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 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
70 Report and patch from Dmitry Banschikov.
72 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
73 Removed SSL_clear() after SSL_new() which led to protocol negotiation
74 failures. We appear to now support TLS1.1+ with Exim.
76 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
77 lets Exim select keys and certificates based upon TLS SNI from client.
78 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
79 before an outbound SMTP session. New log_selector, +tls_sni.
81 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
82 NULL dereference. Report and patch from Alun Jones.
84 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
85 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
86 Not seeing resolver debug output on NetBSD, but suspect this is a
87 resolver implementation change.
89 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
90 Left warnings. Added "eximon gdb" invocation mode.
92 PP/21 Defaulting "accept_8bitmime" to true, not false.
94 PP/22 Added -bw for inetd wait mode support.
96 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
97 locate the relevant includes and libraries. Made this the default.
99 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
100 Bugzilla 1246, report and most of solution from Tomasz Kusy.
102 JH/02 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
103 This may cause build issues on older platforms.
105 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
106 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
107 gnutls_require_mac & gnutls_require_protocols (no longer supported).
108 Added SNI support via GnuTLS too.
109 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
111 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
113 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
115 JH/03 Added expansion variable $tod_epoch_l for a higher-precision time.
121 PP/01 Solaris build fix for Oracle's LDAP libraries.
122 Bugzilla 1109, patch from Stephen Usher.
124 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
126 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
129 TF/02 Fix a couple more cases where we did not log the error message
130 when unlink() failed. See also change 4.74-TF/03.
132 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
133 lock up or crash if it happened to be inside a call to libc when it
134 got a SIGUSR1 from exiwhat.
136 The SIGUSR1 handler appends the current process status to the process
137 log which is later printed by exiwhat. It used to use the general
138 purpose logging code to do this, but several functions it calls are
139 not safe for signals.
141 The new output code in the SIGUSR1 handler is specific to the process
142 log, and simple enough that it's easy to inspect for signal safety.
143 Removing some special cases also simplifies the general logging code.
144 Removing the spurious timestamps from the process log simplifies
147 TF/04 Improved ratelimit ACL condition.
149 The /noupdate option has been deprecated in favour of /readonly which
150 has clearer semantics. The /leaky, /strict, and /readonly update modes
151 are mutually exclusive. The update mode is no longer included in the
152 database key; it just determines when the database is updated. (This
153 means that when you upgrde Exim will forget old rate measurements.)
155 Exim now checks that the per_* options are used with an update mode that
156 makes sense for the current ACL. For example, when Exim is processing a
157 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
158 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
159 must specify per_mail/readonly. If you omit the update mode it defaults to
160 /leaky where that makes sense (as before) or /readonly where required.
162 The /noupdate option is now undocumented but still supported for
163 backwards compatibility. It is equivalent to /readonly except that in
164 ACLs where /readonly is required you may specify /leaky/noupdate or
165 /strict/noupdate which are treated the same as /readonly.
167 A useful new feature is the /count= option. This is a generalization
168 of the per_byte option, so that you can measure the throughput of other
169 aggregate values. For example, the per_byte option is now equivalent
170 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
172 The per_rcpt option has been generalized using the /count= mechanism
173 (though it's more complicated than the per_byte equivalence). When it is
174 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
175 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
176 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
177 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
178 Note that using per_rcpt with a non-readonly update mode in more than
179 one ACL will cause the recipients to be double-counted. (The per_mail
180 and per_byte options don't have this problem.)
182 The handling of very low rates has changed slightly. If the computed rate
183 is less than the event's count (usually one) then this event is the first
184 after a long gap. In this case the rate is set to the same as this event's
185 count, so that the first message of a spam run is counted properly.
187 The major new feature is a mechanism for counting the rate of unique
188 events. The new per_addr option counts the number of different
189 recipients that someone has sent messages to in the last time period. It
190 behaves like per_rcpt if all the recipient addresses are different, but
191 duplicate recipient addresses do not increase the measured rate. Like
192 the /count= option this is a general mechanism, so the per_addr option
193 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
194 example, measure the rate that a client uses different sender addresses
195 with the options per_mail/unique=$sender_address. There are further
196 details in the main documentation.
198 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
200 TF/06 Removed a few PCRE remnants.
202 TF/07 Automatically extract Exim's version number from tags in the git
203 repository when doing development or release builds.
205 PP/02 Raise smtp_cmd_buffer_size to 16kB.
206 Bugzilla 879. Patch from Paul Fisher.
208 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
209 Heavily based on revision 40f9a89a from Simon Arlott's tree.
212 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
214 PP/05 Variable $av_failed, true if the AV scanner deferred.
215 Bugzilla 1078. Patch from John Horne.
217 PP/06 Stop make process more reliably on build failure.
218 Bugzilla 1087. Patch from Heiko Schlittermann.
220 PP/07 Make maildir_use_size_file an _expandable_ boolean.
221 Bugzilla 1089. Patch from Heiko Schlittermann.
223 PP/08 Handle ${run} returning more data than OS pipe buffer size.
224 Bugzilla 1131. Patch from Holger Weiß.
226 PP/09 Handle IPv6 addresses with SPF.
227 Bugzilla 860. Patch from Wolfgang Breyha.
229 PP/10 GnuTLS: support TLS 1.2 & 1.1.
231 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
234 PP/11 match_* no longer expand right-hand-side by default.
235 New compile-time build option, EXPAND_LISTMATCH_RHS.
236 New expansion conditions, "inlist", "inlisti".
238 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
240 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
242 PP/14 fix log_write() format string regression from TF/03.
243 Bugzilla 1152. Patch from Dmitry Isaikin.
249 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
251 PP/02 Harmonised TLS library version reporting; only show if debugging.
252 Layout now matches that introduced for other libraries in 4.74 PP/03.
254 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
256 PP/04 New "dns_use_edns0" global option.
258 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
261 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
262 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
264 TK/01 Updated PolarSSL code to 0.14.2.
265 Bugzilla 1097. Patch from Andreas Metzler.
267 PP/07 Catch divide-by-zero in ${eval:...}.
270 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
273 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
274 format-string attack -- SECURITY: remote arbitrary code execution.
276 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
277 time unintentionally subject to list matching rules, letting the header
278 cause arbitrary Exim lookups (of items which can occur in lists, *not*
279 arbitrary string expansion). This allowed for information disclosure.
281 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
282 INT_MIN/-1 -- value coerced to INT_MAX.
288 NM/01 Workround for PCRE version dependency in version reporting
291 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
292 This fixes portability to compilers other than gcc, notably
293 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
295 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
296 makefiles for portability to HP-UX and POSIX correctness.
298 PP/01 Permit LOOKUP_foo enabling on the make command-line.
299 Also via indented variable definition in the Makefile.
300 (Debugging by Oliver Heesakkers).
302 PP/02 Restore caching of spamd results with expanded spamd_address.
303 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
305 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
306 Improves build reliability. Fix from: Frank Elsner
308 NM/02 Fix wide character breakage in the rfc2047 coding
309 Fixes bug 1064. Patch from Andrey N. Oktyabrski
311 NM/03 Allow underscore in dnslist lookups
312 Fixes bug 1026. Patch from Graeme Fowler
314 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
315 Code patches from Adam Ciarcinski of NetBSD.
317 NM/04 Fixed exiqgrep to cope with mailq missing size issue
320 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
321 is logged, to avoid truncation. Patch from John Horne.
323 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
324 Patch from Jakob Hirsch.
326 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
327 SQL string expansion failure details.
328 Patch from Andrey Oktyabrski.
330 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
331 Patch from Simon Arlott.
333 PP/09 New lookups functionality failed to compile on old gcc which rejects
334 extern declarations in function scope.
335 Patch from Oliver Fleischmann
337 PP/10 Use sig_atomic_t for flags set from signal handlers.
338 Check getgroups() return and improve debugging.
339 Fixed developed for diagnosis in bug 927 (which turned out to be
342 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
343 Patch from Mark Zealey.
345 PP/12 Bugzilla 1056: Improved spamd server selection.
346 Patch from Mark Zealey.
348 PP/13 Bugzilla 1086: Deal with maildir quota file races.
349 Based on patch from Heiko Schlittermann.
351 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
352 Patch from Uwe Doering, sign-off by Michael Haardt.
354 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
355 variable declaration deep within a block. Bug and patch from
358 PP/15 lookups-Makefile IRIX compatibilty coercion.
360 PP/16 Make DISABLE_DKIM build knob functional.
362 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
363 Patch by Simon Arlott
365 TF/03 Fix valgrind.h portability to C89 compilers that do not support
366 variable argument macros. Our copy now differs from upstream.
372 TF/01 Failure to get a lock on a hints database can have serious
373 consequences so log it to the panic log.
375 TF/02 Log LMTP confirmation messages in the same way as SMTP,
376 controlled using the smtp_confirmation log selector.
378 TF/03 Include the error message when we fail to unlink a spool file.
380 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
381 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
382 for maintaining out-of-tree patches for some time.
384 PP/01 Bugzilla 139: Documentation and portability issues.
385 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
386 Handle per-OS dynamic-module compilation flags.
388 PP/02 Let /dev/null have normal permissions.
389 The 4.73 fixes were a little too stringent and complained about the
390 permissions on /dev/null. Exempt it from some checks.
391 Reported by Andreas M. Kirchwitz.
393 PP/03 Report version information for many libraries, including
394 Exim version information for dynamically loaded libraries. Created
395 version.h, now support a version extension string for distributors
396 who patch heavily. Dynamic module ABI change.
398 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
399 privilege escalation vulnerability whereby the Exim run-time user
400 can cause root to append content of the attacker's choosing to
403 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
406 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
407 If dropping privileges for untrusted macros, we disabled normal logging
408 on the basis that it would fail; for the Exim run-time user, this is not
409 the case, and it resulted in successful deliveries going unlogged.
410 Fixed. Reported by Andreas Metzler.
416 PP/01 Date: & Message-Id: revert to normally being appended to a message,
417 only prepend for the Resent-* case. Fixes regression introduced in
418 Exim 4.70 by NM/22 for Bugzilla 607.
420 PP/02 Include check_rfc2047_length in configure.default because we're seeing
421 increasing numbers of administrators be bitten by this.
423 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
425 PP/03 Bugzilla 994: added openssl_options main configuration option.
427 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
429 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
431 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
433 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
434 without a peer certificate, leading to a segfault because of an
435 assumption that peers always have certificates. Be a little more
436 paranoid. Problem reported by Martin Tscholak.
438 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
439 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
440 NB: ClamAV planning to remove STREAM in "middle of 2010".
441 CL also introduces -bmalware, various -d+acl logging additions and
442 more caution in buffer sizes.
444 PP/09 Implemented reverse_ip expansion operator.
446 PP/10 Bugzilla 937: provide a "debug" ACL control.
448 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
450 PP/12 Bugzilla 973: Implement --version.
452 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
454 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
456 PP/15 Bugzilla 816: support multiple condition rules on Routers.
458 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
459 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
460 ignore trailing whitespace.
462 JJ/02 prevent non-panic DKIM error from being sent to paniclog
464 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
467 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
468 Notification from Dr Andrew Aitchison.
470 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
471 ExtendedDetectionInfo response format.
472 Notification from John Horne.
474 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
477 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
478 XSL and documented dependency on system catalogs, with examples of how
481 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
484 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
485 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
486 configuration file which is writeable by the Exim user or group.
488 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
489 of configuration files to cover files specified with the -C option if
490 they are going to be used with root privileges, not just the default
493 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
494 option (effectively making it always true).
496 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
497 files to be used while preserving root privileges.
499 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
500 that rogue child processes cannot use them.
502 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
503 run-time user, instead of root.
505 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
506 Exim run-time user without dropping privileges.
508 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
509 result string, instead of calling string_vformat() twice with the same
512 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
513 for other users. Others should always drop root privileges if they use
514 -C on the command line, even for a whitelisted configure file.
516 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
518 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
524 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
525 $data_path, and $header_path variables; fixed documentation bugs and
528 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
529 exipick to access non-standard spools, including the "frozen" queue
532 NM/01 Bugzilla 965: Support mysql stored procedures.
533 Patch from Alain Williams
535 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
537 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
538 Patch from Andreas Metzler
540 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
541 Patch from Kirill Miazine
543 NM/05 Bugzilla 671: Added umask to procmail example.
545 JJ/03 installed exipick 20100323.0, fixing doc bug
547 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
548 directory. Notification and patch from Dan Rosenberg.
550 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
552 TK/02 Improve log output when DKIM signing operation fails.
554 MH/01 Treat the transport option dkim_domain as a colon separated
555 list, not as a single string, and sign the message with each element,
556 omitting multiple occurences of the same signer.
558 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
559 Bugzilla 985, 986. Patch by Simon Arlott
561 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
562 Patch by Simon Arlott
564 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
565 MBX locking. Notification from Dan Rosenberg.
571 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
573 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
575 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
577 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
579 NM/04 Bugzilla 915: Flag broken perl installation during build.
585 TK/01 Added patch by Johannes Berg that expands the main option
586 "spamd_address" if it starts with a dollar sign.
588 TK/02 Write list of recipients to X-Envelope-Sender header when building
589 the mbox-format spool file for content scanning (suggested by Jakob
592 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
593 (http://www.dcc-servers.net/) support via dccifd. Activated by
594 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
596 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
597 by Mark Daniel Reidel <mr@df.eu>.
599 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
600 When building exim an external PCRE library is now needed -
601 PCRE is a system library on the majority of modern systems.
602 See entry on PCRE_LIBS in EDITME file.
604 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
605 conversation. Added nologin parameter to request.
606 Patch contributed by Kirill Miazine.
608 TF/01 Do not log submission mode rewrites if they do not change the address.
610 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
612 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
613 log files in place. Contributed by Roberto Lima.
615 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
617 TF/03 Bugzilla 615: When checking the local_parts router precondition
618 after a local_part_suffix or local_part_prefix option, Exim now
619 does not use the address's named list lookup cache, since this
620 contains cached lookups for the whole local part.
622 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
623 Robert Millan. Documentation is in experimental-spec.txt.
625 TF/04 Bugzilla 668: Fix parallel build (make -j).
627 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
629 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
630 Patch provided by Jan Srzednicki.
632 TF/05 Leading white space used to be stripped from $spam_report which
633 wrecked the formatting. Now it is preserved.
635 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
636 that they are available at delivery time.
638 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
640 TF/08 TLS error reporting now respects the incoming_interface and
641 incoming_port log selectors.
643 TF/09 Produce a more useful error message if an SMTP transport's hosts
644 setting expands to an empty string.
646 NM/06 Bugzilla 744: EXPN did not work under TLS.
647 Patch provided by Phil Pennock.
649 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
650 Patch provided by Richard Godbee.
652 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
653 acl_smtp_notquit, added index entry.
655 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
656 Patch provided by Eugene Bujak.
658 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
659 accept(). Patch provided by Maxim Dounin.
661 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
662 Patch provided by Phil Pennock.
664 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
666 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
667 Patch provided by Brad "anomie" Jorsch.
669 NM/14 Bugzilla 622: Added timeout setting to transport filter.
670 Patch provided by Dean Brooks.
672 TK/05 Add native DKIM support (does not depend on external libraries).
674 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
675 Patch provided by Graeme Fowler.
677 NM/16 Bugzilla 851: Documentation example syntax fix.
679 NM/17 Changed NOTICE file to remove references to embedded PCRE.
681 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
684 NM/19 Bugzilla 745: TLS version reporting.
685 Patch provided by Phil Pennock.
687 NM/20 Bugzilla 167: bool: condition support.
688 Patch provided by Phil Pennock.
690 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
691 clients. Patch provided by Phil Pennock.
693 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
694 Patch provided by Brad "anomie" Jorsch.
696 NM/23 Bugzilla 687: Fix misparses in eximstats.
697 Patch provided by Heiko Schlittermann.
699 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
700 Patch provided by Heiko Schlittermann.
702 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
703 plus update to original patch.
705 NM/26 Bugzilla 799: Documentation correction for ratelimit.
707 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
708 Patch provided by David Brownlee.
710 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
712 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
714 NM/30 Bugzilla 888: TLS documentation bugfixes.
716 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
718 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
719 Unlike the original bugzilla I have changed all shell scripts in src tree.
721 NM/33 Bugzilla 898: Transport filter timeout fix.
722 Patch by Todd Rinaldo.
724 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
725 Patch by Serge Demonchaux.
727 NM/35 Bugzilla 39: Base64 decode bug fixes.
728 Patch by Jakob Hirsch.
730 NM/36 Bugzilla 909: Correct connect() call in dcc code.
732 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
734 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
736 NM/39 Bugzilla 911: Fixed MakeLinks build script.
742 TK/01 Add preliminary DKIM support. Currently requires a forked version of
743 ALT-N's libdkim that I have put here:
744 http://duncanthrax.net/exim-experimental/
746 Note to Michael Haardt: I had to rename some vars in sieve.c. They
747 were called 'true' and it seems that C99 defines that as a reserved
748 keyword to be used with 'bool' variable types. That means you could
749 not include C99-style headers which use bools without triggering
750 build errors in sieve.c.
752 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
753 as mailq or other aliases. Changed the --help handling significantly
754 to do whats expected. exim_usage() emits usage/help information.
756 SC/01 Added the -bylocaldomain option to eximstats.
758 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
760 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
762 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
768 PH/01 Another patch from the Sieve maintainer.
770 PH/02 When an IPv6 address is converted to a string for single-key lookup
771 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
772 dots are used instead of colons so that keys in lsearch files need not
773 contain colons. This was done some time before quoting was made available
774 in lsearch files. However, iplsearch files do require colons in IPv6 keys
775 (notated using the quote facility) so as to distinguish them from IPv4
776 keys. This meant that lookups for IP addresses in host lists did not work
777 for iplsearch lookups.
779 This has been fixed by arranging for IPv6 addresses to be expressed with
780 colons if the lookup type is iplsearch. This is not incompatible, because
781 previously such lookups could never work.
783 The situation is now rather anomolous, since one *can* have colons in
784 ordinary lsearch keys. However, making the change in all cases is
785 incompatible and would probably break a number of configurations.
787 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
790 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
791 conversion specification without a maximum field width, thereby enabling
792 a rogue spamd server to cause a buffer overflow. While nobody in their
793 right mind would setup Exim to query an untrusted spamd server, an
794 attacker that gains access to a server running spamd could potentially
795 exploit this vulnerability to run arbitrary code as the Exim user.
797 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
798 $primary_hostname instead of what libspf2 thinks the hosts name is.
800 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
801 a directory entry by the name of the lookup key. Previously, if a
802 symlink pointed to a non-existing file or a file in a directory that
803 Exim lacked permissions to read, a lookup for a key matching that
804 symlink would fail. Now it is enough that a matching directory entry
805 exists, symlink or not. (Bugzilla 503.)
807 PH/03 The body_linecount and body_zerocount variables are now exported in the
810 PH/04 Added the $dnslist_matched variable.
812 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
813 This means they are set thereafter only if the connection becomes
816 PH/06 Added the client_condition to authenticators so that some can be skipped
817 by clients under certain conditions.
819 PH/07 The error message for a badly-placed control=no_multiline_responses left
820 "_responses" off the end of the name.
822 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
824 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
825 (without spaces) instead of just copying the configuration text.
827 PH/10 Added the /noupdate option to the ratelimit ACL condition.
829 PH/11 Added $max_received_linelength.
831 PH/12 Added +ignore_defer and +include_defer to host lists.
833 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
834 way in which PCRE > 7.0 is built.
836 PH/14 Implemented queue_only_load_latch.
838 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
839 MAIL command. The effect was to mangle the value on 64-bit systems.
841 PH/16 Another patch from the Sieve maintainer.
843 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
845 PH/18 If a system quota error occurred while trying to create the file for
846 a maildir delivery, the message "Mailbox is full" was not appended to the
847 bounce if the delivery eventually timed out. Change 4.67/27 below applied
848 only to a quota excession during the actual writing of the file.
850 PH/19 It seems that peer DN values may contain newlines (and other non-printing
851 characters?) which causes problems in log lines. The DN values are now
852 passed through string_printing() before being added to log lines.
854 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
855 and InterBase are left for another time.)
857 PH/21 Added message_body_newlines option.
859 PH/22 Guard against possible overflow in moan_check_errorcopy().
861 PH/23 POSIX allows open() to be a macro; guard against that.
863 PH/24 If the recipient of an error message contained an @ in the local part
864 (suitably quoted, of course), incorrect values were put in $domain and
865 $local_part during the evaluation of errors_copy.
871 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
872 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
875 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
876 issue a MAIL command.
878 PH/02 In an ACL statement such as
880 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
882 if a client was not listed at all, or was listed with a value other than
883 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
884 the condition was not true (as it should be), so access was not denied.
885 The bug was that the ! inversion was incorrectly passed on to the second
886 item. This has been fixed.
888 PH/03 Added additional dnslists conditions == and =& which are different from
889 = and & when the dns lookup returns more than one IP address.
891 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
892 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
894 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
895 FSYNC, which compiles an option called disable_fsync that allows for
896 bypassing fsync(). The documentation is heavily laced with warnings.
898 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
900 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
901 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
902 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
903 including adding "make clean"; (3) Added -fPIC when compiling the test
904 dynamically loaded module, to get rid of a warning.
906 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
907 message fails, move_frozen_messages = true and ignore_bounce_errors_after
908 = 0s. The bug is otherwise harmless.
910 PH/07 There was a bug in the dovecot authenticator such that the value of
911 $auth1 could be overwritten, and so not correctly preserved, after a
912 successful authentication. This usually meant that the value preserved by
913 the server_setid option was incorrect.
915 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
917 PH/09 Installed PCRE release 7.0.
919 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
920 run for batched SMTP input. It is now run at the start of every message
921 in the batch. While fixing this I discovered that the process information
922 (output by running exiwhat) was not always getting set for -bs and -bS
923 input. This is fixed, and it now also says "batched" for BSMTP.
925 PH/11 Added control=no_pipelining.
927 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
928 patch, slightly modified), and move the expansion of helo_data till after
929 the connection is made in the smtp transport (so it can use these
932 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
934 PH/14 Added log_selector = +pid.
936 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
938 PH/16 Add ${if forany and ${if forall.
940 PH/17 Added dsn_from option to vary the From: line in DSNs.
942 PH/18 Flush SMTP output before performing a callout, unless control =
943 no_callout_flush is set.
945 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
946 was true (the default) a successful delivery failed to delete the retry
947 item, thus causing premature timeout of the address. The bug is now
950 PH/20 Added hosts_avoid_pipelining to the smtp transport.
952 PH/21 Long custom messages for fakedefer and fakereject are now split up
953 into multiline reponses in the same way that messages for "deny" and
954 other ACL rejections are.
956 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
957 with slight modification.
959 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
960 draft, changing the syntax and factoring some duplicate code".
962 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
963 for deliveries of the second and subsequent messages over the same SMTP
966 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
967 ${reduce, with only minor "tidies".
969 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
971 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
972 expansion side effects.
974 PH/27 When a message times out after an over-quota error from an Exim-imposed
975 quota, the bounce message says "mailbox is full". This message was not
976 being given when it was a system quota that was exceeded. It now should
979 MH/03 Made $recipients available in local_scan(). local_scan() already has
980 better access to the recipient list through recipients_list[], but
981 $recipients can be useful in postmaster-provided expansion strings.
983 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
984 in the case of a MAIL command with additional options following the
985 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
986 were accidentally chopped off.
988 PH/29 SMTP synchronization checks are implemented when a command is read -
989 there is a check that no more input is waiting when there shouldn't be
990 any. However, for some commands, a delay in an ACL can mean that it is
991 some time before the response is written. In this time, more input might
992 arrive, invalidly. So now there are extra checks after an ACL has run for
993 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
994 pipelining has not been advertised.
996 PH/30 MH's patch to allow iscntrl() characters to be list separators.
998 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
999 returned in the SMTP response when smtp_return_error_details was false.
1000 This has been fixed.
1002 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1003 instead of the C I/O that was originally supplied, because problems were
1004 reported on Solaris.
1006 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1007 Exim which did not show up earlier: it was assuming that a call to
1008 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1009 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1010 was a macro that became an assignment, so it seemed to work. This has
1011 changed to a proper function call with a void return, hence the compile
1012 error. Exim's code has been fixed.
1014 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1017 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1019 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1021 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1022 because it turns out to be too big or there is a timeout) there is no
1023 "Completed" line in the log. When some messages of this type were
1024 selected by exigrep, they were listed as "not completed". Others were
1025 picked up by some special patterns. I have improved the selection
1026 criteria to be more general.
1028 PH/38 The host_find_failed option in the manualroute router can now be set
1029 to "ignore", to completely ignore a host whose IP address cannot be
1030 found. If all hosts are ignored, the behaviour is controlled by the new
1031 host_all_ignored option.
1033 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1034 homing or because of multiple MX records with /mx) generated more than
1035 one IP address, and the following item turned out to be the local host,
1036 all the secondary addresses of the first item were incorrectly removed
1037 from the list, along with the local host and any following hosts (which
1038 is what is supposed to happen).
1040 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1041 whoever called Exim into the -H file. In the case of the daemon it was
1042 behaving confusingly. When first started, it used values for whoever
1043 started the daemon, but after a SIGHUP it used the Exim user (because it
1044 calls itself on a restart). I have changed the code so that it now always
1047 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1048 message are rejected with the same error (e.g. no authentication or bad
1049 sender address), and a DATA command is nevertheless sent (as can happen
1050 with PIPELINING or a stupid MUA), the error message that was given to the
1051 RCPT commands is included in the rejection of the DATA command. This is
1052 intended to be helpful for MUAs that show only the final error to their
1055 PH/42 Another patch from the Sieve maintainer.
1057 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1058 Eximstats - Fixed some broken HTML links and added missing column headers
1060 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1061 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1063 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1065 PH/43 Yet another patch from the Sieve maintainer.
1067 PH/44 I found a way to check for a TCP/IP connection going away before sending
1068 the response to the final '.' that terminates a message, but only in the
1069 case where the client has not sent further data following the '.'
1070 (unfortunately, this is allowed). However, in many cases there won't be
1071 any further data because there won't be any more messages to send. A call
1072 to select() can be used: if it shows that the input is "ready", there is
1073 either input waiting, or the socket has been closed. An attempt to read
1074 the next input character can distinguish the two cases. Previously, Exim
1075 would have sent an OK response which the client would never have see.
1076 This could lead to message repetition. This fix should cure that, at
1077 least in a lot of common cases.
1079 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1080 advertised in response to EHLO.
1086 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1087 fixed by 4.65/MH/01 (is this a record?) are fixed:
1089 (i) An empty string was always treated as zero by the numeric comparison
1090 operators. This behaviour has been restored.
1092 (ii) It is documented that the numeric comparison operators always treat
1093 their arguments as decimal numbers. This was broken in that numbers
1094 starting with 0 were being interpreted as octal.
1096 While fixing these problems I realized that there was another issue that
1097 hadn't been noticed. Values of message_size_limit (both the global option
1098 and the transport option) were treated as octal if they started with 0.
1099 The documentation was vague. These values are now always treated as
1100 decimal, and I will make that clear in the documentation.
1106 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1107 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1110 MH/01 Don't check that the operands of numeric comparison operators are
1111 integers when their expansion is in "skipping" mode (fixes bug
1112 introduced by 4.64-PH/07).
1114 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1115 child addresses, Exim now panics and dies. Previously, because the count
1116 is held in a short int, deliveries were likely to be lost. As such a
1117 large number of recipients for a single message is ridiculous
1118 (performance will be very, very poor), I have chosen to impose a limit
1119 rather than extend the field.
1125 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1126 leftover -K file (the existence of which was triggered by #402).
1127 While we were at it, introduced process PID as part of the -K
1128 filename. This should rule out race conditions when creating
1131 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1132 processing considerably. Previous code took too long for large mails,
1133 triggering a timeout which in turn triggers #401.
1135 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1136 in the DK code in transports.c. sendfile() is not really portable,
1137 hence the _LINUX specificness.
1139 TF/01 In the add_headers option to the mail command in an Exim filter,
1140 there was a bug that Exim would claim a syntax error in any
1141 header after the first one which had an odd number of characters
1144 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1145 callout verification, Exim cached a "reject" for the entire domain. This
1146 is correct for most verifications, but it is not correct for a recipient
1147 verification with use_sender or use_postmaster set, because in that case
1148 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1149 case of MAIL FROM:<> rejection from other early rejections (e.g.
1150 rejection of HELO). When verifying a recipient using a non-null MAIL
1151 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1152 Whatever the result of the callout, the value of the domain cache is
1153 left unchanged (for any other kind of callout, getting as far as trying
1154 RCPT means that the domain itself is ok).
1156 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1159 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1160 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1161 session, the daemon ignores ECONNECT errors and logs others; it now
1162 ignores EPIPE as well.
1164 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1165 (quoted-printable decoding).
1167 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1168 later the small subsequent patch to fix an introduced bug.
1170 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1172 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1174 PH/08 An error is now given if message_size_limit is specified negative.
1176 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1177 to be given (somewhat) arbitrary names.
1179 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1182 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1183 miscellaneous code fixes
1185 PH/10 Added the log_reject_target ACL modifier to specify where to log
1188 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1189 hostname. This is wrong, because it relates to the incoming message (and
1190 probably the interface on which it is arriving) and not to the outgoing
1191 callout (which could be using a different interface). This has been
1192 changed to use the value of the helo_data option from the smtp transport
1193 instead - this is what is used when a message is actually being sent. If
1194 there is no remote transport (possible with a router that sets up host
1195 addresses), $smtp_active_hostname is used.
1197 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1198 tweaks were necessary in order to get it to work (see also 21 below):
1199 (a) The code assumed that strncpy() returns a negative number on buffer
1200 overflow, which isn't the case. Replaced with Exim's string_format()
1202 (b) There were several signed/unsigned issues. I just did the minimum
1203 hacking in of casts. There is scope for a larger refactoring.
1204 (c) The code used strcasecmp() which is not a standard C function.
1205 Replaced with Exim's strcmpic() function.
1206 (d) The code set only $1; it now sets $auth1 as well.
1207 (e) A simple test gave the error "authentication client didn't specify
1208 service in request". It would seem that Dovecot has changed its
1209 interface. Fortunately there's a specification; I followed it and
1210 changed what the client sends and it appears to be working now.
1212 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1215 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1216 address A is aliased to B and C, where B exists and C does not. Without
1217 -v the output is "A verified" because verification stops after a
1218 successful redirection if more than one address is generated. However,
1219 with -v the child addresses are also verified. Exim was outputting "A
1220 failed to verify" and then showing the successful verification for C,
1221 with its parentage. It now outputs "B failed to verify", showing B's
1222 parentage before showing the successful verification of C.
1224 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1225 look up a TXT record in a specific list after matching in a combined
1228 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1229 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1230 they consult the DNS. I had assumed they would set it the way they
1231 wanted; and indeed my experiments on Linux seem to show that in some
1232 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1233 To be on the safe side, however, I have now made the interface to
1234 host_find_byname() similar to host_find_bydns(), with an argument
1235 containing the DNS resolver options. The host_find_byname() function now
1236 sets these options at its start, just as host_find_bydns() does. The smtp
1237 transport options dns_qualify_single and dns_search_parents are passed to
1238 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1239 of host_find_byname() use the default settings of RES_DEFNAMES
1240 (qualify_single) but not RES_DNSRCH (search_parents).
1242 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1243 spool_read_header() do less string testing, by means of a preliminary
1244 switch on the second character of optional "-foo" lines. (This is
1245 overdue, caused by the large number of possibilities that now exist.
1246 Originally there were few.) While I was there, I also converted the
1247 str(n)cmp tests so they don't re-test the leading "-" and the first
1248 character, in the hope this might squeeze out yet more improvement.
1250 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1251 flag allowing group syntax was set by the header_syntax check but not
1252 turned off, possible causing trouble later; (2) The flag was not being
1253 set at all for the header_verify test, causing "group"-style headers to
1254 be rejected. I have now set it in this case, and also caused header_
1255 verify to ignore an empty address taken from a group. While doing this, I
1256 came across some other cases where the code for allowing group syntax
1257 while scanning a header line wasn't quite right (mostly, not resetting
1258 the flag correctly in the right place). These bugs could have caused
1259 trouble for malformed header lines. I hope it is now all correct.
1261 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1262 with the "reply" argument non-NULL. The code, however (which originally
1263 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1264 but it didn't always do it. This confused somebody who was copying the
1265 code for some other use. I have removed all the tests.
1267 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1268 feature that was used to support insecure browsers during the U.S. crypto
1269 embargo. It requires special client support, and Exim is probably the
1270 only MTA that supported it -- and would never use it because real RSA is
1271 always available. This code has been removed, because it had the bad
1272 effect of slowing Exim down by computing (never used) parameters for the
1273 RSA_EXPORT functionality.
1275 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1276 authenticator to fail if there's a tab character in the incoming data
1277 (there should never be unless someone is messing about, as it's supposed
1278 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1279 if the connection is using TLS or if the remote IP is the same as the
1280 local IP, and the "valid-client-cert option" if a client certificate has
1283 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1284 authenticators. This can be used for authorization after authentication
1285 succeeds. (In the case of plaintext, it servers for both authentication
1288 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1289 if any retry times were supplied.
1291 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1292 connection, where there is no client IP address to check. In this
1293 situation, the verify now always succeeds.
1295 PH/25 Applied John Jetmore's -Mset patch.
1297 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1299 PH/27 In a string expansion for a processed (not raw) header when multiple
1300 headers of the same name were present, leading whitespace was being
1301 removed from all of them, but trailing whitespace was being removed only
1302 from the last one. Now trailing whitespace is removed from each header
1303 before concatenation. Completely empty headers in a concatenation (as
1304 before) are ignored.
1306 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1307 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1309 PH/29 [Removed. This was a change that I later backed out, and forgot to
1310 correct the ChangeLog entry (that I had efficiently created) before
1311 committing the later change.]
1313 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1314 address errors (4xx response to RCPT) over the same connection as other
1315 messages routed to the same hosts. Such deliveries are always "forced",
1316 so retry times are not inspected. This resulted in far too many retries
1317 for the affected addresses. The effect occurred only when there were more
1318 hosts than the hosts_max_try setting in the smtp transport when it had
1319 the 4xx errors. Those hosts that it had tried were not added to the list
1320 of hosts for which the message was waiting, so if all were tried, there
1321 was no problem. Two fixes have been applied:
1323 (i) If there are any address or message errors in an SMTP delivery, none
1324 of the hosts (tried or untried) are now added to the list of hosts
1325 for which the message is waiting, so the message should not be a
1326 candidate for sending over the same connection that was used for a
1327 successful delivery of some other message. This seems entirely
1328 reasonable: after all the message is NOT "waiting for some host".
1329 This is so "obvious" that I'm not sure why it wasn't done
1330 previously. Hope I haven't missed anything, but it can't do any
1331 harm, as the worst effect is to miss an optimization.
1333 (ii) If, despite (i), such a delivery is accidentally attempted, the
1334 routing retry time is respected, so at least it doesn't keep
1335 hammering the server.
1337 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1338 in ${readsocket because some servers need this prod.
1340 PH/32 Added some extra debug output when updating a wait-xxx database.
1342 PH/33 The hint "could be header name not terminated by colon", which has been
1343 given for certain expansion errors for a long time, was not being given
1344 for the ${if def:h_colon_omitted{... case.
1346 PH/34 The spec says: "With one important exception, whenever a domain list is
1347 being scanned, $domain contains the subject domain." There was at least
1348 one case where this was not true.
1350 PH/35 The error "getsockname() failed: connection reset by peer" was being
1351 written to the panic log as well as the main log, but it isn't really
1352 panic-worthy as it just means the connection died rather early on. I have
1353 removed the panic log writing for the ECONNRESET error when getsockname()
1356 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1357 runs only) independently of the message's sender address. This meant
1358 that, if the 4xx error was in fact related to the sender, a different
1359 message to the same recipient with a different sender could confuse
1360 things. In particualar, this can happen when sending to a greylisting
1361 server, but other circumstances could also provoke similar problems.
1362 I have changed the default so that the retry time for these errors is now
1363 based a combination of the sender and recipient addresses. This change
1364 can be overridden by setting address_retry_include_sender=false in the
1367 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1368 remote server are returned as part of bounce messages. This was not
1369 happening for LMTP over a pipe (the lmtp transport), but now it is the
1370 same for both kinds of LMTP.
1372 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1373 in header lines that were in fact CNAMEs. This is no longer the case.
1375 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1376 and queue runs started by the daemon processed all messages. This has
1377 been fixed so that -R and -S can now usefully be given with -q<time>.
1379 PH/40 Import PCRE release 6.7 (fixes some bugs).
1381 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1383 PH/42 Give an error if -q is specified more than once.
1385 PH/43 Renamed the variables $interface_address and $interface_port as
1386 $received_ip_address and $received_port, to make it clear that these
1387 values apply to message reception, and not to the outgoing interface when
1388 a message is delivered. (The old names remain recognized, of course.)
1390 PH/44 There was no timeout on the connect() call when using a Unix domain
1391 socket in the ${readsocket expansion. There now is.
1393 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1394 be meaningful with "accept".
1396 SC/01 Eximstats V1.43
1397 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1399 SC/02 Eximstats V1.44
1400 Use a glob alias rather than an array ref in the generated
1401 parser. This improves both readability and performance.
1403 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1404 Collect SpamAssassin and rejection statistics.
1405 Don't display local sender or destination tables unless
1406 there is data to show.
1407 Added average volumes into the top table text output.
1409 SC/04 Eximstats V1.46
1410 Collect data on the number of addresses (recipients)
1411 as well as the number of messages.
1413 SC/05 Eximstats V1.47
1414 Added 'Message too big' to the list of mail rejection
1415 reasons (thanks to Marco Gaiarin).
1417 SC/06 Eximstats V1.48
1418 Mainlog lines which have GMT offsets and are too short to
1419 have a flag are now skipped.
1421 SC/07 Eximstats V1.49 (Alain Williams)
1422 Added the -emptyok flag.
1424 SC/08 Eximstats V1.50
1425 Fixes for obtaining the IP address from reject messages.
1427 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1428 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1429 whitesspace changes from 4.64-PH/27
1431 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1434 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1435 are found, allow negative numbers in numeric criteria)
1437 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1439 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1440 to match changes made in 4.64-PH/43
1442 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1444 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1445 "domain" is actually an IP address, and force a failure. This locks out
1446 those revolvers/nameservers that support "A-for-A" lookups, in
1447 contravention of the specifications.
1449 PH/48 When a host name was looked up from an IP address, and the subsequent
1450 forward lookup of the name timed out, the host name was left in
1451 $sender_host_name, contrary to the specification.
1453 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1454 restricted to single-key lookups, Exim was not diagnosing an error if
1455 * or *@ was used with a query-style lookup.
1457 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1459 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1460 long ago, but noone interested enough thought of it. Let's just say that
1461 the "1.1" means that there are some new functions that weren't there at
1462 some point in the past.
1464 PH/51 Error processing for expansion failure of helo_data from an smtp
1465 transport during callout processing was broken.
1467 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1468 tested/used via the -bh/-bhc/-bs options.
1470 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1471 bug, fixed in subsequent PCRE releases).
1473 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1474 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1476 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1482 SC/01 Use a glob alias rather than an array ref in eximstats generated
1483 parser. This improves both readability and performance.
1485 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1486 Don't display local sender or destination tables in eximstats unless
1487 there is data to show.
1488 Added average volumes into the eximstats top table text output.
1490 SC/03 Collect data on the number of addresses (recipients) as well
1491 as the number of messages in eximstats.
1493 TF/01 Correct an error in the documentation for the redirect router. Exim
1494 does (usually) call initgroups() when daemonizing.
1496 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1497 with consistent privilege compared to when running as a daemon.
1499 TF/03 Note in the spec that $authenticated_id is not set for local
1500 submissions from trusted users.
1502 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1503 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1505 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1506 by adding some example configuration directives to the default
1507 configuration file. A little bit of work is required to uncomment the
1508 directives and define how usernames and passwords are checked, but
1509 there is now a framework to start from.
1511 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1512 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1513 without this. I don't know how relevant this is to other LDAP libraries.
1515 PH/02 Add the verb name to the "unknown ACL verb" error.
1517 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1519 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1521 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1522 directory not expanded when it should be if an expanded home directory
1523 was set for the address (which is overridden by the transport).
1525 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1528 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1529 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1530 because it is too late at that time, and has no effect.
1532 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1533 security issue with \' (bugzilla #107). I could not use the
1534 PQescapeStringConn() function, because it needs a PGconn value as one of
1537 PH/08 When testing addresses using -bt, indicate those final addresses that
1538 are duplicates that would not cause an additional delivery. At least one
1539 person was confused, thinking that -bt output corresponded to deliveries.
1540 (Suppressing duplicates isn't a good idea as you lose the information
1541 about possibly different redirections that led to the duplicates.)
1543 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1544 systems where poll() doesn't work, in particular OS X.
1546 PH/10 Added more information to debugging output for retry time not reached.
1548 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1549 operations in malware.c.
1551 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1554 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1555 syslog_duplication set false, log lines that would normally be written
1556 both the the main log and to the reject log were not written to syslog at
1559 PH/14 In the default configuration, change the use of "message" in ACL warn
1560 statements to "add_header".
1562 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1563 not followed by a command (e.g. "seen endif").
1565 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1566 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1569 PH/17 Added extra conditions to the default value of delay_warning_condition
1573 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1574 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1575 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1578 The Auto-Submitted: and various List- headers are standardised, whereas I
1579 don't think Precedence: ever was.
1581 PH/18 Refactored debugging code in route_finduser() to show more information,
1582 in particular, the error code if getpwnam() issues one.
1584 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1585 This is apparently needed in addition to the PH/07 change above to avoid
1586 any possible encoding problems.
1588 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1589 but not after initializing Perl.
1591 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1592 output them only if debugging. By default they are written stderr,
1593 apparently, which is not desirable.
1595 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1598 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1601 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1603 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1604 authenticated or an ident call has been made. Suppress the default
1605 values for $authenticated_id and $authenticated_sender (but permit -oMai
1606 and -oMas) when testing with -bh.
1608 PH/24 Re-jigged the order of the tests in the default configuration so that the
1609 tests for valid domains and recipients precede the DNS black list and CSA
1610 tests, on the grounds that those ones are more expensive.
1612 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1613 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1614 This bug exists in every version of Exim that I still have, right back to
1617 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1618 However, an attempt to turn on case-sensitivity in a regex key by
1619 including (?-i) didn't work because the subject string was already
1620 lowercased, and the effects were non-intuitive. It turns out that a
1621 one-line patch can be used to allow (?-i) to work as expected.
1627 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1628 other effects) broke the use of negated acl sub-conditions.
1630 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1633 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1634 "Deny" causes Exim to reject the incoming connection with a 554 error.
1635 Unfortunately, if there is a major crisis, such as a disk failure,
1636 tcp-wrappers gives "deny", whereas what one would like would be some
1637 kind of temporary error. A kludge has been added to help with this.
1638 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1639 554 error is used if errno is still zero or contains ENOENT (which occurs
1640 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1643 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1645 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1646 errors. Otherwise a message that provokes a temporary error (when other
1647 messages do not) can cause a whole host to time out.
1649 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1650 addresses were routed directly to files or pipes from a redirect router.
1651 File deliveries just didn't batch; pipe deliveries might have suffered
1654 PH/06 A failure to get a lock for a hints database would erroneously always say
1655 "Failed to get write lock", even when it was really a read lock.
1657 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1658 of 0600. This has been changed to use the value of the lockfile_mode
1659 option (which defaults to 0600).
1661 PH/08 Applied small patch from the Sieve maintainer.
1663 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1664 folder from quota calculations, a direct delivery into this folder messed
1665 up the contents of the maildirsize file. This was because the regex was
1666 used only to exclude .Trash (or whatever) when the size of the mailbox
1667 was calculated. There was no check that a delivery was happening into an
1668 excluded directory. This bug has been fixed by ignoring all quota
1669 processing for deliveries into excluded directories.
1671 PH/10 Added the maildirfolder_create_regex option to appendfile.
1677 PH/01 The code for finding all the local interface addresses on a FreeBSD
1678 system running IPv6 was broken. This may well have applied to all BSD
1679 systems, as well as to others that have similar system calls. The broken
1680 code found IPv4 interfaces correctly, but gave incorrect values for the
1681 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1682 that it would not match correctly against @[] and not recognize the IPv6
1685 PH/02 The ipliteral router was not recognizing addresses of the form user@
1686 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1688 PH/03 Added disable_ipv6.
1690 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1691 decoded form, because it is most often used to construct To: headers
1692 lines in autoreplies, and the decoded form may well be syntactically
1693 invalid. However, $reply_address has leading white space removed, and all
1694 newlines turned into spaces so that the autoreply transport does not
1697 PH/05 If group was specified without a user on a router, and no group or user
1698 was specified on a transport, the group from the router was ignored.
1700 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1701 for visible compile-time settings that can be used to change these
1702 numbers, for those that want even more. Backwards compatibility with old
1703 spool files has been maintained. However, going back to a previous Exim
1704 release will lost any variables that are in spool files.
1706 PH/07 Two small changes when running in the test harness: increase delay when
1707 passing a TCP/IP connection to a new process, in case the original
1708 process has to generate a bounce, and remove special handling of
1709 127.0.0.2 (sic), which is no longer necessary.
1711 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1712 be the same on different OS.
1714 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1717 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1718 whether --show-vars was specified or not
1720 JJ/02 exipick: Added support for new ACL variable spool format introduced
1723 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1724 syntactically invalid From: or Reply-to: line, and a filter used this to
1725 generate an autoreply, and therefore failed to obtain an address for the
1726 autoreply, Exim could try to deliver to a non-existent relative file
1727 name, causing unrelated and misleading errors. What now happens is that
1728 it logs this as a hard delivery error, but does not attempt to create a
1731 PH/11 The exinext utility has a -C option for testing purposes, but although
1732 the given file was scanned by exinext itself; it wasn't being passed on
1733 when Exim was called.
1735 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1736 an end-of-file indication when reading a command response.
1738 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1739 compiled. In many other places in Exim, IPv6 addresses are always
1740 recognized, so I have changed this. It also means that IPv4 domain
1741 literals of the form [IPV4:n.n.n.n] are now always recognized.
1743 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1744 used if the router is not running as root, for example, when verifying at
1745 ACL time, or when using -bh. The debugging output from this situation was
1746 non-existent - all you got was a failure to exec. I have made two
1749 (a) Failures to set uid/gid, the current directory, or a process leader
1750 in a subprocess such as that created by queryprogram now generate
1751 suitable debugging ouput when -d is set.
1753 (b) The queryprogram router detects when it is not running as root,
1754 outputs suitable debugging information if -d is set, and then runs
1755 the subprocess without attempting to change uid/gid.
1757 PH/15 Minor change to Makefile for building test_host (undocumented testing
1760 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1761 additional section of a DNS packet that returns MX or SRV records.
1762 Instead, it always explicitly searches for A/AAAA records. This avoids
1763 major problems that occur when a DNS server includes only records of one
1764 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1765 fixed another bug: if SRV records were looked up and the corresponding
1766 address records were *not* found in the additional section, the port
1767 values from the SRV records were lost.
1769 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1770 using the correct key (the original address) when searching the retry
1771 rules in order to find which one to use for generating the retry hint.
1773 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1774 adding the default one. Similarly, if it contains a Reply-To: header, the
1775 errors_reply_to option, if set, is not used.
1777 PH/19 When calculating a retry time, Exim used to measure the "time since
1778 failure" by looking at the "first failed" field in the retry record. Now
1779 it does not use this if it is later than than the arrival time of the
1780 message. Instead it uses the arrival time. This makes for better
1781 behaviour in cases where some deliveries succeed, thus re-setting the
1782 "first failed" field. An example is a quota failure for a huge message
1783 when small messages continue to be delivered. Without this change, the
1784 "time since failure" will always be short, possible causing more frequent
1785 delivery attempts for the huge message than are intended.
1786 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1788 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1789 $1, $2, $3) because the numerical variables can be reset during some
1790 expansion items (e.g. "match"), thereby losing the authentication data.
1792 PH/21 Make -bV show the size of off_t variables so that the test suite can
1793 decide whether to run tests for quotas > 2G.
1795 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1796 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1797 filecount value is greater than 2G or if a quota value is greater than 2G
1798 on a system where the size of off_t is not greater than 4, a panic error
1801 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1802 never match. The debug and -bh output now contains an explicit error
1803 message indicating a malformed IPv4 address or mask.
1805 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1806 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1807 PH/23 above applies.
1809 PH/25 Do not write to syslog when running in the test harness. The only
1810 occasion when this arises is a failure to open the main or panic logs
1811 (for which there is an explicit test).
1813 PH/26 Added the /no_tell option to "control=freeze".
1815 PH/27 If a host name lookup failed very early in a connection, for example, if
1816 the IP address matched host_lookup and the reverse lookup yielded a name
1817 that did not have a forward lookup, an error message of the form "no IP
1818 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1819 could be logged. Now it outputs the IP address instead of "NULL".
1821 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1822 allows the sender and the authenticated sender to be set when
1823 submitting a message from within Exim. Since child_open_exim() is
1824 documented for local_scan(), the new function should be too.
1826 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1827 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1828 results in an empty string is now treated as unset.
1830 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1832 PH/31 Added sender_verify_fail logging option.
1834 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1835 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1836 tidied the source and removed it altogether.
1838 PH/33 When a queue run was abandoned because the load average was too high, a
1839 log line was always written; now it is written only if the queue_run log
1840 selector is set. In addition, the log line for abandonment now contains
1841 information about the queue run such as the pid. This is always present
1842 in "start" and "stop" lines but was omitted from the "abandon" line.
1844 PH/34 Omit spaces between a header name and the colon in the error message that
1845 is given when verify = headers_syntax fails (if there are lots of them,
1846 the message gets confusing).
1848 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1849 names, as there are now some PTR records that contain slashes. This check
1850 is only to protect against broken name servers that fall over on strange
1851 characters, so the fact that it applies to all lookups doesn't matter.
1853 PH/36 Now that the new test suite is complete, we can remove some of the
1854 special code in Exim that was needed for the old test suite. For example,
1855 sorting DNS records because real resolvers return them in an arbitrary
1856 order. The new test suite's fake resolver always returns records in the
1859 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1860 bounces) except when queue_only is set, to avoid logging races between
1861 the different processes.
1863 PH/38 Panic-die if .include specifies a non-absolute path.
1865 PH/39 A tweak to the "H" retry rule from its user.
1867 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1868 a label. They prevented compilation on older perls.
1870 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1871 a warning to be raised on newish perls.
1873 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1874 on queue. Changes to match documented behaviour of showing count of
1875 messages matching specified criteria.
1877 PH/40 Changed the default ident timeout from 30s to 5s.
1879 PH/41 Added support for the use of login_cap features, on those BSD systems
1880 that have them, for controlling the resources used by pipe deliveries.
1882 PH/42 The content-scanning code uses fopen() to create files in which to put
1883 message data. Previously it was not paying any attention to the mode of
1884 the files. Exim runs with umask(0) because the rest of the code creates
1885 files with open(), and sets the required mode explicitly. Thus, these
1886 files were ending up world-writeable. This was not a big issue, because,
1887 being within the spool directory, they were not world-accessible. I have
1888 created a function called modefopen, which takes an additional mode
1889 argument. It sets umask(777), creates the file, chmods it to the required
1890 mode, then resets the umask. All the relevant calls to fopen() in the
1891 content scanning code have been changed to use this function.
1893 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1894 to 24 hours. This avoids potential overflow problems when processing G
1895 and H retry rules. I suspect nobody ever tinkers with this value.
1897 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1899 PH/45 When the plaintext authenticator is running as a client, the server's
1900 challenges are checked to ensure they are valid base64 strings. By
1901 default, the authentication attempt is cancelled if an invalid string is
1902 received. Setting client_ignore_invalid_base64 true ignores these errors.
1903 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1904 they are received. Thus, the responses can be made to depend on the
1905 challenges. If an invalid string is ignored, an empty string is placed in
1908 PH/46 Messages that are created by the autoreply transport now contains a
1909 References: header, in accordance with RFCs 2822 and 3834.
1911 PH/47 Added authenticated_sender_force to the smtp transport.
1913 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1915 PH/49 Installed latest patch from the Sieve maintainer.
1917 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1918 was also set, the appendfile transport was unnecessarily scanning a
1919 directory of message files (e.g. for maildir delivery) to find the count
1920 of files (along with the size), even though it did not need this
1921 information. It now does the scan only if it needs to find either the
1922 size of the count of files.
1924 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1926 PH/52 Two bugs concerned with error handling when the smtp transport is
1929 (i) Exim was not creating retry information for temporary errors given
1930 for individual recipients after the DATA command when the smtp transport
1931 was used in LMTP mode. This meant that they could be retried too
1932 frequently, and not timed out correctly.
1934 (ii) Exim was setting the flag that allows error details to be returned
1935 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1936 recipients that were returned after the DATA command.
1938 PH/53 This is related to PH/52, but is more general: for any failing address,
1939 when detailed error information was permitted to be returned to the
1940 sender, but the error was temporary, then after the final timeout, only
1941 "retry timeout exceeded" was returned. Now it returns the full error as
1942 well as "retry timeout exceeded".
1944 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1945 do this, and (what is worse) MTAs that accept it.
1947 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1948 will now be deprecated.
1950 PH/56 New os.c-cygwin from the Cygwin maintainer.
1952 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1953 formats (previously only available in exim formats via -bpr, -bpru,
1954 and -bpra. Now also available in native and exiqgrep formats)
1956 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1957 with very large, slow to parse queues
1959 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1961 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1963 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1964 responses to authentication challenges, though it was showing the
1965 challenges; (ii) I've removed the CR characters from the debug output for
1968 PH/58 Allow for the insertion of a newline as well as a space when a string
1969 is turned into more than one encoded-word during RFC 2047 encoding. The
1970 Sieve code now uses this.
1972 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1973 data_4xx, lost_connection, tls_required.
1975 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1976 message was being sent as an SMTP response.
1978 PH/61 Add -l and -k options to exicyclog.
1980 PH/62 When verifying, if an address was redirected to one new address, so that
1981 verification continued, and the new address failed or deferred after
1982 having set something in $address_data, the value of $address_data was not
1983 passed back to the ACL. This was different to the case when no
1984 redirection occurred. The value is now passed back in both cases.
1986 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1987 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1988 use login_cap.h, so on its own it isn't the distinguishing feature. The
1989 new name refers directly to the setclassresources() function.
1991 PH/65 Added configuration files for NetBSD3.
1993 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1995 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1996 is preferred over IPv4.
1998 PH/68 The bounce_return_message and bounce_return_body options were not being
1999 honoured for bounces generated during the reception of non-SMTP messages.
2000 In particular, this applied to messages rejected by the ACL. This bug has
2001 been fixed. However, if bounce_return_message is true and bounce_return_
2002 body is false, the headers that are returned for a non-SMTP message
2003 include only those that have been read before the error was detected.
2004 (In the case of an ACL rejection, they have all been read.)
2006 PH/69 The HTML version of the specification is now built in a directory called
2007 spec_html instead of spec.html, because the latter looks like a path with
2008 a MIME-type, and this confuses some software.
2010 PH/70 Catch two compiler warnings in sieve.c.
2012 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2013 function verify_get_ident() calls ip_connect() to connect a socket, but
2014 if the "connect()" function timed out, ip_connect() used to close the
2015 socket. However, verify_get_ident() also closes the socket later, and in
2016 between Exim writes to the log, which may get opened at this point. When
2017 the socket was closed in ip_connect(), the log could get the same file
2018 descriptor number as the socket. This naturally causes chaos. The fix is
2019 not to close the socket in ip_connect(); the socket should be closed by
2020 the function that creates it. There was only one place in the code where
2021 this was missing, in the iplookup router, which I don't think anybody now
2022 uses, but I've fixed it anyway.
2024 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2025 well as to direct DNS lookups. Otherwise the handling of names in host
2026 lists is inconsistent and therefore confusing.
2032 PH/01 Two changes to the default runtime configuration:
2034 (1) Move the checks for relay_from_hosts and authenticated clients from
2035 after to before the (commented out) DNS black list checks.
2037 (2) Add control=submission to the relay_from_hosts and authenticated
2038 clients checks, on the grounds that messages accepted by these
2039 statements are most likely to be submissions.
2041 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2043 (1) Generate an error if the third argument for the ${prvs expansion is
2046 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2049 (3) Reset the variables that are obtained from the first argument of
2050 ${prvscheck and used in the second argument before leaving the code,
2051 because their memory is reclaimed, so using them afterwards may do
2054 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2055 one (it's much easier than Tom thought :-).
2057 (5) Because of (4), we can now allow for the use of $prvscheck_result
2058 inside the third argument.
2060 PH/03 For some reason, the default setting of PATH when running a command from
2061 a pipe transport was just "/usr/bin". I have changed it to
2064 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2065 anything to be listed in the output from -bV.
2067 PH/05 When a filter generated an autoreply, the entire To: header line was
2068 quoted in the delivery log line, like this:
2070 => >A.N.Other <ano@some.domain> <original@ddress> ...
2072 This has been changed so that it extracts the operative address. There
2073 may be more than one such address. If so, they are comma-separated, like
2076 => >ano@some.domain,ona@other.domain <original@ddress> ...
2078 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2079 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2080 not being looked up in the rDNS to get a host name, Exim was showing the
2081 IP address twice in Received: lines, even though the IP addresses were
2082 identical. For example:
2084 Received: from [1.2.3.4] (helo=[1.2.3.4])
2086 However, if the real host name was known, it was omitting the HELO data
2087 if it matched the actual IP address. This has been tidied up so that it
2088 doesn't show the same IP address twice.
2090 PH/07 When both +timestamp and +memory debugging was on, the value given by
2091 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2092 called by the debug printing, thereby overwriting the timestamp buffer.
2093 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2095 PH/08 When the original message was included in an autoreply transport, it
2096 always said "this is a copy of the message, including all the headers",
2097 even if body_only or headers_only was set. It now gives an appropriate
2100 PH/09 Applied a patch from the Sieve maintainer which:
2102 o fixes some comments
2103 o adds the (disabled) notify extension core
2104 o adds some debug output for the result of if/elsif tests
2105 o points to the current vacation draft in the documentation
2106 and documents the missing references header update
2110 o fixes a bug in processing the envelope test (when testing
2111 multiple envelope elements, the last element determinted the
2114 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2115 Electronic Mail") by including:
2117 Auto-submitted: auto-generated
2119 in the messages that it generates (bounce messages and others, such as
2120 warnings). In the case of bounce messages for non-SMTP mesages, there was
2121 also a typo: it was using "Auto_submitted" (underscore instead of
2122 hyphen). Since every message generated by Exim is necessarily in response
2123 to another message, thes have all been changed to:
2125 Auto-Submitted: auto-replied
2127 in accordance with these statements in the RFC:
2129 The auto-replied keyword:
2131 - SHOULD be used on messages sent in direct response to another
2132 message by an automatic process,
2134 - MUST NOT be used on manually-generated messages,
2136 - MAY be used on Delivery Status Notifications (DSNs) and Message
2137 Disposition Notifications (MDNs),
2139 - MUST NOT be used on messages generated by automatic or periodic
2140 processes, except for messages which are automatic responses to
2143 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2144 to the default Received: header definition.
2146 PH/12 Added log selector acl_warn_skipped (default on).
2148 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2149 variables because (a) they are in the wrong storage pool and (b) even if
2150 they were copied, it wouldn't work properly because of the caching.
2152 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2153 checking when decoding. Apparently there are clients that generate
2154 overlong encoded strings. Why am I not surprised?
2156 PH/15 If the first argument of "${if match_address" was not empty, but did not
2157 contain an "@" character, Exim crashed. Now it writes a panic log message
2158 and treats the condition as false.
2160 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2162 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2163 extension "envelope-auth". The code is finished and in agreement with
2164 other implementations, but there is no documentation so far and in fact,
2165 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2166 not changing the active code.
2168 Print executed "if" and "elsif" statements when debugging is used. This
2169 helps a great deal to understand what a filter does.
2171 Document more things not specified clearly in RFC3028. I had all this
2172 sorted out, when out of a sudden new issues came to my mind. Oops."
2174 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2177 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2178 canonical form (as documented). However, after a host name lookup from
2179 the IP address, check_host() was doing a simple string comparison with
2180 addresses acquired from the DNS when checking that the found name did
2181 have the original IP as one of its addresses. Since any found IPv6
2182 addresses are likely to be in abbreviated form, the comparison could
2183 fail. Luckily, there already exists a function for doing the comparison
2184 by converting both addresses to binary, so now that is used instead of
2185 the text comparison.
2187 PH/20 There was another similar case to PH/19, when a complete host name was
2188 given in a host list; looking up its IP address could give an abbreviated
2189 form, whereas the current host's name might or might not be abbreviated.
2190 The same fix has been applied.
2196 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2197 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2200 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2201 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2203 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2205 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2206 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2207 Also, for "Auto-Submitted", treat anything other than "no" as
2208 non-personal, in accordance with RFC 3834. (Previously it treated
2209 anything starting "auto-" as non-personal.)
2211 TF/01 The control=submission/name=... option had a problem with syntax
2212 errors if the name included a slash character. The /name= option
2213 now slurps the rest of the string, so it can include any characters
2214 but it must come last in the list of options (after /sender_retain
2217 PH/05 Some modifications to the interface to the fake nameserver for the new
2225 TK/01 Added the "success_on_redirect" address verification option. See
2226 NewStuff for rationale and an example.
2228 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2230 PH/02 Patch to exigrep to allow it to work on syslog lines.
2232 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2233 fread() to skip over the body file's header line, because in Cygwin the
2234 header line is locked and is inaccessible.
2236 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2237 co-exist for some time) to make it clear that it is the Exim ID that is
2238 referenced, not the Message-ID: header line.
2240 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2241 string_format() function, because snprintf() does not exist on all
2244 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2245 ${stat: expansion item.
2247 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2248 protocol synchronization error", to keep the pedants happy.
2250 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2251 well as for IRIX systems, when gcc is being used. See the host.c source
2254 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2256 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2259 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2260 transport and to the smtp transport in LMTP mode.
2262 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2264 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2265 run a filter in a subprocess. This could lead to confusion in subsequent
2266 lookups in the parent process. There should also be a search_tidyup() at
2267 the end of the subprocess.
2269 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2270 only if the host matched helo_try_verify_hosts, which caused the
2271 verification to occur when the EHLO/HELO command was issued. The ACL just
2272 tested the remembered result. Now, if a previous verification attempt has
2273 not happened, "verify = helo" does it there and then.
2275 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2277 TK/03 Fix log output including CR from clamd.
2279 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2280 exist provoked a memory error which could cause a segfault.
2282 PH/15 Installed PCRE 6.2
2284 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2286 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2287 of the problem. Specifically, suggested +O2 rather than +O1 for the
2290 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2292 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2293 record, the port was not correctly logged when the outgoing_port log
2294 selector was set (it logged the transort's default port).
2296 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2297 fallback_hosts, and "hosts" in the smtp transport.
2299 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2300 host errors such as "Connection refused".
2302 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2303 authentication with radiusclient 0.4.9:
2305 - Error returned from rc_read_config was caught wrongly
2306 - Username/password not passed on to radius server due to wrong length.
2308 The presumption is that some radiusclient API changes for 4.51/PH/17
2309 were not taken care of correctly. The code is still untested by me (my
2310 Linux distribution still has 0.3.2 of radiusclient), but it was
2311 contributed by a Radius user.
2313 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2314 expanding the "port" option of the smtp transport.
2316 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2317 while reading a MIME header. Thanks to Tom Hughes for a patch.
2319 PH/24 Include config.h inside local_scan.h so that configuration settings are
2322 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2323 that in an ACL for RCPT (for example), you can examine exactly what was
2326 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2327 commands, but it was not correctly comparing the address with the actual
2328 client host address. Thus, it would show the EHLO address in Received:
2329 header lines when this was not necessary.
2331 PH/27 Added the % operator to ${eval:}.
2333 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2334 it should be ignoring failures (because with -C, for example, it has lost
2335 privilege). It wasn't ignoring creation failures other than "already
2338 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2341 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2342 because an input line was too long, either on its own, or by virtue of
2343 too many continuations, the temporary file was not being removed, and the
2344 return code was incorrect.
2346 PH/31 Missing "BOOL" in function definition in filtertest.c.
2348 PH/32 Applied Sieve patches from the maintainer.
2350 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2352 PH/33 Added "verify = not_blind".
2354 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2355 Local/Makefile (with some defaults set). These are used in built scripts
2356 such as exicyclog, but they have never been used in the exim_install
2357 script (though there are many overriding facilities there). I have
2358 arranged that the exim_install script now takes note of these two
2361 PH/35 Installed configuration files for Dragonfly.
2363 PH/36 When a locally submitted message by a trusted user did not contain a
2364 From: header, and the sender address was obtained from -f or from an SMTP
2365 MAIL command, and the trusted user did not use -F to supply a sender
2366 name, $originator_name was incorrectly used when constructing a From:
2367 header. Furthermore, $originator_name was used for submission mode
2368 messages from external hosts without From: headers in a similar way,
2369 which is clearly wrong.
2371 PH/37 Added control=suppress_local_fixups.
2373 PH/38 When log_selector = +received_sender was set, and the addition of the
2374 sender made the log line's construction buffer exactly full, or one byte
2375 less than full, an overflow happened when the terminating "\n" was
2378 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2379 when the result of a list match is failure because a DNS lookup failed.
2381 PH/40 RM_COMMAND is now used in the building process.
2383 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2384 the "build-* directories that it finds.
2386 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2387 address in a domain literal was a prefix of an interface address.
2389 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2390 when verifying a sender address, unless rewrite_headers is false.
2392 PH/44 Wrote a long comment about why errors_to addresses are verified as
2393 recipients, not senders.
2395 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2396 the ratelimit ACL was added.
2398 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2400 PH/46 Added extra information about PostgreSQL errors to the error string.
2402 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2403 avoiding the need to install special zones in a real server. This is
2404 backwards compatible; if it can't find the fake resolver, it drops back.
2405 Thus, both old and new test suites can be run.
2407 TF/02 Added util/ratelimit.pl
2409 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2410 clock is set back in time.
2412 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2413 Candler <B.Candler@pobox.com>.
2415 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2416 OK for addresses that are the result of redirections.
2418 PH/48 A number of further additions for the benefit of the new test suite,
2419 including a fake gethostbyname() that interfaces to the fake DNS resolver
2422 TF/06 The fix for widen_domains has also been applied to qualify_single and
2423 search_parents which are the other dnslookup options that can cause
2426 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2429 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2431 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2432 service" if it encountered an MX record with an empty target hostname.
2433 The message is now "an MX or SRV record indicated no SMTP service".
2435 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2436 if the DNS of the sending site is misconfigured. This is quite a
2437 common situation. This change restores the behaviour of treating a
2438 helo verification defer as a failure.
2440 PH/51 If self=fail was set on a router, the bounce message did not include the
2441 actual error message.
2447 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2449 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2450 command itself ended in error, the underlying message about the transport
2451 filter timeout was being overwritten with the pipe command error. Now the
2452 underlying error message should be appended to the second error message.
2454 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2456 PH/02 Reduce the amount of output that "make" produces by default. Full output
2457 can still be requested.
2459 PH/03 The warning log line about a condition test deferring for a "warn" verb
2460 was being output only once per connection, rather than after each
2461 occurrence (because it was using the same function as for successful
2462 "warn" verbs). This seems wrong, so I have changed it.
2464 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2465 it should not have, which might have caused a crash in the right
2466 circumstances, but probably never did.
2468 PH/04 Installed a modified version of Tony Finch's patch to make submission
2469 mode fix the return path as well as the Sender: header line, and to
2470 add a /name= option so that you can make the user's friendly name appear
2473 TF/03 Added the control = fakedefer ACL modifier.
2475 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2476 Mark Lowes for thorough testing.
2478 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2480 TK/03 Merged latest SRS patch from Miles Wilton.
2482 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2483 with the definition in sysexits.h (which is #included earlier).
2484 Fortunately, Exim does not actually use EX_OK. The code used to try to
2485 preserve the sysexits.h value, by assumimg that macro definitions were
2486 scanned for macro replacements. I have been disabused of this notion,
2487 so now the code just undefines EX_OK before #including unistd.h.
2489 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2490 in the smtp transport. When a block could not be written in a single
2491 write() function, the timeout was being re-applied to each part-write.
2492 This seems wrong - if the receiver was accepting one byte at a time it
2493 would take for ever. The timeout is now adjusted when this happens. It
2494 doesn't have to be particularly precise.
2496 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2497 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2499 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2500 without a domain if the check to <postmaster@domain> fails.
2502 SC/01 Eximstats: added -xls and the ability to specify output files
2503 (patch written by Frank Heydlauf).
2505 SC/02 Eximstats: use FileHandles for outputing results.
2507 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2509 SC/04 Eximstats: fixed display of large numbers with -nvr option
2511 SC/05 Eximstats: fixed merging of reports with empty tables.
2513 SC/06 Eximstats: added the -include_original_destination flag
2515 SC/07 Eximstats: removed tabs and trailing whitespace.
2517 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2519 TK/06 MBOX spool code: Add real "From " MBOX separator line
2520 so the .eml file is really in mbox format (even though
2521 most programs do not really care). Patch from Alex Miller.
2523 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2524 The latter is generated from $received_to and is only set if the
2525 message has one envelope recipient. SA can use these headers,
2526 obviously out-of-the-box. Patch from Alex Miller.
2528 PH/08 The ${def test on a variable was returning false if the variable's
2529 value was "0", contrary to what the specification has always said!
2530 The result should be true unless the variable is empty.
2532 PH/09 The syntax error of a character other than { following "${if
2533 def:variable_name" (after optional whitespace) was not being diagnosed.
2534 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2535 accidental colon was present, for example, could give incorrect results.
2537 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2538 result is used (not including appendfile, where other changes are about
2541 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2542 This involved changing a lot of size variables from int to off_t. It
2543 should work with maildirs and everything.
2545 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2546 spamd dying while we are connected to it.
2548 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2551 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2552 function for BDB 4.3.
2554 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2556 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2557 the sources for PCRE have been re-arranged and more files are now
2560 PH/15 The code I had for printing potentially long long variables in PH/11
2561 above was not the best (it lost precision). The length of off_t variables
2562 is now inspected at build time, and an appropriate printing format (%ld
2563 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2564 to be "long long int" or "long int". This is needed for the internal
2565 formatting function string_vformat().
2567 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2568 the configuration file to be ":syslog", then the script "guesses" where
2569 the logs files are, rather than using the compiled in default. In our
2570 case the guess is not the same as the compiled default, so the script
2571 suddenly stopped working when I started to use syslog. The patch checks
2572 to see if log_file_path is "". If so, it attempts to read it from exim
2573 with no configuration file to get the compiled in version, before it
2574 falls back to the previous guessing code."
2576 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2577 implementing BATV in an Exim configuration. See NewStuff for the gory
2580 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2581 Makefile that are specific to HP-UX.
2583 PH/18 If the "use_postmaster" option was set for a recipient callout together
2584 with the "random" option, the postmaster address was used as the MAIL
2585 FROM address for the random test, but not for the subsequent recipient
2586 test. It is now used for both.
2588 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2589 patch removes a few documentation additions to RFC 3028, because the
2590 latest draft now contains them. It adds the new en;ascii-case comparator
2591 and a new error check for 8bit text in MIME parts. Comparator and
2592 require names are now matched exactly. I enabled the subaddress
2593 extension, but it is not well tested yet (read: it works for me)."
2595 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2596 rework some of the code of TK/09 above to avoid the hardwired use of
2597 "%lld" and "long long". Replaced the call to snprintf() with a call to
2600 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2601 records point to non-existent hosts", "retry timeout exceeded", and
2602 "retry time not reached for any host after a long failure period".
2604 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2605 experimental DomainKeys support:
2607 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2608 (2) On an error such as an illegally used "control", the wrong name for
2609 the control was given.
2611 These problems did NOT occur unless DomainKeys support was compiled.
2613 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2615 PH/24 Added ${if match_ip condition.
2617 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2618 databases so that it will be absolutely obvious if a crash occurs in the
2619 DB library. This is a regular occurrence (often caused by mis-matched
2622 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2623 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2624 picked up on a user's system that detects such things. There doesn't seem
2625 to be a gcc warning option for this - only an attribute that has to be
2626 put on the function's prototype. It seems that in Fedora Core 4 they have
2627 set this on a number of new functions. No doubt there will be more in due
2630 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2631 specify a transport. However, if an address that was verified by such a
2632 router was the subject of a callout, Exim crashed because it tried to
2633 read the rcpt_include_affixes from the non-existent transport. Now it
2634 just assumes that the setting of that option is false. This bug was
2635 introduced by 4.51/PH/31.
2637 PH/28 Changed -d+all to exclude +memory, because that information is very
2638 rarely of interest, but it makes the output a lot bigger. People tend to
2639 do -d+all out of habit.
2641 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2642 code in os-type was giving problems when libc.so lives in lib64, like on
2645 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2646 aren't the modern standard, and it seems that some systems' include files
2647 don't always have them. Exim was already checking for some of the newer
2648 ones like T_AAAA, and defining it itself. I've added checks for all the
2649 record types that Exim uses.
2651 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2652 not automatically generating a new one, as it is supposed to. This
2653 prevented TLS from working. If the file did exist, but contained invalid
2654 data, a new version was generated, as expected. It was only the case of a
2655 non-existent file that was broken.
2657 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2658 with a change in libdomainkeys > 0.64.
2660 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2661 from DNS. If the selector record carries the flag, it now has
2662 precedence over the domain-wide flag.
2664 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2666 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2667 the use of a port name that isn't defined in /etc/services) occurred, the
2668 message was deferred as in a normal delivery, and thus remained on the
2669 spool, instead of being failed because of the mua_wrapper setting. This
2670 is now fixed, and I tidied up some of the mua_wrapper messages at the
2673 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2674 the messages in a hash of arrays rather than using individual hashes.
2675 This is a bit cleaner and results in dramatic memory savings, albeit
2676 at a slight CPU cost.
2678 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2679 as requested by Marc Sherman.
2681 SC/10 Eximstats: added histograms for user specified patterns as requested
2684 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2686 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2687 fopen() in the content-scanning modules that did not already have it.
2693 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2694 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2696 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2698 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2700 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2701 to test Sieve filters that use "vacation".
2703 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2704 that changes the way the GnuTLS parameters are stored in the cache file.
2705 The new format can be generated externally. For backward compatibility,
2706 if the data in the cache doesn't make sense, Exim assumes it has read an
2707 old-format file, and it generates new data and writes a new file. This
2708 means that you can't go back to an older release without removing the
2711 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2712 work if there are any delivery delays because "one_time" forces the
2713 parent to be marked "delivered", so its unseen clone is never tried
2714 again. For this reason, Exim now forbids the simultaneous setting of
2717 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2718 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2719 entry if you want to know the details. The fix, however, neglected to
2720 consider the case when local delivery batching is involved. The test for
2721 "previously delivered" was not happening when checking to see if an
2722 address could be batched with a previous (undelivered) one; under
2723 certain circumstances this could lead to multiple deliveries to the same
2726 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2727 in its include files, and this causes problems building Exim.
2729 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2730 header_syntax) but Exim was just ignoring anything given after a slash.
2731 In particular, this caused confusion with an attempt to use "verify =
2732 reverse_host_lookup/defer_ok". An error is now given when options are
2733 supplied for verify items that do not have them. (Maybe reverse_host_
2734 lookup should have a defer_ok option, but that's a different point.)
2736 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2737 defined by RFC 821) to 2048, because there were problems with some AUTH
2738 commands, and RFC 1869 says the size should be increased for extended
2739 SMTP commands that take arguments.
2741 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2744 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2745 "unknown" error; now it says that the functionality isn't in the binary.
2747 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2748 an address' error message when a string expansion fails (syntax or
2749 whatever). Otherwise the password may appear in the log. Following change
2750 PH/42 below, there is no longer a chance of it appearing in a bounce
2753 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2755 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2756 of its addresses was ever tried. (Bugzilla bug #2.)
2758 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2759 the result incorrectly in the debug output. (It correctly added a newline
2760 to what was transported.)
2762 TF/01 Added $received_time.
2764 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2765 commented out examples of how to interface to a virus scanner and to
2766 SpamAssassin. Also added commented examples of av_scanner and
2767 spamd_address settings.
2769 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2770 and controls are allowed in which ACLs. There were a couple of minor
2771 errors. Some of the entries in the conditions table (which is a table of
2772 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2773 negation of where the condition IS allowed.
2775 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2777 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2778 header file does not have a version number, so I've had to invent a new
2779 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2780 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2781 radiusclient), but it was contributed by a Radius user.
2783 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2784 files or directories, for OpenSSL.
2786 PH/19 When an Exim process that is running as root has to create an Exim log
2787 file, it does so in a subprocess that runs as exim:exim so as to get the
2788 ownership right at creation (otherwise, other Exim processes might see
2789 the file with the wrong ownership). There was no test for failure of this
2790 fork() call, which would lead to the process getting stuck as it waited
2791 for a non-existent subprocess. Forks do occasionally fail when resources
2792 run out. I reviewed all the other calls to fork(); they all seem to check
2795 PH/20 When checking for unexpected SMTP input at connect time (before writing
2796 the banner), Exim was not dealing correctly with a non-positive return
2797 from the read() function. If the client had disconnected by this time,
2798 the result was a log entry for a synchronization error with an empty
2799 string after "input=" when read() returned zero. If read() returned -1
2800 (an event I could not check), uninitialized data bytes were printed.
2801 There were reports of junk text (parts of files, etc) appearing after
2804 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2806 PH/22 Added support for macro redefinition, and (re)definition in between
2807 driver and ACL definitions.
2809 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2810 forgetting to use the resulting value; it was using the unexpanded value.
2812 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2813 hadn't been configured. The fix is from Juergen Kreileder, who
2814 understands it better than I do:
2816 "Here's what I see happening with three configured cyrus_sasl
2817 authenticators configured (plain, login, cram-md5):
2819 On startup auth_cyrus_sasl_init() gets called for each of these.
2820 This means three calls to sasl_listmech() without a specified mech_list.
2821 => SASL tests which mechs of all available mechs actually work
2822 => three warnings about OTP not working
2823 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2825 With the patch, sasl_listmech() also gets called three times. But now
2826 SASL's mech_list option is set to the server_mech specified in the the
2827 authenticator. Or in other words, the answer from sasl_listmech()
2828 gets limited to just the mech you're testing for (which is different
2830 => the return list contains just 'plain' or 'login', 'cram-md5' or
2831 nothing depending on the value of ob->server_mech.
2833 I've just tested the patch: Authentication still works fine,
2834 unavailable mechs specified in the exim configuration are still
2835 caught, and the auth.log warnings about OTP are gone."
2837 PH/25 When debugging is enabled, the contents of the command line are added
2838 to the debugging output, even when log_selector=+arguments is not
2841 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2842 answer is "GNU", and only if the return is "GNU/something" is the answer
2845 PH/27 $acl_verify_message is now set immediately after the failure of a
2846 verification in an ACL, and so is available in subsequent modifiers. In
2847 particular, the message can be preserved by coding like this:
2849 warn !verify = sender
2850 set acl_m0 = $acl_verify_message
2852 Previously, $acl_verify_message was set only while expanding "message"
2853 and "log_message" when a very denied access.
2855 PH/28 Modified OS/os.c-Linux with
2857 -#ifndef OS_LOAD_AVERAGE
2858 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2860 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2861 nomenclature these days.)
2863 PH/29 Installed patch from the Sieve maintainer that adds the options
2864 sieve_useraddress and sieve_subaddress to the redirect router.
2866 PH/30 In these circumstances:
2867 . Two addresses routed to the same list of hosts;
2868 . First host does not offer TLS;
2869 . First host accepts first address;
2870 . First host gives temporary error to second address;
2871 . Second host offers TLS and a TLS session is established;
2872 . Second host accepts second address.
2873 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2874 and peerdn, if requested) that were in fact used only for the second
2877 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2878 attention to any local part prefix or suffix that was matched by the
2879 router that accepted the address. It now behaves in the same way as it
2880 does for delivery: the affixes are removed from the local part unless
2881 rcpt_include_affixes is set on the transport.
2883 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2884 timeout during the DATA phase of an incoming message.
2886 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2887 applied a patch sanctioned by the Sieve maintainer.
2889 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2890 the uid or gid is negative. A case of a negative gid caused this to be
2891 noticed. The fix allows for either to be negative.
2893 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2894 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2897 PH/36 The change PH/12 above was broken. Fixed it.
2899 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2900 the grounds that routing the same address twice would always produce the
2901 same answer. This might have been true once, but it is certainly no
2902 longer true now. Routing a child address may depend on the previous
2903 routing that produced that child. Some complicated redirection strategies
2904 went wrong when messages had multiple recipients, and made Exim's
2905 behaviour dependent on the order in which the addresses were given.
2907 I have moved the duplicate checking until after the routing is complete.
2908 Exim scans the addresses that are assigned to local and remote
2909 transports, and removes any duplicates. This means that more work will be
2910 done, as duplicates will always all be routed, but duplicates are
2911 presumably rare, so I don't expect this is of any significance.
2913 For deliveries to pipes, files, and autoreplies, the duplicate checking
2914 still happens during the routing process, since they are not going to be
2917 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2918 It corrects a timeout issue with spamd. This is Ian's comment: "The
2919 background is that sometimes spamd either never reads data from a
2920 connection it has accepted, or it never writes response data. The exiscan
2921 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2922 blindly assumes that writes won't block so it may never time out."
2924 PH/39 Allow G after quota size as well as K and M.
2926 PH/40 The value set for $authenticated_id in an authenticator may not contain
2927 binary zeroes or newlines because the value is written to log lines and
2928 to spool files. There was no check on this. Now the value is run through
2929 the string_printing() function so that such characters are converted to
2930 printable escape sequences.
2932 PH/41 $message_linecount is a new variable that contains the total number of
2933 lines in the message. Compare $body_linecount, which is the count for the
2936 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2937 bounce and delay warning messages, except in certain special cases, which
2940 (a) An SMTP error message from a remote host;
2941 (b) A message specified in a :fail: redirection;
2942 (c) A message specified in a "fail" command in a system filter;
2943 (d) A message specified in a FAIL return from the queryprogram router;
2944 (e) A message specified by the cannot_route_message router option.
2946 In these cases only, Exim does include the error details in bounce and
2947 warning messages. There are also a few cases where bland messages such
2948 as "unrouteable address" or "local delivery error" are given.
2950 PH/43 $value is now also set for the "else" part of a ${run expansion.
2952 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2953 being worked on, but at least Exim now implements the latest version to
2956 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2957 process to complete was treated as a delivery failure, a timeout while
2958 writing the message to the pipe was logged, but erroneously treated as a
2959 successful delivery. Such timeouts include transport filter timeouts. For
2960 consistency with the overall process timeout, these timeouts are now
2961 treated as errors, giving rise to delivery failures by default. However,
2962 there is now a new Boolean option for the pipe transport called
2963 timeout_defer, which, if set TRUE, converts the failures into defers for
2964 both kinds of timeout. A transport filter timeout is now identified in
2967 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2968 systems where "make" and "gmake" are different, calling "gmake" at top
2969 level broke things. I've arranged for the value of $(MAKE) to be passed
2970 from the Makefile to this script so that it can call the same version of
2974 A note about Exim versions 4.44 and 4.50
2975 ----------------------------------------
2977 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2978 changes of various kinds. As a consequence, a big documentation update was
2979 needed. This delayed the release for rather longer than seemed good, especially
2980 in the light of a couple of (minor) security issues. Therefore, the changes
2981 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2982 release. So 4.44 and 4.50 are in effect two different branches that both start
2985 I have left the 4.50 change log unchanged; it contains all the changes since
2986 4.43. The change log for 4.44 is below; many of its items are identical to
2987 those for 4.50. This seems to be the most sensible way to preserve the
2988 historical information.
2994 1. Minor wording change to the doc/README.SIEVE file.
2996 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2997 computation of the current number of files was incorrect.
2999 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3000 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3003 4. Give more explanation in the error message when the command for a transport
3004 filter fails to execute.
3006 5. There are several places where Exim runs a non-Exim command in a
3007 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3008 was being done only for the command run by the queryprogram router. It is
3009 now done for all such subprocesses. The other cases are: ${run, transport
3010 filters, and the commands run by the lmtp and pipe transports.
3012 6. Added CONFIGURE_GROUP build-time option.
3014 7. Some older OS have a limit of 256 on the maximum number of file
3015 descriptors. Exim was using setrlimit() to set 1000 as a large value
3016 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3017 systems. I've change it so that if it can't get 1000, it tries for 256.
3019 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3020 was an oversight, and furthermore, ever since the addition of extra
3021 controls (e.g. 4.43/32), the checks on when to allow different forms of
3022 "control" were broken. There should now be diagnostics for all cases when a
3023 control that does not make sense is encountered.
3025 9. Added the /retain_sender option to "control=submission".
3027 10. $recipients is now available in the predata ACL (oversight).
3029 11. Tidy the search cache before the fork to do a delivery from a message
3030 received from the command line. Otherwise the child will trigger a lookup
3031 failure and thereby defer the delivery if it tries to use (for example) a
3032 cached ldap connection that the parent has called unbind on.
3034 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3035 of $address_data from the recipient verification was clobbered by the
3036 sender verification.
3038 13. The value of address_data from a sender verification is now available in
3039 $sender_address_data in subsequent conditions in the ACL statement.
3041 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3043 15. Added a new option "connect=<time>" to callout options, to set a different
3046 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3047 was its contents. (It was OK if the option was not defined at all.)
3049 17. A "Completed" log line is now written for messages that are removed from
3050 the spool by the -Mrm option.
3052 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3053 information about exactly what failed.
3055 19. Added -dd to debug only the daemon process.
3057 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3058 handles timeouts, both on the server side and network timeouts. Renamed the
3059 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3061 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3062 It is now set to "smtps".
3064 22. $host_address is now set to the target address during the checking of
3065 ignore_target_hosts.
3067 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3068 being passed; this would have caused $sender_host_name to have been used if
3069 matching the list had actually called for a host name (not very likely,
3070 since this list is usually IP addresses). A host name is now passed as
3073 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3074 code that specifies a non-restarting handler (typically sigaction() in
3075 modern systems) in an attempt to fix a rare and obscure crash bug.
3077 25. Narrowed the window for a race in the daemon that could cause it to ignore
3078 SIGCHLD signals. This is not a major problem, because they are used only to
3079 wake it up if nothing else does.
3081 26. A malformed maildirsize file could cause Exim to calculate negative values
3082 for the mailbox size or file count. Odd effects could occur as a result.
3083 The maildirsize information is now recalculated if the size or filecount
3086 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3087 support for a long time. Removed HAVE_SYS_VFS_H.
3089 28. Installed the latest version of exipick from John Jetmore.
3091 29. In an address list, if the pattern was not a regular expression, an empty
3092 subject address (from a bounce message) matched only if the pattern was an
3093 empty string. Non-empty patterns were not even tested. This was the wrong
3094 because it is perfectly reasonable to use an empty address as part of a
3095 database query. An empty address is now tested by patterns that are
3096 lookups. However, all the other forms of pattern expect the subject to
3097 contain a local part and a domain, and therefore, for them, an empty
3098 address still always fails if the pattern is not itself empty.
3100 30. Exim went into a mad DNS loop when attempting to do a callout where the
3101 host was specified on an smtp transport, and looking it up yielded more
3102 than one IP address.
3104 31. Re-factored the code for checking spool and log partition space into a
3105 function that finds that data and another that does the check. The former
3106 is then used to implement four new variables: $spool_space, $log_space,
3107 $spool_inodes, and $log_inodes.
3109 32. The RFC2047 encoding function was originally intended for short strings
3110 such as real names; it was not keeping to the 75-character limit for
3111 encoded words that the RFC imposes. It now respects the limit, and
3112 generates multiple encoded words if necessary. To be on the safe side, I
3113 have increased the buffer size for the ${rfc2047: expansion operator from
3116 33. It is now permitted to omit both strings after an "if" condition; if the
3117 condition is true, the result is "true". As before, when the second string
3118 is omitted, a false condition yields an empty string. This makes it less
3119 cumbersome to write custom ACL and router conditions.
3121 34. Failure to deliver a bounce message always caused it to be frozen, even if
3122 there was an errors_to setting on the router. The errors_to setting is now
3125 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3126 canonical form (fully expanded) before being placed in
3127 $sender_host_address.
3129 36. The table in the code that translates DNS record types into text (T_A to
3130 "A" for instance) was missing entries for NS and CNAME. It is just possible
3131 that this could have caused confusion if both these types were looked up
3132 for the same domain, because the text type is used as part of Exim's
3133 per-process caching. But the chance of anyone hitting this buglet seems
3136 37. The dnsdb lookup has been extended in a number of ways.
3138 (1) There is a new type, "zns", which walks up the domain tree until it
3139 finds some nameserver records. It should be used with care.
3141 (2) There is a new type, "mxh", which is like "mx" except that it returns
3142 just the host names, not the priorities.
3144 (3) It is now possible to give a list of domains (or IP addresses) to be
3145 looked up. The behaviour when one of the lookups defers can be
3146 controlled by a keyword.
3148 (4) It is now possible to specify the separator character for use when
3149 multiple records are returned.
3151 38. The dnslists ACL condition has been extended: it is now possible to supply
3152 a list of IP addresses and/or domains to be looked up in a particular DNS
3155 39. Added log_selector=+queue_time_overall.
3157 40. When running the queue in the test harness, wait just a tad after forking a
3158 delivery process, to get repeatability of debugging output.
3160 41. Include certificate and key file names in error message when GnuTLS fails
3161 to set them up, because the GnuTLS error message doesn't include the name
3162 of the failing file when there is a problem reading it.
3164 42. Allow both -bf and -bF in the same test run.
3166 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3168 44. The "Exiscan patch" is now merged into the mainline Exim source.
3170 45. Sometimes the final signoff response after QUIT could fail to get
3171 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3172 < 0 before doing a fflush(). This bug looks as though it goes back to the
3173 introduction of TLS in release 3.20, but "sometimes" must have been rare
3174 because the tests only now provoked it.
3176 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3177 (this can affect the format of dates).
3179 47. exim_tidydb, when checking for the continued existence of a message for
3180 which it has found a message-specific retry record, was not finding
3181 messages that were in split spool directories. Consequently, it was
3182 deleting retry records that should have stayed in existence.
3184 48. Steve fixed some bugs in eximstats.
3186 49. The SPA authentication driver was not abandoning authentication and moving
3187 on to the next authenticator when an expansion was forced to fail,
3188 contradicting the general specification for all authenticators. Instead it
3189 was generating a temporary error. It now behaves as specified.
3191 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3192 (the order specifies the preference for clients). The order is now AES256,
3193 AES128, 3DES, ARCFOUR128.
3195 51. Small patch to Sieve code - explicitly set From: when generating an
3198 52. Exim crashed if a remote delivery caused a very long error message to be
3199 recorded - for instance if somebody sent an entire SpamAssassin report back
3200 as a large number of 550 error lines. This bug was coincidentally fixed by
3201 increasing the size of one of Exim's internal buffers (big_buffer) that
3202 happened as part of the Exiscan merge. However, to be on the safe side, I
3203 have made the code more robust (and fixed the comments that describe what
3206 53. Now that there can be additional text after "Completed" in log lines (if
3207 the queue_time_overall log selector is set), a one-byte patch to exigrep
3208 was needed to allow it to recognize "Completed" as not the last thing in
3211 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3212 patch that reportedly fixes this has been added. I am not expert enough to
3213 create a test for it. This is what the patch creator wrote:
3215 "I found a little strange behaviour of ldap code when working with
3216 Windows 2003 AD Domain, where users was placed in more than one
3217 Organization Units. When I tried to give exim partial DN, the exit code
3218 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3219 But simultaneously result of request was absolutely normal ldap result,
3220 so I produce this patch..."
3222 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3223 REFERENCE, so I have modified the code to exclude the patch when that macro
3226 55. Some experimental protocols are using DNS PTR records for new purposes. The
3227 keys for these records are domain names, not reversed IP addresses. The
3228 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3229 leaves it alone. Component reversal etc. now happens only for IP addresses.
3232 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3234 57. Double the size of the debug message buffer (to 2048) so that more of very
3235 long debug lines gets shown.
3237 58. The exicyclog utility now does better if the number of log files to keep
3238 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3240 59. Two changes related to the smtp_active_hostname option:
3242 (1) $smtp_active_hostname is now available as a variable.
3243 (2) The default for smtp_banner uses $smtp_active_hostname instead
3244 of $primary_hostname.
3246 60. The host_aton() function is supposed to be passed a string that is known
3247 to be a valid IP address. However, in the case of IPv6 addresses, it was
3248 not checking this. This is a hostage to fortune. Exim now panics and dies
3249 if the condition is not met. A case was found where this could be provoked
3250 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3251 components; fortuitously, this particular loophole had already been fixed
3252 by change 4.50/55 above.
3254 If there are any other similar loopholes, the new check in host_aton()
3255 itself should stop them being exploited. The report I received stated that
3256 data on the command line could provoke the exploit when Exim was running as
3257 exim, but did not say which command line option was involved. All I could
3258 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3259 running as the user.
3262 61. There was a buffer overflow vulnerability in the SPA authentication code
3263 (which came originally from the Samba project). I have added a test to the
3264 spa_base64_to_bits() function which I hope fixes it.
3267 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3268 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3270 63. The daemon start-up calls getloadavg() while still root for those OS that
3271 need the first call to be done as root, but it missed one case: when
3272 deliver_queue_load_max is set with deliver_drop_privilege. This is
3273 necessary for the benefit of the queue runner, because there is no re-exec
3274 when deliver_drop_privilege is set.
3276 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3277 This has been fixed.
3279 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3280 was in use, was not putting the data itself into the right store pool;
3281 consequently, it could be overwritten for a subsequent message in the same
3282 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3285 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3287 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3288 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3289 the function were treating the return as a boolean value, which happened to
3290 work because 0=false and not-0=true, but is not correct code.
3292 68. The host_aton() function was not handling scoped IPv6 addresses (those
3293 with, for example, "%eth0" on the end) correctly.
3295 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3296 negated items (that is, ~something) in unsigned ints. Some compilers
3297 apparently mutter when there is no cast.
3299 70. If an address verification called from an ACL failed, and did not produce a
3300 user-specific message (i.e. there was only a "system" message), nothing was
3301 put in $acl_verify_message. In this situation, it now puts the system
3304 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3305 session; change 4.31/43 added the unwanted input to the log line - except
3306 that it did not do this in the start of session case. It now does.
3308 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3309 This is wrong and can cause the other end to generate a synchronization
3310 error if it is another Exim or anything else that does the synchronization
3311 check. A QUIT command is no longer sent after a timeout.
3313 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3314 during host lookups.
3316 74. The defer_ok option of callout verification was not working if it was used
3317 when verifying addresses in header lines, that is, for this case:
3319 verify = header_sender/callout=defer_ok
3321 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3322 those file descriptors could be used for SMTP connections. If anything
3323 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3324 could be sent to the SMTP client, causing chaos. The daemon now opens
3325 stdin, stdout, and stderr to /dev/null when it puts itself into the
3328 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3329 log by default. The user can override this with suitable Perl magic.
3331 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3332 the log message when discard triggers, was not working for the DATA ACL or
3333 for the non-SMTP ACL.
3335 78. Error message wording change in sieve.c.
3337 79. If smtp_accept_max_per_host was set, the number of connections could be
3338 restricted to fewer than expected, because the daemon was trying to set up
3339 a new connection before checking whether the processes handling previous
3340 connections had finished. The check for completed processes is now done
3341 earlier. On busy systems, this bug wouldn't be noticed because something
3342 else would have woken the daemon, and it would have reaped the completed
3345 80. If a message was submitted locally by a user whose login name contained one
3346 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3347 It caused a spool format error. I have fixed the spool reading code. A
3348 related problem was that the "from" clause in the Received: line became
3349 illegal because of the space(s). It is now covered by ${quote_local_part.
3351 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3354 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3356 83. Patch from Sieve maintainer to fix unterminated string problem in
3357 "vacation" handling.
3359 84. Some minor changes to the Linux configuration files to help with other
3360 OS variants using glibc.
3362 85. One more patch for Sieve to update vacation handling to latest spec.
3365 ----------------------------------------------------
3366 See the note above about the 4.44 and 4.50 releases.
3367 ----------------------------------------------------
3373 1. Change 4.43/35 introduced a bug that caused file counts to be
3374 incorrectly computed when quota_filecount was set in an appendfile
3377 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3378 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3381 3. Give more explanation in the error message when the command for a transport
3382 filter fails to execute.
3384 4. There are several places where Exim runs a non-Exim command in a
3385 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3386 was being done only for the command run by the queryprogram router. It is
3387 now done for all such subprocesses. The other cases are: ${run, transport
3388 filters, and the commands run by the lmtp and pipe transports.
3390 5. Some older OS have a limit of 256 on the maximum number of file
3391 descriptors. Exim was using setrlimit() to set 1000 as a large value
3392 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3393 systems. I've change it so that if it can't get 1000, it tries for 256.
3395 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3396 was an oversight, and furthermore, ever since the addition of extra
3397 controls (e.g. 4.43/32), the checks on when to allow different forms of
3398 "control" were broken. There should now be diagnostics for all cases when a
3399 control that does not make sense is encountered.
3401 7. $recipients is now available in the predata ACL (oversight).
3403 8. Tidy the search cache before the fork to do a delivery from a message
3404 received from the command line. Otherwise the child will trigger a lookup
3405 failure and thereby defer the delivery if it tries to use (for example) a
3406 cached ldap connection that the parent has called unbind on.
3408 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3409 of $address_data from the recipient verification was clobbered by the
3410 sender verification.
3412 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3413 was its contents. (It was OK if the option was not defined at all.)
3415 11. A "Completed" log line is now written for messages that are removed from
3416 the spool by the -Mrm option.
3418 12. $host_address is now set to the target address during the checking of
3419 ignore_target_hosts.
3421 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3422 being passed; this would have caused $sender_host_name to have been used if
3423 matching the list had actually called for a host name (not very likely,
3424 since this list is usually IP addresses). A host name is now passed as
3427 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3428 code that specifies a non-restarting handler (typically sigaction() in
3429 modern systems) in an attempt to fix a rare and obscure crash bug.
3431 15. Narrowed the window for a race in the daemon that could cause it to ignore
3432 SIGCHLD signals. This is not a major problem, because they are used only to
3433 wake it up if nothing else does.
3435 16. A malformed maildirsize file could cause Exim to calculate negative values
3436 for the mailbox size or file count. Odd effects could occur as a result.
3437 The maildirsize information is now recalculated if the size or filecount
3440 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3441 support for a long time. Removed HAVE_SYS_VFS_H.
3443 18. Updated exipick to current release from John Jetmore.
3445 19. Allow an empty sender to be matched against a lookup in an address list.
3446 Previously the only cases considered were a regular expression, or an
3449 20. Exim went into a mad DNS lookup loop when doing a callout where the
3450 host was specified on the transport, if the DNS lookup yielded more than
3453 21. The RFC2047 encoding function was originally intended for short strings
3454 such as real names; it was not keeping to the 75-character limit for
3455 encoded words that the RFC imposes. It now respects the limit, and
3456 generates multiple encoded words if necessary. To be on the safe side, I
3457 have increased the buffer size for the ${rfc2047: expansion operator from
3460 22. Failure to deliver a bounce message always caused it to be frozen, even if
3461 there was an errors_to setting on the router. The errors_to setting is now
3464 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3465 canonical form (fully expanded) before being placed in
3466 $sender_host_address.
3468 24. Updated eximstats to version 1.33
3470 25. Include certificate and key file names in error message when GnuTLS fails
3471 to set them up, because the GnuTLS error message doesn't include the name
3472 of the failing file when there is a problem reading it.
3474 26. Expand error message when OpenSSL has problems setting up cert/key files.
3477 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3478 (this can affect the format of dates).
3480 28. exim_tidydb, when checking for the continued existence of a message for
3481 which it has found a message-specific retry record, was not finding
3482 messages that were in split spool directories. Consequently, it was
3483 deleting retry records that should have stayed in existence.
3485 29. eximstats updated to version 1.35
3486 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3487 1.35 - bugfix such that pie charts by volume are generated correctly
3489 30. The SPA authentication driver was not abandoning authentication and moving
3490 on to the next authenticator when an expansion was forced to fail,
3491 contradicting the general specification for all authenticators. Instead it
3492 was generating a temporary error. It now behaves as specified.
3494 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3495 (the order specifies the preference for clients). The order is now AES256,
3496 AES128, 3DES, ARCFOUR128.
3498 31. Small patch to Sieve code - explicitly set From: when generating an
3501 32. Exim crashed if a remote delivery caused a very long error message to be
3502 recorded - for instance if somebody sent an entire SpamAssassin report back
3503 as a large number of 550 error lines. This bug was coincidentally fixed by
3504 increasing the size of one of Exim's internal buffers (big_buffer) that
3505 happened as part of the Exiscan merge. However, to be on the safe side, I
3506 have made the code more robust (and fixed the comments that describe what
3509 33. Some experimental protocols are using DNS PTR records for new purposes. The
3510 keys for these records are domain names, not reversed IP addresses. The
3511 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3512 leaves it alone. Component reversal etc. now happens only for IP addresses.
3515 34. The host_aton() function is supposed to be passed a string that is known
3516 to be a valid IP address. However, in the case of IPv6 addresses, it was
3517 not checking this. This is a hostage to fortune. Exim now panics and dies
3518 if the condition is not met. A case was found where this could be provoked
3519 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3520 components; fortuitously, this particular loophole had already been fixed
3521 by change 4.50/55 or 4.44/33 above.
3523 If there are any other similar loopholes, the new check in host_aton()
3524 itself should stop them being exploited. The report I received stated that
3525 data on the command line could provoke the exploit when Exim was running as
3526 exim, but did not say which command line option was involved. All I could
3527 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3528 running as the user.
3531 35. There was a buffer overflow vulnerability in the SPA authentication code
3532 (which came originally from the Samba project). I have added a test to the
3533 spa_base64_to_bits() function which I hope fixes it.
3536 36. The daemon start-up calls getloadavg() while still root for those OS that
3537 need the first call to be done as root, but it missed one case: when
3538 deliver_queue_load_max is set with deliver_drop_privilege. This is
3539 necessary for the benefit of the queue runner, because there is no re-exec
3540 when deliver_drop_privilege is set.
3542 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3543 was in use, was not putting the data itself into the right store pool;
3544 consequently, it could be overwritten for a subsequent message in the same
3545 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3548 38. Sometimes the final signoff response after QUIT could fail to get
3549 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3550 < 0 before doing a fflush(). This bug looks as though it goes back to the
3551 introduction of TLS in release 3.20, but "sometimes" must have been rare
3552 because the tests only now provoked it.
3558 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3559 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3560 Now it returns an integer. A number of calls were still expecting a T/F
3561 return. Fortuitously, in all cases, the tests worked in OK situations,
3562 which is the norm. However, things would have gone wrong on any write
3563 failures on the smtp file descriptor. This function is used when sending
3564 messages over SMTP and also when doing verify callouts.
3566 2. When Exim is called to do synchronous delivery of a locally submitted
3567 message (the -odf or -odi options), it no longer closes stderr before doing
3570 3. Implemented the mua_wrapper option.
3572 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3574 5. Implemented the functions header_remove(), header_testname(),
3575 header_add_at_position(), and receive_remove_recipient(), and exported them
3578 6. If an ACL "warn" statement specified the addition of headers, Exim already
3579 inserted X-ACL-Warn: at the start if there was no header name. However, it
3580 was not making this test for the second and subsequent header lines if
3581 there were newlines in the string. This meant that an invalid header could
3582 be inserted if Exim was badly configured.
3584 7. Allow an ACL "warn" statement to add header lines at the start or after all
3585 the Received: headers, as well as at the end.
3587 8. Added the rcpt_4xx retry error code.
3589 9. Added postmaster_mailfrom=xxx to callout verification option.
3591 10. Added mailfrom=xxxx to the callout verification option, for verify=
3594 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3595 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3597 12. Inserted some casts to stop certain compilers warning when using pointer
3598 differences as field lengths or precisions in printf-type calls (mostly
3599 affecting debugging statements).
3601 13. Added optional readline() support for -be (dynamically loaded).
3603 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3604 same clock tick as a message's arrival, so that its received time was the
3605 same as the "first fail" time on the retry record, and that message
3606 remained on the queue past the ultimate address timeout, every queue runner
3607 would try a delivery (because it was past the ultimate address timeout) but
3608 after another failure, the ultimate address timeout, which should have then
3609 bounced the address, did not kick in. This was a "< instead of <=" error;
3610 in most cases the first failure would have been in the next clock tick
3611 after the received time, and all would be well.
3613 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3614 being recognized when the domain list was tested by the match_domain
3615 condition in an expansion string.
3617 16. Added the ${str2b64: operator.
3619 17. Exim was always calling setrlimit() to set a large limit for the number of
3620 processes, without checking whether the existing limit was already
3621 adequate. (It did check for the limit on file descriptors.) Furthermore,
3622 errors from getrlimit() and setrlimit() were being ignored. Now they are
3623 logged to the main and panic logs, but Exim does carry on, to try to do its
3624 job under whatever limits there are.
3626 18. Imported PCRE 5.0.
3628 19. Trivial typo in log message " temporarily refused connection" (the leading
3631 20. If the log selector return_path_on_delivery was set and an address was
3632 redirected to /dev/null, the delivery process crashed because it assumed
3633 that a return path would always be set for a "successful" delivery. In this
3634 case, the whole delivery is bypassed as an optimization, and therefore no
3637 21. Internal re-arrangement: the function for sending a challenge and reading
3638 a response while authentication was assuming a zero-terminated challenge
3639 string. It's now changed to take a pointer and a length, to allow for
3640 binary data in such strings.
3642 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3644 23. Exim was not respecting finduser_retries when seeking the login of the
3645 uid under which it was called; it was always trying 10 times. (The default
3646 setting of finduser_retries is zero.) Also, it was sleeping after the final
3647 failure, which is pointless.
3649 24. Implemented tls_on_connect_ports.
3651 25. Implemented acl_smtp_predata.
3653 26. If the domain in control=submission is set empty, Exim assumes that the
3654 authenticated id is a complete email address when it generates From: or
3655 Sender: header lines.
3657 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3658 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3659 chown and chgrp in /bin and hostname in /usr/bin.
3661 28. Exim was keeping the "process log" file open after each use, just as it
3662 does for the main log. This opens the possibility of it remaining open for
3663 long periods when the USR1 signal hits a daemon. Occasional processlog
3664 errors were reported, that could have been caused by this. Anyway, it seems
3665 much more sensible not to leave this file open at all, so that is what now
3668 29. The long-running daemon process does not normally write to the log once it
3669 has entered its main loop, and it closes the log before doing so. This is
3670 so that log files can straightforwardly be renamed and moved. However,
3671 there are a couple of unusual error situations where the daemon does write
3672 log entries, and I had neglected to close the log afterwards.
3674 30. The text of an SMTP error response that was received during a remote
3675 delivery was being truncated at 512 bytes. This is too short for some of
3676 the long messages that one sometimes sees. I've increased the limit to
3679 31. It is now possible to make retry rules that apply only when a message has a
3680 specific sender, in particular, an empty sender.
3682 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3683 it possible to be selective about when SMTP synchronization is enforced.
3685 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3687 32. Implemented hosts_connection_nolog.
3689 33. Added an ACL for QUIT.
3691 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3694 35. Added mailbox_size and mailbox_filecount to appendfile.
3696 36. Added control = no_multiline_responses to ACLs.
3698 37. There was a bug in the logic of the code that waits for the clock to tick
3699 in the case where the clock went backwards by a substantial amount such
3700 that the microsecond fraction of "now" was more than the microsecond
3701 fraction of "then" (but the whole seconds number was less).
3703 38. Added support for the libradius Radius client library this is found on
3704 FreeBSD (previously only the radiusclient library was supported).
3710 1. When certain lookups returned multiple values in the form name=value, the
3711 quoting of the values was not always being done properly. Specifically:
3712 (a) If the value started with a double quote, but contained no whitespace,
3714 (b) If the value contained whitespace other than a space character (i.e.
3715 tabs or newlines or carriage returns) it was not quoted.
3716 This fix has been applied to the mysql and pgsql lookups by writing a
3717 separate quoting function and calling it from the lookup code. The fix
3718 should probably also be applied to nisplus, ibase and oracle lookups, but
3719 since I cannot test any of those, I have not disturbed their existing code.
3721 2. A hit in the callout cache for a specific address caused a log line with no
3722 reason for rejecting RCPT. Now it says "Previous (cached) callout
3723 verification failure".
3725 3. There was an off-by-one bug in the queryprogram router. An over-long
3726 return line was truncated at 256 instead of 255 characters, thereby
3727 overflowing its buffer with the terminating zero. As well as fixing this, I
3728 have increased the buffer size to 1024 (and made a note to document this).
3730 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3731 when Exim is waiting for an SMTP response from a remote server, Exim
3732 restarts its select() call on the socket, thereby resetting its timeout.
3733 This is not a problem when such interrupts are rare. Somebody set up a cron
3734 job to run exiwhat every 2 minutes, which is less than the normal select()
3735 timeout (5 or 10 minutes). This meant that the select() timeout never
3736 kicked in because it was always reset. I have fixed this by comparing the
3737 time when an interrupt arrives with the time at the start of the first call
3738 to select(). If more time than the timeout has elapsed, the interrupt is
3739 treated as a timeout.
3741 5. Some internal re-factoring in preparation for the addition of Sieve
3742 extensions (by MH). In particular, the "personal" test is moved to a
3743 separate function, and given an option for scanning Cc: and Bcc: (which is
3744 not set for Exim filters).
3746 6. When Exim created an email address using the login of the caller as the
3747 local part (e.g. when creating a From: or Sender: header line), it was not
3748 quoting the local part when it contained special characters such as @.
3750 7. Installed new OpenBSD configuration files.
3752 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3753 try to make them clearer.
3755 9. Callout options, other than the timeout value, were being ignored when
3756 verifying sender addresses in header lines. For example, when using
3758 verify = header_sender/callout=no_cache
3760 the cache was (incorrectly) being used.
3762 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3763 only the Cygwin environment.
3765 11. When return_path_on_delivery was set as a log selector, if different remote
3766 addresses in the same message used different return paths and parallel
3767 remote delivery occurred, the wrong values would sometimes be logged.
3768 (Whenever a remote delivery process finished, the return path value from
3769 the most recently started remote delivery process was logged.)
3771 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3772 lines when AUTH and/or TLS are in use. This is the "received protocol"
3773 field. Exim used to use "asmtp" for authenticated SMTP, without any
3774 indication (in the protocol name) for TLS use. Now it follows the RFC and
3775 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3776 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3777 names appear in log lines as well as in Received: header lines.
3779 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3780 extensions, and comparison tests, and to fix some bugs.
3782 14. Changes to the "personal" filter test:
3784 (1) The test was buggy in that it was just doing the equivalent of
3785 "contains" tests on header lines. For example, if a user's address was
3786 anne@some.where, the "personal" test would incorrectly be true for
3788 To: susanne@some.where
3790 This test is now done by extracting each address from the header in turn,
3791 and checking the entire address. Other tests that are part of "personal"
3792 are now done using regular expressions (for example, to check local parts
3793 of addresses in From: header lines).
3795 (2) The list of non-personal local parts in From: addresses has been
3796 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3797 taken from the Sieve specification recommendations.
3799 (3) If the message contains any header line starting with "List-" it is
3800 treated as non-personal.
3802 (4) The test for "circular" in the Subject: header line has been removed
3803 because it now seems ill-conceived.
3805 15. Minor typos in src/EDITME comments corrected.
3807 16. Installed latest exipick from John Jetmore.
3809 17. If headers_add on a router specified a text string that was too long for
3810 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3811 of string_sprintf() is now avoided.
3813 18. $message_body_size was not set (it was always zero) when running the DATA
3814 ACL and the local_scan() function.
3816 19. For the "mail" command in an Exim filter, no default was being set for
3817 the once_repeat time, causing a random time value to be used if "once" was
3818 specified. (If the value happened to be <= 0, no repeat happened.) The
3819 default is now 0s, meaning "never repeat". The "vacation" command was OK
3820 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3821 (I found it when inspecting the code).
3823 20. There is now an overall timeout for performing a callout verification. It
3824 defaults to 4 times the callout timeout, which applies to individual SMTP
3825 commands during the callout. The overall timeout applies when there is more
3826 than one host that can be tried. The timeout is checked before trying the
3827 next host. This prevents very long delays if there are a large number of
3828 hosts and all are timing out (e.g. when the network connections are timing
3829 out). The value of the overall timeout can be changed by specifying an
3830 additional sub-option for "callout", called "maxwait". For example:
3832 verify = sender/callout=5s,maxwait=20s
3834 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3835 to the end before writing, but this should make it even safer).
3837 22. Exim was forgetting that it had advertised PIPELINING for the second and
3838 subsequent messages on an SMTP connection. It was also not resetting its
3839 memory on STARTTLS and an internal HELO.
3841 23. When Exim logs an SMTP synchronization error within a session, it now
3842 records whether PIPELINING has been advertised or not.
3844 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3845 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3846 rather than long int.
3848 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3850 26. Added the never_mail option to autoreply.
3856 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3857 crash if the getsockname() call failed; this can happen if a connection is
3858 closed very soon after it is established. The problem was simply in the
3859 order in which certain operations were done, causing Exim to try to write
3860 to the SMTP stream before it had set up the file descriptor. The bug has
3861 been fixed by making things happen in the correct order.
3867 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3868 before the connection was closed, thus losing the rejection response.
3870 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3871 some early Solaris releases, but causes trouble in current releases where
3872 socklen_t is defined.
3874 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3877 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3880 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3881 that releasing the top of it at the end releases what was used for sub-
3882 expansions (unless the block got too big). However, discard this block if
3883 the first thing is a variable or header, so that we can use its block when
3884 it is dynamic (useful for very large $message_headers, for example).
3886 6. Lookups now cache *every* query, not just the most recent. A new, separate
3887 store pool is used for this. It can be recovered when all lookup caches are
3888 flushed. Lookups now release memory at the end of their result strings.
3889 This has involved some general refactoring of the lookup sources.
3891 7. Some code has been added to the store_xxx() functions to reduce the amount
3892 of flapping under certain conditions.
3894 8. log_incoming_interface used to affect only the <= reception log lines. Now
3895 it causes the local interface and port to be added to several more SMTP log
3896 lines, for example "SMTP connection from", and rejection lines.
3898 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3900 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3902 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3903 could be overwritten at the end of the current message (or the start of a
3904 new message if it was set in a HELO ACL). The value is now preserved for
3905 the duration of the SMTP connection.
3907 12. If a transport had a headers_rewrite setting, and a matching header line
3908 contained an unqualified address, that address was qualified, even if it
3909 did not match any rewriting rules. The underlying bug was that the values
3910 of the flags that permit the existence of unqualified sender and recipient
3911 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3912 non-local messages, and by -bnq for local messages) were not being
3913 preserved with the message after it was received.
3915 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3916 "next input=" as part of the text comprising the host identity instead of
3917 the correct text. The code was using the same buffer for two different
3918 strings. However, depending on which order the printing function evaluated
3919 its arguments, the bug did not always show up. Under Linux, for example, my
3920 test suite worked just fine.
3922 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3923 doesn't work with some older versions of Perl. It has been changed to "my",
3924 which in any case is probably the better facility to use.
3926 15. A really picky compiler found some instances of statements for creating
3927 error messages that either had too many or two few arguments for the format
3930 16. The size of the buffer for calls to the DNS resolver has been increased
3931 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3932 for addresses that have a lot of PTR records. This alleviates a problem; it
3933 does not fully solve it.
3935 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3936 buffer now truncates the list and logs the incident, which is the same
3937 action as happens when Exim is looking up a host name and its aliases.
3938 Previously in this situation something unpredictable would happen;
3939 sometimes it was "internal error: store_reset failed".
3941 18. If a server dropped the connection unexpectedly when an Exim client was
3942 using GnuTLS and trying to read a response, the client delivery process
3943 crashed while trying to generate an error log message.
3945 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3946 string, for example:
3948 warn message = H1: something\nH2: something
3950 the text was added as a single header line from Exim's point of view
3951 though it ended up OK in the delivered message. However, searching for the
3952 second and subsequent header lines using $h_h2: did not work. This has been
3953 fixed. Similarly, if a system filter added multiple headers in this way,
3954 the routers could not see them.
3956 20. Expanded the error message when iplsearch is called with an invalid key to
3957 suggest using net-iplsearch in a host list.
3959 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3960 ACLs are no longer actually imposed (and a message to that effect is
3963 22. If a "gecos" field in a passwd entry contained escaped characters, in
3964 particular, if it contained a \" sequence, Exim got it wrong when building
3965 a From: or a Sender: header from that name. A second bug also caused
3966 incorrect handling when an unquoted " was present following a character
3967 that needed quoting.
3969 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3970 was not being matched caselessly.
3972 24. Arranged for all hyphens in the exim.8 source to be escaped with
3975 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3976 properly. Recipient callout cache records were still being keyed to include
3977 the sender, even when use_sender was set false. This led to far more
3978 callouts that were necessary. The sender is no longer included in the key
3979 when use_sender is false.
3981 26. Added "control = submission" modifier to ACLs.
3983 27. Added the ${base62d: operator to decode base 62 numbers.
3985 28. dnsdb lookups can now access SRV records.
3987 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3988 the configuration file.
3990 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3991 (-v) mode. This makes the output for a verbose queue run more intelligible.
3993 31. Added a use_postmaster feature to recipient callouts.
3995 32. Added the $body_zerocount variable, containing the number of binary zero
3996 bytes in the message body.
3998 33. The time of last modification of the "new" subdirectory is now used as the
3999 "mailbox time last read" when there is a quota error for a maildir
4002 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4004 35. Added +ignore_unknown as a special item in host lists.
4006 36. Code for decoding IPv6 addresses in host lists is now included, even if
4007 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4008 address was recognized as an IP address, but was then not correctly decoded
4009 into binary, causing unexpected and incorrect effects when compared with
4016 1. Very minor rewording of debugging text in manualroute to say "list of
4017 hosts" instead of "hostlist".
4019 2. If verify=header_syntax was set, and a header line with an unqualified
4020 address (no domain) and a large number of spaces between the end of the
4021 name and the colon was received, the reception process suffered a buffer
4022 overflow, and (when I tested it) crashed. This was caused by some obsolete
4023 code that should have been removed. The fix is to remove it!
4025 3. When running in the test harness, delay a bit after writing a bounce
4026 message to get a bit more predictability in the log output.
4028 4. Added a call to search_tidyup() just before forking a reception process. In
4029 theory, someone could use a lookup in the expansion of smtp_accept_max_
4030 per_host which, without the tidyup, could leave open a database connection.
4032 5. Added the variables $recipient_data and $sender_data which get set from a
4033 lookup success in an ACL "recipients" or "senders" condition, or a router
4034 "senders" option, similar to $domain_data and $local_part_data.
4036 6. Moved the writing of debug_print from before to after the "senders" test
4039 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4040 problems for message scanning, either using a data ACL, or using
4041 local_scan() because the Received: header was not generated till after they
4042 were called (in order to set the time as the time of reception completion).
4043 I have revised the way this works. The header is now generated after the
4044 body is received, but before the ACL or local_scan() are called. After they
4045 are run, the timestamp in the header is updated.
4051 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4052 before starting a queue runner without re-exec. This happened only when
4053 deliver_drop_privilege was set or when the Exim user was set to root. The
4054 effect of the bug was that timeouts during subsequent deliveries caused
4055 crashes instead of being properly handled. The handler is now left at its
4056 default (and expected) setting.
4058 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4059 message, again when deliver_drop_privilege is set or Exim is run as root.
4060 The bug described in (1) was not present in this case, but the tidying up
4061 of the other signals was missing. I have made the two cases consistent.
4063 3. The ignore_target_hosts setting on a manualroute router was being ignored
4064 for hosts that were looked up using the /MX notation.
4066 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4069 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4070 operated on the sender address. After changing the $sender_address to <>
4071 for the sender address verify, Exim was re-instated it as the original
4072 (before rewriting) address, but remembering that it had rewritten it, so it
4073 wasn't rewriting it again. This bug also had the effect of breaking the
4074 sender address verification caching when the sender address was rewritten.
4076 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4077 This has been changed so that if the ip literal address matches
4078 ignore_target_hosts, the router declines.
4080 7. Added expansion conditions match_domain, match_address, and match_local_
4081 part (NOT match_host).
4083 8. The placeholder for the Received: header didn't have a length field set.
4085 9. Added code to Exim itself and to exim_lock to test for a specific race
4086 condition that could lead to file corruption when using MBX delivery. The
4087 issue is with the lockfile that is created in /tmp. If this file is removed
4088 after a process has opened it but before that process has acquired a lock,
4089 there is the potential for a second process to recreate the file and also
4090 acquire a lock. This could lead to two Exim processes writing to the file
4091 at the same time. The added code performs the same test as UW imapd; it
4092 checks after acquiring the lock that its file descriptor still refers to
4093 the same named file.
4095 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4096 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4099 11. Added the smtp_active_hostname option. If used, this will typically be made
4100 to depend on the incoming interface address. Because $interface_address is
4101 not set up until the daemon has forked a reception process, error responses
4102 that can happen earlier (such as "too many connections") no longer contain
4105 12. If an expansion in a condition on a "warn" statement fails because a lookup
4106 defers, the "warn" statement is abandoned, and the next ACL statement is
4107 processed. Previously this caused the whole ACL to be aborted.
4109 13. Added the iplsearch lookup type.
4111 14. Added ident_timeout as a log selector.
4113 15. Added tls_certificate_verified as a log selector.
4115 16. Added a global option tls_require_ciphers (compare the smtp transport
4116 option of the same name). This controls incoming TLS connections.
4118 17. I finally figured out how to make tls_require_ciphers do a similar thing
4119 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4120 before starting the TLS session.
4122 18. Tabs are now shown as \t in -bP output.
4124 19. If the log selector return_path_on_delivery was set, Exim crashed when
4125 bouncing a message because it had too many Received: header lines.
4127 20. If two routers both had headers_remove settings, and the first one included
4128 a superfluous trailing colon, the final name in the first list and the
4129 first name in the second list were incorrectly joined into one item (with a
4130 colon in the middle).
4136 1. Added -C and -D options to the exinext utility, mainly to make it easier
4137 to include in the automated testing, but these could be helpful when
4138 multiple configurations are in use.
4140 2. The exinext utility was not formatting the output nicely when there was
4141 an alternate port involved in the retry record key, nor when there was a
4142 message id as well (for retries that were specific to a specific message
4143 and a specific host). It was also confused by IPv6 addresses, because of
4144 the additional colons they contain. I have fixed the IPv4 problem, and
4145 patched it up to do a reasonable job for IPv6.
4147 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4148 delivery, the log line now contains "pipelined" if PIPELINING was used.
4150 4. An SMTP transport process used to panic and die if the bind() call to set
4151 an explicit outgoing interface failed. This has been changed; it is now
4152 treated in the same way as a connect() failure.
4154 5. A reference to $sender_host_name in the part of a conditional expansion
4155 that was being skipped was still causing a DNS lookup. This no longer
4158 6. The def: expansion condition was not recognizing references to header lines
4159 that used bh_ and bheader_.
4161 7. Added the _cache feature to named lists.
4163 8. The code for checking quota_filecount in the appendfile transport was
4164 allowing one more file than it should have been.
4166 9. For compatibility with Sendmail, the command line option
4174 and sets the incoming protocol and host name (for trusted callers). The
4175 host name and its colon can be omitted when only the protocol is to be set.
4176 Note the Exim already has two private options, -pd and -ps, that refer to
4177 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4178 "s", but I don't think that's a major issue.
4180 10. A number of refactoring changes to the code, none of which should affect
4183 (a) The number of logging options was getting close to filling up the
4184 32-bit word that was used as a bit map. I have split them into two classes:
4185 those that are passed in the argument to log_write(), and those that are
4186 only ever tested independently outside of that function. These are now in
4187 separate 32-bit words, so there is plenty of room for expansion again.
4188 There is no change in the user interface or the logging behaviour.
4190 (b) When building, for example, log lines, the code previously used a
4191 macro that called string_cat() twice, in order to add two strings. This is
4192 not really sufficiently general. Furthermore, there was one instance where
4193 it was actually wrong because one of the argument was used twice, and in
4194 one call a function was used. (As it happened, calling the function twice
4195 did not affect the overall behaviour.) The macro has been replaced by a
4196 function that can join an arbitrary number of extra strings onto a growing
4199 (c) The code for expansion conditions now uses a table and a binary chop
4200 instead of a serial search (which was left over from when there were very
4201 few conditions). Also, it now recognizes conditions like "pam" even when
4202 the relevant support is not compiled in: a suitably worded error message is
4203 given if an attempt is made to use such a condition.
4205 11. Added ${time_interval:xxxxx}.
4207 12. A bug was causing one of the ddress fields not to be passed back correctly
4208 from remote delivery subprocesses. The field in question was not being
4209 subsequently used, so this caused to problems in practice.
4211 13. Added new log selectors queue_time and deliver_time.
4213 14. Might have fixed a bug in maildirsizefile handling that threw up
4214 "unexpected character" debug warnings, and recalculated the data
4215 unnecessarily. In any case, I expanded the warning message to give more
4218 15. Added the message "Restricted characters in address" to the statements in
4219 the default ACL that block characters like @ and % in local parts.
4221 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4222 Three changes have been made:
4224 (a) There was a serious bug; a negative response to MAIL caused the whole
4225 recipient domain to be cached as invalid, thereby blocking all messages
4226 to all local parts at the same domain, from all senders. This bug has
4227 been fixed. The domain is no longer cached after a negative response to
4228 MAIL if the sender used is not empty.
4230 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4233 (c) A new callout option, "use_sender" has been added for people who want
4234 the modified behaviour.
4240 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4243 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4244 indeed breaks things for older releases.
4246 3. Added additional logging to the case where there is a problem reading data
4247 from a filter that is running in a subprocess using a pipe, in order to
4248 try to track down a specific problem.
4250 4. Testing facility fudge: when running in the test harness and attempting
4251 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4252 getting "No route to host". Convert this to a timeout.
4254 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4257 6. Some OS don't have socklen_t but use size_t instead. This affects the
4258 fifth argument of getsockopt() amongst other things. This is now
4259 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4260 can be set for individual OS. I have set it for SunOS5, OSF1, and
4261 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4262 some earlier ones do not.
4264 7. Change 4.30/15 was not doing the test caselessly.
4266 8. The standard form for an IPv6 address literal was being rejected by address
4267 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4268 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4269 this, as well as the form without the "IPv6" on the front (but only when
4270 address literals are enabled, of course).
4272 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4274 10. Exim crashed if a message with an empty sender address specified by -f
4275 encountered a router with an errors_to setting. This could be provoked only
4276 by a command such as
4280 where an empty string was supplied; "<>" did not hit this bug.
4282 11. Installed PCRE release 4.5.
4284 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4285 remained set. It is now erased.
4287 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4288 times from message ids (which are base 36 rather than the normal 62).
4290 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4291 were being counted as actual protocol errors, and logged if the log
4292 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4293 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4294 and DATA following a set of rejected RCPTs do not count as protocol errors.
4295 In other words, Exim assumes they were pipelined, though this may not
4296 actually be the case. Of course, in all cases the client gets an
4297 appropriate error code.
4299 15. If a lookup fails in an ACL condition, a message about the failure may
4300 be available; it is used if testing the ACL cannot continue, because most
4301 such messages specify what the cause of the deferral is. However, some
4302 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4303 that caused an old message to be retained and used if a later statement
4304 caused a defer, replacing the real cause of the deferral.
4306 16. If an IP address had so many PTR records that the DNS lookup buffer
4307 was not large enough to hold them, Exim could crash while trying to process
4308 the truncated data. It now detects and logs this case.
4310 17. Further to 4.21/58, another change has been made: if (and only if) the
4311 first line of a message (the first header line) ends with CRLF, a bare LF
4312 in a subsequent header line has a space inserted after it, so as not to
4313 terminate the header.
4315 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4316 unnecessarily, used atoi() instead of strtol(), and didn't check the
4317 termination when getting file sizes from file names by regex.
4319 19. Completely re-implemented the support for maildirsize files, in the light
4320 of a number of problems with the previous contributed implementation
4321 (4.30/29). In particular:
4323 . If the quota is zero, the maildirsize file is maintained, but no quota is
4326 . If the maildir directory does not exist, it is created before any attempt
4327 to write a maildirsize file.
4329 . The quota value in the file is just a cache; if the quota is changed in
4330 the transport, the new value overrides.
4332 . A regular expression is available for excluding directories from the
4335 20. The autoreply transport checks the characters in options that define the
4336 message's headers; it allows continued headers, but it was checking with
4337 isspace() after an embedded newline instead of explicitly looking for a
4340 21. If all the "regular" hosts to which an address was routed had passed their
4341 expiry times, and had not reached their retry times, the address was
4342 bounced, even if fallback hosts were defined. Now Exim should go on to try
4345 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4346 equivalent code in the SMTP transport. Some hosts send humungous responses
4347 to HELO/EHLO, more than 1024 it seems.
4349 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4350 gives compiler warnings in some environments. I've now done it "properly",
4353 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4354 (because of problems with the built-in one) was declared to return uschar *
4355 instead of char *, causing compiler failure.
4357 25. Fixed a file descriptor leak when processing alias/forward files.
4359 26. Fixed a minor format string issue in dbfn.c.
4361 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4363 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4364 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4367 29. When a sender address is verified, it is cached, to save repeating the test
4368 when there is more than one recipient in a message. However, when the
4369 verification involves a callout, it is possible for different callout
4370 options to be set for different recipients. It is too complicated to keep
4371 track of this in the cache, so now Exim always runs a verification when a
4372 callout is required, relying on the callout cache for the optimization.
4373 The overhead is duplication of the address routing, but this should not be
4376 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4377 to be verified with callout=postmaster, and the main callout worked but the
4378 postmaster check failed, the verification correctly failed. However, if a
4379 subsequent RCPT command asked for sender verification *without* the
4380 postmaster check, incorrect caching caused this verification also to fail,
4383 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4384 it was not caching the DNS options (qualify_single, search_parents) that
4385 were used when the lookup failed. A subsequent lookup with different
4386 options therefore always gave the same answer, though there were cases
4387 where it should not have. (Example: a "domains = !$mx_any" option on a
4388 dnslookup router: the "domains" option is always processed without any
4389 widening, but the router might have qualify_single set.) Now Exim uses the
4390 cached value only when the same options are set.
4392 32. Added John Jetmore's "exipick" utility to the distribution.
4394 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4395 than a timeout (e.g. a certificate is required, and is not provided), an
4396 Exim server now closes the connection immediately. Previously it waited for
4397 the client to close - but if the client is SSL, it seems that they each
4398 wait for each other, leading to a delay before one of them times out.
4400 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4401 maintained 0.8.x compatibility because I don't think many are using it, and
4402 it is clearly obsolete.
4404 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4407 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4408 client certificate was expired. A simple patch fixes this, though I don't
4409 understand the full logic of why the verify callback is called multiple
4412 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4413 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4414 which causes problems with some clients (such as the Certicom SSL Plus
4415 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4416 disables the coutermeasure allowing Eudora to connect."
4418 38. Exim was not checking that a write() to a log file succeeded. This could
4419 lead to Bad Things if a log got too big, in particular if it hit a file
4420 size limit. Exim now panics and dies if it cannot write to a log file, just
4421 as it does if it cannot open a log file.
4423 39. Modified OS/Makefile-Linux so that it now contains
4425 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4427 The two -D definitions ensure that Exim is compiled with large file
4428 support, which makes it possible to handle log files that are bigger than
4431 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4432 instance) a domain was checked against a named list that involved a lookup,
4433 causing $domain_data to be set, then another domain was checked against the
4434 same list, then the first domain was re-checked, the value of $domain_data
4435 after the final check could be wrong. In particular, if the second check
4436 failed, it could be set empty. This bug probably also applied to
4439 41. The strip_trailing_dot option was not being applied to the address given
4440 with the -f command-line option.
4442 42. The code for reading a message's header from the spool was incrementing
4443 $received_count, but never initializing it. This meant that the value was
4444 incorrect (doubled) while delivering a message in the same process in which
4445 it was received. In the most common configuration of Exim, this never
4446 happens - a fresh exec is done - but it can happen when
4447 deliver_drop_privilege is set.
4449 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4450 it now includes up to 150 characters of the unexpected data in the log
4453 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4454 and building data strings. The size of both of these buffers was 10 000
4455 bytes - far larger than anybody would *ever* want, thought I. Needless to
4456 say, somebody hit the limit. I have increased the maximum line length to
4457 20 000 and the maximum data length of concatenated lines to 100 000. I have
4458 also fixed two bugs, because there was no checking on these buffers. Tsk,
4459 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4460 buffer is too small.
4462 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4463 lsearch lookups. Now it does.
4465 46. When parsing a route_list item in a manualroute router, a fixed-length
4466 buffer was used for the list of hosts. I made this 1024 bytes long,
4467 thinking that nobody would ever have a list of hosts that long. Wrong.
4468 Somebody had a whole pile of complicated expansion conditions, and the
4469 string was silently truncated, leading to an expansion error. It turns out
4470 that it is easier to change to an unlimited length (owing to other changes
4471 that have happened since this code was originally written) than to build
4472 structure for giving a limitation error. The length of the item that
4473 expands into the list of hosts is now unlimited.
4475 47. The lsearch lookup could not handle data where the length of text line was
4476 more than 4095 characters. Such lines were truncated, leading to shortened
4477 data being returned. It should now handle lines of any length.
4479 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4480 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4483 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4484 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4485 no longer happens in comment lines. A list of replacements is now placed
4486 at the head of all of the source files, except those whose only change is
4487 to replace PERL_COMMAND in the very first #! line.
4489 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4490 messages on the queue, with a bottom-up merge sort, using code contributed
4491 by Michael Haardt. This should make operations like -bp somewhat faster on
4492 large queues. It won't affect queue runners, except when queue_run_in_order
4495 51. Installed eximstats 1.31 in the distribution.
4497 52. Added support for SRV lookups to the dnslookup router.
4499 53. If an ACL referred to $message_body or $message_body_end, the value was not
4500 reset for any messages that followed in the same SMTP session.
4502 54. The store-handling optimization for building very long strings was not
4503 differentiating between the different store pools. I don't think this
4504 actually made any difference in practice, but I've tidied it.
4506 55. While running the routers to verify a sender address, $sender_address
4507 was still set to the sender address. This is wrong, because when routing to
4508 send a bounce to the sender, it would be empty. Therefore, I have changed
4509 it so that, while verifying a sender address, $sender_address is set to <>.
4510 (There is no change to what happens when verifying a recipient address.)
4512 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4513 target A or AAAA records (if not already returned) without resetting the
4514 qualify_single or search_parents options of the DNS resolver. These are
4515 inappropriate in this case because the targets of MX and SRV records must
4516 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4517 target that, when qualified, matched something in the local domain. These
4518 two options are now turned off when doing these lookups.
4520 57. It seems that at least some releases of Reiserfs (which does not have the
4521 concept of a fixed number of inodes) returns zero and not -1 for the
4522 number of available inodes. This interacted badly with check_spool_inodes,
4523 which assumed that -1 was the "no such thing" setting. What I have done is
4524 to check that the total number of inodes is greater than zero before doing
4525 the test of how many are available.
4527 58. When a "warn" ACL statement has a log_message modifier, the message is
4528 remembered, and not repeated. This is to avoid a lot of repetition when a
4529 message has many recipients that cause the same warning to be written.
4530 Howewer, Exim was preserving the list of already written lines for an
4531 entire SMTP session, which doesn't seem right. The memory is now reset if a
4532 new message is started.
4534 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4535 debugging output unless log_rewrite was also set.
4537 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4538 of (int)(handle) when we know that handle contains (void *)(-1).
4540 61. The Exim daemon panic-logs an error return when it closes the incoming
4541 connection. However "connection reset by peer" seems to be common, and
4542 isn't really an error worthy of noting specially, so that particular error
4545 62. When Exim is trying to find all the local interfaces, it used to panic and
4546 die if the ioctl to get the interface flags failed. However, it seems that
4547 on at least one OS (Solaris 9) it is possible to have an interface that is
4548 included in the list of interfaces, but for which you get a failure error
4549 for this call. This happens when the interface is not "plumbed" into a
4550 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4551 failure of the "get flags" call assumes that the interface is down.
4553 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4554 makes life easier for people who are doing arithmetic on fields extracted
4555 from dates, where you often get leading zeros that should not be
4556 interpreted as octal.
4558 64. Added qualify_domain to the redirect router, to override the global
4561 65. If a pathologically long header line contained very many addresses (the
4562 report of this problem mentioned 10 000) and each of them was rewritten,
4563 Exim could use up a very large amount of memory. (It kept on making new
4564 copies of the header line as it rewrote, and never released the old ones.)
4565 At the expense of a bit more processing, the header rewriting function has
4566 been changed so that it no longer eats memory in this way.
4568 66. The generation of the Received: header has been moved from the time that a
4569 message starts to be received, to the time that it finishes. The timestamp
4570 in the Received: header should now be very close to that of the <= log
4571 line. There are two side-effects of this change:
4573 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4574 logged header lines no longer include the local Received: line, because
4575 it has not yet been created. The same applies to a copy of the message
4576 that is returned to a non-SMTP sender when a message is rejected.
4578 (b) When a filter file is tested using -bf, no additional Received: header
4579 is added to the test message. After some thought, I decided that this
4582 This change does not affect the value of $received_for. It is still set
4583 after address rewriting, but before local_scan() is called.
4585 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4587 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4588 gave an unhelpful panic error message, and a defer error. I have managed to
4589 change this behaviour so that it now rejects any supplied certificate,
4590 which seems right, as the list of acceptable certificates is empty.
4592 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4593 gave an unhelpful defer error. I have not managed to make this reject any
4594 supplied certificates, but the error message it gives is "no certificate
4595 supplied", which is not helpful.
4597 70. exigrep's output now also includes lines that are not associated with any
4598 message, but which match the given pattern. Implemented by a patch from
4599 Martin Sluka, which also tidied up the Perl a bit.
4601 71. Recipient callout verification, like sender verification, was using <> in
4602 the MAIL FROM command. This isn't really the right thing, since the actual
4603 sender may affect whether the remote host accepts the recipient or not. I
4604 have changed it to use the actual sender in the callout; this means that
4605 the cache record is now keyed on a recipient/sender pair, not just the
4606 recipient address. There doesn't seem to be a real danger of callout loops,
4607 since a callout by the remote host to check the sender would use <>.
4608 [SEE ABOVE: changed after hitting problems.]
4610 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4611 temporary errors. However, in the case of such a code being given after
4612 the end of a data transmission (i.e. after ".") Exim was failing to write
4613 a retry record for the message. (Yes, there was some broken host that was
4614 actually sending 8xx at this point.)
4616 73. An unknown lookup type in a host list could cause Exim to panic-die when
4617 the list was checked. (An example that provoked this was putting <; in the
4618 middle of a list instead of at the start.) If this happened during a DATA
4619 ACL check, a -D file could be left lying around. This kind of configuration
4620 error no longer causes Exim to die; instead it causes a defer errror. The
4621 incident is still logged to the main and panic logs.
4623 74. Buglet left over from Exim 3 conversion. The message "too many messages
4624 in one connection" was written to the rejectlog but not the mainlog, except
4625 when address rewriting (yes!) was being logged.
4627 75. Added write_rejectlog option.
4629 76. When a system filter was run not as root (that is, when system_filter_user
4630 was set), the values of the $n variables were not being returned to the
4631 main process; thus, they were not subsequently available in the $sn
4634 77. Added +return_path_on_delivery log selector.
4636 78. A connection timeout was being treated differently from recipients deferred
4637 when testing hosts_max_try with a message that was older than the host's
4638 retry timeout. (The host should not be counted, thus allowing all hosts to
4639 be tried at least once before bouncing.) This may have been the cause of an
4640 occasionally reported bug whereby a message would remain on the queue
4641 longer than the retry timeout, but would be bounced if a delivery was
4642 forced. I say "may" because I never totally pinned down the problem;
4643 setting up timeout/retry tests is difficult. See also the next item.
4645 79. The ultimate address timeout was not being applied to errors that involved
4646 a combination of host plus message (for example, a timeout on a MAIL
4647 command). When an address resolved to a number of possible hosts, and they
4648 were not all tried for each delivery (e.g. because of hosts_max_try), a
4649 message could remain on the queue longer than the retry timeout.
4651 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4654 81. Fixed an obscure SMTP outgoing bug which required at least the following
4655 conditions: (a) there was another message waiting for the same server;
4656 (b) the server returned 5xx to all RCPT commands in the first message so
4657 that the message was not completed; (c) the server dropped the connection
4658 or gave a negative response to the RSET that Exim sends to abort the
4659 transaction. The observed case was a dropped connection after DATA that had
4660 been sent in pipelining mode. That is, the server had advertised PIPELINING
4661 but was not implementing it correctly. The effect of the bug was incorrect
4662 behaviour, such as trying another host, and this could lead to a crash.
4668 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4669 and daemon.c were passed as pointers to ints; they should have been
4670 pointers to socklen_t variables (which are typically unsigned ints).
4672 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4675 3. Fixed a really odd bug that affected only the testing scheme; patching a
4676 certain fixed string in the binary changed the value of another string that
4677 happened to be identical to the end of the original first string.
4679 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4680 name", it returns that address as the IP address. On some operating
4681 systems (e.g. Solaris), it also passes back the IP address string as the
4682 "host name". However, on others (e.g. Linux), it passes back an empty
4683 string. Exim wasn't checking for this, and was changing the host name to an
4684 empty string, assuming it had been canonicized.
4686 5. Although rare, it is permitted to have more than one PTR record for a given
4687 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4688 all the names associated with an address, because they do in Solaris.
4689 However, it seems that they do not in Linux for data that comes from the
4690 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4691 I found this out when I moved to a new Linux workstation and tried to run
4692 the Exim test suite.
4694 To get round this problem I have changed the code so that it now does its
4695 own call to the DNS to look up PTR records when searching for a host name.
4696 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4697 addresses that are only in /etc/hosts are still found.
4699 This behaviour is, however, controlled by an option called host_lookup_
4700 order, which defaults to "bydns:byaddr". If people want to use the other
4701 order, or indeed, just use one or the other means of lookup, they can
4702 specify it in this variable.
4704 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4705 some operating systems, this comes back from gethostbyaddr() as an empty
4706 string, and this is what Exim used to test for. However, it seems that in
4707 other systems, "." is yielded. Exim now tests for this case too.
4709 7. The values of check_spool_space and check_log_space are now held internally
4710 as a number of kilobytes instead of an absolute number of bytes. If a
4711 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4712 kilobyte. This means that much larger values can be stored.
4714 8. Exim monitor: an attempt to get the action menu when not actually pointing
4715 at a message produces an empty menu entitled "No message selected". This
4716 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4717 no entries in it ("Shell widget menu has zero width and/or height"). So I
4718 have added a single, blank menu entry in this case.
4720 9. Added ${quote_local_part.
4722 10. MIME decoding is now applied to the contents of Subject: header lines when
4725 11. Now that a reference to $sender_host_address automatically causes a reverse
4726 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4727 host lookup before query-style lookups in lists that might use this
4728 variable. This has therefore been abolished, and the "net-" prefix is no
4729 longer necessary for query-style lookups.
4731 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4732 have been a typo for LFLAGS, so it has been changed.
4734 13. The install script calls Exim with "-C /dev/null" in order to find the
4735 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4736 to be output. Howeve, since Exim outputs its version number before the
4737 error, it didn't break the script. It just looked ugly. I fixed this by
4738 always allowing "-C /dev/null" if the caller is root.
4740 14. Ignore overlarge ACL variable number when reading spool file - insurance
4741 against a later release with more variables having written the file.
4743 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4744 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4745 this, as well as the form without the "IPv6" on the front.
4747 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4748 OS/Makefile-Darwin file.
4750 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4751 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4753 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4754 variable. However, some OS use ints for inodes. I've added cast to long int
4755 to get rid of the compiler warning.
4757 19. I had forgotten to lock out "/../" in configuration file names when
4758 ALT_CONFIG_PREFIX was set.
4760 20. Routers used for verification do not need to specify transports. However,
4761 if such a router generated a host list, and callout was configured, Exim
4762 crashed, because it could not find a port number from the (non-existent)
4763 transport. It now assumes port 25 in this circumstance.
4765 21. Added the -t option to exigrep.
4767 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4768 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4769 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4770 them is set, LOOKUP_LSEARCH is forced.
4772 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4774 24. Added sender and host information to the "rejected by local_scan()" log
4775 line; previously there was no indication of these.
4777 25. Added .include_if_exists.
4779 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4780 Linux. It turns out that not all file systems support this. Apparently some
4781 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4782 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4783 not supported on the file descriptor, is now ignored when Exim is trying to
4784 sync a directory. This applies only to Linux.
4786 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4788 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4789 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4792 29. Added support for maildirsize files from supplied patch (modified a bit).
4794 30. The use of :fail: followed by an empty string could lead Exim to respond to
4795 sender verification failures with (e.g.):
4797 550 Verification failed for <xxx>
4798 550 Sender verify failed
4800 where the first response line was missing the '-' that indicates it is not
4801 the final line of the response.
4803 31. The loop for finding the name of the user that called Exim had a hardwired
4804 limit of 10; it now uses the value of finduser_retries, which is used for
4805 all other user lookups.
4807 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4810 33. Exim was neglecting to zero errno before one call of strtol() when
4811 expanding a string and expecting an integer value. On some systems this
4812 resulted in spurious "integer overflow" errors. Also, it was casting the
4813 result into an int without checking.
4815 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4816 did not work. The code looks as if it has *never* worked, though it appears
4817 to have been documented since at least releast 1.62. I have made it work.
4819 35. The "timeout_DNS" error in retry rules, also documented since at least
4820 1.62, also never worked. As it isn't clear exactly what this means, and
4821 clearly it isn't a major issue, I have abolished the feature by treating it
4822 as "timeout", and writing a warning to the main and panic logs.
4824 36. The display of retry rules for -brt wasn't always showing the error code
4827 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4828 timeout_connect_A, timeout_connect_MX.
4830 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4831 to the empty sender.
4833 39. The daemon was not analysing the content of -oX till after it had closed
4834 stderr and disconnected from the controlling terminal. This meant that any
4835 syntax errors were only noted on the panic log, and the return code from
4836 the command was 0. By re-arranging the code a little, I've made the
4837 decoding happen first, so such errors now appear on stderr, and the return
4838 code is 1. However, the actual setting up of the sockets still happens in
4839 the disconnected process, so errors there are still only recorded on the
4842 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4843 connections (as happens on some IP stacks) was logged at start up time as
4844 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4845 it from "IPv6 and IPv4", which means that two separate sockets are being
4848 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4849 says whether AF_INET or AF_INET6 was passed as an argument.
4851 42. Exiwhat output was messed up when time zones were included in log
4854 43. Exiwhat now gives more information about the daemon's listening ports,
4855 and whether -tls-on-connect was used.
4857 44. The "port" option of the smtp transport is now expanded.
4859 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4860 silently ignored. Now an error message is written to the main and panic
4863 46. There's a new ACL modifier called "logwrite" which writes to a log file
4864 as soon as it is encountered.
4866 47. Added $local_user_uid and $local_user_gid at routing time.
4868 48. Exim crashed when trying to verify a sender address that was being
4871 49. Exim was recognizing only a space character after ".include". It now also
4872 recognizes a tab character.
4874 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4875 extracting the relevant information from the specification. The man page no
4876 longer contains scrambled data for the -d option, and I've added a section
4877 at the front about calling Exim under different names.
4879 51. Added "extra_headers" argument to the "mail" command in filter files.
4881 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4884 53. Installed eximstats 1.29.
4886 54. Added transport_filter_timeout as a generic transport option.
4888 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4889 Cc: header lines. This was required by RFC 822, but it not required by RFC
4892 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4893 incoming messages that did not have them. Now it does so only if the
4894 message originates locally, that is, if there is no associated remote host
4895 address. When Resent- header lines are present, this applies to the Resent-
4896 lines rather than the non-Resent- lines.
4898 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4899 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4901 58. Messages for configuration errors now include the name of the main
4902 configuration file - useful now that there may be more than one file in a
4903 list (.included file names were always shown).
4905 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4906 for those rare installations that do not start the daemon as root or run it
4907 setuid root. I've cut out the call to initgroups() if the daemon is not
4910 60. The Exim user and group can now be bound into the binary as text strings
4911 that are looked up at the start of Exim's processing.
4913 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4915 62. Added $mailstore_basename variable.
4917 63. Installed patch to sieve.c from Michael Haardt.
4919 64. When Exim failed to open the panic log after failing to open the main log,
4920 the original message it was trying to log was written to stderr and debug
4921 output, but if they were not available (the usual case in production), it
4922 was lost. Now it is written to syslog before the two lines that record the
4923 failures to open the logs.
4925 65. Users' Exim filters run in subprocesses under the user's uid. It is
4926 possible for a "deliver" command or an alias in a "personal" command to
4927 provoke an address rewrite. If logging of address rewriting is configured,
4928 this fails because the process is not running as root or exim. There may be
4929 a better way of dealing with this, but for the moment (because 4.30 needs
4930 to be released), I have disabled address rewrite logging when running a
4931 filter in a non-root, non-exim process.
4937 1. The buildconfig auxiliary program wasn't quoting the value set for
4938 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4939 not defined. This bug was masked in 4.22 by the effect that was fixed in
4942 2. Some messages that were rejected after a message id was allocated were
4943 shown as "incomplete" by exigrep. It no longer does this for messages that
4944 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4946 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4947 have allow_domain_literals set, the ID did not get logged in the <= line.
4948 Domain literals are now always recognized in Message-ID: header lines.
4950 4. The first argument for a ${extract expansion item is the key name or field
4951 number. Leading and trailing spaces in this item were not being ignored,
4952 causing some misleading effects.
4954 5. When deliver_drop_privilege was set, single queue runner processes started
4955 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4956 same command in the process it spins off) were not dropping privilege.
4958 6. When the daemon running as "exim" started a queue runner, it always
4959 re-executed Exim in the spun-off process. This is a waste of effort when
4960 deliver_drop_privilege is set. The new process now just calls the
4961 queue-runner function directly.
4967 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4970 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4971 ignored. Though the use of -oP was forcing the writing of a pid file, it
4972 was always written to the default place.
4974 3. If the message "no IP address found for host xxxx" is generated during
4975 incoming verification, it is now followed by identification of the incoming
4976 connection (so you can more easily find what provoked it).
4978 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4980 5. Added some features to "harden" Exim a bit more against certain attacks:
4982 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4983 be put in Local/Makefile. This is like the never_users runtime option,
4984 but it cannot be overridden. The default setting is "root".
4986 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4987 prefix string with which any file named in a -C command line option
4990 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4991 is retained for -C and -D only if the caller of Exim is root. Without
4992 it, the exim user may also use -C and -D and retain privilege.
4994 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4995 command line option is disabled.
4997 6. Macro names set by the -D option must start with an upper case letter, just
4998 like macro names defined in the configuration file.
5000 7. Added "dereference=" facility to LDAP.
5002 8. Two instances of the typo "uknown" in the source files are fixed.
5004 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5005 the Configure-Makefile script screwed up while processing it.
5007 10. Incorporated PCRE 4.4.
5009 11. The SMTP synchronization check was not operating right at the start of an
5010 SMTP session. For example, it could not catch a HELO sent before the client
5011 waited for the greeting. There is now a check for outstanding input at the
5012 point when the greeting is written. Because of the duplex, asynchronous
5013 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5014 way, but not yet received, when the check is performed.
5016 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5017 on TCP/IP sockets, because this apparently causes some broken clients to
5020 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5021 unchanged) from the Cygwin maintainer.
5023 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5024 was supported instead of testing for mbx. Effectively a typo.
5026 15. The spa authenticator server code was not checking that the input it
5027 received was valid base64.
5029 16. The debug output line for the "set" modifier in ACLs was not showing the
5030 name of the variable that was being set.
5032 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5034 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5035 lookup on its own. Something else was needed to trigger the lookup. For
5036 example, a match in host_lookup or the need for a host name in a host list.
5037 Now, if $sender_host_name is referenced and the host name has not yet been
5038 looked up, a lookup is performed. If the lookup fails, the variable remains
5039 empty, and $host_lookup_failed is set to "1".
5041 19. Added "eqi" as a case-independent comparison operator.
5043 20. The saslauthd authentication condition could segfault if neither service
5044 nor realm was specified.
5046 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5047 error message that was logged was misleading, and incoming SMTP
5048 connections were dropped. The message is now more accurate, and temporary
5049 errors are given to SMTP connections.
5051 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5052 (except RSET) with a 421 error, until QUIT is received. However, it was
5053 failing to send a response to QUIT.
5055 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5056 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5057 ACL is now run after the helo_try_verify_hosts code.
5059 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5060 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5061 case-independent, so other case variants are also recognized). Apparently
5062 some systems use these upper case variants.
5064 25. If more than two messages were waiting for the same host, and a transport
5065 filter was specified for the transport, Exim sent two messages over the
5066 same TCP/IP connection, and then failed with "socket operation on non-
5067 socket" when it tried to send the third.
5069 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5071 27. The extern definition of crypt16() in expand.c was not being excluded when
5072 the OS had its own crypt16() function.
5074 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5075 as a preferred synonym for return_size_limit, both as an option and as an
5078 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5080 30. Changed the default configuration ACL to relax the local part checking rule
5081 for addresses that are not in any local domains. For these addresses,
5082 slashes and pipe symbols are allowed within local parts, but the sequence
5083 /../ is explicitly forbidden.
5085 31. SPA server authentication was not clearing the challenge buffer before
5088 32. log_message in a "warn" ACL statement was writing to the reject log as
5089 well as to the main log, which contradicts the documentation and doesn't
5090 seem right (because no rejection is happening). So I have stopped it.
5092 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5093 However, I am unable to do any testing of this.
5095 34. Fixed an infelicity in the appendfile transport. When checking directories
5096 for a mailbox, to see if any needed to be created, it was accidentally
5097 using path names with one or more superfluous leading slashes; tracing
5098 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5100 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5101 contents are added to the log line that is written for every discarded
5102 recipient. (Previously a log_message setting was ignored.)
5104 36. The ${quote: operator now quotes the string if it is empty.
5106 37. The install script runs exim in order to find its version number. If for
5107 some reason other than non-existence or emptiness, which it checks, it
5108 could not run './exim', it was installing it with an empty version number,
5109 i.e. as "exim-". This error state is now caught, and the installation is
5112 38. An argument was missing from the function that creates an error message
5113 when Exim fails to connect to the socket for saslauthd authentication.
5114 This could cause Exim to crash, or give a corrupted message.
5116 39. Added isip, isip4, and isip6 to ${if conditions.
5118 40. The ACL variables $acl_xx are now saved with the message, and can be
5119 accessed later in routers, transports, and filters.
5121 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5122 strings in the file are not string-expanded.
5124 42. If a MAIL command specified a SIZE value that was too large to fit into an
5125 int variable, the check against message_size_limit failed. Such values are
5126 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5127 day this will have to be increased, but I don't think I want to be around
5128 when emails are that large.
5135 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5136 iconv() is not standard in FreeBSD.
5138 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5139 IPv6 enabled. The observed symptom was a segmentation fault on return from
5140 the function os_common_find_running_interfaces() in src/os.c.
5142 3. In the check_special_case() function in daemon.c I had used "errno" as an
5143 argument name, which causes warnings on some systems. This was basically a
5144 typo, since it was named "eno" in the comments!
5146 4. The code that waits for the clock to tick (at a resolution of some fraction
5147 of a second) so as to ensure message-id uniqueness was always waiting for
5148 at least one whole tick, when it could have waited for less. [This is
5149 almost certainly not relevant at current processor speeds, where it is
5150 unlikely to ever wait at all. But we try to future-proof.]
5152 5. The function that sleeps for a time interval that includes fractions of a
5153 second contained a race. It did not block SIGALRM between setting the
5154 timer, and suspending (a couple of lines later). If the interval was short
5155 and the sigsuspend() was delayed until after it had expired, the suspension
5156 never ended. On busy systems this could lead to processes getting stuck for
5159 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5160 process, before it forks any delivery processes. The open lookup caching
5161 mechanism meant that the open file or database connection was passed into
5162 the delivery process. The problem was that delivery processes always tidy
5163 up cached lookup data. This could cause a problem for the next delivery
5164 process started by the queue runner, because the external queue runner
5165 process does not know about the closure. So the next delivery process
5166 still has data in the lookup cache. In the case of a file lookup, there was
5167 no problem because closing a file descriptor in a subprocess doesn't affect
5168 the parent. However, if the lookup was caching a connection to a database,
5169 the connection was closed, and the second delivery process was likely to
5170 see errors such as "PGSQL: query failed: server closed the connection
5171 unexpectedly". The problem has been fixed by closing all cached lookups
5172 in a queue runner before running a delivery process.
5174 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5175 seem to have the "const" qualifier which it has on other OS. I've
5178 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5179 *of the same type* (client or server) with the same public name that an
5180 error should be diagnosed.
5182 9. When Exim looked up a host name for an IP address, but failed to find the
5183 original IP address when looking up the host name (a safety check), it
5184 output the message "<ip address> does not match any IP for NULL", which was
5185 confusing, to say the least. The bug was that the host name should have
5186 appeared instead of "NULL".
5188 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5189 user that is built into the binary, and the -C or -D options is used, root
5190 privilege is dropped before the configuration file is read. In addition,
5191 logging is switched to stderr instead of the normal log files. If the
5192 configuration then re-defines the Exim user, the unprivileged environment
5193 is probably not what is expected, so Exim logs a panic warning message (but
5196 However, if deliver_drop_privilege is set, the unprivileged state may well
5197 be exactly what is intended, so the warning has been cut out in that case,
5198 and Exim is allowed to try to write to its normal log files.
5204 1. smtp_return_error_details was not giving details for temporary sender
5205 or receiver verification errors.
5207 2. Diagnose a configuration error if two authenticators have the same public
5210 3. Exim used not to create the message log file for a message until the first
5211 delivery attempt. This could be confusing when incoming messages were held
5212 for policy or load reasons. The message log file is now created at the time
5213 the message is received, and an initial "Received" line is written to it.
5215 4. The automatically generated man page for command line options had a minor
5216 bug that caused no ill effects; however, a more serious problem was that
5217 the procedure for building the man page automatically didn't always
5218 operate. Consequently, release 4.20 contains an out-of-date version. This
5219 shouldn't happen again.
5221 5. When building Exim with embedded Perl support, the script that builds the
5222 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5223 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5225 6. The freeze_tell option was not being used for messages that were frozen on
5226 arrival, either by an ACL or by local_scan().
5228 7. Added the smtp_incomplete_transaction log selector.
5230 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5231 was accepting AUTH without a new EHLO.
5233 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5234 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5237 10. Logging of TCP/IP connections (when configured) now happens in the main
5238 daemon process instead of the child process, so that the TCP/IP connection
5239 count is more accurate (but it can never be perfect).
5241 11. The use of "drop" in a nested ACL was not being handled correctly in the
5242 outer ACL. Now, if condition failure induced by the nested "drop" causes
5243 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5244 or "require"), the connection is dropped.
5246 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5247 that yield "discard" can now be used with an "accept" or a "discard" verb,
5248 but an error is generated for any others (because I can't see a useful way
5249 to define what should happen).
5251 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5252 are now processed in the same way as lines in the Exim configuration file.
5253 In particular, continuation lines are supported.
5255 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5257 15. Added -ti meaning -t -i.
5259 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5260 domains, and warn by logging if others are found.
5262 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5263 returned from GIFCONF when Exim is trying to find the list of interfaces on
5264 a host. The code in os.c has been modified to copy each ifreq to an aligned
5265 structure in all cases.
5267 Also, in some cases, the returned ifreq's were being copied to a 'struct
5268 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5269 means the last couple of bytes of an IPv6 address could be chopped if the
5270 ifreq contained only a normal sockaddr (14 bytes storage).
5272 18. Named domain lists were not supported in the hosts_treat_as_local option.
5273 An entry such as +xxxx was not recognized, and was treated as a literal
5276 19. Ensure that header lines added by a DATA ACL are included in the reject log
5277 if the ACL subsequently rejects the message.
5279 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5280 MD5 (which is deprecated).
5282 21. When testing a filter file using -bf, Exim was writing a message when it
5283 took the sender from a "From " line in the message, but it was not doing so
5284 when it took $return_path from a Return-Path: header line. It now does.
5286 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5287 with a valid header line field name (a series of printing characters
5288 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5290 23. Changed "disc" in the source to "disk" to conform to the documentation and
5291 the book and for uniformity.
5293 24. Ignore Sendmail's -Ooption=value command line item.
5295 25. When execve() failed while trying to run a command in a pipe transport,
5296 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5297 could be confused with a return value of 69 from the command itself. This
5298 has been changed to 127, the value the shell returns if it is asked to run
5299 a non-existent command. The wording for the related log line suggests a
5300 non-existent command as the problem.
5302 26. If received_header_text expands to an empty string, do not add a Received:
5303 header line to the message. (Well, it adds a token one on the spool, but
5304 marks it "old" so that it doesn't get used or transmitted.)
5306 27. Installed eximstats 1.28 (addition of -nt option).
5308 28. There was no check for failure on the call to getsockname() in the daemon
5309 code. This can fail if there is a shortage of resources on the system, with
5310 ENOMEM, for example. A temporary error is now given on failure.
5312 29. Contrary to the C standard, it seems that in some environments, the
5313 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5314 program. Exim now does this explicitly; it affects the formatting of
5315 timestamps using strftime().
5317 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5318 complaints. I've now initialized all the variables, to avoid this.
5320 32. Header lines added by a system filter were not being "seen" during
5321 transport-time rewrites.
5323 33. The info_callback() function passed to OpenSSL is set up with type void
5324 (*)(SSL *, int, int), as described somewhere. However, when calling the
5325 function (actually a macro) that sets it up, the type void(*)() is
5326 expected. I've put in a cast to prevent warnings from picky compilers.
5328 34. If a DNS black list lookup found a CNAME record, but there were no A
5329 records associated with the domain it pointed at, Exim crashed.
5331 35. If a DNS black list lookup returned more than one A record, Exim ignored
5332 all but the first. It now scans all returned addresses if a particular IP
5333 value is being sought. In this situation, the contents of the
5334 $dnslist_value variable are a list of all the addresses, separated by a
5337 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5338 to accept a host name and all its aliases if the forward lookup for any of
5339 them yielded the IP address of the incoming connection. Now it accepts only
5340 those names whose forward lookup yields the correct IP address. Any other
5341 names are discarded. This closes a loophole whereby a rogue DNS
5342 administrator could create reverse DNS records to break through a
5343 wildcarded host restriction in an ACL.
5345 37. If a user filter or a system filter that ran in a subprocess used any of
5346 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5347 the wrong values were passed to the pipe command ($thisaddress had the
5348 value of $0, $0 had the value of $1, etc). This bug was introduced by
5349 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5351 38. Improved the line breaking for long SMTP error messages from ACLs.
5352 Previously, if there was no break point between 40 and 75 characters, Exim
5353 left the rest of the message alone. Two changes have been made: (a) I've
5354 reduced the minimum length to 35 characters; (b) if it can't find a break
5355 point between 35 and 75 characters, it looks ahead and uses the first one
5356 that it finds. This may give the occasional overlong line, but at least the
5357 remaining text gets split now.
5359 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5360 file descriptors might be low, and that setting 1000 would always raise it.
5361 It turns out that in some environments, the limit is already over 1000 and
5362 that lowering it causes trouble. So now Exim takes care not to decrease it.
5364 40. When delivering a message, the value of $return_path is set to $sender_
5365 address at the start of routing (routers may change the value). By an
5366 oversight, this default was not being set up when an address was tested by
5367 -bt or -bv, which affected the outcome if any router or filter referred to
5370 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5371 log only when "message" or "log_message" are set. However, if one of the
5372 conditions was an address verification, or a call to a nested ACL, the
5373 messages generated by the underlying test were being passed through. This
5374 no longer happens. The underlying message is available in $acl_verify_
5375 message for both "message" and "log_message" expansions, so it can be
5376 passed through if needed.
5378 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5379 new expansion $bh_ to give the encoded byte string without charset
5380 translation. Translation happens only if iconv() is available; HAVE_ICONV
5381 indicates this at build time. HEADERS_CHARSET gives the charset to
5382 translate to; headers_charset can change it in the configuration, and
5383 "headers charset" can change it in an individual filter file.
5385 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5386 that creates RFC 2047 encoded "words" labels them as that charset instead
5387 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5388 expansion operator; (ii) when Exim creates a From: line for a local
5389 message; (iii) when a header line is rewritten to include a "phrase" part.
5391 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5392 buggy, causing it to skip the first lines of messages whose message ID
5393 ended in 'D'. This would not have bitten before Exim release 4.14, because
5394 message IDs were unlikely to end in 'D' before then. The effect was to have
5395 incorrect size information for certain domains.
5397 45. #include "config.h" was missing at the start of the crypt16.c module. This
5398 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5401 46. If there was a timeout during a "random" callout check, Exim treated it as
5402 a failure of the random address, and carried on sending RSET and the real
5403 address. If the delay was just some slowness somewhere, the response to the
5404 original RCPT would be taken as a response to RSET and so on, causing
5405 mayhem of various kinds.
5407 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5408 when I implemented it. It didn't allow for the fact that some option values
5409 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5410 the right test for positive values.
5412 48. Domain names in DNS records are case-independent. Exim always looks them up
5413 in lower case. Some resolvers return domain names in exactly the case they
5414 appear in the zone file, that is, they may contain uppercase letters. Not
5415 all resolvers do this - some return always lower case. Exim was treating a
5416 change of case by a resolver as a change of domain, similar to a widening
5417 of a domain abbreviation. This triggered its re-routing code and so it was
5418 trying to route what was effectively the same domain again. This normally
5419 caused routing to fail (because the router wouldn't handle the domain
5420 twice). Now Exim checks for this case specially, and just changes the
5421 casing of the domain that it ultimately uses when it transmits the message
5424 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5427 50. If a filter generated a file delivery with a non-absolute name (possible if
5428 no home directory exists for the router), the forbid_file option was not
5431 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5432 the existing equality matching.
5434 52. Exim was using ints instead of ino_t variables in some places where it was
5435 dealing with inode numbers.
5437 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5438 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5439 called TMPDIR, and if it finds it is different, it changes its value.
5441 54. The smtp_printf() function is now made available to local_scan() so
5442 additional output lines can be written before returning. There is also an
5443 smtp_fflush() function to enable the detection of a dropped connection.
5444 The variables smtp_input and smtp_batched_input are exported to
5447 55. Changed the default runtime configuration: the message "Unknown user"
5448 has been removed from the ACL, and instead placed on the localuser router,
5449 using the cannot_route_message feature. This means that any verification
5450 failures that generate their own messages won't get overridden. Similarly,
5451 the "Unrouteable address" message that was in the ACL for unverifiable
5452 relay addresses has also been removed.
5454 56. Added hosts_avoid_esmtp to the smtp transport.
5456 57. The exicyclog script was not checking for the esoteric option
5457 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5458 will work only if exicyclog is run under the appropriate euid.
5460 58. Following a discussion on the list, the rules by which Exim recognises line
5461 endings on incoming messages have been changed. The -dropcr and drop_cr
5462 options are now no-ops, retained only for backwards compatibility. The
5463 following line terminators are recognized: LF CRLF CR. However, special
5464 processing applies to CR:
5466 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5467 nor a local message in the state where . is a terminator.
5469 (ii) If a bare CR is encountered in a header line, an extra space is added
5470 after the line terminator so as not to end the header. The reasoning
5471 behind this is that bare CRs in header lines are most likely either
5472 to be mistakes, or people trying to play silly games.
5474 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5475 was being incorrectly given as 18 bytes larger than it should have been.
5476 This is a VOB (very old bug).
5478 60. This may never have affected anything current, but just in case it has:
5479 When the local host is found other than at the start of a list of hosts,
5480 the local host, those with the same MX, and any that follow, are discarded.
5481 When the list in question was part of a longer list of hosts, the following
5482 hosts (not currently being processed) were also being discarded. This no
5483 longer happens. I'm not sure if this situation could ever has previously
5486 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5489 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5490 header. This is something that is recommended in a new Internet Draft, and
5491 is something that is documented as being done by Sendmail. There are two
5492 possible values. For messages generated by the autoreply transport, Exim
5495 Auto-Submitted: auto-replied
5497 whereas for all other generated messages (e.g. bounces) it adds
5499 Auto-Submitted: auto-generated
5501 63. The "personal" condition in filters now includes a test for the
5502 Auto-Submitted: header. If it contains the string "auto-" the message it
5503 not considered personal.
5505 64. Added rcpt_include_affixes as a generic transport option.
5507 65. Added queue_only_override (default true).
5509 66. Added the syslog_duplication option.
5511 67. If what should have been the first header line of a message consisted of
5512 a space followed by a colon, Exim was mis-interpreting it as a header line.
5513 It isn't of course - it is syntactically invalid and should therefore be
5514 treated as the start of the message body. The misbehaviour could have
5515 caused a number of strange effects, including loss of data in subsequent
5516 header lines, and spool format errors.
5518 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5519 client host had authenticated. This control can now be exercised by an ACL
5520 for more flexibility.
5522 69. By default, callouts do not happen when testing with -bh. There is now a
5523 variant, -bhc, which does actually run the callout code, including
5524 consulting and updating the callout cache.
5526 70. Added support for saslauthd authentication, courtesy of Alexander
5529 71. If statvfs() failed on the spool or log directories while checking their
5530 size for availability, Exim confusingly gave the error "space shortage".
5531 Furthermore, in debugging mode it crashed with a floating point exception.
5532 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5533 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5534 serious problem, Exim now writes to the main and panic logs when this
5535 happens, with details of the failure. It then refuses to accept the
5536 incoming message, giving the message "spool directory problem" or "log
5537 directory problem" with a 421 code for SMTP messages.
5539 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5540 0, 1, and 2 exist, because some OS complain for execs without them (see
5541 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5542 descriptors. However, the code omitted to check that the open succeeded,
5543 causing mysterious errors if for some reason the permissions on /dev/null
5544 got screwed. Now Exim writes a message to the main and panic logs, and
5545 bombs out if it can't open /dev/null.
5547 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5548 interact so that it is all more flexible. It is supposed to remain
5549 backwards compatible. Also added extra_local_interfaces.
5551 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5552 to bomb out with an assertion failure - to the client this appears as a
5553 connection drop. This problem occurs in the part of the code that was taken
5554 from the Samba project. Fortunately, the assertion is in a very simple
5555 function, so I have fixed this by reproducing the function inline in the
5556 one place where it is called, and arranging for authentication to fail
5557 instead of killing the process with assert().
5559 75. The SPA client code was not working when the server requested OEM rather
5560 than Unicode encoding.
5562 76. Added code to make require_files with a specific uid setting more usable in
5563 the case where statting the file as root fails - usually a non-root-mounted
5564 NFS file system. When this happens and the failure is EACCES, Exim now
5565 forks a subprocess and does the per-uid checking as the relevant uid.
5567 77. Added process_log_path.
5569 78. If log_file_path was not explicitly set, a setting of check_log_space or
5570 check_log_inodes was ignored.
5572 79. If a space check for the spool or log partitions fails, the incident is now
5573 logged. Of course, in the latter case the data may get lost...
5575 80. Added the %p formatting code to string_format() so that it can be used to
5576 print addresses in debug_print(). Adjusted all the address printing in the
5577 debugging in store.c to use %p rather than %d.
5579 81. There was a concern that a line of code in smtp_in.c could overflow a
5580 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5581 initially expressed, the concern was not well-founded, because trailing
5582 spaces are removed early. However, if the trailing spaces were followed by
5583 a NULL, they did not get removed, so the overflow was possible. Two fixes
5586 (a) I re-wrote the offending code in a cleaner fashion.
5587 (b) If an incoming SMTP command contains a NULL character, it is rejected
5590 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5591 runs initgroups(), so that if the Exim user is in any additional groups,
5592 they will be used during message reception.
5598 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob