1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage no indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
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.
96 PP/01 Solaris build fix for Oracle's LDAP libraries.
97 Bugzilla 1109, patch from Stephen Usher.
99 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
101 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
104 TF/02 Fix a couple more cases where we did not log the error message
105 when unlink() failed. See also change 4.74-TF/03.
107 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
108 lock up or crash if it happened to be inside a call to libc when it
109 got a SIGUSR1 from exiwhat.
111 The SIGUSR1 handler appends the current process status to the process
112 log which is later printed by exiwhat. It used to use the general
113 purpose logging code to do this, but several functions it calls are
114 not safe for signals.
116 The new output code in the SIGUSR1 handler is specific to the process
117 log, and simple enough that it's easy to inspect for signal safety.
118 Removing some special cases also simplifies the general logging code.
119 Removing the spurious timestamps from the process log simplifies
122 TF/04 Improved ratelimit ACL condition.
124 The /noupdate option has been deprecated in favour of /readonly which
125 has clearer semantics. The /leaky, /strict, and /readonly update modes
126 are mutually exclusive. The update mode is no longer included in the
127 database key; it just determines when the database is updated. (This
128 means that when you upgrde Exim will forget old rate measurements.)
130 Exim now checks that the per_* options are used with an update mode that
131 makes sense for the current ACL. For example, when Exim is processing a
132 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
133 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
134 must specify per_mail/readonly. If you omit the update mode it defaults to
135 /leaky where that makes sense (as before) or /readonly where required.
137 The /noupdate option is now undocumented but still supported for
138 backwards compatibility. It is equivalent to /readonly except that in
139 ACLs where /readonly is required you may specify /leaky/noupdate or
140 /strict/noupdate which are treated the same as /readonly.
142 A useful new feature is the /count= option. This is a generalization
143 of the per_byte option, so that you can measure the throughput of other
144 aggregate values. For example, the per_byte option is now equivalent
145 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
147 The per_rcpt option has been generalized using the /count= mechanism
148 (though it's more complicated than the per_byte equivalence). When it is
149 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
150 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
151 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
152 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
153 Note that using per_rcpt with a non-readonly update mode in more than
154 one ACL will cause the recipients to be double-counted. (The per_mail
155 and per_byte options don't have this problem.)
157 The handling of very low rates has changed slightly. If the computed rate
158 is less than the event's count (usually one) then this event is the first
159 after a long gap. In this case the rate is set to the same as this event's
160 count, so that the first message of a spam run is counted properly.
162 The major new feature is a mechanism for counting the rate of unique
163 events. The new per_addr option counts the number of different
164 recipients that someone has sent messages to in the last time period. It
165 behaves like per_rcpt if all the recipient addresses are different, but
166 duplicate recipient addresses do not increase the measured rate. Like
167 the /count= option this is a general mechanism, so the per_addr option
168 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
169 example, measure the rate that a client uses different sender addresses
170 with the options per_mail/unique=$sender_address. There are further
171 details in the main documentation.
173 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
175 TF/06 Removed a few PCRE remnants.
177 TF/07 Automatically extract Exim's version number from tags in the git
178 repository when doing development or release builds.
180 PP/02 Raise smtp_cmd_buffer_size to 16kB.
181 Bugzilla 879. Patch from Paul Fisher.
183 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
184 Heavily based on revision 40f9a89a from Simon Arlott's tree.
187 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
189 PP/05 Variable $av_failed, true if the AV scanner deferred.
190 Bugzilla 1078. Patch from John Horne.
192 PP/06 Stop make process more reliably on build failure.
193 Bugzilla 1087. Patch from Heiko Schlittermann.
195 PP/07 Make maildir_use_size_file an _expandable_ boolean.
196 Bugzilla 1089. Patch from Heiko Schlittermann.
198 PP/08 Handle ${run} returning more data than OS pipe buffer size.
199 Bugzilla 1131. Patch from Holger Weiß.
201 PP/09 Handle IPv6 addresses with SPF.
202 Bugzilla 860. Patch from Wolfgang Breyha.
204 PP/10 GnuTLS: support TLS 1.2 & 1.1.
206 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
209 PP/11 match_* no longer expand right-hand-side by default.
210 New compile-time build option, EXPAND_LISTMATCH_RHS.
211 New expansion conditions, "inlist", "inlisti".
213 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
215 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
217 PP/14 fix log_write() format string regression from TF/03.
218 Bugzilla 1152. Patch from Dmitry Isaikin.
224 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
226 PP/02 Harmonised TLS library version reporting; only show if debugging.
227 Layout now matches that introduced for other libraries in 4.74 PP/03.
229 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
231 PP/04 New "dns_use_edns0" global option.
233 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
236 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
237 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
239 TK/01 Updated PolarSSL code to 0.14.2.
240 Bugzilla 1097. Patch from Andreas Metzler.
242 PP/07 Catch divide-by-zero in ${eval:...}.
245 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
248 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
249 format-string attack -- SECURITY: remote arbitrary code execution.
251 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
252 time unintentionally subject to list matching rules, letting the header
253 cause arbitrary Exim lookups (of items which can occur in lists, *not*
254 arbitrary string expansion). This allowed for information disclosure.
256 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
257 INT_MIN/-1 -- value coerced to INT_MAX.
263 NM/01 Workround for PCRE version dependency in version reporting
266 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
267 This fixes portability to compilers other than gcc, notably
268 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
270 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
271 makefiles for portability to HP-UX and POSIX correctness.
273 PP/01 Permit LOOKUP_foo enabling on the make command-line.
274 Also via indented variable definition in the Makefile.
275 (Debugging by Oliver Heesakkers).
277 PP/02 Restore caching of spamd results with expanded spamd_address.
278 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
280 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
281 Improves build reliability. Fix from: Frank Elsner
283 NM/02 Fix wide character breakage in the rfc2047 coding
284 Fixes bug 1064. Patch from Andrey N. Oktyabrski
286 NM/03 Allow underscore in dnslist lookups
287 Fixes bug 1026. Patch from Graeme Fowler
289 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
290 Code patches from Adam Ciarcinski of NetBSD.
292 NM/04 Fixed exiqgrep to cope with mailq missing size issue
295 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
296 is logged, to avoid truncation. Patch from John Horne.
298 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
299 Patch from Jakob Hirsch.
301 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
302 SQL string expansion failure details.
303 Patch from Andrey Oktyabrski.
305 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
306 Patch from Simon Arlott.
308 PP/09 New lookups functionality failed to compile on old gcc which rejects
309 extern declarations in function scope.
310 Patch from Oliver Fleischmann
312 PP/10 Use sig_atomic_t for flags set from signal handlers.
313 Check getgroups() return and improve debugging.
314 Fixed developed for diagnosis in bug 927 (which turned out to be
317 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
318 Patch from Mark Zealey.
320 PP/12 Bugzilla 1056: Improved spamd server selection.
321 Patch from Mark Zealey.
323 PP/13 Bugzilla 1086: Deal with maildir quota file races.
324 Based on patch from Heiko Schlittermann.
326 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
327 Patch from Uwe Doering, sign-off by Michael Haardt.
329 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
330 variable declaration deep within a block. Bug and patch from
333 PP/15 lookups-Makefile IRIX compatibilty coercion.
335 PP/16 Make DISABLE_DKIM build knob functional.
337 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
338 Patch by Simon Arlott
340 TF/03 Fix valgrind.h portability to C89 compilers that do not support
341 variable argument macros. Our copy now differs from upstream.
347 TF/01 Failure to get a lock on a hints database can have serious
348 consequences so log it to the panic log.
350 TF/02 Log LMTP confirmation messages in the same way as SMTP,
351 controlled using the smtp_confirmation log selector.
353 TF/03 Include the error message when we fail to unlink a spool file.
355 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
356 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
357 for maintaining out-of-tree patches for some time.
359 PP/01 Bugzilla 139: Documentation and portability issues.
360 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
361 Handle per-OS dynamic-module compilation flags.
363 PP/02 Let /dev/null have normal permissions.
364 The 4.73 fixes were a little too stringent and complained about the
365 permissions on /dev/null. Exempt it from some checks.
366 Reported by Andreas M. Kirchwitz.
368 PP/03 Report version information for many libraries, including
369 Exim version information for dynamically loaded libraries. Created
370 version.h, now support a version extension string for distributors
371 who patch heavily. Dynamic module ABI change.
373 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
374 privilege escalation vulnerability whereby the Exim run-time user
375 can cause root to append content of the attacker's choosing to
378 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
381 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
382 If dropping privileges for untrusted macros, we disabled normal logging
383 on the basis that it would fail; for the Exim run-time user, this is not
384 the case, and it resulted in successful deliveries going unlogged.
385 Fixed. Reported by Andreas Metzler.
391 PP/01 Date: & Message-Id: revert to normally being appended to a message,
392 only prepend for the Resent-* case. Fixes regression introduced in
393 Exim 4.70 by NM/22 for Bugzilla 607.
395 PP/02 Include check_rfc2047_length in configure.default because we're seeing
396 increasing numbers of administrators be bitten by this.
398 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
400 PP/03 Bugzilla 994: added openssl_options main configuration option.
402 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
404 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
406 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
408 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
409 without a peer certificate, leading to a segfault because of an
410 assumption that peers always have certificates. Be a little more
411 paranoid. Problem reported by Martin Tscholak.
413 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
414 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
415 NB: ClamAV planning to remove STREAM in "middle of 2010".
416 CL also introduces -bmalware, various -d+acl logging additions and
417 more caution in buffer sizes.
419 PP/09 Implemented reverse_ip expansion operator.
421 PP/10 Bugzilla 937: provide a "debug" ACL control.
423 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
425 PP/12 Bugzilla 973: Implement --version.
427 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
429 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
431 PP/15 Bugzilla 816: support multiple condition rules on Routers.
433 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
434 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
435 ignore trailing whitespace.
437 JJ/02 prevent non-panic DKIM error from being sent to paniclog
439 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
442 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
443 Notification from Dr Andrew Aitchison.
445 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
446 ExtendedDetectionInfo response format.
447 Notification from John Horne.
449 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
452 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
453 XSL and documented dependency on system catalogs, with examples of how
456 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
459 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
460 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
461 configuration file which is writeable by the Exim user or group.
463 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
464 of configuration files to cover files specified with the -C option if
465 they are going to be used with root privileges, not just the default
468 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
469 option (effectively making it always true).
471 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
472 files to be used while preserving root privileges.
474 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
475 that rogue child processes cannot use them.
477 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
478 run-time user, instead of root.
480 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
481 Exim run-time user without dropping privileges.
483 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
484 result string, instead of calling string_vformat() twice with the same
487 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
488 for other users. Others should always drop root privileges if they use
489 -C on the command line, even for a whitelisted configure file.
491 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
493 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
499 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
500 $data_path, and $header_path variables; fixed documentation bugs and
503 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
504 exipick to access non-standard spools, including the "frozen" queue
507 NM/01 Bugzilla 965: Support mysql stored procedures.
508 Patch from Alain Williams
510 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
512 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
513 Patch from Andreas Metzler
515 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
516 Patch from Kirill Miazine
518 NM/05 Bugzilla 671: Added umask to procmail example.
520 JJ/03 installed exipick 20100323.0, fixing doc bug
522 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
523 directory. Notification and patch from Dan Rosenberg.
525 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
527 TK/02 Improve log output when DKIM signing operation fails.
529 MH/01 Treat the transport option dkim_domain as a colon separated
530 list, not as a single string, and sign the message with each element,
531 omitting multiple occurences of the same signer.
533 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
534 Bugzilla 985, 986. Patch by Simon Arlott
536 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
537 Patch by Simon Arlott
539 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
540 MBX locking. Notification from Dan Rosenberg.
546 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
548 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
550 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
552 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
554 NM/04 Bugzilla 915: Flag broken perl installation during build.
560 TK/01 Added patch by Johannes Berg that expands the main option
561 "spamd_address" if it starts with a dollar sign.
563 TK/02 Write list of recipients to X-Envelope-Sender header when building
564 the mbox-format spool file for content scanning (suggested by Jakob
567 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
568 (http://www.dcc-servers.net/) support via dccifd. Activated by
569 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
571 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
572 by Mark Daniel Reidel <mr@df.eu>.
574 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
575 When building exim an external PCRE library is now needed -
576 PCRE is a system library on the majority of modern systems.
577 See entry on PCRE_LIBS in EDITME file.
579 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
580 conversation. Added nologin parameter to request.
581 Patch contributed by Kirill Miazine.
583 TF/01 Do not log submission mode rewrites if they do not change the address.
585 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
587 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
588 log files in place. Contributed by Roberto Lima.
590 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
592 TF/03 Bugzilla 615: When checking the local_parts router precondition
593 after a local_part_suffix or local_part_prefix option, Exim now
594 does not use the address's named list lookup cache, since this
595 contains cached lookups for the whole local part.
597 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
598 Robert Millan. Documentation is in experimental-spec.txt.
600 TF/04 Bugzilla 668: Fix parallel build (make -j).
602 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
604 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
605 Patch provided by Jan Srzednicki.
607 TF/05 Leading white space used to be stripped from $spam_report which
608 wrecked the formatting. Now it is preserved.
610 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
611 that they are available at delivery time.
613 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
615 TF/08 TLS error reporting now respects the incoming_interface and
616 incoming_port log selectors.
618 TF/09 Produce a more useful error message if an SMTP transport's hosts
619 setting expands to an empty string.
621 NM/06 Bugzilla 744: EXPN did not work under TLS.
622 Patch provided by Phil Pennock.
624 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
625 Patch provided by Richard Godbee.
627 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
628 acl_smtp_notquit, added index entry.
630 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
631 Patch provided by Eugene Bujak.
633 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
634 accept(). Patch provided by Maxim Dounin.
636 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
637 Patch provided by Phil Pennock.
639 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
641 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
642 Patch provided by Brad "anomie" Jorsch.
644 NM/14 Bugzilla 622: Added timeout setting to transport filter.
645 Patch provided by Dean Brooks.
647 TK/05 Add native DKIM support (does not depend on external libraries).
649 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
650 Patch provided by Graeme Fowler.
652 NM/16 Bugzilla 851: Documentation example syntax fix.
654 NM/17 Changed NOTICE file to remove references to embedded PCRE.
656 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
659 NM/19 Bugzilla 745: TLS version reporting.
660 Patch provided by Phil Pennock.
662 NM/20 Bugzilla 167: bool: condition support.
663 Patch provided by Phil Pennock.
665 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
666 clients. Patch provided by Phil Pennock.
668 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
669 Patch provided by Brad "anomie" Jorsch.
671 NM/23 Bugzilla 687: Fix misparses in eximstats.
672 Patch provided by Heiko Schlittermann.
674 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
675 Patch provided by Heiko Schlittermann.
677 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
678 plus update to original patch.
680 NM/26 Bugzilla 799: Documentation correction for ratelimit.
682 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
683 Patch provided by David Brownlee.
685 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
687 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
689 NM/30 Bugzilla 888: TLS documentation bugfixes.
691 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
693 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
694 Unlike the original bugzilla I have changed all shell scripts in src tree.
696 NM/33 Bugzilla 898: Transport filter timeout fix.
697 Patch by Todd Rinaldo.
699 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
700 Patch by Serge Demonchaux.
702 NM/35 Bugzilla 39: Base64 decode bug fixes.
703 Patch by Jakob Hirsch.
705 NM/36 Bugzilla 909: Correct connect() call in dcc code.
707 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
709 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
711 NM/39 Bugzilla 911: Fixed MakeLinks build script.
717 TK/01 Add preliminary DKIM support. Currently requires a forked version of
718 ALT-N's libdkim that I have put here:
719 http://duncanthrax.net/exim-experimental/
721 Note to Michael Haardt: I had to rename some vars in sieve.c. They
722 were called 'true' and it seems that C99 defines that as a reserved
723 keyword to be used with 'bool' variable types. That means you could
724 not include C99-style headers which use bools without triggering
725 build errors in sieve.c.
727 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
728 as mailq or other aliases. Changed the --help handling significantly
729 to do whats expected. exim_usage() emits usage/help information.
731 SC/01 Added the -bylocaldomain option to eximstats.
733 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
735 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
737 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
743 PH/01 Another patch from the Sieve maintainer.
745 PH/02 When an IPv6 address is converted to a string for single-key lookup
746 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
747 dots are used instead of colons so that keys in lsearch files need not
748 contain colons. This was done some time before quoting was made available
749 in lsearch files. However, iplsearch files do require colons in IPv6 keys
750 (notated using the quote facility) so as to distinguish them from IPv4
751 keys. This meant that lookups for IP addresses in host lists did not work
752 for iplsearch lookups.
754 This has been fixed by arranging for IPv6 addresses to be expressed with
755 colons if the lookup type is iplsearch. This is not incompatible, because
756 previously such lookups could never work.
758 The situation is now rather anomolous, since one *can* have colons in
759 ordinary lsearch keys. However, making the change in all cases is
760 incompatible and would probably break a number of configurations.
762 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
765 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
766 conversion specification without a maximum field width, thereby enabling
767 a rogue spamd server to cause a buffer overflow. While nobody in their
768 right mind would setup Exim to query an untrusted spamd server, an
769 attacker that gains access to a server running spamd could potentially
770 exploit this vulnerability to run arbitrary code as the Exim user.
772 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
773 $primary_hostname instead of what libspf2 thinks the hosts name is.
775 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
776 a directory entry by the name of the lookup key. Previously, if a
777 symlink pointed to a non-existing file or a file in a directory that
778 Exim lacked permissions to read, a lookup for a key matching that
779 symlink would fail. Now it is enough that a matching directory entry
780 exists, symlink or not. (Bugzilla 503.)
782 PH/03 The body_linecount and body_zerocount variables are now exported in the
785 PH/04 Added the $dnslist_matched variable.
787 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
788 This means they are set thereafter only if the connection becomes
791 PH/06 Added the client_condition to authenticators so that some can be skipped
792 by clients under certain conditions.
794 PH/07 The error message for a badly-placed control=no_multiline_responses left
795 "_responses" off the end of the name.
797 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
799 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
800 (without spaces) instead of just copying the configuration text.
802 PH/10 Added the /noupdate option to the ratelimit ACL condition.
804 PH/11 Added $max_received_linelength.
806 PH/12 Added +ignore_defer and +include_defer to host lists.
808 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
809 way in which PCRE > 7.0 is built.
811 PH/14 Implemented queue_only_load_latch.
813 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
814 MAIL command. The effect was to mangle the value on 64-bit systems.
816 PH/16 Another patch from the Sieve maintainer.
818 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
820 PH/18 If a system quota error occurred while trying to create the file for
821 a maildir delivery, the message "Mailbox is full" was not appended to the
822 bounce if the delivery eventually timed out. Change 4.67/27 below applied
823 only to a quota excession during the actual writing of the file.
825 PH/19 It seems that peer DN values may contain newlines (and other non-printing
826 characters?) which causes problems in log lines. The DN values are now
827 passed through string_printing() before being added to log lines.
829 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
830 and InterBase are left for another time.)
832 PH/21 Added message_body_newlines option.
834 PH/22 Guard against possible overflow in moan_check_errorcopy().
836 PH/23 POSIX allows open() to be a macro; guard against that.
838 PH/24 If the recipient of an error message contained an @ in the local part
839 (suitably quoted, of course), incorrect values were put in $domain and
840 $local_part during the evaluation of errors_copy.
846 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
847 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
850 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
851 issue a MAIL command.
853 PH/02 In an ACL statement such as
855 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
857 if a client was not listed at all, or was listed with a value other than
858 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
859 the condition was not true (as it should be), so access was not denied.
860 The bug was that the ! inversion was incorrectly passed on to the second
861 item. This has been fixed.
863 PH/03 Added additional dnslists conditions == and =& which are different from
864 = and & when the dns lookup returns more than one IP address.
866 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
867 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
869 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
870 FSYNC, which compiles an option called disable_fsync that allows for
871 bypassing fsync(). The documentation is heavily laced with warnings.
873 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
875 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
876 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
877 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
878 including adding "make clean"; (3) Added -fPIC when compiling the test
879 dynamically loaded module, to get rid of a warning.
881 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
882 message fails, move_frozen_messages = true and ignore_bounce_errors_after
883 = 0s. The bug is otherwise harmless.
885 PH/07 There was a bug in the dovecot authenticator such that the value of
886 $auth1 could be overwritten, and so not correctly preserved, after a
887 successful authentication. This usually meant that the value preserved by
888 the server_setid option was incorrect.
890 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
892 PH/09 Installed PCRE release 7.0.
894 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
895 run for batched SMTP input. It is now run at the start of every message
896 in the batch. While fixing this I discovered that the process information
897 (output by running exiwhat) was not always getting set for -bs and -bS
898 input. This is fixed, and it now also says "batched" for BSMTP.
900 PH/11 Added control=no_pipelining.
902 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
903 patch, slightly modified), and move the expansion of helo_data till after
904 the connection is made in the smtp transport (so it can use these
907 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
909 PH/14 Added log_selector = +pid.
911 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
913 PH/16 Add ${if forany and ${if forall.
915 PH/17 Added dsn_from option to vary the From: line in DSNs.
917 PH/18 Flush SMTP output before performing a callout, unless control =
918 no_callout_flush is set.
920 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
921 was true (the default) a successful delivery failed to delete the retry
922 item, thus causing premature timeout of the address. The bug is now
925 PH/20 Added hosts_avoid_pipelining to the smtp transport.
927 PH/21 Long custom messages for fakedefer and fakereject are now split up
928 into multiline reponses in the same way that messages for "deny" and
929 other ACL rejections are.
931 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
932 with slight modification.
934 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
935 draft, changing the syntax and factoring some duplicate code".
937 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
938 for deliveries of the second and subsequent messages over the same SMTP
941 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
942 ${reduce, with only minor "tidies".
944 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
946 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
947 expansion side effects.
949 PH/27 When a message times out after an over-quota error from an Exim-imposed
950 quota, the bounce message says "mailbox is full". This message was not
951 being given when it was a system quota that was exceeded. It now should
954 MH/03 Made $recipients available in local_scan(). local_scan() already has
955 better access to the recipient list through recipients_list[], but
956 $recipients can be useful in postmaster-provided expansion strings.
958 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
959 in the case of a MAIL command with additional options following the
960 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
961 were accidentally chopped off.
963 PH/29 SMTP synchronization checks are implemented when a command is read -
964 there is a check that no more input is waiting when there shouldn't be
965 any. However, for some commands, a delay in an ACL can mean that it is
966 some time before the response is written. In this time, more input might
967 arrive, invalidly. So now there are extra checks after an ACL has run for
968 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
969 pipelining has not been advertised.
971 PH/30 MH's patch to allow iscntrl() characters to be list separators.
973 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
974 returned in the SMTP response when smtp_return_error_details was false.
977 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
978 instead of the C I/O that was originally supplied, because problems were
981 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
982 Exim which did not show up earlier: it was assuming that a call to
983 SSL_CTX_set_info_callback() might give an error value. In fact, there is
984 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
985 was a macro that became an assignment, so it seemed to work. This has
986 changed to a proper function call with a void return, hence the compile
987 error. Exim's code has been fixed.
989 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
992 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
994 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
996 PH/37 If a message is not accepted after it has had an id assigned (e.g.
997 because it turns out to be too big or there is a timeout) there is no
998 "Completed" line in the log. When some messages of this type were
999 selected by exigrep, they were listed as "not completed". Others were
1000 picked up by some special patterns. I have improved the selection
1001 criteria to be more general.
1003 PH/38 The host_find_failed option in the manualroute router can now be set
1004 to "ignore", to completely ignore a host whose IP address cannot be
1005 found. If all hosts are ignored, the behaviour is controlled by the new
1006 host_all_ignored option.
1008 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1009 homing or because of multiple MX records with /mx) generated more than
1010 one IP address, and the following item turned out to be the local host,
1011 all the secondary addresses of the first item were incorrectly removed
1012 from the list, along with the local host and any following hosts (which
1013 is what is supposed to happen).
1015 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1016 whoever called Exim into the -H file. In the case of the daemon it was
1017 behaving confusingly. When first started, it used values for whoever
1018 started the daemon, but after a SIGHUP it used the Exim user (because it
1019 calls itself on a restart). I have changed the code so that it now always
1022 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1023 message are rejected with the same error (e.g. no authentication or bad
1024 sender address), and a DATA command is nevertheless sent (as can happen
1025 with PIPELINING or a stupid MUA), the error message that was given to the
1026 RCPT commands is included in the rejection of the DATA command. This is
1027 intended to be helpful for MUAs that show only the final error to their
1030 PH/42 Another patch from the Sieve maintainer.
1032 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1033 Eximstats - Fixed some broken HTML links and added missing column headers
1035 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1036 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1038 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1040 PH/43 Yet another patch from the Sieve maintainer.
1042 PH/44 I found a way to check for a TCP/IP connection going away before sending
1043 the response to the final '.' that terminates a message, but only in the
1044 case where the client has not sent further data following the '.'
1045 (unfortunately, this is allowed). However, in many cases there won't be
1046 any further data because there won't be any more messages to send. A call
1047 to select() can be used: if it shows that the input is "ready", there is
1048 either input waiting, or the socket has been closed. An attempt to read
1049 the next input character can distinguish the two cases. Previously, Exim
1050 would have sent an OK response which the client would never have see.
1051 This could lead to message repetition. This fix should cure that, at
1052 least in a lot of common cases.
1054 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1055 advertised in response to EHLO.
1061 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1062 fixed by 4.65/MH/01 (is this a record?) are fixed:
1064 (i) An empty string was always treated as zero by the numeric comparison
1065 operators. This behaviour has been restored.
1067 (ii) It is documented that the numeric comparison operators always treat
1068 their arguments as decimal numbers. This was broken in that numbers
1069 starting with 0 were being interpreted as octal.
1071 While fixing these problems I realized that there was another issue that
1072 hadn't been noticed. Values of message_size_limit (both the global option
1073 and the transport option) were treated as octal if they started with 0.
1074 The documentation was vague. These values are now always treated as
1075 decimal, and I will make that clear in the documentation.
1081 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1082 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1085 MH/01 Don't check that the operands of numeric comparison operators are
1086 integers when their expansion is in "skipping" mode (fixes bug
1087 introduced by 4.64-PH/07).
1089 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1090 child addresses, Exim now panics and dies. Previously, because the count
1091 is held in a short int, deliveries were likely to be lost. As such a
1092 large number of recipients for a single message is ridiculous
1093 (performance will be very, very poor), I have chosen to impose a limit
1094 rather than extend the field.
1100 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1101 leftover -K file (the existence of which was triggered by #402).
1102 While we were at it, introduced process PID as part of the -K
1103 filename. This should rule out race conditions when creating
1106 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1107 processing considerably. Previous code took too long for large mails,
1108 triggering a timeout which in turn triggers #401.
1110 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1111 in the DK code in transports.c. sendfile() is not really portable,
1112 hence the _LINUX specificness.
1114 TF/01 In the add_headers option to the mail command in an Exim filter,
1115 there was a bug that Exim would claim a syntax error in any
1116 header after the first one which had an odd number of characters
1119 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1120 callout verification, Exim cached a "reject" for the entire domain. This
1121 is correct for most verifications, but it is not correct for a recipient
1122 verification with use_sender or use_postmaster set, because in that case
1123 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1124 case of MAIL FROM:<> rejection from other early rejections (e.g.
1125 rejection of HELO). When verifying a recipient using a non-null MAIL
1126 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1127 Whatever the result of the callout, the value of the domain cache is
1128 left unchanged (for any other kind of callout, getting as far as trying
1129 RCPT means that the domain itself is ok).
1131 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1134 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1135 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1136 session, the daemon ignores ECONNECT errors and logs others; it now
1137 ignores EPIPE as well.
1139 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1140 (quoted-printable decoding).
1142 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1143 later the small subsequent patch to fix an introduced bug.
1145 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1147 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1149 PH/08 An error is now given if message_size_limit is specified negative.
1151 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1152 to be given (somewhat) arbitrary names.
1154 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1157 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1158 miscellaneous code fixes
1160 PH/10 Added the log_reject_target ACL modifier to specify where to log
1163 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1164 hostname. This is wrong, because it relates to the incoming message (and
1165 probably the interface on which it is arriving) and not to the outgoing
1166 callout (which could be using a different interface). This has been
1167 changed to use the value of the helo_data option from the smtp transport
1168 instead - this is what is used when a message is actually being sent. If
1169 there is no remote transport (possible with a router that sets up host
1170 addresses), $smtp_active_hostname is used.
1172 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1173 tweaks were necessary in order to get it to work (see also 21 below):
1174 (a) The code assumed that strncpy() returns a negative number on buffer
1175 overflow, which isn't the case. Replaced with Exim's string_format()
1177 (b) There were several signed/unsigned issues. I just did the minimum
1178 hacking in of casts. There is scope for a larger refactoring.
1179 (c) The code used strcasecmp() which is not a standard C function.
1180 Replaced with Exim's strcmpic() function.
1181 (d) The code set only $1; it now sets $auth1 as well.
1182 (e) A simple test gave the error "authentication client didn't specify
1183 service in request". It would seem that Dovecot has changed its
1184 interface. Fortunately there's a specification; I followed it and
1185 changed what the client sends and it appears to be working now.
1187 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1190 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1191 address A is aliased to B and C, where B exists and C does not. Without
1192 -v the output is "A verified" because verification stops after a
1193 successful redirection if more than one address is generated. However,
1194 with -v the child addresses are also verified. Exim was outputting "A
1195 failed to verify" and then showing the successful verification for C,
1196 with its parentage. It now outputs "B failed to verify", showing B's
1197 parentage before showing the successful verification of C.
1199 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1200 look up a TXT record in a specific list after matching in a combined
1203 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1204 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1205 they consult the DNS. I had assumed they would set it the way they
1206 wanted; and indeed my experiments on Linux seem to show that in some
1207 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1208 To be on the safe side, however, I have now made the interface to
1209 host_find_byname() similar to host_find_bydns(), with an argument
1210 containing the DNS resolver options. The host_find_byname() function now
1211 sets these options at its start, just as host_find_bydns() does. The smtp
1212 transport options dns_qualify_single and dns_search_parents are passed to
1213 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1214 of host_find_byname() use the default settings of RES_DEFNAMES
1215 (qualify_single) but not RES_DNSRCH (search_parents).
1217 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1218 spool_read_header() do less string testing, by means of a preliminary
1219 switch on the second character of optional "-foo" lines. (This is
1220 overdue, caused by the large number of possibilities that now exist.
1221 Originally there were few.) While I was there, I also converted the
1222 str(n)cmp tests so they don't re-test the leading "-" and the first
1223 character, in the hope this might squeeze out yet more improvement.
1225 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1226 flag allowing group syntax was set by the header_syntax check but not
1227 turned off, possible causing trouble later; (2) The flag was not being
1228 set at all for the header_verify test, causing "group"-style headers to
1229 be rejected. I have now set it in this case, and also caused header_
1230 verify to ignore an empty address taken from a group. While doing this, I
1231 came across some other cases where the code for allowing group syntax
1232 while scanning a header line wasn't quite right (mostly, not resetting
1233 the flag correctly in the right place). These bugs could have caused
1234 trouble for malformed header lines. I hope it is now all correct.
1236 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1237 with the "reply" argument non-NULL. The code, however (which originally
1238 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1239 but it didn't always do it. This confused somebody who was copying the
1240 code for some other use. I have removed all the tests.
1242 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1243 feature that was used to support insecure browsers during the U.S. crypto
1244 embargo. It requires special client support, and Exim is probably the
1245 only MTA that supported it -- and would never use it because real RSA is
1246 always available. This code has been removed, because it had the bad
1247 effect of slowing Exim down by computing (never used) parameters for the
1248 RSA_EXPORT functionality.
1250 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1251 authenticator to fail if there's a tab character in the incoming data
1252 (there should never be unless someone is messing about, as it's supposed
1253 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1254 if the connection is using TLS or if the remote IP is the same as the
1255 local IP, and the "valid-client-cert option" if a client certificate has
1258 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1259 authenticators. This can be used for authorization after authentication
1260 succeeds. (In the case of plaintext, it servers for both authentication
1263 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1264 if any retry times were supplied.
1266 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1267 connection, where there is no client IP address to check. In this
1268 situation, the verify now always succeeds.
1270 PH/25 Applied John Jetmore's -Mset patch.
1272 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1274 PH/27 In a string expansion for a processed (not raw) header when multiple
1275 headers of the same name were present, leading whitespace was being
1276 removed from all of them, but trailing whitespace was being removed only
1277 from the last one. Now trailing whitespace is removed from each header
1278 before concatenation. Completely empty headers in a concatenation (as
1279 before) are ignored.
1281 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1282 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1284 PH/29 [Removed. This was a change that I later backed out, and forgot to
1285 correct the ChangeLog entry (that I had efficiently created) before
1286 committing the later change.]
1288 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1289 address errors (4xx response to RCPT) over the same connection as other
1290 messages routed to the same hosts. Such deliveries are always "forced",
1291 so retry times are not inspected. This resulted in far too many retries
1292 for the affected addresses. The effect occurred only when there were more
1293 hosts than the hosts_max_try setting in the smtp transport when it had
1294 the 4xx errors. Those hosts that it had tried were not added to the list
1295 of hosts for which the message was waiting, so if all were tried, there
1296 was no problem. Two fixes have been applied:
1298 (i) If there are any address or message errors in an SMTP delivery, none
1299 of the hosts (tried or untried) are now added to the list of hosts
1300 for which the message is waiting, so the message should not be a
1301 candidate for sending over the same connection that was used for a
1302 successful delivery of some other message. This seems entirely
1303 reasonable: after all the message is NOT "waiting for some host".
1304 This is so "obvious" that I'm not sure why it wasn't done
1305 previously. Hope I haven't missed anything, but it can't do any
1306 harm, as the worst effect is to miss an optimization.
1308 (ii) If, despite (i), such a delivery is accidentally attempted, the
1309 routing retry time is respected, so at least it doesn't keep
1310 hammering the server.
1312 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1313 in ${readsocket because some servers need this prod.
1315 PH/32 Added some extra debug output when updating a wait-xxx database.
1317 PH/33 The hint "could be header name not terminated by colon", which has been
1318 given for certain expansion errors for a long time, was not being given
1319 for the ${if def:h_colon_omitted{... case.
1321 PH/34 The spec says: "With one important exception, whenever a domain list is
1322 being scanned, $domain contains the subject domain." There was at least
1323 one case where this was not true.
1325 PH/35 The error "getsockname() failed: connection reset by peer" was being
1326 written to the panic log as well as the main log, but it isn't really
1327 panic-worthy as it just means the connection died rather early on. I have
1328 removed the panic log writing for the ECONNRESET error when getsockname()
1331 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1332 runs only) independently of the message's sender address. This meant
1333 that, if the 4xx error was in fact related to the sender, a different
1334 message to the same recipient with a different sender could confuse
1335 things. In particualar, this can happen when sending to a greylisting
1336 server, but other circumstances could also provoke similar problems.
1337 I have changed the default so that the retry time for these errors is now
1338 based a combination of the sender and recipient addresses. This change
1339 can be overridden by setting address_retry_include_sender=false in the
1342 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1343 remote server are returned as part of bounce messages. This was not
1344 happening for LMTP over a pipe (the lmtp transport), but now it is the
1345 same for both kinds of LMTP.
1347 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1348 in header lines that were in fact CNAMEs. This is no longer the case.
1350 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1351 and queue runs started by the daemon processed all messages. This has
1352 been fixed so that -R and -S can now usefully be given with -q<time>.
1354 PH/40 Import PCRE release 6.7 (fixes some bugs).
1356 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1358 PH/42 Give an error if -q is specified more than once.
1360 PH/43 Renamed the variables $interface_address and $interface_port as
1361 $received_ip_address and $received_port, to make it clear that these
1362 values apply to message reception, and not to the outgoing interface when
1363 a message is delivered. (The old names remain recognized, of course.)
1365 PH/44 There was no timeout on the connect() call when using a Unix domain
1366 socket in the ${readsocket expansion. There now is.
1368 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1369 be meaningful with "accept".
1371 SC/01 Eximstats V1.43
1372 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1374 SC/02 Eximstats V1.44
1375 Use a glob alias rather than an array ref in the generated
1376 parser. This improves both readability and performance.
1378 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1379 Collect SpamAssassin and rejection statistics.
1380 Don't display local sender or destination tables unless
1381 there is data to show.
1382 Added average volumes into the top table text output.
1384 SC/04 Eximstats V1.46
1385 Collect data on the number of addresses (recipients)
1386 as well as the number of messages.
1388 SC/05 Eximstats V1.47
1389 Added 'Message too big' to the list of mail rejection
1390 reasons (thanks to Marco Gaiarin).
1392 SC/06 Eximstats V1.48
1393 Mainlog lines which have GMT offsets and are too short to
1394 have a flag are now skipped.
1396 SC/07 Eximstats V1.49 (Alain Williams)
1397 Added the -emptyok flag.
1399 SC/08 Eximstats V1.50
1400 Fixes for obtaining the IP address from reject messages.
1402 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1403 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1404 whitesspace changes from 4.64-PH/27
1406 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1409 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1410 are found, allow negative numbers in numeric criteria)
1412 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1414 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1415 to match changes made in 4.64-PH/43
1417 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1419 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1420 "domain" is actually an IP address, and force a failure. This locks out
1421 those revolvers/nameservers that support "A-for-A" lookups, in
1422 contravention of the specifications.
1424 PH/48 When a host name was looked up from an IP address, and the subsequent
1425 forward lookup of the name timed out, the host name was left in
1426 $sender_host_name, contrary to the specification.
1428 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1429 restricted to single-key lookups, Exim was not diagnosing an error if
1430 * or *@ was used with a query-style lookup.
1432 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1434 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1435 long ago, but noone interested enough thought of it. Let's just say that
1436 the "1.1" means that there are some new functions that weren't there at
1437 some point in the past.
1439 PH/51 Error processing for expansion failure of helo_data from an smtp
1440 transport during callout processing was broken.
1442 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1443 tested/used via the -bh/-bhc/-bs options.
1445 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1446 bug, fixed in subsequent PCRE releases).
1448 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1449 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1451 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1457 SC/01 Use a glob alias rather than an array ref in eximstats generated
1458 parser. This improves both readability and performance.
1460 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1461 Don't display local sender or destination tables in eximstats unless
1462 there is data to show.
1463 Added average volumes into the eximstats top table text output.
1465 SC/03 Collect data on the number of addresses (recipients) as well
1466 as the number of messages in eximstats.
1468 TF/01 Correct an error in the documentation for the redirect router. Exim
1469 does (usually) call initgroups() when daemonizing.
1471 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1472 with consistent privilege compared to when running as a daemon.
1474 TF/03 Note in the spec that $authenticated_id is not set for local
1475 submissions from trusted users.
1477 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1478 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1480 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1481 by adding some example configuration directives to the default
1482 configuration file. A little bit of work is required to uncomment the
1483 directives and define how usernames and passwords are checked, but
1484 there is now a framework to start from.
1486 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1487 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1488 without this. I don't know how relevant this is to other LDAP libraries.
1490 PH/02 Add the verb name to the "unknown ACL verb" error.
1492 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1494 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1496 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1497 directory not expanded when it should be if an expanded home directory
1498 was set for the address (which is overridden by the transport).
1500 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1503 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1504 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1505 because it is too late at that time, and has no effect.
1507 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1508 security issue with \' (bugzilla #107). I could not use the
1509 PQescapeStringConn() function, because it needs a PGconn value as one of
1512 PH/08 When testing addresses using -bt, indicate those final addresses that
1513 are duplicates that would not cause an additional delivery. At least one
1514 person was confused, thinking that -bt output corresponded to deliveries.
1515 (Suppressing duplicates isn't a good idea as you lose the information
1516 about possibly different redirections that led to the duplicates.)
1518 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1519 systems where poll() doesn't work, in particular OS X.
1521 PH/10 Added more information to debugging output for retry time not reached.
1523 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1524 operations in malware.c.
1526 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1529 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1530 syslog_duplication set false, log lines that would normally be written
1531 both the the main log and to the reject log were not written to syslog at
1534 PH/14 In the default configuration, change the use of "message" in ACL warn
1535 statements to "add_header".
1537 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1538 not followed by a command (e.g. "seen endif").
1540 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1541 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1544 PH/17 Added extra conditions to the default value of delay_warning_condition
1548 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1549 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1550 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1553 The Auto-Submitted: and various List- headers are standardised, whereas I
1554 don't think Precedence: ever was.
1556 PH/18 Refactored debugging code in route_finduser() to show more information,
1557 in particular, the error code if getpwnam() issues one.
1559 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1560 This is apparently needed in addition to the PH/07 change above to avoid
1561 any possible encoding problems.
1563 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1564 but not after initializing Perl.
1566 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1567 output them only if debugging. By default they are written stderr,
1568 apparently, which is not desirable.
1570 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1573 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1576 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1578 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1579 authenticated or an ident call has been made. Suppress the default
1580 values for $authenticated_id and $authenticated_sender (but permit -oMai
1581 and -oMas) when testing with -bh.
1583 PH/24 Re-jigged the order of the tests in the default configuration so that the
1584 tests for valid domains and recipients precede the DNS black list and CSA
1585 tests, on the grounds that those ones are more expensive.
1587 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1588 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1589 This bug exists in every version of Exim that I still have, right back to
1592 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1593 However, an attempt to turn on case-sensitivity in a regex key by
1594 including (?-i) didn't work because the subject string was already
1595 lowercased, and the effects were non-intuitive. It turns out that a
1596 one-line patch can be used to allow (?-i) to work as expected.
1602 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1603 other effects) broke the use of negated acl sub-conditions.
1605 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1608 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1609 "Deny" causes Exim to reject the incoming connection with a 554 error.
1610 Unfortunately, if there is a major crisis, such as a disk failure,
1611 tcp-wrappers gives "deny", whereas what one would like would be some
1612 kind of temporary error. A kludge has been added to help with this.
1613 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1614 554 error is used if errno is still zero or contains ENOENT (which occurs
1615 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1618 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1620 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1621 errors. Otherwise a message that provokes a temporary error (when other
1622 messages do not) can cause a whole host to time out.
1624 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1625 addresses were routed directly to files or pipes from a redirect router.
1626 File deliveries just didn't batch; pipe deliveries might have suffered
1629 PH/06 A failure to get a lock for a hints database would erroneously always say
1630 "Failed to get write lock", even when it was really a read lock.
1632 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1633 of 0600. This has been changed to use the value of the lockfile_mode
1634 option (which defaults to 0600).
1636 PH/08 Applied small patch from the Sieve maintainer.
1638 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1639 folder from quota calculations, a direct delivery into this folder messed
1640 up the contents of the maildirsize file. This was because the regex was
1641 used only to exclude .Trash (or whatever) when the size of the mailbox
1642 was calculated. There was no check that a delivery was happening into an
1643 excluded directory. This bug has been fixed by ignoring all quota
1644 processing for deliveries into excluded directories.
1646 PH/10 Added the maildirfolder_create_regex option to appendfile.
1652 PH/01 The code for finding all the local interface addresses on a FreeBSD
1653 system running IPv6 was broken. This may well have applied to all BSD
1654 systems, as well as to others that have similar system calls. The broken
1655 code found IPv4 interfaces correctly, but gave incorrect values for the
1656 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1657 that it would not match correctly against @[] and not recognize the IPv6
1660 PH/02 The ipliteral router was not recognizing addresses of the form user@
1661 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1663 PH/03 Added disable_ipv6.
1665 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1666 decoded form, because it is most often used to construct To: headers
1667 lines in autoreplies, and the decoded form may well be syntactically
1668 invalid. However, $reply_address has leading white space removed, and all
1669 newlines turned into spaces so that the autoreply transport does not
1672 PH/05 If group was specified without a user on a router, and no group or user
1673 was specified on a transport, the group from the router was ignored.
1675 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1676 for visible compile-time settings that can be used to change these
1677 numbers, for those that want even more. Backwards compatibility with old
1678 spool files has been maintained. However, going back to a previous Exim
1679 release will lost any variables that are in spool files.
1681 PH/07 Two small changes when running in the test harness: increase delay when
1682 passing a TCP/IP connection to a new process, in case the original
1683 process has to generate a bounce, and remove special handling of
1684 127.0.0.2 (sic), which is no longer necessary.
1686 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1687 be the same on different OS.
1689 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1692 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1693 whether --show-vars was specified or not
1695 JJ/02 exipick: Added support for new ACL variable spool format introduced
1698 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1699 syntactically invalid From: or Reply-to: line, and a filter used this to
1700 generate an autoreply, and therefore failed to obtain an address for the
1701 autoreply, Exim could try to deliver to a non-existent relative file
1702 name, causing unrelated and misleading errors. What now happens is that
1703 it logs this as a hard delivery error, but does not attempt to create a
1706 PH/11 The exinext utility has a -C option for testing purposes, but although
1707 the given file was scanned by exinext itself; it wasn't being passed on
1708 when Exim was called.
1710 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1711 an end-of-file indication when reading a command response.
1713 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1714 compiled. In many other places in Exim, IPv6 addresses are always
1715 recognized, so I have changed this. It also means that IPv4 domain
1716 literals of the form [IPV4:n.n.n.n] are now always recognized.
1718 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1719 used if the router is not running as root, for example, when verifying at
1720 ACL time, or when using -bh. The debugging output from this situation was
1721 non-existent - all you got was a failure to exec. I have made two
1724 (a) Failures to set uid/gid, the current directory, or a process leader
1725 in a subprocess such as that created by queryprogram now generate
1726 suitable debugging ouput when -d is set.
1728 (b) The queryprogram router detects when it is not running as root,
1729 outputs suitable debugging information if -d is set, and then runs
1730 the subprocess without attempting to change uid/gid.
1732 PH/15 Minor change to Makefile for building test_host (undocumented testing
1735 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1736 additional section of a DNS packet that returns MX or SRV records.
1737 Instead, it always explicitly searches for A/AAAA records. This avoids
1738 major problems that occur when a DNS server includes only records of one
1739 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1740 fixed another bug: if SRV records were looked up and the corresponding
1741 address records were *not* found in the additional section, the port
1742 values from the SRV records were lost.
1744 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1745 using the correct key (the original address) when searching the retry
1746 rules in order to find which one to use for generating the retry hint.
1748 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1749 adding the default one. Similarly, if it contains a Reply-To: header, the
1750 errors_reply_to option, if set, is not used.
1752 PH/19 When calculating a retry time, Exim used to measure the "time since
1753 failure" by looking at the "first failed" field in the retry record. Now
1754 it does not use this if it is later than than the arrival time of the
1755 message. Instead it uses the arrival time. This makes for better
1756 behaviour in cases where some deliveries succeed, thus re-setting the
1757 "first failed" field. An example is a quota failure for a huge message
1758 when small messages continue to be delivered. Without this change, the
1759 "time since failure" will always be short, possible causing more frequent
1760 delivery attempts for the huge message than are intended.
1761 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1763 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1764 $1, $2, $3) because the numerical variables can be reset during some
1765 expansion items (e.g. "match"), thereby losing the authentication data.
1767 PH/21 Make -bV show the size of off_t variables so that the test suite can
1768 decide whether to run tests for quotas > 2G.
1770 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1771 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1772 filecount value is greater than 2G or if a quota value is greater than 2G
1773 on a system where the size of off_t is not greater than 4, a panic error
1776 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1777 never match. The debug and -bh output now contains an explicit error
1778 message indicating a malformed IPv4 address or mask.
1780 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1781 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1782 PH/23 above applies.
1784 PH/25 Do not write to syslog when running in the test harness. The only
1785 occasion when this arises is a failure to open the main or panic logs
1786 (for which there is an explicit test).
1788 PH/26 Added the /no_tell option to "control=freeze".
1790 PH/27 If a host name lookup failed very early in a connection, for example, if
1791 the IP address matched host_lookup and the reverse lookup yielded a name
1792 that did not have a forward lookup, an error message of the form "no IP
1793 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1794 could be logged. Now it outputs the IP address instead of "NULL".
1796 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1797 allows the sender and the authenticated sender to be set when
1798 submitting a message from within Exim. Since child_open_exim() is
1799 documented for local_scan(), the new function should be too.
1801 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1802 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1803 results in an empty string is now treated as unset.
1805 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1807 PH/31 Added sender_verify_fail logging option.
1809 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1810 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1811 tidied the source and removed it altogether.
1813 PH/33 When a queue run was abandoned because the load average was too high, a
1814 log line was always written; now it is written only if the queue_run log
1815 selector is set. In addition, the log line for abandonment now contains
1816 information about the queue run such as the pid. This is always present
1817 in "start" and "stop" lines but was omitted from the "abandon" line.
1819 PH/34 Omit spaces between a header name and the colon in the error message that
1820 is given when verify = headers_syntax fails (if there are lots of them,
1821 the message gets confusing).
1823 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1824 names, as there are now some PTR records that contain slashes. This check
1825 is only to protect against broken name servers that fall over on strange
1826 characters, so the fact that it applies to all lookups doesn't matter.
1828 PH/36 Now that the new test suite is complete, we can remove some of the
1829 special code in Exim that was needed for the old test suite. For example,
1830 sorting DNS records because real resolvers return them in an arbitrary
1831 order. The new test suite's fake resolver always returns records in the
1834 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1835 bounces) except when queue_only is set, to avoid logging races between
1836 the different processes.
1838 PH/38 Panic-die if .include specifies a non-absolute path.
1840 PH/39 A tweak to the "H" retry rule from its user.
1842 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1843 a label. They prevented compilation on older perls.
1845 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1846 a warning to be raised on newish perls.
1848 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1849 on queue. Changes to match documented behaviour of showing count of
1850 messages matching specified criteria.
1852 PH/40 Changed the default ident timeout from 30s to 5s.
1854 PH/41 Added support for the use of login_cap features, on those BSD systems
1855 that have them, for controlling the resources used by pipe deliveries.
1857 PH/42 The content-scanning code uses fopen() to create files in which to put
1858 message data. Previously it was not paying any attention to the mode of
1859 the files. Exim runs with umask(0) because the rest of the code creates
1860 files with open(), and sets the required mode explicitly. Thus, these
1861 files were ending up world-writeable. This was not a big issue, because,
1862 being within the spool directory, they were not world-accessible. I have
1863 created a function called modefopen, which takes an additional mode
1864 argument. It sets umask(777), creates the file, chmods it to the required
1865 mode, then resets the umask. All the relevant calls to fopen() in the
1866 content scanning code have been changed to use this function.
1868 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1869 to 24 hours. This avoids potential overflow problems when processing G
1870 and H retry rules. I suspect nobody ever tinkers with this value.
1872 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1874 PH/45 When the plaintext authenticator is running as a client, the server's
1875 challenges are checked to ensure they are valid base64 strings. By
1876 default, the authentication attempt is cancelled if an invalid string is
1877 received. Setting client_ignore_invalid_base64 true ignores these errors.
1878 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1879 they are received. Thus, the responses can be made to depend on the
1880 challenges. If an invalid string is ignored, an empty string is placed in
1883 PH/46 Messages that are created by the autoreply transport now contains a
1884 References: header, in accordance with RFCs 2822 and 3834.
1886 PH/47 Added authenticated_sender_force to the smtp transport.
1888 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1890 PH/49 Installed latest patch from the Sieve maintainer.
1892 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1893 was also set, the appendfile transport was unnecessarily scanning a
1894 directory of message files (e.g. for maildir delivery) to find the count
1895 of files (along with the size), even though it did not need this
1896 information. It now does the scan only if it needs to find either the
1897 size of the count of files.
1899 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1901 PH/52 Two bugs concerned with error handling when the smtp transport is
1904 (i) Exim was not creating retry information for temporary errors given
1905 for individual recipients after the DATA command when the smtp transport
1906 was used in LMTP mode. This meant that they could be retried too
1907 frequently, and not timed out correctly.
1909 (ii) Exim was setting the flag that allows error details to be returned
1910 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1911 recipients that were returned after the DATA command.
1913 PH/53 This is related to PH/52, but is more general: for any failing address,
1914 when detailed error information was permitted to be returned to the
1915 sender, but the error was temporary, then after the final timeout, only
1916 "retry timeout exceeded" was returned. Now it returns the full error as
1917 well as "retry timeout exceeded".
1919 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1920 do this, and (what is worse) MTAs that accept it.
1922 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1923 will now be deprecated.
1925 PH/56 New os.c-cygwin from the Cygwin maintainer.
1927 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1928 formats (previously only available in exim formats via -bpr, -bpru,
1929 and -bpra. Now also available in native and exiqgrep formats)
1931 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1932 with very large, slow to parse queues
1934 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1936 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1938 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1939 responses to authentication challenges, though it was showing the
1940 challenges; (ii) I've removed the CR characters from the debug output for
1943 PH/58 Allow for the insertion of a newline as well as a space when a string
1944 is turned into more than one encoded-word during RFC 2047 encoding. The
1945 Sieve code now uses this.
1947 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1948 data_4xx, lost_connection, tls_required.
1950 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1951 message was being sent as an SMTP response.
1953 PH/61 Add -l and -k options to exicyclog.
1955 PH/62 When verifying, if an address was redirected to one new address, so that
1956 verification continued, and the new address failed or deferred after
1957 having set something in $address_data, the value of $address_data was not
1958 passed back to the ACL. This was different to the case when no
1959 redirection occurred. The value is now passed back in both cases.
1961 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1962 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1963 use login_cap.h, so on its own it isn't the distinguishing feature. The
1964 new name refers directly to the setclassresources() function.
1966 PH/65 Added configuration files for NetBSD3.
1968 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1970 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1971 is preferred over IPv4.
1973 PH/68 The bounce_return_message and bounce_return_body options were not being
1974 honoured for bounces generated during the reception of non-SMTP messages.
1975 In particular, this applied to messages rejected by the ACL. This bug has
1976 been fixed. However, if bounce_return_message is true and bounce_return_
1977 body is false, the headers that are returned for a non-SMTP message
1978 include only those that have been read before the error was detected.
1979 (In the case of an ACL rejection, they have all been read.)
1981 PH/69 The HTML version of the specification is now built in a directory called
1982 spec_html instead of spec.html, because the latter looks like a path with
1983 a MIME-type, and this confuses some software.
1985 PH/70 Catch two compiler warnings in sieve.c.
1987 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1988 function verify_get_ident() calls ip_connect() to connect a socket, but
1989 if the "connect()" function timed out, ip_connect() used to close the
1990 socket. However, verify_get_ident() also closes the socket later, and in
1991 between Exim writes to the log, which may get opened at this point. When
1992 the socket was closed in ip_connect(), the log could get the same file
1993 descriptor number as the socket. This naturally causes chaos. The fix is
1994 not to close the socket in ip_connect(); the socket should be closed by
1995 the function that creates it. There was only one place in the code where
1996 this was missing, in the iplookup router, which I don't think anybody now
1997 uses, but I've fixed it anyway.
1999 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2000 well as to direct DNS lookups. Otherwise the handling of names in host
2001 lists is inconsistent and therefore confusing.
2007 PH/01 Two changes to the default runtime configuration:
2009 (1) Move the checks for relay_from_hosts and authenticated clients from
2010 after to before the (commented out) DNS black list checks.
2012 (2) Add control=submission to the relay_from_hosts and authenticated
2013 clients checks, on the grounds that messages accepted by these
2014 statements are most likely to be submissions.
2016 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2018 (1) Generate an error if the third argument for the ${prvs expansion is
2021 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2024 (3) Reset the variables that are obtained from the first argument of
2025 ${prvscheck and used in the second argument before leaving the code,
2026 because their memory is reclaimed, so using them afterwards may do
2029 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2030 one (it's much easier than Tom thought :-).
2032 (5) Because of (4), we can now allow for the use of $prvscheck_result
2033 inside the third argument.
2035 PH/03 For some reason, the default setting of PATH when running a command from
2036 a pipe transport was just "/usr/bin". I have changed it to
2039 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2040 anything to be listed in the output from -bV.
2042 PH/05 When a filter generated an autoreply, the entire To: header line was
2043 quoted in the delivery log line, like this:
2045 => >A.N.Other <ano@some.domain> <original@ddress> ...
2047 This has been changed so that it extracts the operative address. There
2048 may be more than one such address. If so, they are comma-separated, like
2051 => >ano@some.domain,ona@other.domain <original@ddress> ...
2053 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2054 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2055 not being looked up in the rDNS to get a host name, Exim was showing the
2056 IP address twice in Received: lines, even though the IP addresses were
2057 identical. For example:
2059 Received: from [1.2.3.4] (helo=[1.2.3.4])
2061 However, if the real host name was known, it was omitting the HELO data
2062 if it matched the actual IP address. This has been tidied up so that it
2063 doesn't show the same IP address twice.
2065 PH/07 When both +timestamp and +memory debugging was on, the value given by
2066 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2067 called by the debug printing, thereby overwriting the timestamp buffer.
2068 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2070 PH/08 When the original message was included in an autoreply transport, it
2071 always said "this is a copy of the message, including all the headers",
2072 even if body_only or headers_only was set. It now gives an appropriate
2075 PH/09 Applied a patch from the Sieve maintainer which:
2077 o fixes some comments
2078 o adds the (disabled) notify extension core
2079 o adds some debug output for the result of if/elsif tests
2080 o points to the current vacation draft in the documentation
2081 and documents the missing references header update
2085 o fixes a bug in processing the envelope test (when testing
2086 multiple envelope elements, the last element determinted the
2089 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2090 Electronic Mail") by including:
2092 Auto-submitted: auto-generated
2094 in the messages that it generates (bounce messages and others, such as
2095 warnings). In the case of bounce messages for non-SMTP mesages, there was
2096 also a typo: it was using "Auto_submitted" (underscore instead of
2097 hyphen). Since every message generated by Exim is necessarily in response
2098 to another message, thes have all been changed to:
2100 Auto-Submitted: auto-replied
2102 in accordance with these statements in the RFC:
2104 The auto-replied keyword:
2106 - SHOULD be used on messages sent in direct response to another
2107 message by an automatic process,
2109 - MUST NOT be used on manually-generated messages,
2111 - MAY be used on Delivery Status Notifications (DSNs) and Message
2112 Disposition Notifications (MDNs),
2114 - MUST NOT be used on messages generated by automatic or periodic
2115 processes, except for messages which are automatic responses to
2118 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2119 to the default Received: header definition.
2121 PH/12 Added log selector acl_warn_skipped (default on).
2123 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2124 variables because (a) they are in the wrong storage pool and (b) even if
2125 they were copied, it wouldn't work properly because of the caching.
2127 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2128 checking when decoding. Apparently there are clients that generate
2129 overlong encoded strings. Why am I not surprised?
2131 PH/15 If the first argument of "${if match_address" was not empty, but did not
2132 contain an "@" character, Exim crashed. Now it writes a panic log message
2133 and treats the condition as false.
2135 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2137 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2138 extension "envelope-auth". The code is finished and in agreement with
2139 other implementations, but there is no documentation so far and in fact,
2140 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2141 not changing the active code.
2143 Print executed "if" and "elsif" statements when debugging is used. This
2144 helps a great deal to understand what a filter does.
2146 Document more things not specified clearly in RFC3028. I had all this
2147 sorted out, when out of a sudden new issues came to my mind. Oops."
2149 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2152 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2153 canonical form (as documented). However, after a host name lookup from
2154 the IP address, check_host() was doing a simple string comparison with
2155 addresses acquired from the DNS when checking that the found name did
2156 have the original IP as one of its addresses. Since any found IPv6
2157 addresses are likely to be in abbreviated form, the comparison could
2158 fail. Luckily, there already exists a function for doing the comparison
2159 by converting both addresses to binary, so now that is used instead of
2160 the text comparison.
2162 PH/20 There was another similar case to PH/19, when a complete host name was
2163 given in a host list; looking up its IP address could give an abbreviated
2164 form, whereas the current host's name might or might not be abbreviated.
2165 The same fix has been applied.
2171 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2172 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2175 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2176 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2178 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2180 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2181 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2182 Also, for "Auto-Submitted", treat anything other than "no" as
2183 non-personal, in accordance with RFC 3834. (Previously it treated
2184 anything starting "auto-" as non-personal.)
2186 TF/01 The control=submission/name=... option had a problem with syntax
2187 errors if the name included a slash character. The /name= option
2188 now slurps the rest of the string, so it can include any characters
2189 but it must come last in the list of options (after /sender_retain
2192 PH/05 Some modifications to the interface to the fake nameserver for the new
2200 TK/01 Added the "success_on_redirect" address verification option. See
2201 NewStuff for rationale and an example.
2203 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2205 PH/02 Patch to exigrep to allow it to work on syslog lines.
2207 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2208 fread() to skip over the body file's header line, because in Cygwin the
2209 header line is locked and is inaccessible.
2211 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2212 co-exist for some time) to make it clear that it is the Exim ID that is
2213 referenced, not the Message-ID: header line.
2215 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2216 string_format() function, because snprintf() does not exist on all
2219 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2220 ${stat: expansion item.
2222 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2223 protocol synchronization error", to keep the pedants happy.
2225 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2226 well as for IRIX systems, when gcc is being used. See the host.c source
2229 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2231 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2234 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2235 transport and to the smtp transport in LMTP mode.
2237 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2239 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2240 run a filter in a subprocess. This could lead to confusion in subsequent
2241 lookups in the parent process. There should also be a search_tidyup() at
2242 the end of the subprocess.
2244 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2245 only if the host matched helo_try_verify_hosts, which caused the
2246 verification to occur when the EHLO/HELO command was issued. The ACL just
2247 tested the remembered result. Now, if a previous verification attempt has
2248 not happened, "verify = helo" does it there and then.
2250 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2252 TK/03 Fix log output including CR from clamd.
2254 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2255 exist provoked a memory error which could cause a segfault.
2257 PH/15 Installed PCRE 6.2
2259 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2261 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2262 of the problem. Specifically, suggested +O2 rather than +O1 for the
2265 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2267 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2268 record, the port was not correctly logged when the outgoing_port log
2269 selector was set (it logged the transort's default port).
2271 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2272 fallback_hosts, and "hosts" in the smtp transport.
2274 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2275 host errors such as "Connection refused".
2277 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2278 authentication with radiusclient 0.4.9:
2280 - Error returned from rc_read_config was caught wrongly
2281 - Username/password not passed on to radius server due to wrong length.
2283 The presumption is that some radiusclient API changes for 4.51/PH/17
2284 were not taken care of correctly. The code is still untested by me (my
2285 Linux distribution still has 0.3.2 of radiusclient), but it was
2286 contributed by a Radius user.
2288 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2289 expanding the "port" option of the smtp transport.
2291 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2292 while reading a MIME header. Thanks to Tom Hughes for a patch.
2294 PH/24 Include config.h inside local_scan.h so that configuration settings are
2297 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2298 that in an ACL for RCPT (for example), you can examine exactly what was
2301 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2302 commands, but it was not correctly comparing the address with the actual
2303 client host address. Thus, it would show the EHLO address in Received:
2304 header lines when this was not necessary.
2306 PH/27 Added the % operator to ${eval:}.
2308 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2309 it should be ignoring failures (because with -C, for example, it has lost
2310 privilege). It wasn't ignoring creation failures other than "already
2313 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2316 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2317 because an input line was too long, either on its own, or by virtue of
2318 too many continuations, the temporary file was not being removed, and the
2319 return code was incorrect.
2321 PH/31 Missing "BOOL" in function definition in filtertest.c.
2323 PH/32 Applied Sieve patches from the maintainer.
2325 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2327 PH/33 Added "verify = not_blind".
2329 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2330 Local/Makefile (with some defaults set). These are used in built scripts
2331 such as exicyclog, but they have never been used in the exim_install
2332 script (though there are many overriding facilities there). I have
2333 arranged that the exim_install script now takes note of these two
2336 PH/35 Installed configuration files for Dragonfly.
2338 PH/36 When a locally submitted message by a trusted user did not contain a
2339 From: header, and the sender address was obtained from -f or from an SMTP
2340 MAIL command, and the trusted user did not use -F to supply a sender
2341 name, $originator_name was incorrectly used when constructing a From:
2342 header. Furthermore, $originator_name was used for submission mode
2343 messages from external hosts without From: headers in a similar way,
2344 which is clearly wrong.
2346 PH/37 Added control=suppress_local_fixups.
2348 PH/38 When log_selector = +received_sender was set, and the addition of the
2349 sender made the log line's construction buffer exactly full, or one byte
2350 less than full, an overflow happened when the terminating "\n" was
2353 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2354 when the result of a list match is failure because a DNS lookup failed.
2356 PH/40 RM_COMMAND is now used in the building process.
2358 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2359 the "build-* directories that it finds.
2361 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2362 address in a domain literal was a prefix of an interface address.
2364 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2365 when verifying a sender address, unless rewrite_headers is false.
2367 PH/44 Wrote a long comment about why errors_to addresses are verified as
2368 recipients, not senders.
2370 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2371 the ratelimit ACL was added.
2373 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2375 PH/46 Added extra information about PostgreSQL errors to the error string.
2377 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2378 avoiding the need to install special zones in a real server. This is
2379 backwards compatible; if it can't find the fake resolver, it drops back.
2380 Thus, both old and new test suites can be run.
2382 TF/02 Added util/ratelimit.pl
2384 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2385 clock is set back in time.
2387 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2388 Candler <B.Candler@pobox.com>.
2390 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2391 OK for addresses that are the result of redirections.
2393 PH/48 A number of further additions for the benefit of the new test suite,
2394 including a fake gethostbyname() that interfaces to the fake DNS resolver
2397 TF/06 The fix for widen_domains has also been applied to qualify_single and
2398 search_parents which are the other dnslookup options that can cause
2401 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2404 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2406 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2407 service" if it encountered an MX record with an empty target hostname.
2408 The message is now "an MX or SRV record indicated no SMTP service".
2410 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2411 if the DNS of the sending site is misconfigured. This is quite a
2412 common situation. This change restores the behaviour of treating a
2413 helo verification defer as a failure.
2415 PH/51 If self=fail was set on a router, the bounce message did not include the
2416 actual error message.
2422 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2424 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2425 command itself ended in error, the underlying message about the transport
2426 filter timeout was being overwritten with the pipe command error. Now the
2427 underlying error message should be appended to the second error message.
2429 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2431 PH/02 Reduce the amount of output that "make" produces by default. Full output
2432 can still be requested.
2434 PH/03 The warning log line about a condition test deferring for a "warn" verb
2435 was being output only once per connection, rather than after each
2436 occurrence (because it was using the same function as for successful
2437 "warn" verbs). This seems wrong, so I have changed it.
2439 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2440 it should not have, which might have caused a crash in the right
2441 circumstances, but probably never did.
2443 PH/04 Installed a modified version of Tony Finch's patch to make submission
2444 mode fix the return path as well as the Sender: header line, and to
2445 add a /name= option so that you can make the user's friendly name appear
2448 TF/03 Added the control = fakedefer ACL modifier.
2450 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2451 Mark Lowes for thorough testing.
2453 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2455 TK/03 Merged latest SRS patch from Miles Wilton.
2457 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2458 with the definition in sysexits.h (which is #included earlier).
2459 Fortunately, Exim does not actually use EX_OK. The code used to try to
2460 preserve the sysexits.h value, by assumimg that macro definitions were
2461 scanned for macro replacements. I have been disabused of this notion,
2462 so now the code just undefines EX_OK before #including unistd.h.
2464 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2465 in the smtp transport. When a block could not be written in a single
2466 write() function, the timeout was being re-applied to each part-write.
2467 This seems wrong - if the receiver was accepting one byte at a time it
2468 would take for ever. The timeout is now adjusted when this happens. It
2469 doesn't have to be particularly precise.
2471 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2472 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2474 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2475 without a domain if the check to <postmaster@domain> fails.
2477 SC/01 Eximstats: added -xls and the ability to specify output files
2478 (patch written by Frank Heydlauf).
2480 SC/02 Eximstats: use FileHandles for outputing results.
2482 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2484 SC/04 Eximstats: fixed display of large numbers with -nvr option
2486 SC/05 Eximstats: fixed merging of reports with empty tables.
2488 SC/06 Eximstats: added the -include_original_destination flag
2490 SC/07 Eximstats: removed tabs and trailing whitespace.
2492 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2494 TK/06 MBOX spool code: Add real "From " MBOX separator line
2495 so the .eml file is really in mbox format (even though
2496 most programs do not really care). Patch from Alex Miller.
2498 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2499 The latter is generated from $received_to and is only set if the
2500 message has one envelope recipient. SA can use these headers,
2501 obviously out-of-the-box. Patch from Alex Miller.
2503 PH/08 The ${def test on a variable was returning false if the variable's
2504 value was "0", contrary to what the specification has always said!
2505 The result should be true unless the variable is empty.
2507 PH/09 The syntax error of a character other than { following "${if
2508 def:variable_name" (after optional whitespace) was not being diagnosed.
2509 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2510 accidental colon was present, for example, could give incorrect results.
2512 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2513 result is used (not including appendfile, where other changes are about
2516 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2517 This involved changing a lot of size variables from int to off_t. It
2518 should work with maildirs and everything.
2520 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2521 spamd dying while we are connected to it.
2523 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2526 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2527 function for BDB 4.3.
2529 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2531 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2532 the sources for PCRE have been re-arranged and more files are now
2535 PH/15 The code I had for printing potentially long long variables in PH/11
2536 above was not the best (it lost precision). The length of off_t variables
2537 is now inspected at build time, and an appropriate printing format (%ld
2538 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2539 to be "long long int" or "long int". This is needed for the internal
2540 formatting function string_vformat().
2542 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2543 the configuration file to be ":syslog", then the script "guesses" where
2544 the logs files are, rather than using the compiled in default. In our
2545 case the guess is not the same as the compiled default, so the script
2546 suddenly stopped working when I started to use syslog. The patch checks
2547 to see if log_file_path is "". If so, it attempts to read it from exim
2548 with no configuration file to get the compiled in version, before it
2549 falls back to the previous guessing code."
2551 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2552 implementing BATV in an Exim configuration. See NewStuff for the gory
2555 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2556 Makefile that are specific to HP-UX.
2558 PH/18 If the "use_postmaster" option was set for a recipient callout together
2559 with the "random" option, the postmaster address was used as the MAIL
2560 FROM address for the random test, but not for the subsequent recipient
2561 test. It is now used for both.
2563 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2564 patch removes a few documentation additions to RFC 3028, because the
2565 latest draft now contains them. It adds the new en;ascii-case comparator
2566 and a new error check for 8bit text in MIME parts. Comparator and
2567 require names are now matched exactly. I enabled the subaddress
2568 extension, but it is not well tested yet (read: it works for me)."
2570 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2571 rework some of the code of TK/09 above to avoid the hardwired use of
2572 "%lld" and "long long". Replaced the call to snprintf() with a call to
2575 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2576 records point to non-existent hosts", "retry timeout exceeded", and
2577 "retry time not reached for any host after a long failure period".
2579 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2580 experimental DomainKeys support:
2582 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2583 (2) On an error such as an illegally used "control", the wrong name for
2584 the control was given.
2586 These problems did NOT occur unless DomainKeys support was compiled.
2588 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2590 PH/24 Added ${if match_ip condition.
2592 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2593 databases so that it will be absolutely obvious if a crash occurs in the
2594 DB library. This is a regular occurrence (often caused by mis-matched
2597 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2598 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2599 picked up on a user's system that detects such things. There doesn't seem
2600 to be a gcc warning option for this - only an attribute that has to be
2601 put on the function's prototype. It seems that in Fedora Core 4 they have
2602 set this on a number of new functions. No doubt there will be more in due
2605 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2606 specify a transport. However, if an address that was verified by such a
2607 router was the subject of a callout, Exim crashed because it tried to
2608 read the rcpt_include_affixes from the non-existent transport. Now it
2609 just assumes that the setting of that option is false. This bug was
2610 introduced by 4.51/PH/31.
2612 PH/28 Changed -d+all to exclude +memory, because that information is very
2613 rarely of interest, but it makes the output a lot bigger. People tend to
2614 do -d+all out of habit.
2616 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2617 code in os-type was giving problems when libc.so lives in lib64, like on
2620 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2621 aren't the modern standard, and it seems that some systems' include files
2622 don't always have them. Exim was already checking for some of the newer
2623 ones like T_AAAA, and defining it itself. I've added checks for all the
2624 record types that Exim uses.
2626 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2627 not automatically generating a new one, as it is supposed to. This
2628 prevented TLS from working. If the file did exist, but contained invalid
2629 data, a new version was generated, as expected. It was only the case of a
2630 non-existent file that was broken.
2632 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2633 with a change in libdomainkeys > 0.64.
2635 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2636 from DNS. If the selector record carries the flag, it now has
2637 precedence over the domain-wide flag.
2639 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2641 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2642 the use of a port name that isn't defined in /etc/services) occurred, the
2643 message was deferred as in a normal delivery, and thus remained on the
2644 spool, instead of being failed because of the mua_wrapper setting. This
2645 is now fixed, and I tidied up some of the mua_wrapper messages at the
2648 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2649 the messages in a hash of arrays rather than using individual hashes.
2650 This is a bit cleaner and results in dramatic memory savings, albeit
2651 at a slight CPU cost.
2653 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2654 as requested by Marc Sherman.
2656 SC/10 Eximstats: added histograms for user specified patterns as requested
2659 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2661 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2662 fopen() in the content-scanning modules that did not already have it.
2668 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2669 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2671 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2673 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2675 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2676 to test Sieve filters that use "vacation".
2678 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2679 that changes the way the GnuTLS parameters are stored in the cache file.
2680 The new format can be generated externally. For backward compatibility,
2681 if the data in the cache doesn't make sense, Exim assumes it has read an
2682 old-format file, and it generates new data and writes a new file. This
2683 means that you can't go back to an older release without removing the
2686 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2687 work if there are any delivery delays because "one_time" forces the
2688 parent to be marked "delivered", so its unseen clone is never tried
2689 again. For this reason, Exim now forbids the simultaneous setting of
2692 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2693 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2694 entry if you want to know the details. The fix, however, neglected to
2695 consider the case when local delivery batching is involved. The test for
2696 "previously delivered" was not happening when checking to see if an
2697 address could be batched with a previous (undelivered) one; under
2698 certain circumstances this could lead to multiple deliveries to the same
2701 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2702 in its include files, and this causes problems building Exim.
2704 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2705 header_syntax) but Exim was just ignoring anything given after a slash.
2706 In particular, this caused confusion with an attempt to use "verify =
2707 reverse_host_lookup/defer_ok". An error is now given when options are
2708 supplied for verify items that do not have them. (Maybe reverse_host_
2709 lookup should have a defer_ok option, but that's a different point.)
2711 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2712 defined by RFC 821) to 2048, because there were problems with some AUTH
2713 commands, and RFC 1869 says the size should be increased for extended
2714 SMTP commands that take arguments.
2716 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2719 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2720 "unknown" error; now it says that the functionality isn't in the binary.
2722 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2723 an address' error message when a string expansion fails (syntax or
2724 whatever). Otherwise the password may appear in the log. Following change
2725 PH/42 below, there is no longer a chance of it appearing in a bounce
2728 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2730 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2731 of its addresses was ever tried. (Bugzilla bug #2.)
2733 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2734 the result incorrectly in the debug output. (It correctly added a newline
2735 to what was transported.)
2737 TF/01 Added $received_time.
2739 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2740 commented out examples of how to interface to a virus scanner and to
2741 SpamAssassin. Also added commented examples of av_scanner and
2742 spamd_address settings.
2744 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2745 and controls are allowed in which ACLs. There were a couple of minor
2746 errors. Some of the entries in the conditions table (which is a table of
2747 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2748 negation of where the condition IS allowed.
2750 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2752 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2753 header file does not have a version number, so I've had to invent a new
2754 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2755 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2756 radiusclient), but it was contributed by a Radius user.
2758 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2759 files or directories, for OpenSSL.
2761 PH/19 When an Exim process that is running as root has to create an Exim log
2762 file, it does so in a subprocess that runs as exim:exim so as to get the
2763 ownership right at creation (otherwise, other Exim processes might see
2764 the file with the wrong ownership). There was no test for failure of this
2765 fork() call, which would lead to the process getting stuck as it waited
2766 for a non-existent subprocess. Forks do occasionally fail when resources
2767 run out. I reviewed all the other calls to fork(); they all seem to check
2770 PH/20 When checking for unexpected SMTP input at connect time (before writing
2771 the banner), Exim was not dealing correctly with a non-positive return
2772 from the read() function. If the client had disconnected by this time,
2773 the result was a log entry for a synchronization error with an empty
2774 string after "input=" when read() returned zero. If read() returned -1
2775 (an event I could not check), uninitialized data bytes were printed.
2776 There were reports of junk text (parts of files, etc) appearing after
2779 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2781 PH/22 Added support for macro redefinition, and (re)definition in between
2782 driver and ACL definitions.
2784 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2785 forgetting to use the resulting value; it was using the unexpanded value.
2787 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2788 hadn't been configured. The fix is from Juergen Kreileder, who
2789 understands it better than I do:
2791 "Here's what I see happening with three configured cyrus_sasl
2792 authenticators configured (plain, login, cram-md5):
2794 On startup auth_cyrus_sasl_init() gets called for each of these.
2795 This means three calls to sasl_listmech() without a specified mech_list.
2796 => SASL tests which mechs of all available mechs actually work
2797 => three warnings about OTP not working
2798 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2800 With the patch, sasl_listmech() also gets called three times. But now
2801 SASL's mech_list option is set to the server_mech specified in the the
2802 authenticator. Or in other words, the answer from sasl_listmech()
2803 gets limited to just the mech you're testing for (which is different
2805 => the return list contains just 'plain' or 'login', 'cram-md5' or
2806 nothing depending on the value of ob->server_mech.
2808 I've just tested the patch: Authentication still works fine,
2809 unavailable mechs specified in the exim configuration are still
2810 caught, and the auth.log warnings about OTP are gone."
2812 PH/25 When debugging is enabled, the contents of the command line are added
2813 to the debugging output, even when log_selector=+arguments is not
2816 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2817 answer is "GNU", and only if the return is "GNU/something" is the answer
2820 PH/27 $acl_verify_message is now set immediately after the failure of a
2821 verification in an ACL, and so is available in subsequent modifiers. In
2822 particular, the message can be preserved by coding like this:
2824 warn !verify = sender
2825 set acl_m0 = $acl_verify_message
2827 Previously, $acl_verify_message was set only while expanding "message"
2828 and "log_message" when a very denied access.
2830 PH/28 Modified OS/os.c-Linux with
2832 -#ifndef OS_LOAD_AVERAGE
2833 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2835 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2836 nomenclature these days.)
2838 PH/29 Installed patch from the Sieve maintainer that adds the options
2839 sieve_useraddress and sieve_subaddress to the redirect router.
2841 PH/30 In these circumstances:
2842 . Two addresses routed to the same list of hosts;
2843 . First host does not offer TLS;
2844 . First host accepts first address;
2845 . First host gives temporary error to second address;
2846 . Second host offers TLS and a TLS session is established;
2847 . Second host accepts second address.
2848 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2849 and peerdn, if requested) that were in fact used only for the second
2852 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2853 attention to any local part prefix or suffix that was matched by the
2854 router that accepted the address. It now behaves in the same way as it
2855 does for delivery: the affixes are removed from the local part unless
2856 rcpt_include_affixes is set on the transport.
2858 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2859 timeout during the DATA phase of an incoming message.
2861 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2862 applied a patch sanctioned by the Sieve maintainer.
2864 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2865 the uid or gid is negative. A case of a negative gid caused this to be
2866 noticed. The fix allows for either to be negative.
2868 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2869 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2872 PH/36 The change PH/12 above was broken. Fixed it.
2874 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2875 the grounds that routing the same address twice would always produce the
2876 same answer. This might have been true once, but it is certainly no
2877 longer true now. Routing a child address may depend on the previous
2878 routing that produced that child. Some complicated redirection strategies
2879 went wrong when messages had multiple recipients, and made Exim's
2880 behaviour dependent on the order in which the addresses were given.
2882 I have moved the duplicate checking until after the routing is complete.
2883 Exim scans the addresses that are assigned to local and remote
2884 transports, and removes any duplicates. This means that more work will be
2885 done, as duplicates will always all be routed, but duplicates are
2886 presumably rare, so I don't expect this is of any significance.
2888 For deliveries to pipes, files, and autoreplies, the duplicate checking
2889 still happens during the routing process, since they are not going to be
2892 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2893 It corrects a timeout issue with spamd. This is Ian's comment: "The
2894 background is that sometimes spamd either never reads data from a
2895 connection it has accepted, or it never writes response data. The exiscan
2896 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2897 blindly assumes that writes won't block so it may never time out."
2899 PH/39 Allow G after quota size as well as K and M.
2901 PH/40 The value set for $authenticated_id in an authenticator may not contain
2902 binary zeroes or newlines because the value is written to log lines and
2903 to spool files. There was no check on this. Now the value is run through
2904 the string_printing() function so that such characters are converted to
2905 printable escape sequences.
2907 PH/41 $message_linecount is a new variable that contains the total number of
2908 lines in the message. Compare $body_linecount, which is the count for the
2911 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2912 bounce and delay warning messages, except in certain special cases, which
2915 (a) An SMTP error message from a remote host;
2916 (b) A message specified in a :fail: redirection;
2917 (c) A message specified in a "fail" command in a system filter;
2918 (d) A message specified in a FAIL return from the queryprogram router;
2919 (e) A message specified by the cannot_route_message router option.
2921 In these cases only, Exim does include the error details in bounce and
2922 warning messages. There are also a few cases where bland messages such
2923 as "unrouteable address" or "local delivery error" are given.
2925 PH/43 $value is now also set for the "else" part of a ${run expansion.
2927 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2928 being worked on, but at least Exim now implements the latest version to
2931 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2932 process to complete was treated as a delivery failure, a timeout while
2933 writing the message to the pipe was logged, but erroneously treated as a
2934 successful delivery. Such timeouts include transport filter timeouts. For
2935 consistency with the overall process timeout, these timeouts are now
2936 treated as errors, giving rise to delivery failures by default. However,
2937 there is now a new Boolean option for the pipe transport called
2938 timeout_defer, which, if set TRUE, converts the failures into defers for
2939 both kinds of timeout. A transport filter timeout is now identified in
2942 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2943 systems where "make" and "gmake" are different, calling "gmake" at top
2944 level broke things. I've arranged for the value of $(MAKE) to be passed
2945 from the Makefile to this script so that it can call the same version of
2949 A note about Exim versions 4.44 and 4.50
2950 ----------------------------------------
2952 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2953 changes of various kinds. As a consequence, a big documentation update was
2954 needed. This delayed the release for rather longer than seemed good, especially
2955 in the light of a couple of (minor) security issues. Therefore, the changes
2956 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2957 release. So 4.44 and 4.50 are in effect two different branches that both start
2960 I have left the 4.50 change log unchanged; it contains all the changes since
2961 4.43. The change log for 4.44 is below; many of its items are identical to
2962 those for 4.50. This seems to be the most sensible way to preserve the
2963 historical information.
2969 1. Minor wording change to the doc/README.SIEVE file.
2971 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2972 computation of the current number of files was incorrect.
2974 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2975 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2978 4. Give more explanation in the error message when the command for a transport
2979 filter fails to execute.
2981 5. There are several places where Exim runs a non-Exim command in a
2982 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2983 was being done only for the command run by the queryprogram router. It is
2984 now done for all such subprocesses. The other cases are: ${run, transport
2985 filters, and the commands run by the lmtp and pipe transports.
2987 6. Added CONFIGURE_GROUP build-time option.
2989 7. Some older OS have a limit of 256 on the maximum number of file
2990 descriptors. Exim was using setrlimit() to set 1000 as a large value
2991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2992 systems. I've change it so that if it can't get 1000, it tries for 256.
2994 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2995 was an oversight, and furthermore, ever since the addition of extra
2996 controls (e.g. 4.43/32), the checks on when to allow different forms of
2997 "control" were broken. There should now be diagnostics for all cases when a
2998 control that does not make sense is encountered.
3000 9. Added the /retain_sender option to "control=submission".
3002 10. $recipients is now available in the predata ACL (oversight).
3004 11. Tidy the search cache before the fork to do a delivery from a message
3005 received from the command line. Otherwise the child will trigger a lookup
3006 failure and thereby defer the delivery if it tries to use (for example) a
3007 cached ldap connection that the parent has called unbind on.
3009 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3010 of $address_data from the recipient verification was clobbered by the
3011 sender verification.
3013 13. The value of address_data from a sender verification is now available in
3014 $sender_address_data in subsequent conditions in the ACL statement.
3016 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3018 15. Added a new option "connect=<time>" to callout options, to set a different
3021 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3022 was its contents. (It was OK if the option was not defined at all.)
3024 17. A "Completed" log line is now written for messages that are removed from
3025 the spool by the -Mrm option.
3027 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3028 information about exactly what failed.
3030 19. Added -dd to debug only the daemon process.
3032 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3033 handles timeouts, both on the server side and network timeouts. Renamed the
3034 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3036 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3037 It is now set to "smtps".
3039 22. $host_address is now set to the target address during the checking of
3040 ignore_target_hosts.
3042 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3043 being passed; this would have caused $sender_host_name to have been used if
3044 matching the list had actually called for a host name (not very likely,
3045 since this list is usually IP addresses). A host name is now passed as
3048 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3049 code that specifies a non-restarting handler (typically sigaction() in
3050 modern systems) in an attempt to fix a rare and obscure crash bug.
3052 25. Narrowed the window for a race in the daemon that could cause it to ignore
3053 SIGCHLD signals. This is not a major problem, because they are used only to
3054 wake it up if nothing else does.
3056 26. A malformed maildirsize file could cause Exim to calculate negative values
3057 for the mailbox size or file count. Odd effects could occur as a result.
3058 The maildirsize information is now recalculated if the size or filecount
3061 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3062 support for a long time. Removed HAVE_SYS_VFS_H.
3064 28. Installed the latest version of exipick from John Jetmore.
3066 29. In an address list, if the pattern was not a regular expression, an empty
3067 subject address (from a bounce message) matched only if the pattern was an
3068 empty string. Non-empty patterns were not even tested. This was the wrong
3069 because it is perfectly reasonable to use an empty address as part of a
3070 database query. An empty address is now tested by patterns that are
3071 lookups. However, all the other forms of pattern expect the subject to
3072 contain a local part and a domain, and therefore, for them, an empty
3073 address still always fails if the pattern is not itself empty.
3075 30. Exim went into a mad DNS loop when attempting to do a callout where the
3076 host was specified on an smtp transport, and looking it up yielded more
3077 than one IP address.
3079 31. Re-factored the code for checking spool and log partition space into a
3080 function that finds that data and another that does the check. The former
3081 is then used to implement four new variables: $spool_space, $log_space,
3082 $spool_inodes, and $log_inodes.
3084 32. The RFC2047 encoding function was originally intended for short strings
3085 such as real names; it was not keeping to the 75-character limit for
3086 encoded words that the RFC imposes. It now respects the limit, and
3087 generates multiple encoded words if necessary. To be on the safe side, I
3088 have increased the buffer size for the ${rfc2047: expansion operator from
3091 33. It is now permitted to omit both strings after an "if" condition; if the
3092 condition is true, the result is "true". As before, when the second string
3093 is omitted, a false condition yields an empty string. This makes it less
3094 cumbersome to write custom ACL and router conditions.
3096 34. Failure to deliver a bounce message always caused it to be frozen, even if
3097 there was an errors_to setting on the router. The errors_to setting is now
3100 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3101 canonical form (fully expanded) before being placed in
3102 $sender_host_address.
3104 36. The table in the code that translates DNS record types into text (T_A to
3105 "A" for instance) was missing entries for NS and CNAME. It is just possible
3106 that this could have caused confusion if both these types were looked up
3107 for the same domain, because the text type is used as part of Exim's
3108 per-process caching. But the chance of anyone hitting this buglet seems
3111 37. The dnsdb lookup has been extended in a number of ways.
3113 (1) There is a new type, "zns", which walks up the domain tree until it
3114 finds some nameserver records. It should be used with care.
3116 (2) There is a new type, "mxh", which is like "mx" except that it returns
3117 just the host names, not the priorities.
3119 (3) It is now possible to give a list of domains (or IP addresses) to be
3120 looked up. The behaviour when one of the lookups defers can be
3121 controlled by a keyword.
3123 (4) It is now possible to specify the separator character for use when
3124 multiple records are returned.
3126 38. The dnslists ACL condition has been extended: it is now possible to supply
3127 a list of IP addresses and/or domains to be looked up in a particular DNS
3130 39. Added log_selector=+queue_time_overall.
3132 40. When running the queue in the test harness, wait just a tad after forking a
3133 delivery process, to get repeatability of debugging output.
3135 41. Include certificate and key file names in error message when GnuTLS fails
3136 to set them up, because the GnuTLS error message doesn't include the name
3137 of the failing file when there is a problem reading it.
3139 42. Allow both -bf and -bF in the same test run.
3141 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3143 44. The "Exiscan patch" is now merged into the mainline Exim source.
3145 45. Sometimes the final signoff response after QUIT could fail to get
3146 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3147 < 0 before doing a fflush(). This bug looks as though it goes back to the
3148 introduction of TLS in release 3.20, but "sometimes" must have been rare
3149 because the tests only now provoked it.
3151 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3152 (this can affect the format of dates).
3154 47. exim_tidydb, when checking for the continued existence of a message for
3155 which it has found a message-specific retry record, was not finding
3156 messages that were in split spool directories. Consequently, it was
3157 deleting retry records that should have stayed in existence.
3159 48. Steve fixed some bugs in eximstats.
3161 49. The SPA authentication driver was not abandoning authentication and moving
3162 on to the next authenticator when an expansion was forced to fail,
3163 contradicting the general specification for all authenticators. Instead it
3164 was generating a temporary error. It now behaves as specified.
3166 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3167 (the order specifies the preference for clients). The order is now AES256,
3168 AES128, 3DES, ARCFOUR128.
3170 51. Small patch to Sieve code - explicitly set From: when generating an
3173 52. Exim crashed if a remote delivery caused a very long error message to be
3174 recorded - for instance if somebody sent an entire SpamAssassin report back
3175 as a large number of 550 error lines. This bug was coincidentally fixed by
3176 increasing the size of one of Exim's internal buffers (big_buffer) that
3177 happened as part of the Exiscan merge. However, to be on the safe side, I
3178 have made the code more robust (and fixed the comments that describe what
3181 53. Now that there can be additional text after "Completed" in log lines (if
3182 the queue_time_overall log selector is set), a one-byte patch to exigrep
3183 was needed to allow it to recognize "Completed" as not the last thing in
3186 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3187 patch that reportedly fixes this has been added. I am not expert enough to
3188 create a test for it. This is what the patch creator wrote:
3190 "I found a little strange behaviour of ldap code when working with
3191 Windows 2003 AD Domain, where users was placed in more than one
3192 Organization Units. When I tried to give exim partial DN, the exit code
3193 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3194 But simultaneously result of request was absolutely normal ldap result,
3195 so I produce this patch..."
3197 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3198 REFERENCE, so I have modified the code to exclude the patch when that macro
3201 55. Some experimental protocols are using DNS PTR records for new purposes. The
3202 keys for these records are domain names, not reversed IP addresses. The
3203 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3204 leaves it alone. Component reversal etc. now happens only for IP addresses.
3207 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3209 57. Double the size of the debug message buffer (to 2048) so that more of very
3210 long debug lines gets shown.
3212 58. The exicyclog utility now does better if the number of log files to keep
3213 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3215 59. Two changes related to the smtp_active_hostname option:
3217 (1) $smtp_active_hostname is now available as a variable.
3218 (2) The default for smtp_banner uses $smtp_active_hostname instead
3219 of $primary_hostname.
3221 60. The host_aton() function is supposed to be passed a string that is known
3222 to be a valid IP address. However, in the case of IPv6 addresses, it was
3223 not checking this. This is a hostage to fortune. Exim now panics and dies
3224 if the condition is not met. A case was found where this could be provoked
3225 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3226 components; fortuitously, this particular loophole had already been fixed
3227 by change 4.50/55 above.
3229 If there are any other similar loopholes, the new check in host_aton()
3230 itself should stop them being exploited. The report I received stated that
3231 data on the command line could provoke the exploit when Exim was running as
3232 exim, but did not say which command line option was involved. All I could
3233 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3234 running as the user.
3237 61. There was a buffer overflow vulnerability in the SPA authentication code
3238 (which came originally from the Samba project). I have added a test to the
3239 spa_base64_to_bits() function which I hope fixes it.
3242 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3243 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3245 63. The daemon start-up calls getloadavg() while still root for those OS that
3246 need the first call to be done as root, but it missed one case: when
3247 deliver_queue_load_max is set with deliver_drop_privilege. This is
3248 necessary for the benefit of the queue runner, because there is no re-exec
3249 when deliver_drop_privilege is set.
3251 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3252 This has been fixed.
3254 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3255 was in use, was not putting the data itself into the right store pool;
3256 consequently, it could be overwritten for a subsequent message in the same
3257 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3260 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3262 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3263 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3264 the function were treating the return as a boolean value, which happened to
3265 work because 0=false and not-0=true, but is not correct code.
3267 68. The host_aton() function was not handling scoped IPv6 addresses (those
3268 with, for example, "%eth0" on the end) correctly.
3270 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3271 negated items (that is, ~something) in unsigned ints. Some compilers
3272 apparently mutter when there is no cast.
3274 70. If an address verification called from an ACL failed, and did not produce a
3275 user-specific message (i.e. there was only a "system" message), nothing was
3276 put in $acl_verify_message. In this situation, it now puts the system
3279 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3280 session; change 4.31/43 added the unwanted input to the log line - except
3281 that it did not do this in the start of session case. It now does.
3283 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3284 This is wrong and can cause the other end to generate a synchronization
3285 error if it is another Exim or anything else that does the synchronization
3286 check. A QUIT command is no longer sent after a timeout.
3288 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3289 during host lookups.
3291 74. The defer_ok option of callout verification was not working if it was used
3292 when verifying addresses in header lines, that is, for this case:
3294 verify = header_sender/callout=defer_ok
3296 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3297 those file descriptors could be used for SMTP connections. If anything
3298 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3299 could be sent to the SMTP client, causing chaos. The daemon now opens
3300 stdin, stdout, and stderr to /dev/null when it puts itself into the
3303 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3304 log by default. The user can override this with suitable Perl magic.
3306 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3307 the log message when discard triggers, was not working for the DATA ACL or
3308 for the non-SMTP ACL.
3310 78. Error message wording change in sieve.c.
3312 79. If smtp_accept_max_per_host was set, the number of connections could be
3313 restricted to fewer than expected, because the daemon was trying to set up
3314 a new connection before checking whether the processes handling previous
3315 connections had finished. The check for completed processes is now done
3316 earlier. On busy systems, this bug wouldn't be noticed because something
3317 else would have woken the daemon, and it would have reaped the completed
3320 80. If a message was submitted locally by a user whose login name contained one
3321 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3322 It caused a spool format error. I have fixed the spool reading code. A
3323 related problem was that the "from" clause in the Received: line became
3324 illegal because of the space(s). It is now covered by ${quote_local_part.
3326 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3329 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3331 83. Patch from Sieve maintainer to fix unterminated string problem in
3332 "vacation" handling.
3334 84. Some minor changes to the Linux configuration files to help with other
3335 OS variants using glibc.
3337 85. One more patch for Sieve to update vacation handling to latest spec.
3340 ----------------------------------------------------
3341 See the note above about the 4.44 and 4.50 releases.
3342 ----------------------------------------------------
3348 1. Change 4.43/35 introduced a bug that caused file counts to be
3349 incorrectly computed when quota_filecount was set in an appendfile
3352 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3353 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3356 3. Give more explanation in the error message when the command for a transport
3357 filter fails to execute.
3359 4. There are several places where Exim runs a non-Exim command in a
3360 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3361 was being done only for the command run by the queryprogram router. It is
3362 now done for all such subprocesses. The other cases are: ${run, transport
3363 filters, and the commands run by the lmtp and pipe transports.
3365 5. Some older OS have a limit of 256 on the maximum number of file
3366 descriptors. Exim was using setrlimit() to set 1000 as a large value
3367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3368 systems. I've change it so that if it can't get 1000, it tries for 256.
3370 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3371 was an oversight, and furthermore, ever since the addition of extra
3372 controls (e.g. 4.43/32), the checks on when to allow different forms of
3373 "control" were broken. There should now be diagnostics for all cases when a
3374 control that does not make sense is encountered.
3376 7. $recipients is now available in the predata ACL (oversight).
3378 8. Tidy the search cache before the fork to do a delivery from a message
3379 received from the command line. Otherwise the child will trigger a lookup
3380 failure and thereby defer the delivery if it tries to use (for example) a
3381 cached ldap connection that the parent has called unbind on.
3383 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3384 of $address_data from the recipient verification was clobbered by the
3385 sender verification.
3387 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3388 was its contents. (It was OK if the option was not defined at all.)
3390 11. A "Completed" log line is now written for messages that are removed from
3391 the spool by the -Mrm option.
3393 12. $host_address is now set to the target address during the checking of
3394 ignore_target_hosts.
3396 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3397 being passed; this would have caused $sender_host_name to have been used if
3398 matching the list had actually called for a host name (not very likely,
3399 since this list is usually IP addresses). A host name is now passed as
3402 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3403 code that specifies a non-restarting handler (typically sigaction() in
3404 modern systems) in an attempt to fix a rare and obscure crash bug.
3406 15. Narrowed the window for a race in the daemon that could cause it to ignore
3407 SIGCHLD signals. This is not a major problem, because they are used only to
3408 wake it up if nothing else does.
3410 16. A malformed maildirsize file could cause Exim to calculate negative values
3411 for the mailbox size or file count. Odd effects could occur as a result.
3412 The maildirsize information is now recalculated if the size or filecount
3415 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3416 support for a long time. Removed HAVE_SYS_VFS_H.
3418 18. Updated exipick to current release from John Jetmore.
3420 19. Allow an empty sender to be matched against a lookup in an address list.
3421 Previously the only cases considered were a regular expression, or an
3424 20. Exim went into a mad DNS lookup loop when doing a callout where the
3425 host was specified on the transport, if the DNS lookup yielded more than
3428 21. The RFC2047 encoding function was originally intended for short strings
3429 such as real names; it was not keeping to the 75-character limit for
3430 encoded words that the RFC imposes. It now respects the limit, and
3431 generates multiple encoded words if necessary. To be on the safe side, I
3432 have increased the buffer size for the ${rfc2047: expansion operator from
3435 22. Failure to deliver a bounce message always caused it to be frozen, even if
3436 there was an errors_to setting on the router. The errors_to setting is now
3439 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3440 canonical form (fully expanded) before being placed in
3441 $sender_host_address.
3443 24. Updated eximstats to version 1.33
3445 25. Include certificate and key file names in error message when GnuTLS fails
3446 to set them up, because the GnuTLS error message doesn't include the name
3447 of the failing file when there is a problem reading it.
3449 26. Expand error message when OpenSSL has problems setting up cert/key files.
3452 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3453 (this can affect the format of dates).
3455 28. exim_tidydb, when checking for the continued existence of a message for
3456 which it has found a message-specific retry record, was not finding
3457 messages that were in split spool directories. Consequently, it was
3458 deleting retry records that should have stayed in existence.
3460 29. eximstats updated to version 1.35
3461 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3462 1.35 - bugfix such that pie charts by volume are generated correctly
3464 30. The SPA authentication driver was not abandoning authentication and moving
3465 on to the next authenticator when an expansion was forced to fail,
3466 contradicting the general specification for all authenticators. Instead it
3467 was generating a temporary error. It now behaves as specified.
3469 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3470 (the order specifies the preference for clients). The order is now AES256,
3471 AES128, 3DES, ARCFOUR128.
3473 31. Small patch to Sieve code - explicitly set From: when generating an
3476 32. Exim crashed if a remote delivery caused a very long error message to be
3477 recorded - for instance if somebody sent an entire SpamAssassin report back
3478 as a large number of 550 error lines. This bug was coincidentally fixed by
3479 increasing the size of one of Exim's internal buffers (big_buffer) that
3480 happened as part of the Exiscan merge. However, to be on the safe side, I
3481 have made the code more robust (and fixed the comments that describe what
3484 33. Some experimental protocols are using DNS PTR records for new purposes. The
3485 keys for these records are domain names, not reversed IP addresses. The
3486 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3487 leaves it alone. Component reversal etc. now happens only for IP addresses.
3490 34. The host_aton() function is supposed to be passed a string that is known
3491 to be a valid IP address. However, in the case of IPv6 addresses, it was
3492 not checking this. This is a hostage to fortune. Exim now panics and dies
3493 if the condition is not met. A case was found where this could be provoked
3494 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3495 components; fortuitously, this particular loophole had already been fixed
3496 by change 4.50/55 or 4.44/33 above.
3498 If there are any other similar loopholes, the new check in host_aton()
3499 itself should stop them being exploited. The report I received stated that
3500 data on the command line could provoke the exploit when Exim was running as
3501 exim, but did not say which command line option was involved. All I could
3502 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3503 running as the user.
3506 35. There was a buffer overflow vulnerability in the SPA authentication code
3507 (which came originally from the Samba project). I have added a test to the
3508 spa_base64_to_bits() function which I hope fixes it.
3511 36. The daemon start-up calls getloadavg() while still root for those OS that
3512 need the first call to be done as root, but it missed one case: when
3513 deliver_queue_load_max is set with deliver_drop_privilege. This is
3514 necessary for the benefit of the queue runner, because there is no re-exec
3515 when deliver_drop_privilege is set.
3517 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3518 was in use, was not putting the data itself into the right store pool;
3519 consequently, it could be overwritten for a subsequent message in the same
3520 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3523 38. Sometimes the final signoff response after QUIT could fail to get
3524 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3525 < 0 before doing a fflush(). This bug looks as though it goes back to the
3526 introduction of TLS in release 3.20, but "sometimes" must have been rare
3527 because the tests only now provoked it.
3533 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3534 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3535 Now it returns an integer. A number of calls were still expecting a T/F
3536 return. Fortuitously, in all cases, the tests worked in OK situations,
3537 which is the norm. However, things would have gone wrong on any write
3538 failures on the smtp file descriptor. This function is used when sending
3539 messages over SMTP and also when doing verify callouts.
3541 2. When Exim is called to do synchronous delivery of a locally submitted
3542 message (the -odf or -odi options), it no longer closes stderr before doing
3545 3. Implemented the mua_wrapper option.
3547 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3549 5. Implemented the functions header_remove(), header_testname(),
3550 header_add_at_position(), and receive_remove_recipient(), and exported them
3553 6. If an ACL "warn" statement specified the addition of headers, Exim already
3554 inserted X-ACL-Warn: at the start if there was no header name. However, it
3555 was not making this test for the second and subsequent header lines if
3556 there were newlines in the string. This meant that an invalid header could
3557 be inserted if Exim was badly configured.
3559 7. Allow an ACL "warn" statement to add header lines at the start or after all
3560 the Received: headers, as well as at the end.
3562 8. Added the rcpt_4xx retry error code.
3564 9. Added postmaster_mailfrom=xxx to callout verification option.
3566 10. Added mailfrom=xxxx to the callout verification option, for verify=
3569 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3570 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3572 12. Inserted some casts to stop certain compilers warning when using pointer
3573 differences as field lengths or precisions in printf-type calls (mostly
3574 affecting debugging statements).
3576 13. Added optional readline() support for -be (dynamically loaded).
3578 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3579 same clock tick as a message's arrival, so that its received time was the
3580 same as the "first fail" time on the retry record, and that message
3581 remained on the queue past the ultimate address timeout, every queue runner
3582 would try a delivery (because it was past the ultimate address timeout) but
3583 after another failure, the ultimate address timeout, which should have then
3584 bounced the address, did not kick in. This was a "< instead of <=" error;
3585 in most cases the first failure would have been in the next clock tick
3586 after the received time, and all would be well.
3588 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3589 being recognized when the domain list was tested by the match_domain
3590 condition in an expansion string.
3592 16. Added the ${str2b64: operator.
3594 17. Exim was always calling setrlimit() to set a large limit for the number of
3595 processes, without checking whether the existing limit was already
3596 adequate. (It did check for the limit on file descriptors.) Furthermore,
3597 errors from getrlimit() and setrlimit() were being ignored. Now they are
3598 logged to the main and panic logs, but Exim does carry on, to try to do its
3599 job under whatever limits there are.
3601 18. Imported PCRE 5.0.
3603 19. Trivial typo in log message " temporarily refused connection" (the leading
3606 20. If the log selector return_path_on_delivery was set and an address was
3607 redirected to /dev/null, the delivery process crashed because it assumed
3608 that a return path would always be set for a "successful" delivery. In this
3609 case, the whole delivery is bypassed as an optimization, and therefore no
3612 21. Internal re-arrangement: the function for sending a challenge and reading
3613 a response while authentication was assuming a zero-terminated challenge
3614 string. It's now changed to take a pointer and a length, to allow for
3615 binary data in such strings.
3617 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3619 23. Exim was not respecting finduser_retries when seeking the login of the
3620 uid under which it was called; it was always trying 10 times. (The default
3621 setting of finduser_retries is zero.) Also, it was sleeping after the final
3622 failure, which is pointless.
3624 24. Implemented tls_on_connect_ports.
3626 25. Implemented acl_smtp_predata.
3628 26. If the domain in control=submission is set empty, Exim assumes that the
3629 authenticated id is a complete email address when it generates From: or
3630 Sender: header lines.
3632 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3633 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3634 chown and chgrp in /bin and hostname in /usr/bin.
3636 28. Exim was keeping the "process log" file open after each use, just as it
3637 does for the main log. This opens the possibility of it remaining open for
3638 long periods when the USR1 signal hits a daemon. Occasional processlog
3639 errors were reported, that could have been caused by this. Anyway, it seems
3640 much more sensible not to leave this file open at all, so that is what now
3643 29. The long-running daemon process does not normally write to the log once it
3644 has entered its main loop, and it closes the log before doing so. This is
3645 so that log files can straightforwardly be renamed and moved. However,
3646 there are a couple of unusual error situations where the daemon does write
3647 log entries, and I had neglected to close the log afterwards.
3649 30. The text of an SMTP error response that was received during a remote
3650 delivery was being truncated at 512 bytes. This is too short for some of
3651 the long messages that one sometimes sees. I've increased the limit to
3654 31. It is now possible to make retry rules that apply only when a message has a
3655 specific sender, in particular, an empty sender.
3657 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3658 it possible to be selective about when SMTP synchronization is enforced.
3660 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3662 32. Implemented hosts_connection_nolog.
3664 33. Added an ACL for QUIT.
3666 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3669 35. Added mailbox_size and mailbox_filecount to appendfile.
3671 36. Added control = no_multiline_responses to ACLs.
3673 37. There was a bug in the logic of the code that waits for the clock to tick
3674 in the case where the clock went backwards by a substantial amount such
3675 that the microsecond fraction of "now" was more than the microsecond
3676 fraction of "then" (but the whole seconds number was less).
3678 38. Added support for the libradius Radius client library this is found on
3679 FreeBSD (previously only the radiusclient library was supported).
3685 1. When certain lookups returned multiple values in the form name=value, the
3686 quoting of the values was not always being done properly. Specifically:
3687 (a) If the value started with a double quote, but contained no whitespace,
3689 (b) If the value contained whitespace other than a space character (i.e.
3690 tabs or newlines or carriage returns) it was not quoted.
3691 This fix has been applied to the mysql and pgsql lookups by writing a
3692 separate quoting function and calling it from the lookup code. The fix
3693 should probably also be applied to nisplus, ibase and oracle lookups, but
3694 since I cannot test any of those, I have not disturbed their existing code.
3696 2. A hit in the callout cache for a specific address caused a log line with no
3697 reason for rejecting RCPT. Now it says "Previous (cached) callout
3698 verification failure".
3700 3. There was an off-by-one bug in the queryprogram router. An over-long
3701 return line was truncated at 256 instead of 255 characters, thereby
3702 overflowing its buffer with the terminating zero. As well as fixing this, I
3703 have increased the buffer size to 1024 (and made a note to document this).
3705 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3706 when Exim is waiting for an SMTP response from a remote server, Exim
3707 restarts its select() call on the socket, thereby resetting its timeout.
3708 This is not a problem when such interrupts are rare. Somebody set up a cron
3709 job to run exiwhat every 2 minutes, which is less than the normal select()
3710 timeout (5 or 10 minutes). This meant that the select() timeout never
3711 kicked in because it was always reset. I have fixed this by comparing the
3712 time when an interrupt arrives with the time at the start of the first call
3713 to select(). If more time than the timeout has elapsed, the interrupt is
3714 treated as a timeout.
3716 5. Some internal re-factoring in preparation for the addition of Sieve
3717 extensions (by MH). In particular, the "personal" test is moved to a
3718 separate function, and given an option for scanning Cc: and Bcc: (which is
3719 not set for Exim filters).
3721 6. When Exim created an email address using the login of the caller as the
3722 local part (e.g. when creating a From: or Sender: header line), it was not
3723 quoting the local part when it contained special characters such as @.
3725 7. Installed new OpenBSD configuration files.
3727 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3728 try to make them clearer.
3730 9. Callout options, other than the timeout value, were being ignored when
3731 verifying sender addresses in header lines. For example, when using
3733 verify = header_sender/callout=no_cache
3735 the cache was (incorrectly) being used.
3737 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3738 only the Cygwin environment.
3740 11. When return_path_on_delivery was set as a log selector, if different remote
3741 addresses in the same message used different return paths and parallel
3742 remote delivery occurred, the wrong values would sometimes be logged.
3743 (Whenever a remote delivery process finished, the return path value from
3744 the most recently started remote delivery process was logged.)
3746 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3747 lines when AUTH and/or TLS are in use. This is the "received protocol"
3748 field. Exim used to use "asmtp" for authenticated SMTP, without any
3749 indication (in the protocol name) for TLS use. Now it follows the RFC and
3750 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3751 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3752 names appear in log lines as well as in Received: header lines.
3754 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3755 extensions, and comparison tests, and to fix some bugs.
3757 14. Changes to the "personal" filter test:
3759 (1) The test was buggy in that it was just doing the equivalent of
3760 "contains" tests on header lines. For example, if a user's address was
3761 anne@some.where, the "personal" test would incorrectly be true for
3763 To: susanne@some.where
3765 This test is now done by extracting each address from the header in turn,
3766 and checking the entire address. Other tests that are part of "personal"
3767 are now done using regular expressions (for example, to check local parts
3768 of addresses in From: header lines).
3770 (2) The list of non-personal local parts in From: addresses has been
3771 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3772 taken from the Sieve specification recommendations.
3774 (3) If the message contains any header line starting with "List-" it is
3775 treated as non-personal.
3777 (4) The test for "circular" in the Subject: header line has been removed
3778 because it now seems ill-conceived.
3780 15. Minor typos in src/EDITME comments corrected.
3782 16. Installed latest exipick from John Jetmore.
3784 17. If headers_add on a router specified a text string that was too long for
3785 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3786 of string_sprintf() is now avoided.
3788 18. $message_body_size was not set (it was always zero) when running the DATA
3789 ACL and the local_scan() function.
3791 19. For the "mail" command in an Exim filter, no default was being set for
3792 the once_repeat time, causing a random time value to be used if "once" was
3793 specified. (If the value happened to be <= 0, no repeat happened.) The
3794 default is now 0s, meaning "never repeat". The "vacation" command was OK
3795 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3796 (I found it when inspecting the code).
3798 20. There is now an overall timeout for performing a callout verification. It
3799 defaults to 4 times the callout timeout, which applies to individual SMTP
3800 commands during the callout. The overall timeout applies when there is more
3801 than one host that can be tried. The timeout is checked before trying the
3802 next host. This prevents very long delays if there are a large number of
3803 hosts and all are timing out (e.g. when the network connections are timing
3804 out). The value of the overall timeout can be changed by specifying an
3805 additional sub-option for "callout", called "maxwait". For example:
3807 verify = sender/callout=5s,maxwait=20s
3809 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3810 to the end before writing, but this should make it even safer).
3812 22. Exim was forgetting that it had advertised PIPELINING for the second and
3813 subsequent messages on an SMTP connection. It was also not resetting its
3814 memory on STARTTLS and an internal HELO.
3816 23. When Exim logs an SMTP synchronization error within a session, it now
3817 records whether PIPELINING has been advertised or not.
3819 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3820 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3821 rather than long int.
3823 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3825 26. Added the never_mail option to autoreply.
3831 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3832 crash if the getsockname() call failed; this can happen if a connection is
3833 closed very soon after it is established. The problem was simply in the
3834 order in which certain operations were done, causing Exim to try to write
3835 to the SMTP stream before it had set up the file descriptor. The bug has
3836 been fixed by making things happen in the correct order.
3842 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3843 before the connection was closed, thus losing the rejection response.
3845 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3846 some early Solaris releases, but causes trouble in current releases where
3847 socklen_t is defined.
3849 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3852 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3855 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3856 that releasing the top of it at the end releases what was used for sub-
3857 expansions (unless the block got too big). However, discard this block if
3858 the first thing is a variable or header, so that we can use its block when
3859 it is dynamic (useful for very large $message_headers, for example).
3861 6. Lookups now cache *every* query, not just the most recent. A new, separate
3862 store pool is used for this. It can be recovered when all lookup caches are
3863 flushed. Lookups now release memory at the end of their result strings.
3864 This has involved some general refactoring of the lookup sources.
3866 7. Some code has been added to the store_xxx() functions to reduce the amount
3867 of flapping under certain conditions.
3869 8. log_incoming_interface used to affect only the <= reception log lines. Now
3870 it causes the local interface and port to be added to several more SMTP log
3871 lines, for example "SMTP connection from", and rejection lines.
3873 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3875 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3877 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3878 could be overwritten at the end of the current message (or the start of a
3879 new message if it was set in a HELO ACL). The value is now preserved for
3880 the duration of the SMTP connection.
3882 12. If a transport had a headers_rewrite setting, and a matching header line
3883 contained an unqualified address, that address was qualified, even if it
3884 did not match any rewriting rules. The underlying bug was that the values
3885 of the flags that permit the existence of unqualified sender and recipient
3886 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3887 non-local messages, and by -bnq for local messages) were not being
3888 preserved with the message after it was received.
3890 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3891 "next input=" as part of the text comprising the host identity instead of
3892 the correct text. The code was using the same buffer for two different
3893 strings. However, depending on which order the printing function evaluated
3894 its arguments, the bug did not always show up. Under Linux, for example, my
3895 test suite worked just fine.
3897 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3898 doesn't work with some older versions of Perl. It has been changed to "my",
3899 which in any case is probably the better facility to use.
3901 15. A really picky compiler found some instances of statements for creating
3902 error messages that either had too many or two few arguments for the format
3905 16. The size of the buffer for calls to the DNS resolver has been increased
3906 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3907 for addresses that have a lot of PTR records. This alleviates a problem; it
3908 does not fully solve it.
3910 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3911 buffer now truncates the list and logs the incident, which is the same
3912 action as happens when Exim is looking up a host name and its aliases.
3913 Previously in this situation something unpredictable would happen;
3914 sometimes it was "internal error: store_reset failed".
3916 18. If a server dropped the connection unexpectedly when an Exim client was
3917 using GnuTLS and trying to read a response, the client delivery process
3918 crashed while trying to generate an error log message.
3920 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3921 string, for example:
3923 warn message = H1: something\nH2: something
3925 the text was added as a single header line from Exim's point of view
3926 though it ended up OK in the delivered message. However, searching for the
3927 second and subsequent header lines using $h_h2: did not work. This has been
3928 fixed. Similarly, if a system filter added multiple headers in this way,
3929 the routers could not see them.
3931 20. Expanded the error message when iplsearch is called with an invalid key to
3932 suggest using net-iplsearch in a host list.
3934 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3935 ACLs are no longer actually imposed (and a message to that effect is
3938 22. If a "gecos" field in a passwd entry contained escaped characters, in
3939 particular, if it contained a \" sequence, Exim got it wrong when building
3940 a From: or a Sender: header from that name. A second bug also caused
3941 incorrect handling when an unquoted " was present following a character
3942 that needed quoting.
3944 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3945 was not being matched caselessly.
3947 24. Arranged for all hyphens in the exim.8 source to be escaped with
3950 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3951 properly. Recipient callout cache records were still being keyed to include
3952 the sender, even when use_sender was set false. This led to far more
3953 callouts that were necessary. The sender is no longer included in the key
3954 when use_sender is false.
3956 26. Added "control = submission" modifier to ACLs.
3958 27. Added the ${base62d: operator to decode base 62 numbers.
3960 28. dnsdb lookups can now access SRV records.
3962 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3963 the configuration file.
3965 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3966 (-v) mode. This makes the output for a verbose queue run more intelligible.
3968 31. Added a use_postmaster feature to recipient callouts.
3970 32. Added the $body_zerocount variable, containing the number of binary zero
3971 bytes in the message body.
3973 33. The time of last modification of the "new" subdirectory is now used as the
3974 "mailbox time last read" when there is a quota error for a maildir
3977 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3979 35. Added +ignore_unknown as a special item in host lists.
3981 36. Code for decoding IPv6 addresses in host lists is now included, even if
3982 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3983 address was recognized as an IP address, but was then not correctly decoded
3984 into binary, causing unexpected and incorrect effects when compared with
3991 1. Very minor rewording of debugging text in manualroute to say "list of
3992 hosts" instead of "hostlist".
3994 2. If verify=header_syntax was set, and a header line with an unqualified
3995 address (no domain) and a large number of spaces between the end of the
3996 name and the colon was received, the reception process suffered a buffer
3997 overflow, and (when I tested it) crashed. This was caused by some obsolete
3998 code that should have been removed. The fix is to remove it!
4000 3. When running in the test harness, delay a bit after writing a bounce
4001 message to get a bit more predictability in the log output.
4003 4. Added a call to search_tidyup() just before forking a reception process. In
4004 theory, someone could use a lookup in the expansion of smtp_accept_max_
4005 per_host which, without the tidyup, could leave open a database connection.
4007 5. Added the variables $recipient_data and $sender_data which get set from a
4008 lookup success in an ACL "recipients" or "senders" condition, or a router
4009 "senders" option, similar to $domain_data and $local_part_data.
4011 6. Moved the writing of debug_print from before to after the "senders" test
4014 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4015 problems for message scanning, either using a data ACL, or using
4016 local_scan() because the Received: header was not generated till after they
4017 were called (in order to set the time as the time of reception completion).
4018 I have revised the way this works. The header is now generated after the
4019 body is received, but before the ACL or local_scan() are called. After they
4020 are run, the timestamp in the header is updated.
4026 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4027 before starting a queue runner without re-exec. This happened only when
4028 deliver_drop_privilege was set or when the Exim user was set to root. The
4029 effect of the bug was that timeouts during subsequent deliveries caused
4030 crashes instead of being properly handled. The handler is now left at its
4031 default (and expected) setting.
4033 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4034 message, again when deliver_drop_privilege is set or Exim is run as root.
4035 The bug described in (1) was not present in this case, but the tidying up
4036 of the other signals was missing. I have made the two cases consistent.
4038 3. The ignore_target_hosts setting on a manualroute router was being ignored
4039 for hosts that were looked up using the /MX notation.
4041 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4044 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4045 operated on the sender address. After changing the $sender_address to <>
4046 for the sender address verify, Exim was re-instated it as the original
4047 (before rewriting) address, but remembering that it had rewritten it, so it
4048 wasn't rewriting it again. This bug also had the effect of breaking the
4049 sender address verification caching when the sender address was rewritten.
4051 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4052 This has been changed so that if the ip literal address matches
4053 ignore_target_hosts, the router declines.
4055 7. Added expansion conditions match_domain, match_address, and match_local_
4056 part (NOT match_host).
4058 8. The placeholder for the Received: header didn't have a length field set.
4060 9. Added code to Exim itself and to exim_lock to test for a specific race
4061 condition that could lead to file corruption when using MBX delivery. The
4062 issue is with the lockfile that is created in /tmp. If this file is removed
4063 after a process has opened it but before that process has acquired a lock,
4064 there is the potential for a second process to recreate the file and also
4065 acquire a lock. This could lead to two Exim processes writing to the file
4066 at the same time. The added code performs the same test as UW imapd; it
4067 checks after acquiring the lock that its file descriptor still refers to
4068 the same named file.
4070 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4071 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4074 11. Added the smtp_active_hostname option. If used, this will typically be made
4075 to depend on the incoming interface address. Because $interface_address is
4076 not set up until the daemon has forked a reception process, error responses
4077 that can happen earlier (such as "too many connections") no longer contain
4080 12. If an expansion in a condition on a "warn" statement fails because a lookup
4081 defers, the "warn" statement is abandoned, and the next ACL statement is
4082 processed. Previously this caused the whole ACL to be aborted.
4084 13. Added the iplsearch lookup type.
4086 14. Added ident_timeout as a log selector.
4088 15. Added tls_certificate_verified as a log selector.
4090 16. Added a global option tls_require_ciphers (compare the smtp transport
4091 option of the same name). This controls incoming TLS connections.
4093 17. I finally figured out how to make tls_require_ciphers do a similar thing
4094 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4095 before starting the TLS session.
4097 18. Tabs are now shown as \t in -bP output.
4099 19. If the log selector return_path_on_delivery was set, Exim crashed when
4100 bouncing a message because it had too many Received: header lines.
4102 20. If two routers both had headers_remove settings, and the first one included
4103 a superfluous trailing colon, the final name in the first list and the
4104 first name in the second list were incorrectly joined into one item (with a
4105 colon in the middle).
4111 1. Added -C and -D options to the exinext utility, mainly to make it easier
4112 to include in the automated testing, but these could be helpful when
4113 multiple configurations are in use.
4115 2. The exinext utility was not formatting the output nicely when there was
4116 an alternate port involved in the retry record key, nor when there was a
4117 message id as well (for retries that were specific to a specific message
4118 and a specific host). It was also confused by IPv6 addresses, because of
4119 the additional colons they contain. I have fixed the IPv4 problem, and
4120 patched it up to do a reasonable job for IPv6.
4122 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4123 delivery, the log line now contains "pipelined" if PIPELINING was used.
4125 4. An SMTP transport process used to panic and die if the bind() call to set
4126 an explicit outgoing interface failed. This has been changed; it is now
4127 treated in the same way as a connect() failure.
4129 5. A reference to $sender_host_name in the part of a conditional expansion
4130 that was being skipped was still causing a DNS lookup. This no longer
4133 6. The def: expansion condition was not recognizing references to header lines
4134 that used bh_ and bheader_.
4136 7. Added the _cache feature to named lists.
4138 8. The code for checking quota_filecount in the appendfile transport was
4139 allowing one more file than it should have been.
4141 9. For compatibility with Sendmail, the command line option
4149 and sets the incoming protocol and host name (for trusted callers). The
4150 host name and its colon can be omitted when only the protocol is to be set.
4151 Note the Exim already has two private options, -pd and -ps, that refer to
4152 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4153 "s", but I don't think that's a major issue.
4155 10. A number of refactoring changes to the code, none of which should affect
4158 (a) The number of logging options was getting close to filling up the
4159 32-bit word that was used as a bit map. I have split them into two classes:
4160 those that are passed in the argument to log_write(), and those that are
4161 only ever tested independently outside of that function. These are now in
4162 separate 32-bit words, so there is plenty of room for expansion again.
4163 There is no change in the user interface or the logging behaviour.
4165 (b) When building, for example, log lines, the code previously used a
4166 macro that called string_cat() twice, in order to add two strings. This is
4167 not really sufficiently general. Furthermore, there was one instance where
4168 it was actually wrong because one of the argument was used twice, and in
4169 one call a function was used. (As it happened, calling the function twice
4170 did not affect the overall behaviour.) The macro has been replaced by a
4171 function that can join an arbitrary number of extra strings onto a growing
4174 (c) The code for expansion conditions now uses a table and a binary chop
4175 instead of a serial search (which was left over from when there were very
4176 few conditions). Also, it now recognizes conditions like "pam" even when
4177 the relevant support is not compiled in: a suitably worded error message is
4178 given if an attempt is made to use such a condition.
4180 11. Added ${time_interval:xxxxx}.
4182 12. A bug was causing one of the ddress fields not to be passed back correctly
4183 from remote delivery subprocesses. The field in question was not being
4184 subsequently used, so this caused to problems in practice.
4186 13. Added new log selectors queue_time and deliver_time.
4188 14. Might have fixed a bug in maildirsizefile handling that threw up
4189 "unexpected character" debug warnings, and recalculated the data
4190 unnecessarily. In any case, I expanded the warning message to give more
4193 15. Added the message "Restricted characters in address" to the statements in
4194 the default ACL that block characters like @ and % in local parts.
4196 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4197 Three changes have been made:
4199 (a) There was a serious bug; a negative response to MAIL caused the whole
4200 recipient domain to be cached as invalid, thereby blocking all messages
4201 to all local parts at the same domain, from all senders. This bug has
4202 been fixed. The domain is no longer cached after a negative response to
4203 MAIL if the sender used is not empty.
4205 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4208 (c) A new callout option, "use_sender" has been added for people who want
4209 the modified behaviour.
4215 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4218 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4219 indeed breaks things for older releases.
4221 3. Added additional logging to the case where there is a problem reading data
4222 from a filter that is running in a subprocess using a pipe, in order to
4223 try to track down a specific problem.
4225 4. Testing facility fudge: when running in the test harness and attempting
4226 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4227 getting "No route to host". Convert this to a timeout.
4229 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4232 6. Some OS don't have socklen_t but use size_t instead. This affects the
4233 fifth argument of getsockopt() amongst other things. This is now
4234 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4235 can be set for individual OS. I have set it for SunOS5, OSF1, and
4236 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4237 some earlier ones do not.
4239 7. Change 4.30/15 was not doing the test caselessly.
4241 8. The standard form for an IPv6 address literal was being rejected by address
4242 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4243 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4244 this, as well as the form without the "IPv6" on the front (but only when
4245 address literals are enabled, of course).
4247 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4249 10. Exim crashed if a message with an empty sender address specified by -f
4250 encountered a router with an errors_to setting. This could be provoked only
4251 by a command such as
4255 where an empty string was supplied; "<>" did not hit this bug.
4257 11. Installed PCRE release 4.5.
4259 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4260 remained set. It is now erased.
4262 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4263 times from message ids (which are base 36 rather than the normal 62).
4265 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4266 were being counted as actual protocol errors, and logged if the log
4267 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4268 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4269 and DATA following a set of rejected RCPTs do not count as protocol errors.
4270 In other words, Exim assumes they were pipelined, though this may not
4271 actually be the case. Of course, in all cases the client gets an
4272 appropriate error code.
4274 15. If a lookup fails in an ACL condition, a message about the failure may
4275 be available; it is used if testing the ACL cannot continue, because most
4276 such messages specify what the cause of the deferral is. However, some
4277 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4278 that caused an old message to be retained and used if a later statement
4279 caused a defer, replacing the real cause of the deferral.
4281 16. If an IP address had so many PTR records that the DNS lookup buffer
4282 was not large enough to hold them, Exim could crash while trying to process
4283 the truncated data. It now detects and logs this case.
4285 17. Further to 4.21/58, another change has been made: if (and only if) the
4286 first line of a message (the first header line) ends with CRLF, a bare LF
4287 in a subsequent header line has a space inserted after it, so as not to
4288 terminate the header.
4290 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4291 unnecessarily, used atoi() instead of strtol(), and didn't check the
4292 termination when getting file sizes from file names by regex.
4294 19. Completely re-implemented the support for maildirsize files, in the light
4295 of a number of problems with the previous contributed implementation
4296 (4.30/29). In particular:
4298 . If the quota is zero, the maildirsize file is maintained, but no quota is
4301 . If the maildir directory does not exist, it is created before any attempt
4302 to write a maildirsize file.
4304 . The quota value in the file is just a cache; if the quota is changed in
4305 the transport, the new value overrides.
4307 . A regular expression is available for excluding directories from the
4310 20. The autoreply transport checks the characters in options that define the
4311 message's headers; it allows continued headers, but it was checking with
4312 isspace() after an embedded newline instead of explicitly looking for a
4315 21. If all the "regular" hosts to which an address was routed had passed their
4316 expiry times, and had not reached their retry times, the address was
4317 bounced, even if fallback hosts were defined. Now Exim should go on to try
4320 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4321 equivalent code in the SMTP transport. Some hosts send humungous responses
4322 to HELO/EHLO, more than 1024 it seems.
4324 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4325 gives compiler warnings in some environments. I've now done it "properly",
4328 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4329 (because of problems with the built-in one) was declared to return uschar *
4330 instead of char *, causing compiler failure.
4332 25. Fixed a file descriptor leak when processing alias/forward files.
4334 26. Fixed a minor format string issue in dbfn.c.
4336 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4338 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4339 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4342 29. When a sender address is verified, it is cached, to save repeating the test
4343 when there is more than one recipient in a message. However, when the
4344 verification involves a callout, it is possible for different callout
4345 options to be set for different recipients. It is too complicated to keep
4346 track of this in the cache, so now Exim always runs a verification when a
4347 callout is required, relying on the callout cache for the optimization.
4348 The overhead is duplication of the address routing, but this should not be
4351 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4352 to be verified with callout=postmaster, and the main callout worked but the
4353 postmaster check failed, the verification correctly failed. However, if a
4354 subsequent RCPT command asked for sender verification *without* the
4355 postmaster check, incorrect caching caused this verification also to fail,
4358 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4359 it was not caching the DNS options (qualify_single, search_parents) that
4360 were used when the lookup failed. A subsequent lookup with different
4361 options therefore always gave the same answer, though there were cases
4362 where it should not have. (Example: a "domains = !$mx_any" option on a
4363 dnslookup router: the "domains" option is always processed without any
4364 widening, but the router might have qualify_single set.) Now Exim uses the
4365 cached value only when the same options are set.
4367 32. Added John Jetmore's "exipick" utility to the distribution.
4369 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4370 than a timeout (e.g. a certificate is required, and is not provided), an
4371 Exim server now closes the connection immediately. Previously it waited for
4372 the client to close - but if the client is SSL, it seems that they each
4373 wait for each other, leading to a delay before one of them times out.
4375 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4376 maintained 0.8.x compatibility because I don't think many are using it, and
4377 it is clearly obsolete.
4379 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4382 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4383 client certificate was expired. A simple patch fixes this, though I don't
4384 understand the full logic of why the verify callback is called multiple
4387 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4388 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4389 which causes problems with some clients (such as the Certicom SSL Plus
4390 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4391 disables the coutermeasure allowing Eudora to connect."
4393 38. Exim was not checking that a write() to a log file succeeded. This could
4394 lead to Bad Things if a log got too big, in particular if it hit a file
4395 size limit. Exim now panics and dies if it cannot write to a log file, just
4396 as it does if it cannot open a log file.
4398 39. Modified OS/Makefile-Linux so that it now contains
4400 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4402 The two -D definitions ensure that Exim is compiled with large file
4403 support, which makes it possible to handle log files that are bigger than
4406 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4407 instance) a domain was checked against a named list that involved a lookup,
4408 causing $domain_data to be set, then another domain was checked against the
4409 same list, then the first domain was re-checked, the value of $domain_data
4410 after the final check could be wrong. In particular, if the second check
4411 failed, it could be set empty. This bug probably also applied to
4414 41. The strip_trailing_dot option was not being applied to the address given
4415 with the -f command-line option.
4417 42. The code for reading a message's header from the spool was incrementing
4418 $received_count, but never initializing it. This meant that the value was
4419 incorrect (doubled) while delivering a message in the same process in which
4420 it was received. In the most common configuration of Exim, this never
4421 happens - a fresh exec is done - but it can happen when
4422 deliver_drop_privilege is set.
4424 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4425 it now includes up to 150 characters of the unexpected data in the log
4428 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4429 and building data strings. The size of both of these buffers was 10 000
4430 bytes - far larger than anybody would *ever* want, thought I. Needless to
4431 say, somebody hit the limit. I have increased the maximum line length to
4432 20 000 and the maximum data length of concatenated lines to 100 000. I have
4433 also fixed two bugs, because there was no checking on these buffers. Tsk,
4434 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4435 buffer is too small.
4437 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4438 lsearch lookups. Now it does.
4440 46. When parsing a route_list item in a manualroute router, a fixed-length
4441 buffer was used for the list of hosts. I made this 1024 bytes long,
4442 thinking that nobody would ever have a list of hosts that long. Wrong.
4443 Somebody had a whole pile of complicated expansion conditions, and the
4444 string was silently truncated, leading to an expansion error. It turns out
4445 that it is easier to change to an unlimited length (owing to other changes
4446 that have happened since this code was originally written) than to build
4447 structure for giving a limitation error. The length of the item that
4448 expands into the list of hosts is now unlimited.
4450 47. The lsearch lookup could not handle data where the length of text line was
4451 more than 4095 characters. Such lines were truncated, leading to shortened
4452 data being returned. It should now handle lines of any length.
4454 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4455 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4458 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4459 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4460 no longer happens in comment lines. A list of replacements is now placed
4461 at the head of all of the source files, except those whose only change is
4462 to replace PERL_COMMAND in the very first #! line.
4464 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4465 messages on the queue, with a bottom-up merge sort, using code contributed
4466 by Michael Haardt. This should make operations like -bp somewhat faster on
4467 large queues. It won't affect queue runners, except when queue_run_in_order
4470 51. Installed eximstats 1.31 in the distribution.
4472 52. Added support for SRV lookups to the dnslookup router.
4474 53. If an ACL referred to $message_body or $message_body_end, the value was not
4475 reset for any messages that followed in the same SMTP session.
4477 54. The store-handling optimization for building very long strings was not
4478 differentiating between the different store pools. I don't think this
4479 actually made any difference in practice, but I've tidied it.
4481 55. While running the routers to verify a sender address, $sender_address
4482 was still set to the sender address. This is wrong, because when routing to
4483 send a bounce to the sender, it would be empty. Therefore, I have changed
4484 it so that, while verifying a sender address, $sender_address is set to <>.
4485 (There is no change to what happens when verifying a recipient address.)
4487 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4488 target A or AAAA records (if not already returned) without resetting the
4489 qualify_single or search_parents options of the DNS resolver. These are
4490 inappropriate in this case because the targets of MX and SRV records must
4491 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4492 target that, when qualified, matched something in the local domain. These
4493 two options are now turned off when doing these lookups.
4495 57. It seems that at least some releases of Reiserfs (which does not have the
4496 concept of a fixed number of inodes) returns zero and not -1 for the
4497 number of available inodes. This interacted badly with check_spool_inodes,
4498 which assumed that -1 was the "no such thing" setting. What I have done is
4499 to check that the total number of inodes is greater than zero before doing
4500 the test of how many are available.
4502 58. When a "warn" ACL statement has a log_message modifier, the message is
4503 remembered, and not repeated. This is to avoid a lot of repetition when a
4504 message has many recipients that cause the same warning to be written.
4505 Howewer, Exim was preserving the list of already written lines for an
4506 entire SMTP session, which doesn't seem right. The memory is now reset if a
4507 new message is started.
4509 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4510 debugging output unless log_rewrite was also set.
4512 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4513 of (int)(handle) when we know that handle contains (void *)(-1).
4515 61. The Exim daemon panic-logs an error return when it closes the incoming
4516 connection. However "connection reset by peer" seems to be common, and
4517 isn't really an error worthy of noting specially, so that particular error
4520 62. When Exim is trying to find all the local interfaces, it used to panic and
4521 die if the ioctl to get the interface flags failed. However, it seems that
4522 on at least one OS (Solaris 9) it is possible to have an interface that is
4523 included in the list of interfaces, but for which you get a failure error
4524 for this call. This happens when the interface is not "plumbed" into a
4525 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4526 failure of the "get flags" call assumes that the interface is down.
4528 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4529 makes life easier for people who are doing arithmetic on fields extracted
4530 from dates, where you often get leading zeros that should not be
4531 interpreted as octal.
4533 64. Added qualify_domain to the redirect router, to override the global
4536 65. If a pathologically long header line contained very many addresses (the
4537 report of this problem mentioned 10 000) and each of them was rewritten,
4538 Exim could use up a very large amount of memory. (It kept on making new
4539 copies of the header line as it rewrote, and never released the old ones.)
4540 At the expense of a bit more processing, the header rewriting function has
4541 been changed so that it no longer eats memory in this way.
4543 66. The generation of the Received: header has been moved from the time that a
4544 message starts to be received, to the time that it finishes. The timestamp
4545 in the Received: header should now be very close to that of the <= log
4546 line. There are two side-effects of this change:
4548 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4549 logged header lines no longer include the local Received: line, because
4550 it has not yet been created. The same applies to a copy of the message
4551 that is returned to a non-SMTP sender when a message is rejected.
4553 (b) When a filter file is tested using -bf, no additional Received: header
4554 is added to the test message. After some thought, I decided that this
4557 This change does not affect the value of $received_for. It is still set
4558 after address rewriting, but before local_scan() is called.
4560 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4562 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4563 gave an unhelpful panic error message, and a defer error. I have managed to
4564 change this behaviour so that it now rejects any supplied certificate,
4565 which seems right, as the list of acceptable certificates is empty.
4567 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4568 gave an unhelpful defer error. I have not managed to make this reject any
4569 supplied certificates, but the error message it gives is "no certificate
4570 supplied", which is not helpful.
4572 70. exigrep's output now also includes lines that are not associated with any
4573 message, but which match the given pattern. Implemented by a patch from
4574 Martin Sluka, which also tidied up the Perl a bit.
4576 71. Recipient callout verification, like sender verification, was using <> in
4577 the MAIL FROM command. This isn't really the right thing, since the actual
4578 sender may affect whether the remote host accepts the recipient or not. I
4579 have changed it to use the actual sender in the callout; this means that
4580 the cache record is now keyed on a recipient/sender pair, not just the
4581 recipient address. There doesn't seem to be a real danger of callout loops,
4582 since a callout by the remote host to check the sender would use <>.
4583 [SEE ABOVE: changed after hitting problems.]
4585 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4586 temporary errors. However, in the case of such a code being given after
4587 the end of a data transmission (i.e. after ".") Exim was failing to write
4588 a retry record for the message. (Yes, there was some broken host that was
4589 actually sending 8xx at this point.)
4591 73. An unknown lookup type in a host list could cause Exim to panic-die when
4592 the list was checked. (An example that provoked this was putting <; in the
4593 middle of a list instead of at the start.) If this happened during a DATA
4594 ACL check, a -D file could be left lying around. This kind of configuration
4595 error no longer causes Exim to die; instead it causes a defer errror. The
4596 incident is still logged to the main and panic logs.
4598 74. Buglet left over from Exim 3 conversion. The message "too many messages
4599 in one connection" was written to the rejectlog but not the mainlog, except
4600 when address rewriting (yes!) was being logged.
4602 75. Added write_rejectlog option.
4604 76. When a system filter was run not as root (that is, when system_filter_user
4605 was set), the values of the $n variables were not being returned to the
4606 main process; thus, they were not subsequently available in the $sn
4609 77. Added +return_path_on_delivery log selector.
4611 78. A connection timeout was being treated differently from recipients deferred
4612 when testing hosts_max_try with a message that was older than the host's
4613 retry timeout. (The host should not be counted, thus allowing all hosts to
4614 be tried at least once before bouncing.) This may have been the cause of an
4615 occasionally reported bug whereby a message would remain on the queue
4616 longer than the retry timeout, but would be bounced if a delivery was
4617 forced. I say "may" because I never totally pinned down the problem;
4618 setting up timeout/retry tests is difficult. See also the next item.
4620 79. The ultimate address timeout was not being applied to errors that involved
4621 a combination of host plus message (for example, a timeout on a MAIL
4622 command). When an address resolved to a number of possible hosts, and they
4623 were not all tried for each delivery (e.g. because of hosts_max_try), a
4624 message could remain on the queue longer than the retry timeout.
4626 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4629 81. Fixed an obscure SMTP outgoing bug which required at least the following
4630 conditions: (a) there was another message waiting for the same server;
4631 (b) the server returned 5xx to all RCPT commands in the first message so
4632 that the message was not completed; (c) the server dropped the connection
4633 or gave a negative response to the RSET that Exim sends to abort the
4634 transaction. The observed case was a dropped connection after DATA that had
4635 been sent in pipelining mode. That is, the server had advertised PIPELINING
4636 but was not implementing it correctly. The effect of the bug was incorrect
4637 behaviour, such as trying another host, and this could lead to a crash.
4643 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4644 and daemon.c were passed as pointers to ints; they should have been
4645 pointers to socklen_t variables (which are typically unsigned ints).
4647 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4650 3. Fixed a really odd bug that affected only the testing scheme; patching a
4651 certain fixed string in the binary changed the value of another string that
4652 happened to be identical to the end of the original first string.
4654 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4655 name", it returns that address as the IP address. On some operating
4656 systems (e.g. Solaris), it also passes back the IP address string as the
4657 "host name". However, on others (e.g. Linux), it passes back an empty
4658 string. Exim wasn't checking for this, and was changing the host name to an
4659 empty string, assuming it had been canonicized.
4661 5. Although rare, it is permitted to have more than one PTR record for a given
4662 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4663 all the names associated with an address, because they do in Solaris.
4664 However, it seems that they do not in Linux for data that comes from the
4665 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4666 I found this out when I moved to a new Linux workstation and tried to run
4667 the Exim test suite.
4669 To get round this problem I have changed the code so that it now does its
4670 own call to the DNS to look up PTR records when searching for a host name.
4671 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4672 addresses that are only in /etc/hosts are still found.
4674 This behaviour is, however, controlled by an option called host_lookup_
4675 order, which defaults to "bydns:byaddr". If people want to use the other
4676 order, or indeed, just use one or the other means of lookup, they can
4677 specify it in this variable.
4679 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4680 some operating systems, this comes back from gethostbyaddr() as an empty
4681 string, and this is what Exim used to test for. However, it seems that in
4682 other systems, "." is yielded. Exim now tests for this case too.
4684 7. The values of check_spool_space and check_log_space are now held internally
4685 as a number of kilobytes instead of an absolute number of bytes. If a
4686 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4687 kilobyte. This means that much larger values can be stored.
4689 8. Exim monitor: an attempt to get the action menu when not actually pointing
4690 at a message produces an empty menu entitled "No message selected". This
4691 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4692 no entries in it ("Shell widget menu has zero width and/or height"). So I
4693 have added a single, blank menu entry in this case.
4695 9. Added ${quote_local_part.
4697 10. MIME decoding is now applied to the contents of Subject: header lines when
4700 11. Now that a reference to $sender_host_address automatically causes a reverse
4701 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4702 host lookup before query-style lookups in lists that might use this
4703 variable. This has therefore been abolished, and the "net-" prefix is no
4704 longer necessary for query-style lookups.
4706 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4707 have been a typo for LFLAGS, so it has been changed.
4709 13. The install script calls Exim with "-C /dev/null" in order to find the
4710 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4711 to be output. Howeve, since Exim outputs its version number before the
4712 error, it didn't break the script. It just looked ugly. I fixed this by
4713 always allowing "-C /dev/null" if the caller is root.
4715 14. Ignore overlarge ACL variable number when reading spool file - insurance
4716 against a later release with more variables having written the file.
4718 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4719 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4720 this, as well as the form without the "IPv6" on the front.
4722 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4723 OS/Makefile-Darwin file.
4725 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4726 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4728 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4729 variable. However, some OS use ints for inodes. I've added cast to long int
4730 to get rid of the compiler warning.
4732 19. I had forgotten to lock out "/../" in configuration file names when
4733 ALT_CONFIG_PREFIX was set.
4735 20. Routers used for verification do not need to specify transports. However,
4736 if such a router generated a host list, and callout was configured, Exim
4737 crashed, because it could not find a port number from the (non-existent)
4738 transport. It now assumes port 25 in this circumstance.
4740 21. Added the -t option to exigrep.
4742 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4743 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4744 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4745 them is set, LOOKUP_LSEARCH is forced.
4747 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4749 24. Added sender and host information to the "rejected by local_scan()" log
4750 line; previously there was no indication of these.
4752 25. Added .include_if_exists.
4754 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4755 Linux. It turns out that not all file systems support this. Apparently some
4756 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4757 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4758 not supported on the file descriptor, is now ignored when Exim is trying to
4759 sync a directory. This applies only to Linux.
4761 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4763 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4764 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4767 29. Added support for maildirsize files from supplied patch (modified a bit).
4769 30. The use of :fail: followed by an empty string could lead Exim to respond to
4770 sender verification failures with (e.g.):
4772 550 Verification failed for <xxx>
4773 550 Sender verify failed
4775 where the first response line was missing the '-' that indicates it is not
4776 the final line of the response.
4778 31. The loop for finding the name of the user that called Exim had a hardwired
4779 limit of 10; it now uses the value of finduser_retries, which is used for
4780 all other user lookups.
4782 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4785 33. Exim was neglecting to zero errno before one call of strtol() when
4786 expanding a string and expecting an integer value. On some systems this
4787 resulted in spurious "integer overflow" errors. Also, it was casting the
4788 result into an int without checking.
4790 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4791 did not work. The code looks as if it has *never* worked, though it appears
4792 to have been documented since at least releast 1.62. I have made it work.
4794 35. The "timeout_DNS" error in retry rules, also documented since at least
4795 1.62, also never worked. As it isn't clear exactly what this means, and
4796 clearly it isn't a major issue, I have abolished the feature by treating it
4797 as "timeout", and writing a warning to the main and panic logs.
4799 36. The display of retry rules for -brt wasn't always showing the error code
4802 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4803 timeout_connect_A, timeout_connect_MX.
4805 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4806 to the empty sender.
4808 39. The daemon was not analysing the content of -oX till after it had closed
4809 stderr and disconnected from the controlling terminal. This meant that any
4810 syntax errors were only noted on the panic log, and the return code from
4811 the command was 0. By re-arranging the code a little, I've made the
4812 decoding happen first, so such errors now appear on stderr, and the return
4813 code is 1. However, the actual setting up of the sockets still happens in
4814 the disconnected process, so errors there are still only recorded on the
4817 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4818 connections (as happens on some IP stacks) was logged at start up time as
4819 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4820 it from "IPv6 and IPv4", which means that two separate sockets are being
4823 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4824 says whether AF_INET or AF_INET6 was passed as an argument.
4826 42. Exiwhat output was messed up when time zones were included in log
4829 43. Exiwhat now gives more information about the daemon's listening ports,
4830 and whether -tls-on-connect was used.
4832 44. The "port" option of the smtp transport is now expanded.
4834 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4835 silently ignored. Now an error message is written to the main and panic
4838 46. There's a new ACL modifier called "logwrite" which writes to a log file
4839 as soon as it is encountered.
4841 47. Added $local_user_uid and $local_user_gid at routing time.
4843 48. Exim crashed when trying to verify a sender address that was being
4846 49. Exim was recognizing only a space character after ".include". It now also
4847 recognizes a tab character.
4849 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4850 extracting the relevant information from the specification. The man page no
4851 longer contains scrambled data for the -d option, and I've added a section
4852 at the front about calling Exim under different names.
4854 51. Added "extra_headers" argument to the "mail" command in filter files.
4856 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4859 53. Installed eximstats 1.29.
4861 54. Added transport_filter_timeout as a generic transport option.
4863 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4864 Cc: header lines. This was required by RFC 822, but it not required by RFC
4867 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4868 incoming messages that did not have them. Now it does so only if the
4869 message originates locally, that is, if there is no associated remote host
4870 address. When Resent- header lines are present, this applies to the Resent-
4871 lines rather than the non-Resent- lines.
4873 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4874 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4876 58. Messages for configuration errors now include the name of the main
4877 configuration file - useful now that there may be more than one file in a
4878 list (.included file names were always shown).
4880 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4881 for those rare installations that do not start the daemon as root or run it
4882 setuid root. I've cut out the call to initgroups() if the daemon is not
4885 60. The Exim user and group can now be bound into the binary as text strings
4886 that are looked up at the start of Exim's processing.
4888 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4890 62. Added $mailstore_basename variable.
4892 63. Installed patch to sieve.c from Michael Haardt.
4894 64. When Exim failed to open the panic log after failing to open the main log,
4895 the original message it was trying to log was written to stderr and debug
4896 output, but if they were not available (the usual case in production), it
4897 was lost. Now it is written to syslog before the two lines that record the
4898 failures to open the logs.
4900 65. Users' Exim filters run in subprocesses under the user's uid. It is
4901 possible for a "deliver" command or an alias in a "personal" command to
4902 provoke an address rewrite. If logging of address rewriting is configured,
4903 this fails because the process is not running as root or exim. There may be
4904 a better way of dealing with this, but for the moment (because 4.30 needs
4905 to be released), I have disabled address rewrite logging when running a
4906 filter in a non-root, non-exim process.
4912 1. The buildconfig auxiliary program wasn't quoting the value set for
4913 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4914 not defined. This bug was masked in 4.22 by the effect that was fixed in
4917 2. Some messages that were rejected after a message id was allocated were
4918 shown as "incomplete" by exigrep. It no longer does this for messages that
4919 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4921 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4922 have allow_domain_literals set, the ID did not get logged in the <= line.
4923 Domain literals are now always recognized in Message-ID: header lines.
4925 4. The first argument for a ${extract expansion item is the key name or field
4926 number. Leading and trailing spaces in this item were not being ignored,
4927 causing some misleading effects.
4929 5. When deliver_drop_privilege was set, single queue runner processes started
4930 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4931 same command in the process it spins off) were not dropping privilege.
4933 6. When the daemon running as "exim" started a queue runner, it always
4934 re-executed Exim in the spun-off process. This is a waste of effort when
4935 deliver_drop_privilege is set. The new process now just calls the
4936 queue-runner function directly.
4942 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4945 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4946 ignored. Though the use of -oP was forcing the writing of a pid file, it
4947 was always written to the default place.
4949 3. If the message "no IP address found for host xxxx" is generated during
4950 incoming verification, it is now followed by identification of the incoming
4951 connection (so you can more easily find what provoked it).
4953 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4955 5. Added some features to "harden" Exim a bit more against certain attacks:
4957 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4958 be put in Local/Makefile. This is like the never_users runtime option,
4959 but it cannot be overridden. The default setting is "root".
4961 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4962 prefix string with which any file named in a -C command line option
4965 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4966 is retained for -C and -D only if the caller of Exim is root. Without
4967 it, the exim user may also use -C and -D and retain privilege.
4969 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4970 command line option is disabled.
4972 6. Macro names set by the -D option must start with an upper case letter, just
4973 like macro names defined in the configuration file.
4975 7. Added "dereference=" facility to LDAP.
4977 8. Two instances of the typo "uknown" in the source files are fixed.
4979 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4980 the Configure-Makefile script screwed up while processing it.
4982 10. Incorporated PCRE 4.4.
4984 11. The SMTP synchronization check was not operating right at the start of an
4985 SMTP session. For example, it could not catch a HELO sent before the client
4986 waited for the greeting. There is now a check for outstanding input at the
4987 point when the greeting is written. Because of the duplex, asynchronous
4988 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4989 way, but not yet received, when the check is performed.
4991 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4992 on TCP/IP sockets, because this apparently causes some broken clients to
4995 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4996 unchanged) from the Cygwin maintainer.
4998 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4999 was supported instead of testing for mbx. Effectively a typo.
5001 15. The spa authenticator server code was not checking that the input it
5002 received was valid base64.
5004 16. The debug output line for the "set" modifier in ACLs was not showing the
5005 name of the variable that was being set.
5007 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5009 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5010 lookup on its own. Something else was needed to trigger the lookup. For
5011 example, a match in host_lookup or the need for a host name in a host list.
5012 Now, if $sender_host_name is referenced and the host name has not yet been
5013 looked up, a lookup is performed. If the lookup fails, the variable remains
5014 empty, and $host_lookup_failed is set to "1".
5016 19. Added "eqi" as a case-independent comparison operator.
5018 20. The saslauthd authentication condition could segfault if neither service
5019 nor realm was specified.
5021 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5022 error message that was logged was misleading, and incoming SMTP
5023 connections were dropped. The message is now more accurate, and temporary
5024 errors are given to SMTP connections.
5026 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5027 (except RSET) with a 421 error, until QUIT is received. However, it was
5028 failing to send a response to QUIT.
5030 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5031 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5032 ACL is now run after the helo_try_verify_hosts code.
5034 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5035 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5036 case-independent, so other case variants are also recognized). Apparently
5037 some systems use these upper case variants.
5039 25. If more than two messages were waiting for the same host, and a transport
5040 filter was specified for the transport, Exim sent two messages over the
5041 same TCP/IP connection, and then failed with "socket operation on non-
5042 socket" when it tried to send the third.
5044 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5046 27. The extern definition of crypt16() in expand.c was not being excluded when
5047 the OS had its own crypt16() function.
5049 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5050 as a preferred synonym for return_size_limit, both as an option and as an
5053 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5055 30. Changed the default configuration ACL to relax the local part checking rule
5056 for addresses that are not in any local domains. For these addresses,
5057 slashes and pipe symbols are allowed within local parts, but the sequence
5058 /../ is explicitly forbidden.
5060 31. SPA server authentication was not clearing the challenge buffer before
5063 32. log_message in a "warn" ACL statement was writing to the reject log as
5064 well as to the main log, which contradicts the documentation and doesn't
5065 seem right (because no rejection is happening). So I have stopped it.
5067 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5068 However, I am unable to do any testing of this.
5070 34. Fixed an infelicity in the appendfile transport. When checking directories
5071 for a mailbox, to see if any needed to be created, it was accidentally
5072 using path names with one or more superfluous leading slashes; tracing
5073 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5075 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5076 contents are added to the log line that is written for every discarded
5077 recipient. (Previously a log_message setting was ignored.)
5079 36. The ${quote: operator now quotes the string if it is empty.
5081 37. The install script runs exim in order to find its version number. If for
5082 some reason other than non-existence or emptiness, which it checks, it
5083 could not run './exim', it was installing it with an empty version number,
5084 i.e. as "exim-". This error state is now caught, and the installation is
5087 38. An argument was missing from the function that creates an error message
5088 when Exim fails to connect to the socket for saslauthd authentication.
5089 This could cause Exim to crash, or give a corrupted message.
5091 39. Added isip, isip4, and isip6 to ${if conditions.
5093 40. The ACL variables $acl_xx are now saved with the message, and can be
5094 accessed later in routers, transports, and filters.
5096 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5097 strings in the file are not string-expanded.
5099 42. If a MAIL command specified a SIZE value that was too large to fit into an
5100 int variable, the check against message_size_limit failed. Such values are
5101 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5102 day this will have to be increased, but I don't think I want to be around
5103 when emails are that large.
5110 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5111 iconv() is not standard in FreeBSD.
5113 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5114 IPv6 enabled. The observed symptom was a segmentation fault on return from
5115 the function os_common_find_running_interfaces() in src/os.c.
5117 3. In the check_special_case() function in daemon.c I had used "errno" as an
5118 argument name, which causes warnings on some systems. This was basically a
5119 typo, since it was named "eno" in the comments!
5121 4. The code that waits for the clock to tick (at a resolution of some fraction
5122 of a second) so as to ensure message-id uniqueness was always waiting for
5123 at least one whole tick, when it could have waited for less. [This is
5124 almost certainly not relevant at current processor speeds, where it is
5125 unlikely to ever wait at all. But we try to future-proof.]
5127 5. The function that sleeps for a time interval that includes fractions of a
5128 second contained a race. It did not block SIGALRM between setting the
5129 timer, and suspending (a couple of lines later). If the interval was short
5130 and the sigsuspend() was delayed until after it had expired, the suspension
5131 never ended. On busy systems this could lead to processes getting stuck for
5134 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5135 process, before it forks any delivery processes. The open lookup caching
5136 mechanism meant that the open file or database connection was passed into
5137 the delivery process. The problem was that delivery processes always tidy
5138 up cached lookup data. This could cause a problem for the next delivery
5139 process started by the queue runner, because the external queue runner
5140 process does not know about the closure. So the next delivery process
5141 still has data in the lookup cache. In the case of a file lookup, there was
5142 no problem because closing a file descriptor in a subprocess doesn't affect
5143 the parent. However, if the lookup was caching a connection to a database,
5144 the connection was closed, and the second delivery process was likely to
5145 see errors such as "PGSQL: query failed: server closed the connection
5146 unexpectedly". The problem has been fixed by closing all cached lookups
5147 in a queue runner before running a delivery process.
5149 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5150 seem to have the "const" qualifier which it has on other OS. I've
5153 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5154 *of the same type* (client or server) with the same public name that an
5155 error should be diagnosed.
5157 9. When Exim looked up a host name for an IP address, but failed to find the
5158 original IP address when looking up the host name (a safety check), it
5159 output the message "<ip address> does not match any IP for NULL", which was
5160 confusing, to say the least. The bug was that the host name should have
5161 appeared instead of "NULL".
5163 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5164 user that is built into the binary, and the -C or -D options is used, root
5165 privilege is dropped before the configuration file is read. In addition,
5166 logging is switched to stderr instead of the normal log files. If the
5167 configuration then re-defines the Exim user, the unprivileged environment
5168 is probably not what is expected, so Exim logs a panic warning message (but
5171 However, if deliver_drop_privilege is set, the unprivileged state may well
5172 be exactly what is intended, so the warning has been cut out in that case,
5173 and Exim is allowed to try to write to its normal log files.
5179 1. smtp_return_error_details was not giving details for temporary sender
5180 or receiver verification errors.
5182 2. Diagnose a configuration error if two authenticators have the same public
5185 3. Exim used not to create the message log file for a message until the first
5186 delivery attempt. This could be confusing when incoming messages were held
5187 for policy or load reasons. The message log file is now created at the time
5188 the message is received, and an initial "Received" line is written to it.
5190 4. The automatically generated man page for command line options had a minor
5191 bug that caused no ill effects; however, a more serious problem was that
5192 the procedure for building the man page automatically didn't always
5193 operate. Consequently, release 4.20 contains an out-of-date version. This
5194 shouldn't happen again.
5196 5. When building Exim with embedded Perl support, the script that builds the
5197 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5198 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5200 6. The freeze_tell option was not being used for messages that were frozen on
5201 arrival, either by an ACL or by local_scan().
5203 7. Added the smtp_incomplete_transaction log selector.
5205 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5206 was accepting AUTH without a new EHLO.
5208 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5209 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5212 10. Logging of TCP/IP connections (when configured) now happens in the main
5213 daemon process instead of the child process, so that the TCP/IP connection
5214 count is more accurate (but it can never be perfect).
5216 11. The use of "drop" in a nested ACL was not being handled correctly in the
5217 outer ACL. Now, if condition failure induced by the nested "drop" causes
5218 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5219 or "require"), the connection is dropped.
5221 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5222 that yield "discard" can now be used with an "accept" or a "discard" verb,
5223 but an error is generated for any others (because I can't see a useful way
5224 to define what should happen).
5226 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5227 are now processed in the same way as lines in the Exim configuration file.
5228 In particular, continuation lines are supported.
5230 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5232 15. Added -ti meaning -t -i.
5234 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5235 domains, and warn by logging if others are found.
5237 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5238 returned from GIFCONF when Exim is trying to find the list of interfaces on
5239 a host. The code in os.c has been modified to copy each ifreq to an aligned
5240 structure in all cases.
5242 Also, in some cases, the returned ifreq's were being copied to a 'struct
5243 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5244 means the last couple of bytes of an IPv6 address could be chopped if the
5245 ifreq contained only a normal sockaddr (14 bytes storage).
5247 18. Named domain lists were not supported in the hosts_treat_as_local option.
5248 An entry such as +xxxx was not recognized, and was treated as a literal
5251 19. Ensure that header lines added by a DATA ACL are included in the reject log
5252 if the ACL subsequently rejects the message.
5254 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5255 MD5 (which is deprecated).
5257 21. When testing a filter file using -bf, Exim was writing a message when it
5258 took the sender from a "From " line in the message, but it was not doing so
5259 when it took $return_path from a Return-Path: header line. It now does.
5261 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5262 with a valid header line field name (a series of printing characters
5263 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5265 23. Changed "disc" in the source to "disk" to conform to the documentation and
5266 the book and for uniformity.
5268 24. Ignore Sendmail's -Ooption=value command line item.
5270 25. When execve() failed while trying to run a command in a pipe transport,
5271 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5272 could be confused with a return value of 69 from the command itself. This
5273 has been changed to 127, the value the shell returns if it is asked to run
5274 a non-existent command. The wording for the related log line suggests a
5275 non-existent command as the problem.
5277 26. If received_header_text expands to an empty string, do not add a Received:
5278 header line to the message. (Well, it adds a token one on the spool, but
5279 marks it "old" so that it doesn't get used or transmitted.)
5281 27. Installed eximstats 1.28 (addition of -nt option).
5283 28. There was no check for failure on the call to getsockname() in the daemon
5284 code. This can fail if there is a shortage of resources on the system, with
5285 ENOMEM, for example. A temporary error is now given on failure.
5287 29. Contrary to the C standard, it seems that in some environments, the
5288 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5289 program. Exim now does this explicitly; it affects the formatting of
5290 timestamps using strftime().
5292 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5293 complaints. I've now initialized all the variables, to avoid this.
5295 32. Header lines added by a system filter were not being "seen" during
5296 transport-time rewrites.
5298 33. The info_callback() function passed to OpenSSL is set up with type void
5299 (*)(SSL *, int, int), as described somewhere. However, when calling the
5300 function (actually a macro) that sets it up, the type void(*)() is
5301 expected. I've put in a cast to prevent warnings from picky compilers.
5303 34. If a DNS black list lookup found a CNAME record, but there were no A
5304 records associated with the domain it pointed at, Exim crashed.
5306 35. If a DNS black list lookup returned more than one A record, Exim ignored
5307 all but the first. It now scans all returned addresses if a particular IP
5308 value is being sought. In this situation, the contents of the
5309 $dnslist_value variable are a list of all the addresses, separated by a
5312 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5313 to accept a host name and all its aliases if the forward lookup for any of
5314 them yielded the IP address of the incoming connection. Now it accepts only
5315 those names whose forward lookup yields the correct IP address. Any other
5316 names are discarded. This closes a loophole whereby a rogue DNS
5317 administrator could create reverse DNS records to break through a
5318 wildcarded host restriction in an ACL.
5320 37. If a user filter or a system filter that ran in a subprocess used any of
5321 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5322 the wrong values were passed to the pipe command ($thisaddress had the
5323 value of $0, $0 had the value of $1, etc). This bug was introduced by
5324 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5326 38. Improved the line breaking for long SMTP error messages from ACLs.
5327 Previously, if there was no break point between 40 and 75 characters, Exim
5328 left the rest of the message alone. Two changes have been made: (a) I've
5329 reduced the minimum length to 35 characters; (b) if it can't find a break
5330 point between 35 and 75 characters, it looks ahead and uses the first one
5331 that it finds. This may give the occasional overlong line, but at least the
5332 remaining text gets split now.
5334 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5335 file descriptors might be low, and that setting 1000 would always raise it.
5336 It turns out that in some environments, the limit is already over 1000 and
5337 that lowering it causes trouble. So now Exim takes care not to decrease it.
5339 40. When delivering a message, the value of $return_path is set to $sender_
5340 address at the start of routing (routers may change the value). By an
5341 oversight, this default was not being set up when an address was tested by
5342 -bt or -bv, which affected the outcome if any router or filter referred to
5345 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5346 log only when "message" or "log_message" are set. However, if one of the
5347 conditions was an address verification, or a call to a nested ACL, the
5348 messages generated by the underlying test were being passed through. This
5349 no longer happens. The underlying message is available in $acl_verify_
5350 message for both "message" and "log_message" expansions, so it can be
5351 passed through if needed.
5353 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5354 new expansion $bh_ to give the encoded byte string without charset
5355 translation. Translation happens only if iconv() is available; HAVE_ICONV
5356 indicates this at build time. HEADERS_CHARSET gives the charset to
5357 translate to; headers_charset can change it in the configuration, and
5358 "headers charset" can change it in an individual filter file.
5360 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5361 that creates RFC 2047 encoded "words" labels them as that charset instead
5362 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5363 expansion operator; (ii) when Exim creates a From: line for a local
5364 message; (iii) when a header line is rewritten to include a "phrase" part.
5366 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5367 buggy, causing it to skip the first lines of messages whose message ID
5368 ended in 'D'. This would not have bitten before Exim release 4.14, because
5369 message IDs were unlikely to end in 'D' before then. The effect was to have
5370 incorrect size information for certain domains.
5372 45. #include "config.h" was missing at the start of the crypt16.c module. This
5373 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5376 46. If there was a timeout during a "random" callout check, Exim treated it as
5377 a failure of the random address, and carried on sending RSET and the real
5378 address. If the delay was just some slowness somewhere, the response to the
5379 original RCPT would be taken as a response to RSET and so on, causing
5380 mayhem of various kinds.
5382 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5383 when I implemented it. It didn't allow for the fact that some option values
5384 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5385 the right test for positive values.
5387 48. Domain names in DNS records are case-independent. Exim always looks them up
5388 in lower case. Some resolvers return domain names in exactly the case they
5389 appear in the zone file, that is, they may contain uppercase letters. Not
5390 all resolvers do this - some return always lower case. Exim was treating a
5391 change of case by a resolver as a change of domain, similar to a widening
5392 of a domain abbreviation. This triggered its re-routing code and so it was
5393 trying to route what was effectively the same domain again. This normally
5394 caused routing to fail (because the router wouldn't handle the domain
5395 twice). Now Exim checks for this case specially, and just changes the
5396 casing of the domain that it ultimately uses when it transmits the message
5399 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5402 50. If a filter generated a file delivery with a non-absolute name (possible if
5403 no home directory exists for the router), the forbid_file option was not
5406 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5407 the existing equality matching.
5409 52. Exim was using ints instead of ino_t variables in some places where it was
5410 dealing with inode numbers.
5412 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5413 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5414 called TMPDIR, and if it finds it is different, it changes its value.
5416 54. The smtp_printf() function is now made available to local_scan() so
5417 additional output lines can be written before returning. There is also an
5418 smtp_fflush() function to enable the detection of a dropped connection.
5419 The variables smtp_input and smtp_batched_input are exported to
5422 55. Changed the default runtime configuration: the message "Unknown user"
5423 has been removed from the ACL, and instead placed on the localuser router,
5424 using the cannot_route_message feature. This means that any verification
5425 failures that generate their own messages won't get overridden. Similarly,
5426 the "Unrouteable address" message that was in the ACL for unverifiable
5427 relay addresses has also been removed.
5429 56. Added hosts_avoid_esmtp to the smtp transport.
5431 57. The exicyclog script was not checking for the esoteric option
5432 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5433 will work only if exicyclog is run under the appropriate euid.
5435 58. Following a discussion on the list, the rules by which Exim recognises line
5436 endings on incoming messages have been changed. The -dropcr and drop_cr
5437 options are now no-ops, retained only for backwards compatibility. The
5438 following line terminators are recognized: LF CRLF CR. However, special
5439 processing applies to CR:
5441 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5442 nor a local message in the state where . is a terminator.
5444 (ii) If a bare CR is encountered in a header line, an extra space is added
5445 after the line terminator so as not to end the header. The reasoning
5446 behind this is that bare CRs in header lines are most likely either
5447 to be mistakes, or people trying to play silly games.
5449 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5450 was being incorrectly given as 18 bytes larger than it should have been.
5451 This is a VOB (very old bug).
5453 60. This may never have affected anything current, but just in case it has:
5454 When the local host is found other than at the start of a list of hosts,
5455 the local host, those with the same MX, and any that follow, are discarded.
5456 When the list in question was part of a longer list of hosts, the following
5457 hosts (not currently being processed) were also being discarded. This no
5458 longer happens. I'm not sure if this situation could ever has previously
5461 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5464 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5465 header. This is something that is recommended in a new Internet Draft, and
5466 is something that is documented as being done by Sendmail. There are two
5467 possible values. For messages generated by the autoreply transport, Exim
5470 Auto-Submitted: auto-replied
5472 whereas for all other generated messages (e.g. bounces) it adds
5474 Auto-Submitted: auto-generated
5476 63. The "personal" condition in filters now includes a test for the
5477 Auto-Submitted: header. If it contains the string "auto-" the message it
5478 not considered personal.
5480 64. Added rcpt_include_affixes as a generic transport option.
5482 65. Added queue_only_override (default true).
5484 66. Added the syslog_duplication option.
5486 67. If what should have been the first header line of a message consisted of
5487 a space followed by a colon, Exim was mis-interpreting it as a header line.
5488 It isn't of course - it is syntactically invalid and should therefore be
5489 treated as the start of the message body. The misbehaviour could have
5490 caused a number of strange effects, including loss of data in subsequent
5491 header lines, and spool format errors.
5493 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5494 client host had authenticated. This control can now be exercised by an ACL
5495 for more flexibility.
5497 69. By default, callouts do not happen when testing with -bh. There is now a
5498 variant, -bhc, which does actually run the callout code, including
5499 consulting and updating the callout cache.
5501 70. Added support for saslauthd authentication, courtesy of Alexander
5504 71. If statvfs() failed on the spool or log directories while checking their
5505 size for availability, Exim confusingly gave the error "space shortage".
5506 Furthermore, in debugging mode it crashed with a floating point exception.
5507 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5508 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5509 serious problem, Exim now writes to the main and panic logs when this
5510 happens, with details of the failure. It then refuses to accept the
5511 incoming message, giving the message "spool directory problem" or "log
5512 directory problem" with a 421 code for SMTP messages.
5514 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5515 0, 1, and 2 exist, because some OS complain for execs without them (see
5516 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5517 descriptors. However, the code omitted to check that the open succeeded,
5518 causing mysterious errors if for some reason the permissions on /dev/null
5519 got screwed. Now Exim writes a message to the main and panic logs, and
5520 bombs out if it can't open /dev/null.
5522 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5523 interact so that it is all more flexible. It is supposed to remain
5524 backwards compatible. Also added extra_local_interfaces.
5526 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5527 to bomb out with an assertion failure - to the client this appears as a
5528 connection drop. This problem occurs in the part of the code that was taken
5529 from the Samba project. Fortunately, the assertion is in a very simple
5530 function, so I have fixed this by reproducing the function inline in the
5531 one place where it is called, and arranging for authentication to fail
5532 instead of killing the process with assert().
5534 75. The SPA client code was not working when the server requested OEM rather
5535 than Unicode encoding.
5537 76. Added code to make require_files with a specific uid setting more usable in
5538 the case where statting the file as root fails - usually a non-root-mounted
5539 NFS file system. When this happens and the failure is EACCES, Exim now
5540 forks a subprocess and does the per-uid checking as the relevant uid.
5542 77. Added process_log_path.
5544 78. If log_file_path was not explicitly set, a setting of check_log_space or
5545 check_log_inodes was ignored.
5547 79. If a space check for the spool or log partitions fails, the incident is now
5548 logged. Of course, in the latter case the data may get lost...
5550 80. Added the %p formatting code to string_format() so that it can be used to
5551 print addresses in debug_print(). Adjusted all the address printing in the
5552 debugging in store.c to use %p rather than %d.
5554 81. There was a concern that a line of code in smtp_in.c could overflow a
5555 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5556 initially expressed, the concern was not well-founded, because trailing
5557 spaces are removed early. However, if the trailing spaces were followed by
5558 a NULL, they did not get removed, so the overflow was possible. Two fixes
5561 (a) I re-wrote the offending code in a cleaner fashion.
5562 (b) If an incoming SMTP command contains a NULL character, it is rejected
5565 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5566 runs initgroups(), so that if the Exim user is in any additional groups,
5567 they will be used during message reception.
5573 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob