the SPA authentication code
3536 (which came originally from the Samba project). I have added a test to the
3537 spa_base64_to_bits() function which I hope fixes it.
3540 36. The daemon start-up calls getloadavg() while still root for those OS that
3541 need the first call to be done as root, but it missed one case: when
3542 deliver_queue_load_max is set with deliver_drop_privilege. This is
3543 necessary for the benefit of the queue runner, because there is no re-exec
3544 when deliver_drop_privilege is set.
3546 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3547 was in use, was not putting the data itself into the right store pool;
3548 consequently, it could be overwritten for a subsequent message in the same
3549 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3552 38. Sometimes the final signoff response after QUIT could fail to get
3553 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3554 < 0 before doing a fflush(). This bug looks as though it goes back to the
3555 introduction of TLS in release 3.20, but "sometimes" must have been rare
3556 because the tests only now provoked it.
3562 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3563 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3564 Now it returns an integer. A number of calls were still expecting a T/F
3565 return. Fortuitously, in all cases, the tests worked in OK situations,
3566 which is the norm. However, things would have gone wrong on any write
3567 failures on the smtp file descriptor. This function is used when sending
3568 messages over SMTP and also when doing verify callouts.
3570 2. When Exim is called to do synchronous delivery of a locally submitted
3571 message (the -odf or -odi options), it no longer closes stderr before doing
3574 3. Implemented the mua_wrapper option.
3576 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3578 5. Implemented the functions header_remove(), header_testname(),
3579 header_add_at_position(), and receive_remove_recipient(), and exported them
3582 6. If an ACL "warn" statement specified the addition of headers, Exim already
3583 inserted X-ACL-Warn: at the start if there was no header name. However, it
3584 was not making this test for the second and subsequent header lines if
3585 there were newlines in the string. This meant that an invalid header could
3586 be inserted if Exim was badly configured.
3588 7. Allow an ACL "warn" statement to add header lines at the start or after all
3589 the Received: headers, as well as at the end.
3591 8. Added the rcpt_4xx retry error code.
3593 9. Added postmaster_mailfrom=xxx to callout verification option.
3595 10. Added mailfrom=xxxx to the callout verification option, for verify=
3598 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3599 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3601 12. Inserted some casts to stop certain compilers warning when using pointer
3602 differences as field lengths or precisions in printf-type calls (mostly
3603 affecting debugging statements).
3605 13. Added optional readline() support for -be (dynamically loaded).
3607 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3608 same clock tick as a message's arrival, so that its received time was the
3609 same as the "first fail" time on the retry record, and that message
3610 remained on the queue past the ultimate address timeout, every queue runner
3611 would try a delivery (because it was past the ultimate address timeout) but
3612 after another failure, the ultimate address timeout, which should have then
3613 bounced the address, did not kick in. This was a "< instead of <=" error;
3614 in most cases the first failure would have been in the next clock tick
3615 after the received time, and all would be well.
3617 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3618 being recognized when the domain list was tested by the match_domain
3619 condition in an expansion string.
3621 16. Added the ${str2b64: operator.
3623 17. Exim was always calling setrlimit() to set a large limit for the number of
3624 processes, without checking whether the existing limit was already
3625 adequate. (It did check for the limit on file descriptors.) Furthermore,
3626 errors from getrlimit() and setrlimit() were being ignored. Now they are
3627 logged to the main and panic logs, but Exim does carry on, to try to do its
3628 job under whatever limits there are.
3630 18. Imported PCRE 5.0.
3632 19. Trivial typo in log message " temporarily refused connection" (the leading
3635 20. If the log selector return_path_on_delivery was set and an address was
3636 redirected to /dev/null, the delivery process crashed because it assumed
3637 that a return path would always be set for a "successful" delivery. In this
3638 case, the whole delivery is bypassed as an optimization, and therefore no
3641 21. Internal re-arrangement: the function for sending a challenge and reading
3642 a response while authentication was assuming a zero-terminated challenge
3643 string. It's now changed to take a pointer and a length, to allow for
3644 binary data in such strings.
3646 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3648 23. Exim was not respecting finduser_retries when seeking the login of the
3649 uid under which it was called; it was always trying 10 times. (The default
3650 setting of finduser_retries is zero.) Also, it was sleeping after the final
3651 failure, which is pointless.
3653 24. Implemented tls_on_connect_ports.
3655 25. Implemented acl_smtp_predata.
3657 26. If the domain in control=submission is set empty, Exim assumes that the
3658 authenticated id is a complete email address when it generates From: or
3659 Sender: header lines.
3661 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3662 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3663 chown and chgrp in /bin and hostname in /usr/bin.
3665 28. Exim was keeping the "process log" file open after each use, just as it
3666 does for the main log. This opens the possibility of it remaining open for
3667 long periods when the USR1 signal hits a daemon. Occasional processlog
3668 errors were reported, that could have been caused by this. Anyway, it seems
3669 much more sensible not to leave this file open at all, so that is what now
3672 29. The long-running daemon process does not normally write to the log once it
3673 has entered its main loop, and it closes the log before doing so. This is
3674 so that log files can straightforwardly be renamed and moved. However,
3675 there are a couple of unusual error situations where the daemon does write
3676 log entries, and I had neglected to close the log afterwards.
3678 30. The text of an SMTP error response that was received during a remote
3679 delivery was being truncated at 512 bytes. This is too short for some of
3680 the long messages that one sometimes sees. I've increased the limit to
3683 31. It is now possible to make retry rules that apply only when a message has a
3684 specific sender, in particular, an empty sender.
3686 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3687 it possible to be selective about when SMTP synchronization is enforced.
3689 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3691 32. Implemented hosts_connection_nolog.
3693 33. Added an ACL for QUIT.
3695 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3698 35. Added mailbox_size and mailbox_filecount to appendfile.
3700 36. Added control = no_multiline_responses to ACLs.
3702 37. There was a bug in the logic of the code that waits for the clock to tick
3703 in the case where the clock went backwards by a substantial amount such
3704 that the microsecond fraction of "now" was more than the microsecond
3705 fraction of "then" (but the whole seconds number was less).
3707 38. Added support for the libradius Radius client library this is found on
3708 FreeBSD (previously only the radiusclient library was supported).
3714 1. When certain lookups returned multiple values in the form name=value, the
3715 quoting of the values was not always being done properly. Specifically:
3716 (a) If the value started with a double quote, but contained no whitespace,
3718 (b) If the value contained whitespace other than a space character (i.e.
3719 tabs or newlines or carriage returns) it was not quoted.
3720 This fix has been applied to the mysql and pgsql lookups by writing a
3721 separate quoting function and calling it from the lookup code. The fix
3722 should probably also be applied to nisplus, ibase and oracle lookups, but
3723 since I cannot test any of those, I have not disturbed their existing code.
3725 2. A hit in the callout cache for a specific address caused a log line with no
3726 reason for rejecting RCPT. Now it says "Previous (cached) callout
3727 verification failure".
3729 3. There was an off-by-one bug in the queryprogram router. An over-long
3730 return line was truncated at 256 instead of 255 characters, thereby
3731 overflowing its buffer with the terminating zero. As well as fixing this, I
3732 have increased the buffer size to 1024 (and made a note to document this).
3734 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3735 when Exim is waiting for an SMTP response from a remote server, Exim
3736 restarts its select() call on the socket, thereby resetting its timeout.
3737 This is not a problem when such interrupts are rare. Somebody set up a cron
3738 job to run exiwhat every 2 minutes, which is less than the normal select()
3739 timeout (5 or 10 minutes). This meant that the select() timeout never
3740 kicked in because it was always reset. I have fixed this by comparing the
3741 time when an interrupt arrives with the time at the start of the first call
3742 to select(). If more time than the timeout has elapsed, the interrupt is
3743 treated as a timeout.
3745 5. Some internal re-factoring in preparation for the addition of Sieve
3746 extensions (by MH). In particular, the "personal" test is moved to a
3747 separate function, and given an option for scanning Cc: and Bcc: (which is
3748 not set for Exim filters).
3750 6. When Exim created an email address using the login of the caller as the
3751 local part (e.g. when creating a From: or Sender: header line), it was not
3752 quoting the local part when it contained special characters such as @.
3754 7. Installed new OpenBSD configuration files.
3756 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3757 try to make them clearer.
3759 9. Callout options, other than the timeout value, were being ignored when
3760 verifying sender addresses in header lines. For example, when using
3762 verify = header_sender/callout=no_cache
3764 the cache was (incorrectly) being used.
3766 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3767 only the Cygwin environment.
3769 11. When return_path_on_delivery was set as a log selector, if different remote
3770 addresses in the same message used different return paths and parallel
3771 remote delivery occurred, the wrong values would sometimes be logged.
3772 (Whenever a remote delivery process finished, the return path value from
3773 the most recently started remote delivery process was logged.)
3775 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3776 lines when AUTH and/or TLS are in use. This is the "received protocol"
3777 field. Exim used to use "asmtp" for authenticated SMTP, without any
3778 indication (in the protocol name) for TLS use. Now it follows the RFC and
3779 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3780 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3781 names appear in log lines as well as in Received: header lines.
3783 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3784 extensions, and comparison tests, and to fix some bugs.
3786 14. Changes to the "personal" filter test:
3788 (1) The test was buggy in that it was just doing the equivalent of
3789 "contains" tests on header lines. For example, if a user's address was
3790 anne@some.where, the "personal" test would incorrectly be true for
3792 To: susanne@some.where
3794 This test is now done by extracting each address from the header in turn,
3795 and checking the entire address. Other tests that are part of "personal"
3796 are now done using regular expressions (for example, to check local parts
3797 of addresses in From: header lines).
3799 (2) The list of non-personal local parts in From: addresses has been
3800 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3801 taken from the Sieve specification recommendations.
3803 (3) If the message contains any header line starting with "List-" it is
3804 treated as non-personal.
3806 (4) The test for "circular" in the Subject: header line has been removed
3807 because it now seems ill-conceived.
3809 15. Minor typos in src/EDITME comments corrected.
3811 16. Installed latest exipick from John Jetmore.
3813 17. If headers_add on a router specified a text string that was too long for
3814 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3815 of string_sprintf() is now avoided.
3817 18. $message_body_size was not set (it was always zero) when running the DATA
3818 ACL and the local_scan() function.
3820 19. For the "mail" command in an Exim filter, no default was being set for
3821 the once_repeat time, causing a random time value to be used if "once" was
3822 specified. (If the value happened to be <= 0, no repeat happened.) The
3823 default is now 0s, meaning "never repeat". The "vacation" command was OK
3824 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3825 (I found it when inspecting the code).
3827 20. There is now an overall timeout for performing a callout verification. It
3828 defaults to 4 times the callout timeout, which applies to individual SMTP
3829 commands during the callout. The overall timeout applies when there is more
3830 than one host that can be tried. The timeout is checked before trying the
3831 next host. This prevents very long delays if there are a large number of
3832 hosts and all are timing out (e.g. when the network connections are timing
3833 out). The value of the overall timeout can be changed by specifying an
3834 additional sub-option for "callout", called "maxwait". For example:
3836 verify = sender/callout=5s,maxwait=20s
3838 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3839 to the end before writing, but this should make it even safer).
3841 22. Exim was forgetting that it had advertised PIPELINING for the second and
3842 subsequent messages on an SMTP connection. It was also not resetting its
3843 memory on STARTTLS and an internal HELO.
3845 23. When Exim logs an SMTP synchronization error within a session, it now
3846 records whether PIPELINING has been advertised or not.
3848 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3849 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3850 rather than long int.
3852 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3854 26. Added the never_mail option to autoreply.
3860 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3861 crash if the getsockname() call failed; this can happen if a connection is
3862 closed very soon after it is established. The problem was simply in the
3863 order in which certain operations were done, causing Exim to try to write
3864 to the SMTP stream before it had set up the file descriptor. The bug has
3865 been fixed by making things happen in the correct order.
3871 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3872 before the connection was closed, thus losing the rejection response.
3874 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3875 some early Solaris releases, but causes trouble in current releases where
3876 socklen_t is defined.
3878 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3881 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3884 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3885 that releasing the top of it at the end releases what was used for sub-
3886 expansions (unless the block got too big). However, discard this block if
3887 the first thing is a variable or header, so that we can use its block when
3888 it is dynamic (useful for very large $message_headers, for example).
3890 6. Lookups now cache *every* query, not just the most recent. A new, separate
3891 store pool is used for this. It can be recovered when all lookup caches are
3892 flushed. Lookups now release memory at the end of their result strings.
3893 This has involved some general refactoring of the lookup sources.
3895 7. Some code has been added to the store_xxx() functions to reduce the amount
3896 of flapping under certain conditions.
3898 8. log_incoming_interface used to affect only the <= reception log lines. Now
3899 it causes the local interface and port to be added to several more SMTP log
3900 lines, for example "SMTP connection from", and rejection lines.
3902 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3904 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3906 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3907 could be overwritten at the end of the current message (or the start of a
3908 new message if it was set in a HELO ACL). The value is now preserved for
3909 the duration of the SMTP connection.
3911 12. If a transport had a headers_rewrite setting, and a matching header line
3912 contained an unqualified address, that address was qualified, even if it
3913 did not match any rewriting rules. The underlying bug was that the values
3914 of the flags that permit the existence of unqualified sender and recipient
3915 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3916 non-local messages, and by -bnq for local messages) were not being
3917 preserved with the message after it was received.
3919 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3920 "next input=" as part of the text comprising the host identity instead of
3921 the correct text. The code was using the same buffer for two different
3922 strings. However, depending on which order the printing function evaluated
3923 its arguments, the bug did not always show up. Under Linux, for example, my
3924 test suite worked just fine.
3926 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3927 doesn't work with some older versions of Perl. It has been changed to "my",
3928 which in any case is probably the better facility to use.
3930 15. A really picky compiler found some instances of statements for creating
3931 error messages that either had too many or two few arguments for the format
3934 16. The size of the buffer for calls to the DNS resolver has been increased
3935 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3936 for addresses that have a lot of PTR records. This alleviates a problem; it
3937 does not fully solve it.
3939 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3940 buffer now truncates the list and logs the incident, which is the same
3941 action as happens when Exim is looking up a host name and its aliases.
3942 Previously in this situation something unpredictable would happen;
3943 sometimes it was "internal error: store_reset failed".
3945 18. If a server dropped the connection unexpectedly when an Exim client was
3946 using GnuTLS and trying to read a response, the client delivery process
3947 crashed while trying to generate an error log message.
3949 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3950 string, for example:
3952 warn message = H1: something\nH2: something
3954 the text was added as a single header line from Exim's point of view
3955 though it ended up OK in the delivered message. However, searching for the
3956 second and subsequent header lines using $h_h2: did not work. This has been
3957 fixed. Similarly, if a system filter added multiple headers in this way,
3958 the routers could not see them.
3960 20. Expanded the error message when iplsearch is called with an invalid key to
3961 suggest using net-iplsearch in a host list.
3963 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3964 ACLs are no longer actually imposed (and a message to that effect is
3967 22. If a "gecos" field in a passwd entry contained escaped characters, in
3968 particular, if it contained a \" sequence, Exim got it wrong when building
3969 a From: or a Sender: header from that name. A second bug also caused
3970 incorrect handling when an unquoted " was present following a character
3971 that needed quoting.
3973 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3974 was not being matched caselessly.
3976 24. Arranged for all hyphens in the exim.8 source to be escaped with
3979 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3980 properly. Recipient callout cache records were still being keyed to include
3981 the sender, even when use_sender was set false. This led to far more
3982 callouts that were necessary. The sender is no longer included in the key
3983 when use_sender is false.
3985 26. Added "control = submission" modifier to ACLs.
3987 27. Added the ${base62d: operator to decode base 62 numbers.
3989 28. dnsdb lookups can now access SRV records.
3991 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3992 the configuration file.
3994 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3995 (-v) mode. This makes the output for a verbose queue run more intelligible.
3997 31. Added a use_postmaster feature to recipient callouts.
3999 32. Added the $body_zerocount variable, containing the number of binary zero
4000 bytes in the message body.
4002 33. The time of last modification of the "new" subdirectory is now used as the
4003 "mailbox time last read" when there is a quota error for a maildir
4006 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4008 35. Added +ignore_unknown as a special item in host lists.
4010 36. Code for decoding IPv6 addresses in host lists is now included, even if
4011 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4012 address was recognized as an IP address, but was then not correctly decoded
4013 into binary, causing unexpected and incorrect effects when compared with
4020 1. Very minor rewording of debugging text in manualroute to say "list of
4021 hosts" instead of "hostlist".
4023 2. If verify=header_syntax was set, and a header line with an unqualified
4024 address (no domain) and a large number of spaces between the end of the
4025 name and the colon was received, the reception process suffered a buffer
4026 overflow, and (when I tested it) crashed. This was caused by some obsolete
4027 code that should have been removed. The fix is to remove it!
4029 3. When running in the test harness, delay a bit after writing a bounce
4030 message to get a bit more predictability in the log output.
4032 4. Added a call to search_tidyup() just before forking a reception process. In
4033 theory, someone could use a lookup in the expansion of smtp_accept_max_
4034 per_host which, without the tidyup, could leave open a database connection.
4036 5. Added the variables $recipient_data and $sender_data which get set from a
4037 lookup success in an ACL "recipients" or "senders" condition, or a router
4038 "senders" option, similar to $domain_data and $local_part_data.
4040 6. Moved the writing of debug_print from before to after the "senders" test
4043 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4044 problems for message scanning, either using a data ACL, or using
4045 local_scan() because the Received: header was not generated till after they
4046 were called (in order to set the time as the time of reception completion).
4047 I have revised the way this works. The header is now generated after the
4048 body is received, but before the ACL or local_scan() are called. After they
4049 are run, the timestamp in the header is updated.
4055 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4056 before starting a queue runner without re-exec. This happened only when
4057 deliver_drop_privilege was set or when the Exim user was set to root. The
4058 effect of the bug was that timeouts during subsequent deliveries caused
4059 crashes instead of being properly handled. The handler is now left at its
4060 default (and expected) setting.
4062 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4063 message, again when deliver_drop_privilege is set or Exim is run as root.
4064 The bug described in (1) was not present in this case, but the tidying up
4065 of the other signals was missing. I have made the two cases consistent.
4067 3. The ignore_target_hosts setting on a manualroute router was being ignored
4068 for hosts that were looked up using the /MX notation.
4070 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4073 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4074 operated on the sender address. After changing the $sender_address to <>
4075 for the sender address verify, Exim was re-instated it as the original
4076 (before rewriting) address, but remembering that it had rewritten it, so it
4077 wasn't rewriting it again. This bug also had the effect of breaking the
4078 sender address verification caching when the sender address was rewritten.
4080 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4081 This has been changed so that if the ip literal address matches
4082 ignore_target_hosts, the router declines.
4084 7. Added expansion conditions match_domain, match_address, and match_local_
4085 part (NOT match_host).
4087 8. The placeholder for the Received: header didn't have a length field set.
4089 9. Added code to Exim itself and to exim_lock to test for a specific race
4090 condition that could lead to file corruption when using MBX delivery. The
4091 issue is with the lockfile that is created in /tmp. If this file is removed
4092 after a process has opened it but before that process has acquired a lock,
4093 there is the potential for a second process to recreate the file and also
4094 acquire a lock. This could lead to two Exim processes writing to the file
4095 at the same time. The added code performs the same test as UW imapd; it
4096 checks after acquiring the lock that its file descriptor still refers to
4097 the same named file.
4099 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4100 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4103 11. Added the smtp_active_hostname option. If used, this will typically be made
4104 to depend on the incoming interface address. Because $interface_address is
4105 not set up until the daemon has forked a reception process, error responses
4106 that can happen earlier (such as "too many connections") no longer contain
4109 12. If an expansion in a condition on a "warn" statement fails because a lookup
4110 defers, the "warn" statement is abandoned, and the next ACL statement is
4111 processed. Previously this caused the whole ACL to be aborted.
4113 13. Added the iplsearch lookup type.
4115 14. Added ident_timeout as a log selector.
4117 15. Added tls_certificate_verified as a log selector.
4119 16. Added a global option tls_require_ciphers (compare the smtp transport
4120 option of the same name). This controls incoming TLS connections.
4122 17. I finally figured out how to make tls_require_ciphers do a similar thing
4123 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4124 before starting the TLS session.
4126 18. Tabs are now shown as \t in -bP output.
4128 19. If the log selector return_path_on_delivery was set, Exim crashed when
4129 bouncing a message because it had too many Received: header lines.
4131 20. If two routers both had headers_remove settings, and the first one included
4132 a superfluous trailing colon, the final name in the first list and the
4133 first name in the second list were incorrectly joined into one item (with a
4134 colon in the middle).
4140 1. Added -C and -D options to the exinext utility, mainly to make it easier
4141 to include in the automated testing, but these could be helpful when
4142 multiple configurations are in use.
4144 2. The exinext utility was not formatting the output nicely when there was
4145 an alternate port involved in the retry record key, nor when there was a
4146 message id as well (for retries that were specific to a specific message
4147 and a specific host). It was also confused by IPv6 addresses, because of
4148 the additional colons they contain. I have fixed the IPv4 problem, and
4149 patched it up to do a reasonable job for IPv6.
4151 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4152 delivery, the log line now contains "pipelined" if PIPELINING was used.
4154 4. An SMTP transport process used to panic and die if the bind() call to set
4155 an explicit outgoing interface failed. This has been changed; it is now
4156 treated in the same way as a connect() failure.
4158 5. A reference to $sender_host_name in the part of a conditional expansion
4159 that was being skipped was still causing a DNS lookup. This no longer
4162 6. The def: expansion condition was not recognizing references to header lines
4163 that used bh_ and bheader_.
4165 7. Added the _cache feature to named lists.
4167 8. The code for checking quota_filecount in the appendfile transport was
4168 allowing one more file than it should have been.
4170 9. For compatibility with Sendmail, the command line option
4178 and sets the incoming protocol and host name (for trusted callers). The
4179 host name and its colon can be omitted when only the protocol is to be set.
4180 Note the Exim already has two private options, -pd and -ps, that refer to
4181 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4182 "s", but I don't think that's a major issue.
4184 10. A number of refactoring changes to the code, none of which should affect
4187 (a) The number of logging options was getting close to filling up the
4188 32-bit word that was used as a bit map. I have split them into two classes:
4189 those that are passed in the argument to log_write(), and those that are
4190 only ever tested independently outside of that function. These are now in
4191 separate 32-bit words, so there is plenty of room for expansion again.
4192 There is no change in the user interface or the logging behaviour.
4194 (b) When building, for example, log lines, the code previously used a
4195 macro that called string_cat() twice, in order to add two strings. This is
4196 not really sufficiently general. Furthermore, there was one instance where
4197 it was actually wrong because one of the argument was used twice, and in
4198 one call a function was used. (As it happened, calling the function twice
4199 did not affect the overall behaviour.) The macro has been replaced by a
4200 function that can join an arbitrary number of extra strings onto a growing
4203 (c) The code for expansion conditions now uses a table and a binary chop
4204 instead of a serial search (which was left over from when there were very
4205 few conditions). Also, it now recognizes conditions like "pam" even when
4206 the relevant support is not compiled in: a suitably worded error message is
4207 given if an attempt is made to use such a condition.
4209 11. Added ${time_interval:xxxxx}.
4211 12. A bug was causing one of the ddress fields not to be passed back correctly
4212 from remote delivery subprocesses. The field in question was not being
4213 subsequently used, so this caused to problems in practice.
4215 13. Added new log selectors queue_time and deliver_time.
4217 14. Might have fixed a bug in maildirsizefile handling that threw up
4218 "unexpected character" debug warnings, and recalculated the data
4219 unnecessarily. In any case, I expanded the warning message to give more
4222 15. Added the message "Restricted characters in address" to the statements in
4223 the default ACL that block characters like @ and % in local parts.
4225 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4226 Three changes have been made:
4228 (a) There was a serious bug; a negative response to MAIL caused the whole
4229 recipient domain to be cached as invalid, thereby blocking all messages
4230 to all local parts at the same domain, from all senders. This bug has
4231 been fixed. The domain is no longer cached after a negative response to
4232 MAIL if the sender used is not empty.
4234 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4237 (c) A new callout option, "use_sender" has been added for people who want
4238 the modified behaviour.
4244 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4247 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4248 indeed breaks things for older releases.
4250 3. Added additional logging to the case where there is a problem reading data
4251 from a filter that is running in a subprocess using a pipe, in order to
4252 try to track down a specific problem.
4254 4. Testing facility fudge: when running in the test harness and attempting
4255 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4256 getting "No route to host". Convert this to a timeout.
4258 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4261 6. Some OS don't have socklen_t but use size_t instead. This affects the
4262 fifth argument of getsockopt() amongst other things. This is now
4263 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4264 can be set for individual OS. I have set it for SunOS5, OSF1, and
4265 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4266 some earlier ones do not.
4268 7. Change 4.30/15 was not doing the test caselessly.
4270 8. The standard form for an IPv6 address literal was being rejected by address
4271 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4272 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4273 this, as well as the form without the "IPv6" on the front (but only when
4274 address literals are enabled, of course).
4276 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4278 10. Exim crashed if a message with an empty sender address specified by -f
4279 encountered a router with an errors_to setting. This could be provoked only
4280 by a command such as
4284 where an empty string was supplied; "<>" did not hit this bug.
4286 11. Installed PCRE release 4.5.
4288 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4289 remained set. It is now erased.
4291 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4292 times from message ids (which are base 36 rather than the normal 62).
4294 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4295 were being counted as actual protocol errors, and logged if the log
4296 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4297 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4298 and DATA following a set of rejected RCPTs do not count as protocol errors.
4299 In other words, Exim assumes they were pipelined, though this may not
4300 actually be the case. Of course, in all cases the client gets an
4301 appropriate error code.
4303 15. If a lookup fails in an ACL condition, a message about the failure may
4304 be available; it is used if testing the ACL cannot continue, because most
4305 such messages specify what the cause of the deferral is. However, some
4306 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4307 that caused an old message to be retained and used if a later statement
4308 caused a defer, replacing the real cause of the deferral.
4310 16. If an IP address had so many PTR records that the DNS lookup buffer
4311 was not large enough to hold them, Exim could crash while trying to process
4312 the truncated data. It now detects and logs this case.
4314 17. Further to 4.21/58, another change has been made: if (and only if) the
4315 first line of a message (the first header line) ends with CRLF, a bare LF
4316 in a subsequent header line has a space inserted after it, so as not to
4317 terminate the header.
4319 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4320 unnecessarily, used atoi() instead of strtol(), and didn't check the
4321 termination when getting file sizes from file names by regex.
4323 19. Completely re-implemented the support for maildirsize files, in the light
4324 of a number of problems with the previous contributed implementation
4325 (4.30/29). In particular:
4327 . If the quota is zero, the maildirsize file is maintained, but no quota is
4330 . If the maildir directory does not exist, it is created before any attempt
4331 to write a maildirsize file.
4333 . The quota value in the file is just a cache; if the quota is changed in
4334 the transport, the new value overrides.
4336 . A regular expression is available for excluding directories from the
4339 20. The autoreply transport checks the characters in options that define the
4340 message's headers; it allows continued headers, but it was checking with
4341 isspace() after an embedded newline instead of explicitly looking for a
4344 21. If all the "regular" hosts to which an address was routed had passed their
4345 expiry times, and had not reached their retry times, the address was
4346 bounced, even if fallback hosts were defined. Now Exim should go on to try
4349 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4350 equivalent code in the SMTP transport. Some hosts send humungous responses
4351 to HELO/EHLO, more than 1024 it seems.
4353 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4354 gives compiler warnings in some environments. I've now done it "properly",
4357 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4358 (because of problems with the built-in one) was declared to return uschar *
4359 instead of char *, causing compiler failure.
4361 25. Fixed a file descriptor leak when processing alias/forward files.
4363 26. Fixed a minor format string issue in dbfn.c.
4365 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4367 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4368 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4371 29. When a sender address is verified, it is cached, to save repeating the test
4372 when there is more than one recipient in a message. However, when the
4373 verification involves a callout, it is possible for different callout
4374 options to be set for different recipients. It is too complicated to keep
4375 track of this in the cache, so now Exim always runs a verification when a
4376 callout is required, relying on the callout cache for the optimization.
4377 The overhead is duplication of the address routing, but this should not be
4380 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4381 to be verified with callout=postmaster, and the main callout worked but the
4382 postmaster check failed, the verification correctly failed. However, if a
4383 subsequent RCPT command asked for sender verification *without* the
4384 postmaster check, incorrect caching caused this verification also to fail,
4387 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4388 it was not caching the DNS options (qualify_single, search_parents) that
4389 were used when the lookup failed. A subsequent lookup with different
4390 options therefore always gave the same answer, though there were cases
4391 where it should not have. (Example: a "domains = !$mx_any" option on a
4392 dnslookup router: the "domains" option is always processed without any
4393 widening, but the router might have qualify_single set.) Now Exim uses the
4394 cached value only when the same options are set.
4396 32. Added John Jetmore's "exipick" utility to the distribution.
4398 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4399 than a timeout (e.g. a certificate is required, and is not provided), an
4400 Exim server now closes the connection immediately. Previously it waited for
4401 the client to close - but if the client is SSL, it seems that they each
4402 wait for each other, leading to a delay before one of them times out.
4404 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4405 maintained 0.8.x compatibility because I don't think many are using it, and
4406 it is clearly obsolete.
4408 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4411 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4412 client certificate was expired. A simple patch fixes this, though I don't
4413 understand the full logic of why the verify callback is called multiple
4416 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4417 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4418 which causes problems with some clients (such as the Certicom SSL Plus
4419 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4420 disables the coutermeasure allowing Eudora to connect."
4422 38. Exim was not checking that a write() to a log file succeeded. This could
4423 lead to Bad Things if a log got too big, in particular if it hit a file
4424 size limit. Exim now panics and dies if it cannot write to a log file, just
4425 as it does if it cannot open a log file.
4427 39. Modified OS/Makefile-Linux so that it now contains
4429 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4431 The two -D definitions ensure that Exim is compiled with large file
4432 support, which makes it possible to handle log files that are bigger than
4435 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4436 instance) a domain was checked against a named list that involved a lookup,
4437 causing $domain_data to be set, then another domain was checked against the
4438 same list, then the first domain was re-checked, the value of $domain_data
4439 after the final check could be wrong. In particular, if the second check
4440 failed, it could be set empty. This bug probably also applied to
4443 41. The strip_trailing_dot option was not being applied to the address given
4444 with the -f command-line option.
4446 42. The code for reading a message's header from the spool was incrementing
4447 $received_count, but never initializing it. This meant that the value was
4448 incorrect (doubled) while delivering a message in the same process in which
4449 it was received. In the most common configuration of Exim, this never
4450 happens - a fresh exec is done - but it can happen when
4451 deliver_drop_privilege is set.
4453 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4454 it now includes up to 150 characters of the unexpected data in the log
4457 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4458 and building data strings. The size of both of these buffers was 10 000
4459 bytes - far larger than anybody would *ever* want, thought I. Needless to
4460 say, somebody hit the limit. I have increased the maximum line length to
4461 20 000 and the maximum data length of concatenated lines to 100 000. I have
4462 also fixed two bugs, because there was no checking on these buffers. Tsk,
4463 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4464 buffer is too small.
4466 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4467 lsearch lookups. Now it does.
4469 46. When parsing a route_list item in a manualroute router, a fixed-length
4470 buffer was used for the list of hosts. I made this 1024 bytes long,
4471 thinking that nobody would ever have a list of hosts that long. Wrong.
4472 Somebody had a whole pile of complicated expansion conditions, and the
4473 string was silently truncated, leading to an expansion error. It turns out
4474 that it is easier to change to an unlimited length (owing to other changes
4475 that have happened since this code was originally written) than to build
4476 structure for giving a limitation error. The length of the item that
4477 expands into the list of hosts is now unlimited.
4479 47. The lsearch lookup could not handle data where the length of text line was
4480 more than 4095 characters. Such lines were truncated, leading to shortened
4481 data being returned. It should now handle lines of any length.
4483 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4484 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4487 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4488 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4489 no longer happens in comment lines. A list of replacements is now placed
4490 at the head of all of the source files, except those whose only change is
4491 to replace PERL_COMMAND in the very first #! line.
4493 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4494 messages on the queue, with a bottom-up merge sort, using code contributed
4495 by Michael Haardt. This should make operations like -bp somewhat faster on
4496 large queues. It won't affect queue runners, except when queue_run_in_order
4499 51. Installed eximstats 1.31 in the distribution.
4501 52. Added support for SRV lookups to the dnslookup router.
4503 53. If an ACL referred to $message_body or $message_body_end, the value was not
4504 reset for any messages that followed in the same SMTP session.
4506 54. The store-handling optimization for building very long strings was not
4507 differentiating between the different store pools. I don't think this
4508 actually made any difference in practice, but I've tidied it.
4510 55. While running the routers to verify a sender address, $sender_address
4511 was still set to the sender address. This is wrong, because when routing to
4512 send a bounce to the sender, it would be empty. Therefore, I have changed
4513 it so that, while verifying a sender address, $sender_address is set to <>.
4514 (There is no change to what happens when verifying a recipient address.)
4516 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4517 target A or AAAA records (if not already returned) without resetting the
4518 qualify_single or search_parents options of the DNS resolver. These are
4519 inappropriate in this case because the targets of MX and SRV records must
4520 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4521 target that, when qualified, matched something in the local domain. These
4522 two options are now turned off when doing these lookups.
4524 57. It seems that at least some releases of Reiserfs (which does not have the
4525 concept of a fixed number of inodes) returns zero and not -1 for the
4526 number of available inodes. This interacted badly with check_spool_inodes,
4527 which assumed that -1 was the "no such thing" setting. What I have done is
4528 to check that the total number of inodes is greater than zero before doing
4529 the test of how many are available.
4531 58. When a "warn" ACL statement has a log_message modifier, the message is
4532 remembered, and not repeated. This is to avoid a lot of repetition when a
4533 message has many recipients that cause the same warning to be written.
4534 Howewer, Exim was preserving the list of already written lines for an
4535 entire SMTP session, which doesn't seem right. The memory is now reset if a
4536 new message is started.
4538 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4539 debugging output unless log_rewrite was also set.
4541 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4542 of (int)(handle) when we know that handle contains (void *)(-1).
4544 61. The Exim daemon panic-logs an error return when it closes the incoming
4545 connection. However "connection reset by peer" seems to be common, and
4546 isn't really an error worthy of noting specially, so that particular error
4549 62. When Exim is trying to find all the local interfaces, it used to panic and
4550 die if the ioctl to get the interface flags failed. However, it seems that
4551 on at least one OS (Solaris 9) it is possible to have an interface that is
4552 included in the list of interfaces, but for which you get a failure error
4553 for this call. This happens when the interface is not "plumbed" into a
4554 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4555 failure of the "get flags" call assumes that the interface is down.
4557 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4558 makes life easier for people who are doing arithmetic on fields extracted
4559 from dates, where you often get leading zeros that should not be
4560 interpreted as octal.
4562 64. Added qualify_domain to the redirect router, to override the global
4565 65. If a pathologically long header line contained very many addresses (the
4566 report of this problem mentioned 10 000) and each of them was rewritten,
4567 Exim could use up a very large amount of memory. (It kept on making new
4568 copies of the header line as it rewrote, and never released the old ones.)
4569 At the expense of a bit more processing, the header rewriting function has
4570 been changed so that it no longer eats memory in this way.
4572 66. The generation of the Received: header has been moved from the time that a
4573 message starts to be received, to the time that it finishes. The timestamp
4574 in the Received: header should now be very close to that of the <= log
4575 line. There are two side-effects of this change:
4577 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4578 logged header lines no longer include the local Received: line, because
4579 it has not yet been created. The same applies to a copy of the message
4580 that is returned to a non-SMTP sender when a message is rejected.
4582 (b) When a filter file is tested using -bf, no additional Received: header
4583 is added to the test message. After some thought, I decided that this
4586 This change does not affect the value of $received_for. It is still set
4587 after address rewriting, but before local_scan() is called.
4589 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4591 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4592 gave an unhelpful panic error message, and a defer error. I have managed to
4593 change this behaviour so that it now rejects any supplied certificate,
4594 which seems right, as the list of acceptable certificates is empty.
4596 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4597 gave an unhelpful defer error. I have not managed to make this reject any
4598 supplied certificates, but the error message it gives is "no certificate
4599 supplied", which is not helpful.
4601 70. exigrep's output now also includes lines that are not associated with any
4602 message, but which match the given pattern. Implemented by a patch from
4603 Martin Sluka, which also tidied up the Perl a bit.
4605 71. Recipient callout verification, like sender verification, was using <> in
4606 the MAIL FROM command. This isn't really the right thing, since the actual
4607 sender may affect whether the remote host accepts the recipient or not. I
4608 have changed it to use the actual sender in the callout; this means that
4609 the cache record is now keyed on a recipient/sender pair, not just the
4610 recipient address. There doesn't seem to be a real danger of callout loops,
4611 since a callout by the remote host to check the sender would use <>.
4612 [SEE ABOVE: changed after hitting problems.]
4614 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4615 temporary errors. However, in the case of such a code being given after
4616 the end of a data transmission (i.e. after ".") Exim was failing to write
4617 a retry record for the message. (Yes, there was some broken host that was
4618 actually sending 8xx at this point.)
4620 73. An unknown lookup type in a host list could cause Exim to panic-die when
4621 the list was checked. (An example that provoked this was putting <; in the
4622 middle of a list instead of at the start.) If this happened during a DATA
4623 ACL check, a -D file could be left lying around. This kind of configuration
4624 error no longer causes Exim to die; instead it causes a defer errror. The
4625 incident is still logged to the main and panic logs.
4627 74. Buglet left over from Exim 3 conversion. The message "too many messages
4628 in one connection" was written to the rejectlog but not the mainlog, except
4629 when address rewriting (yes!) was being logged.
4631 75. Added write_rejectlog option.
4633 76. When a system filter was run not as root (that is, when system_filter_user
4634 was set), the values of the $n variables were not being returned to the
4635 main process; thus, they were not subsequently available in the $sn
4638 77. Added +return_path_on_delivery log selector.
4640 78. A connection timeout was being treated differently from recipients deferred
4641 when testing hosts_max_try with a message that was older than the host's
4642 retry timeout. (The host should not be counted, thus allowing all hosts to
4643 be tried at least once before bouncing.) This may have been the cause of an
4644 occasionally reported bug whereby a message would remain on the queue
4645 longer than the retry timeout, but would be bounced if a delivery was
4646 forced. I say "may" because I never totally pinned down the problem;
4647 setting up timeout/retry tests is difficult. See also the next item.
4649 79. The ultimate address timeout was not being applied to errors that involved
4650 a combination of host plus message (for example, a timeout on a MAIL
4651 command). When an address resolved to a number of possible hosts, and they
4652 were not all tried for each delivery (e.g. because of hosts_max_try), a
4653 message could remain on the queue longer than the retry timeout.
4655 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4658 81. Fixed an obscure SMTP outgoing bug which required at least the following
4659 conditions: (a) there was another message waiting for the same server;
4660 (b) the server returned 5xx to all RCPT commands in the first message so
4661 that the message was not completed; (c) the server dropped the connection
4662 or gave a negative response to the RSET that Exim sends to abort the
4663 transaction. The observed case was a dropped connection after DATA that had
4664 been sent in pipelining mode. That is, the server had advertised PIPELINING
4665 but was not implementing it correctly. The effect of the bug was incorrect
4666 behaviour, such as trying another host, and this could lead to a crash.
4672 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4673 and daemon.c were passed as pointers to ints; they should have been
4674 pointers to socklen_t variables (which are typically unsigned ints).
4676 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4679 3. Fixed a really odd bug that affected only the testing scheme; patching a
4680 certain fixed string in the binary changed the value of another string that
4681 happened to be identical to the end of the original first string.
4683 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4684 name", it returns that address as the IP address. On some operating
4685 systems (e.g. Solaris), it also passes back the IP address string as the
4686 "host name". However, on others (e.g. Linux), it passes back an empty
4687 string. Exim wasn't checking for this, and was changing the host name to an
4688 empty string, assuming it had been canonicized.
4690 5. Although rare, it is permitted to have more than one PTR record for a given
4691 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4692 all the names associated with an address, because they do in Solaris.
4693 However, it seems that they do not in Linux for data that comes from the
4694 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4695 I found this out when I moved to a new Linux workstation and tried to run
4696 the Exim test suite.
4698 To get round this problem I have changed the code so that it now does its
4699 own call to the DNS to look up PTR records when searching for a host name.
4700 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4701 addresses that are only in /etc/hosts are still found.
4703 This behaviour is, however, controlled by an option called host_lookup_
4704 order, which defaults to "bydns:byaddr". If people want to use the other
4705 order, or indeed, just use one or the other means of lookup, they can
4706 specify it in this variable.
4708 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4709 some operating systems, this comes back from gethostbyaddr() as an empty
4710 string, and this is what Exim used to test for. However, it seems that in
4711 other systems, "." is yielded. Exim now tests for this case too.
4713 7. The values of check_spool_space and check_log_space are now held internally
4714 as a number of kilobytes instead of an absolute number of bytes. If a
4715 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4716 kilobyte. This means that much larger values can be stored.
4718 8. Exim monitor: an attempt to get the action menu when not actually pointing
4719 at a message produces an empty menu entitled "No message selected". This
4720 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4721 no entries in it ("Shell widget menu has zero width and/or height"). So I
4722 have added a single, blank menu entry in this case.
4724 9. Added ${quote_local_part.
4726 10. MIME decoding is now applied to the contents of Subject: header lines when
4729 11. Now that a reference to $sender_host_address automatically causes a reverse
4730 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4731 host lookup before query-style lookups in lists that might use this
4732 variable. This has therefore been abolished, and the "net-" prefix is no
4733 longer necessary for query-style lookups.
4735 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4736 have been a typo for LFLAGS, so it has been changed.
4738 13. The install script calls Exim with "-C /dev/null" in order to find the
4739 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4740 to be output. Howeve, since Exim outputs its version number before the
4741 error, it didn't break the script. It just looked ugly. I fixed this by
4742 always allowing "-C /dev/null" if the caller is root.
4744 14. Ignore overlarge ACL variable number when reading spool file - insurance
4745 against a later release with more variables having written the file.
4747 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4748 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4749 this, as well as the form without the "IPv6" on the front.
4751 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4752 OS/Makefile-Darwin file.
4754 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4755 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4757 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4758 variable. However, some OS use ints for inodes. I've added cast to long int
4759 to get rid of the compiler warning.
4761 19. I had forgotten to lock out "/../" in configuration file names when
4762 ALT_CONFIG_PREFIX was set.
4764 20. Routers used for verification do not need to specify transports. However,
4765 if such a router generated a host list, and callout was configured, Exim
4766 crashed, because it could not find a port number from the (non-existent)
4767 transport. It now assumes port 25 in this circumstance.
4769 21. Added the -t option to exigrep.
4771 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4772 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4773 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4774 them is set, LOOKUP_LSEARCH is forced.
4776 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4778 24. Added sender and host information to the "rejected by local_scan()" log
4779 line; previously there was no indication of these.
4781 25. Added .include_if_exists.
4783 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4784 Linux. It turns out that not all file systems support this. Apparently some
4785 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4786 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4787 not supported on the file descriptor, is now ignored when Exim is trying to
4788 sync a directory. This applies only to Linux.
4790 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4792 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4793 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4796 29. Added support for maildirsize files from supplied patch (modified a bit).
4798 30. The use of :fail: followed by an empty string could lead Exim to respond to
4799 sender verification failures with (e.g.):
4801 550 Verification failed for <xxx>
4802 550 Sender verify failed
4804 where the first response line was missing the '-' that indicates it is not
4805 the final line of the response.
4807 31. The loop for finding the name of the user that called Exim had a hardwired
4808 limit of 10; it now uses the value of finduser_retries, which is used for
4809 all other user lookups.
4811 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4814 33. Exim was neglecting to zero errno before one call of strtol() when
4815 expanding a string and expecting an integer value. On some systems this
4816 resulted in spurious "integer overflow" errors. Also, it was casting the
4817 result into an int without checking.
4819 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4820 did not work. The code looks as if it has *never* worked, though it appears
4821 to have been documented since at least releast 1.62. I have made it work.
4823 35. The "timeout_DNS" error in retry rules, also documented since at least
4824 1.62, also never worked. As it isn't clear exactly what this means, and
4825 clearly it isn't a major issue, I have abolished the feature by treating it
4826 as "timeout", and writing a warning to the main and panic logs.
4828 36. The display of retry rules for -brt wasn't always showing the error code
4831 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4832 timeout_connect_A, timeout_connect_MX.
4834 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4835 to the empty sender.
4837 39. The daemon was not analysing the content of -oX till after it had closed
4838 stderr and disconnected from the controlling terminal. This meant that any
4839 syntax errors were only noted on the panic log, and the return code from
4840 the command was 0. By re-arranging the code a little, I've made the
4841 decoding happen first, so such errors now appear on stderr, and the return
4842 code is 1. However, the actual setting up of the sockets still happens in
4843 the disconnected process, so errors there are still only recorded on the
4846 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4847 connections (as happens on some IP stacks) was logged at start up time as
4848 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4849 it from "IPv6 and IPv4", which means that two separate sockets are being
4852 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4853 says whether AF_INET or AF_INET6 was passed as an argument.
4855 42. Exiwhat output was messed up when time zones were included in log
4858 43. Exiwhat now gives more information about the daemon's listening ports,
4859 and whether -tls-on-connect was used.
4861 44. The "port" option of the smtp transport is now expanded.
4863 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4864 silently ignored. Now an error message is written to the main and panic
4867 46. There's a new ACL modifier called "logwrite" which writes to a log file
4868 as soon as it is encountered.
4870 47. Added $local_user_uid and $local_user_gid at routing time.
4872 48. Exim crashed when trying to verify a sender address that was being
4875 49. Exim was recognizing only a space character after ".include". It now also
4876 recognizes a tab character.
4878 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4879 extracting the relevant information from the specification. The man page no
4880 longer contains scrambled data for the -d option, and I've added a section
4881 at the front about calling Exim under different names.
4883 51. Added "extra_headers" argument to the "mail" command in filter files.
4885 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4888 53. Installed eximstats 1.29.
4890 54. Added transport_filter_timeout as a generic transport option.
4892 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4893 Cc: header lines. This was required by RFC 822, but it not required by RFC
4896 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4897 incoming messages that did not have them. Now it does so only if the
4898 message originates locally, that is, if there is no associated remote host
4899 address. When Resent- header lines are present, this applies to the Resent-
4900 lines rather than the non-Resent- lines.
4902 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4903 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4905 58. Messages for configuration errors now include the name of the main
4906 configuration file - useful now that there may be more than one file in a
4907 list (.included file names were always shown).
4909 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4910 for those rare installations that do not start the daemon as root or run it
4911 setuid root. I've cut out the call to initgroups() if the daemon is not
4914 60. The Exim user and group can now be bound into the binary as text strings
4915 that are looked up at the start of Exim's processing.
4917 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4919 62. Added $mailstore_basename variable.
4921 63. Installed patch to sieve.c from Michael Haardt.
4923 64. When Exim failed to open the panic log after failing to open the main log,
4924 the original message it was trying to log was written to stderr and debug
4925 output, but if they were not available (the usual case in production), it
4926 was lost. Now it is written to syslog before the two lines that record the
4927 failures to open the logs.
4929 65. Users' Exim filters run in subprocesses under the user's uid. It is
4930 possible for a "deliver" command or an alias in a "personal" command to
4931 provoke an address rewrite. If logging of address rewriting is configured,
4932 this fails because the process is not running as root or exim. There may be
4933 a better way of dealing with this, but for the moment (because 4.30 needs
4934 to be released), I have disabled address rewrite logging when running a
4935 filter in a non-root, non-exim process.
4941 1. The buildconfig auxiliary program wasn't quoting the value set for
4942 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4943 not defined. This bug was masked in 4.22 by the effect that was fixed in
4946 2. Some messages that were rejected after a message id was allocated were
4947 shown as "incomplete" by exigrep. It no longer does this for messages that
4948 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4950 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4951 have allow_domain_literals set, the ID did not get logged in the <= line.
4952 Domain literals are now always recognized in Message-ID: header lines.
4954 4. The first argument for a ${extract expansion item is the key name or field
4955 number. Leading and trailing spaces in this item were not being ignored,
4956 causing some misleading effects.
4958 5. When deliver_drop_privilege was set, single queue runner processes started
4959 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4960 same command in the process it spins off) were not dropping privilege.
4962 6. When the daemon running as "exim" started a queue runner, it always
4963 re-executed Exim in the spun-off process. This is a waste of effort when
4964 deliver_drop_privilege is set. The new process now just calls the
4965 queue-runner function directly.
4971 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4974 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4975 ignored. Though the use of -oP was forcing the writing of a pid file, it
4976 was always written to the default place.
4978 3. If the message "no IP address found for host xxxx" is generated during
4979 incoming verification, it is now followed by identification of the incoming
4980 connection (so you can more easily find what provoked it).
4982 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4984 5. Added some features to "harden" Exim a bit more against certain attacks:
4986 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4987 be put in Local/Makefile. This is like the never_users runtime option,
4988 but it cannot be overridden. The default setting is "root".
4990 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4991 prefix string with which any file named in a -C command line option
4994 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4995 is retained for -C and -D only if the caller of Exim is root. Without
4996 it, the exim user may also use -C and -D and retain privilege.
4998 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4999 command line option is disabled.
5001 6. Macro names set by the -D option must start with an upper case letter, just
5002 like macro names defined in the configuration file.
5004 7. Added "dereference=" facility to LDAP.
5006 8. Two instances of the typo "uknown" in the source files are fixed.
5008 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5009 the Configure-Makefile script screwed up while processing it.
5011 10. Incorporated PCRE 4.4.
5013 11. The SMTP synchronization check was not operating right at the start of an
5014 SMTP session. For example, it could not catch a HELO sent before the client
5015 waited for the greeting. There is now a check for outstanding input at the
5016 point when the greeting is written. Because of the duplex, asynchronous
5017 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5018 way, but not yet received, when the check is performed.
5020 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5021 on TCP/IP sockets, because this apparently causes some broken clients to
5024 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5025 unchanged) from the Cygwin maintainer.
5027 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5028 was supported instead of testing for mbx. Effectively a typo.
5030 15. The spa authenticator server code was not checking that the input it
5031 received was valid base64.
5033 16. The debug output line for the "set" modifier in ACLs was not showing the
5034 name of the variable that was being set.
5036 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5038 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5039 lookup on its own. Something else was needed to trigger the lookup. For
5040 example, a match in host_lookup or the need for a host name in a host list.
5041 Now, if $sender_host_name is referenced and the host name has not yet been
5042 looked up, a lookup is performed. If the lookup fails, the variable remains
5043 empty, and $host_lookup_failed is set to "1".
5045 19. Added "eqi" as a case-independent comparison operator.
5047 20. The saslauthd authentication condition could segfault if neither service
5048 nor realm was specified.
5050 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5051 error message that was logged was misleading, and incoming SMTP
5052 connections were dropped. The message is now more accurate, and temporary
5053 errors are given to SMTP connections.
5055 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5056 (except RSET) with a 421 error, until QUIT is received. However, it was
5057 failing to send a response to QUIT.
5059 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5060 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5061 ACL is now run after the helo_try_verify_hosts code.
5063 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5064 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5065 case-independent, so other case variants are also recognized). Apparently
5066 some systems use these upper case variants.
5068 25. If more than two messages were waiting for the same host, and a transport
5069 filter was specified for the transport, Exim sent two messages over the
5070 same TCP/IP connection, and then failed with "socket operation on non-
5071 socket" when it tried to send the third.
5073 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5075 27. The extern definition of crypt16() in expand.c was not being excluded when
5076 the OS had its own crypt16() function.
5078 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5079 as a preferred synonym for return_size_limit, both as an option and as an
5082 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5084 30. Changed the default configuration ACL to relax the local part checking rule
5085 for addresses that are not in any local domains. For these addresses,
5086 slashes and pipe symbols are allowed within local parts, but the sequence
5087 /../ is explicitly forbidden.
5089 31. SPA server authentication was not clearing the challenge buffer before
5092 32. log_message in a "warn" ACL statement was writing to the reject log as
5093 well as to the main log, which contradicts the documentation and doesn't
5094 seem right (because no rejection is happening). So I have stopped it.
5096 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5097 However, I am unable to do any testing of this.
5099 34. Fixed an infelicity in the appendfile transport. When checking directories
5100 for a mailbox, to see if any needed to be created, it was accidentally
5101 using path names with one or more superfluous leading slashes; tracing
5102 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5104 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5105 contents are added to the log line that is written for every discarded
5106 recipient. (Previously a log_message setting was ignored.)
5108 36. The ${quote: operator now quotes the string if it is empty.
5110 37. The install script runs exim in order to find its version number. If for
5111 some reason other than non-existence or emptiness, which it checks, it
5112 could not run './exim', it was installing it with an empty version number,
5113 i.e. as "exim-". This error state is now caught, and the installation is
5116 38. An argument was missing from the function that creates an error message
5117 when Exim fails to connect to the socket for saslauthd authentication.
5118 This could cause Exim to crash, or give a corrupted message.
5120 39. Added isip, isip4, and isip6 to ${if conditions.
5122 40. The ACL variables $acl_xx are now saved with the message, and can be
5123 accessed later in routers, transports, and filters.
5125 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5126 strings in the file are not string-expanded.
5128 42. If a MAIL command specified a SIZE value that was too large to fit into an
5129 int variable, the check against message_size_limit failed. Such values are
5130 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5131 day this will have to be increased, but I don't think I want to be around
5132 when emails are that large.
5139 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5140 iconv() is not standard in FreeBSD.
5142 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5143 IPv6 enabled. The observed symptom was a segmentation fault on return from
5144 the function os_common_find_running_interfaces() in src/os.c.
5146 3. In the check_special_case() function in daemon.c I had used "errno" as an
5147 argument name, which causes warnings on some systems. This was basically a
5148 typo, since it was named "eno" in the comments!
5150 4. The code that waits for the clock to tick (at a resolution of some fraction
5151 of a second) so as to ensure message-id uniqueness was always waiting for
5152 at least one whole tick, when it could have waited for less. [This is
5153 almost certainly not relevant at current processor speeds, where it is
5154 unlikely to ever wait at all. But we try to future-proof.]
5156 5. The function that sleeps for a time interval that includes fractions of a
5157 second contained a race. It did not block SIGALRM between setting the
5158 timer, and suspending (a couple of lines later). If the interval was short
5159 and the sigsuspend() was delayed until after it had expired, the suspension
5160 never ended. On busy systems this could lead to processes getting stuck for
5163 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5164 process, before it forks any delivery processes. The open lookup caching
5165 mechanism meant that the open file or database connection was passed into
5166 the delivery process. The problem was that delivery processes always tidy
5167 up cached lookup data. This could cause a problem for the next delivery
5168 process started by the queue runner, because the external queue runner
5169 process does not know about the closure. So the next delivery process
5170 still has data in the lookup cache. In the case of a file lookup, there was
5171 no problem because closing a file descriptor in a subprocess doesn't affect
5172 the parent. However, if the lookup was caching a connection to a database,
5173 the connection was closed, and the second delivery process was likely to
5174 see errors such as "PGSQL: query failed: server closed the connection
5175 unexpectedly". The problem has been fixed by closing all cached lookups
5176 in a queue runner before running a delivery process.
5178 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5179 seem to have the "const" qualifier which it has on other OS. I've
5182 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5183 *of the same type* (client or server) with the same public name that an
5184 error should be diagnosed.
5186 9. When Exim looked up a host name for an IP address, but failed to find the
5187 original IP address when looking up the host name (a safety check), it
5188 output the message "<ip address> does not match any IP for NULL", which was
5189 confusing, to say the least. The bug was that the host name should have
5190 appeared instead of "NULL".
5192 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5193 user that is built into the binary, and the -C or -D options is used, root
5194 privilege is dropped before the configuration file is read. In addition,
5195 logging is switched to stderr instead of the normal log files. If the
5196 configuration then re-defines the Exim user, the unprivileged environment
5197 is probably not what is expected, so Exim logs a panic warning message (but
5200 However, if deliver_drop_privilege is set, the unprivileged state may well
5201 be exactly what is intended, so the warning has been cut out in that case,
5202 and Exim is allowed to try to write to its normal log files.
5208 1. smtp_return_error_details was not giving details for temporary sender
5209 or receiver verification errors.
5211 2. Diagnose a configuration error if two authenticators have the same public
5214 3. Exim used not to create the message log file for a message until the first
5215 delivery attempt. This could be confusing when incoming messages were held
5216 for policy or load reasons. The message log file is now created at the time
5217 the message is received, and an initial "Received" line is written to it.
5219 4. The automatically generated man page for command line options had a minor
5220 bug that caused no ill effects; however, a more serious problem was that
5221 the procedure for building the man page automatically didn't always
5222 operate. Consequently, release 4.20 contains an out-of-date version. This
5223 shouldn't happen again.
5225 5. When building Exim with embedded Perl support, the script that builds the
5226 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5227 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5229 6. The freeze_tell option was not being used for messages that were frozen on
5230 arrival, either by an ACL or by local_scan().
5232 7. Added the smtp_incomplete_transaction log selector.
5234 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5235 was accepting AUTH without a new EHLO.
5237 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5238 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5241 10. Logging of TCP/IP connections (when configured) now happens in the main
5242 daemon process instead of the child process, so that the TCP/IP connection
5243 count is more accurate (but it can never be perfect).
5245 11. The use of "drop" in a nested ACL was not being handled correctly in the
5246 outer ACL. Now, if condition failure induced by the nested "drop" causes
5247 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5248 or "require"), the connection is dropped.
5250 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5251 that yield "discard" can now be used with an "accept" or a "discard" verb,
5252 but an error is generated for any others (because I can't see a useful way
5253 to define what should happen).
5255 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5256 are now processed in the same way as lines in the Exim configuration file.
5257 In particular, continuation lines are supported.
5259 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5261 15. Added -ti meaning -t -i.
5263 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5264 domains, and warn by logging if others are found.
5266 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5267 returned from GIFCONF when Exim is trying to find the list of interfaces on
5268 a host. The code in os.c has been modified to copy each ifreq to an aligned
5269 structure in all cases.
5271 Also, in some cases, the returned ifreq's were being copied to a 'struct
5272 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5273 means the last couple of bytes of an IPv6 address could be chopped if the
5274 ifreq contained only a normal sockaddr (14 bytes storage).
5276 18. Named domain lists were not supported in the hosts_treat_as_local option.
5277 An entry such as +xxxx was not recognized, and was treated as a literal
5280 19. Ensure that header lines added by a DATA ACL are included in the reject log
5281 if the ACL subsequently rejects the message.
5283 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5284 MD5 (which is deprecated).
5286 21. When testing a filter file using -bf, Exim was writing a message when it
5287 took the sender from a "From " line in the message, but it was not doing so
5288 when it took $return_path from a Return-Path: header line. It now does.
5290 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5291 with a valid header line field name (a series of printing characters
5292 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5294 23. Changed "disc" in the source to "disk" to conform to the documentation and
5295 the book and for uniformity.
5297 24. Ignore Sendmail's -Ooption=value command line item.
5299 25. When execve() failed while trying to run a command in a pipe transport,
5300 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5301 could be confused with a return value of 69 from the command itself. This
5302 has been changed to 127, the value the shell returns if it is asked to run
5303 a non-existent command. The wording for the related log line suggests a
5304 non-existent command as the problem.
5306 26. If received_header_text expands to an empty string, do not add a Received:
5307 header line to the message. (Well, it adds a token one on the spool, but
5308 marks it "old" so that it doesn't get used or transmitted.)
5310 27. Installed eximstats 1.28 (addition of -nt option).
5312 28. There was no check for failure on the call to getsockname() in the daemon
5313 code. This can fail if there is a shortage of resources on the system, with
5314 ENOMEM, for example. A temporary error is now given on failure.
5316 29. Contrary to the C standard, it seems that in some environments, the
5317 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5318 program. Exim now does this explicitly; it affects the formatting of
5319 timestamps using strftime().
5321 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5322 complaints. I've now initialized all the variables, to avoid this.
5324 32. Header lines added by a system filter were not being "seen" during
5325 transport-time rewrites.
5327 33. The info_callback() function passed to OpenSSL is set up with type void
5328 (*)(SSL *, int, int), as described somewhere. However, when calling the
5329 function (actually a macro) that sets it up, the type void(*)() is
5330 expected. I've put in a cast to prevent warnings from picky compilers.
5332 34. If a DNS black list lookup found a CNAME record, but there were no A
5333 records associated with the domain it pointed at, Exim crashed.
5335 35. If a DNS black list lookup returned more than one A record, Exim ignored
5336 all but the first. It now scans all returned addresses if a particular IP
5337 value is being sought. In this situation, the contents of the
5338 $dnslist_value variable are a list of all the addresses, separated by a
5341 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5342 to accept a host name and all its aliases if the forward lookup for any of
5343 them yielded the IP address of the incoming connection. Now it accepts only
5344 those names whose forward lookup yields the correct IP address. Any other
5345 names are discarded. This closes a loophole whereby a rogue DNS
5346 administrator could create reverse DNS records to break through a
5347 wildcarded host restriction in an ACL.
5349 37. If a user filter or a system filter that ran in a subprocess used any of
5350 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5351 the wrong values were passed to the pipe command ($thisaddress had the
5352 value of $0, $0 had the value of $1, etc). This bug was introduced by
5353 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5355 38. Improved the line breaking for long SMTP error messages from ACLs.
5356 Previously, if there was no break point between 40 and 75 characters, Exim
5357 left the rest of the message alone. Two changes have been made: (a) I've
5358 reduced the minimum length to 35 characters; (b) if it can't find a break
5359 point between 35 and 75 characters, it looks ahead and uses the first one
5360 that it finds. This may give the occasional overlong line, but at least the
5361 remaining text gets split now.
5363 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5364 file descriptors might be low, and that setting 1000 would always raise it.
5365 It turns out that in some environments, the limit is already over 1000 and
5366 that lowering it causes trouble. So now Exim takes care not to decrease it.
5368 40. When delivering a message, the value of $return_path is set to $sender_
5369 address at the start of routing (routers may change the value). By an
5370 oversight, this default was not being set up when an address was tested by
5371 -bt or -bv, which affected the outcome if any router or filter referred to
5374 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5375 log only when "message" or "log_message" are set. However, if one of the
5376 conditions was an address verification, or a call to a nested ACL, the
5377 messages generated by the underlying test were being passed through. This
5378 no longer happens. The underlying message is available in $acl_verify_
5379 message for both "message" and "log_message" expansions, so it can be
5380 passed through if needed.
5382 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5383 new expansion $bh_ to give the encoded byte string without charset
5384 translation. Translation happens only if iconv() is available; HAVE_ICONV
5385 indicates this at build time. HEADERS_CHARSET gives the charset to
5386 translate to; headers_charset can change it in the configuration, and
5387 "headers charset" can change it in an individual filter file.
5389 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5390 that creates RFC 2047 encoded "words" labels them as that charset instead
5391 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5392 expansion operator; (ii) when Exim creates a From: line for a local
5393 message; (iii) when a header line is rewritten to include a "phrase" part.
5395 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5396 buggy, causing it to skip the first lines of messages whose message ID
5397 ended in 'D'. This would not have bitten before Exim release 4.14, because
5398 message IDs were unlikely to end in 'D' before then. The effect was to have
5399 incorrect size information for certain domains.
5401 45. #include "config.h" was missing at the start of the crypt16.c module. This
5402 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5405 46. If there was a timeout during a "random" callout check, Exim treated it as
5406 a failure of the random address, and carried on sending RSET and the real
5407 address. If the delay was just some slowness somewhere, the response to the
5408 original RCPT would be taken as a response to RSET and so on, causing
5409 mayhem of various kinds.
5411 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5412 when I implemented it. It didn't allow for the fact that some option values
5413 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5414 the right test for positive values.
5416 48. Domain names in DNS records are case-independent. Exim always looks them up
5417 in lower case. Some resolvers return domain names in exactly the case they
5418 appear in the zone file, that is, they may contain uppercase letters. Not
5419 all resolvers do this - some return always lower case. Exim was treating a
5420 change of case by a resolver as a change of domain, similar to a widening
5421 of a domain abbreviation. This triggered its re-routing code and so it was
5422 trying to route what was effectively the same domain again. This normally
5423 caused routing to fail (because the router wouldn't handle the domain
5424 twice). Now Exim checks for this case specially, and just changes the
5425 casing of the domain that it ultimately uses when it transmits the message
5428 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5431 50. If a filter generated a file delivery with a non-absolute name (possible if
5432 no home directory exists for the router), the forbid_file option was not
5435 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5436 the existing equality matching.
5438 52. Exim was using ints instead of ino_t variables in some places where it was
5439 dealing with inode numbers.
5441 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5442 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5443 called TMPDIR, and if it finds it is different, it changes its value.
5445 54. The smtp_printf() function is now made available to local_scan() so
5446 additional output lines can be written before returning. There is also an
5447 smtp_fflush() function to enable the detection of a dropped connection.
5448 The variables smtp_input and smtp_batched_input are exported to
5451 55. Changed the default runtime configuration: the message "Unknown user"
5452 has been removed from the ACL, and instead placed on the localuser router,
5453 using the cannot_route_message feature. This means that any verification
5454 failures that generate their own messages won't get overridden. Similarly,
5455 the "Unrouteable address" message that was in the ACL for unverifiable
5456 relay addresses has also been removed.
5458 56. Added hosts_avoid_esmtp to the smtp transport.
5460 57. The exicyclog script was not checking for the esoteric option
5461 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5462 will work only if exicyclog is run under the appropriate euid.
5464 58. Following a discussion on the list, the rules by which Exim recognises line
5465 endings on incoming messages have been changed. The -dropcr and drop_cr
5466 options are now no-ops, retained only for backwards compatibility. The
5467 following line terminators are recognized: LF CRLF CR. However, special
5468 processing applies to CR:
5470 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5471 nor a local message in the state where . is a terminator.
5473 (ii) If a bare CR is encountered in a header line, an extra space is added
5474 after the line terminator so as not to end the header. The reasoning
5475 behind this is that bare CRs in header lines are most likely either
5476 to be mistakes, or people trying to play silly games.
5478 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5479 was being incorrectly given as 18 bytes larger than it should have been.
5480 This is a VOB (very old bug).
5482 60. This may never have affected anything current, but just in case it has:
5483 When the local host is found other than at the start of a list of hosts,
5484 the local host, those with the same MX, and any that follow, are discarded.
5485 When the list in question was part of a longer list of hosts, the following
5486 hosts (not currently being processed) were also being discarded. This no
5487 longer happens. I'm not sure if this situation could ever has previously
5490 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5493 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5494 header. This is something that is recommended in a new Internet Draft, and
5495 is something that is documented as being done by Sendmail. There are two
5496 possible values. For messages generated by the autoreply transport, Exim
5499 Auto-Submitted: auto-replied
5501 whereas for all other generated messages (e.g. bounces) it adds
5503 Auto-Submitted: auto-generated
5505 63. The "personal" condition in filters now includes a test for the
5506 Auto-Submitted: header. If it contains the string "auto-" the message it
5507 not considered personal.
5509 64. Added rcpt_include_affixes as a generic transport option.
5511 65. Added queue_only_override (default true).
5513 66. Added the syslog_duplication option.
5515 67. If what should have been the first header line of a message consisted of
5516 a space followed by a colon, Exim was mis-interpreting it as a header line.
5517 It isn't of course - it is syntactically invalid and should therefore be
5518 treated as the start of the message body. The misbehaviour could have
5519 caused a number of strange effects, including loss of data in subsequent
5520 header lines, and spool format errors.
5522 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5523 client host had authenticated. This control can now be exercised by an ACL
5524 for more flexibility.
5526 69. By default, callouts do not happen when testing with -bh. There is now a
5527 variant, -bhc, which does actually run the callout code, including
5528 consulting and updating the callout cache.
5530 70. Added support for saslauthd authentication, courtesy of Alexander
5533 71. If statvfs() failed on the spool or log directories while checking their
5534 size for availability, Exim confusingly gave the error "space shortage".
5535 Furthermore, in debugging mode it crashed with a floating point exception.
5536 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5537 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5538 serious problem, Exim now writes to the main and panic logs when this
5539 happens, with details of the failure. It then refuses to accept the
5540 incoming message, giving the message "spool directory problem" or "log
5541 directory problem" with a 421 code for SMTP messages.
5543 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5544 0, 1, and 2 exist, because some OS complain for execs without them (see
5545 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5546 descriptors. However, the code omitted to check that the open succeeded,
5547 causing mysterious errors if for some reason the permissions on /dev/null
5548 got screwed. Now Exim writes a message to the main and panic logs, and
5549 bombs out if it can't open /dev/null.
5551 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5552 interact so that it is all more flexible. It is supposed to remain
5553 backwards compatible. Also added extra_local_interfaces.
5555 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5556 to bomb out with an assertion failure - to the client this appears as a
5557 connection drop. This problem occurs in the part of the code that was taken
5558 from the Samba project. Fortunately, the assertion is in a very simple
5559 function, so I have fixed this by reproducing the function inline in the
5560 one place where it is called, and arranging for authentication to fail
5561 instead of killing the process with assert().
5563 75. The SPA client code was not working when the server requested OEM rather
5564 than Unicode encoding.
5566 76. Added code to make require_files with a specific uid setting more usable in
5567 the case where statting the file as root fails - usually a non-root-mounted
5568 NFS file system. When this happens and the failure is EACCES, Exim now
5569 forks a subprocess and does the per-uid checking as the relevant uid.
5571 77. Added process_log_path.
5573 78. If log_file_path was not explicitly set, a setting of check_log_space or
5574 check_log_inodes was ignored.
5576 79. If a space check for the spool or log partitions fails, the incident is now
5577 logged. Of course, in the latter case the data may get lost...
5579 80. Added the %p formatting code to string_format() so that it can be used to
5580 print addresses in debug_print(). Adjusted all the address printing in the
5581 debugging in store.c to use %p rather than %d.
5583 81. There was a concern that a line of code in smtp_in.c could overflow a
5584 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5585 initially expressed, the concern was not well-founded, because trailing
5586 spaces are removed early. However, if the trailing spaces were followed by
5587 a NULL, they did not get removed, so the overflow was possible. Two fixes
5590 (a) I re-wrote the offending code in a cleaner fashion.
5591 (b) If an incoming SMTP command contains a NULL character, it is rejected
5594 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5595 runs initgroups(), so that if the Exim user is in any additional groups,
5596 they will be used during message reception.
5602 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob