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.
92 PP/21 Defaulting "accept_8bitmime" to true, not false.
94 PP/22 Added -bw for inetd wait mode support.
96 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
97 locate the relevant includes and libraries. Made this the default.
103 PP/01 Solaris build fix for Oracle's LDAP libraries.
104 Bugzilla 1109, patch from Stephen Usher.
106 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
108 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
111 TF/02 Fix a couple more cases where we did not log the error message
112 when unlink() failed. See also change 4.74-TF/03.
114 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
115 lock up or crash if it happened to be inside a call to libc when it
116 got a SIGUSR1 from exiwhat.
118 The SIGUSR1 handler appends the current process status to the process
119 log which is later printed by exiwhat. It used to use the general
120 purpose logging code to do this, but several functions it calls are
121 not safe for signals.
123 The new output code in the SIGUSR1 handler is specific to the process
124 log, and simple enough that it's easy to inspect for signal safety.
125 Removing some special cases also simplifies the general logging code.
126 Removing the spurious timestamps from the process log simplifies
129 TF/04 Improved ratelimit ACL condition.
131 The /noupdate option has been deprecated in favour of /readonly which
132 has clearer semantics. The /leaky, /strict, and /readonly update modes
133 are mutually exclusive. The update mode is no longer included in the
134 database key; it just determines when the database is updated. (This
135 means that when you upgrde Exim will forget old rate measurements.)
137 Exim now checks that the per_* options are used with an update mode that
138 makes sense for the current ACL. For example, when Exim is processing a
139 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
140 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
141 must specify per_mail/readonly. If you omit the update mode it defaults to
142 /leaky where that makes sense (as before) or /readonly where required.
144 The /noupdate option is now undocumented but still supported for
145 backwards compatibility. It is equivalent to /readonly except that in
146 ACLs where /readonly is required you may specify /leaky/noupdate or
147 /strict/noupdate which are treated the same as /readonly.
149 A useful new feature is the /count= option. This is a generalization
150 of the per_byte option, so that you can measure the throughput of other
151 aggregate values. For example, the per_byte option is now equivalent
152 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
154 The per_rcpt option has been generalized using the /count= mechanism
155 (though it's more complicated than the per_byte equivalence). When it is
156 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
157 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
158 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
159 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
160 Note that using per_rcpt with a non-readonly update mode in more than
161 one ACL will cause the recipients to be double-counted. (The per_mail
162 and per_byte options don't have this problem.)
164 The handling of very low rates has changed slightly. If the computed rate
165 is less than the event's count (usually one) then this event is the first
166 after a long gap. In this case the rate is set to the same as this event's
167 count, so that the first message of a spam run is counted properly.
169 The major new feature is a mechanism for counting the rate of unique
170 events. The new per_addr option counts the number of different
171 recipients that someone has sent messages to in the last time period. It
172 behaves like per_rcpt if all the recipient addresses are different, but
173 duplicate recipient addresses do not increase the measured rate. Like
174 the /count= option this is a general mechanism, so the per_addr option
175 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
176 example, measure the rate that a client uses different sender addresses
177 with the options per_mail/unique=$sender_address. There are further
178 details in the main documentation.
180 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
182 TF/06 Removed a few PCRE remnants.
184 TF/07 Automatically extract Exim's version number from tags in the git
185 repository when doing development or release builds.
187 PP/02 Raise smtp_cmd_buffer_size to 16kB.
188 Bugzilla 879. Patch from Paul Fisher.
190 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
191 Heavily based on revision 40f9a89a from Simon Arlott's tree.
194 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
196 PP/05 Variable $av_failed, true if the AV scanner deferred.
197 Bugzilla 1078. Patch from John Horne.
199 PP/06 Stop make process more reliably on build failure.
200 Bugzilla 1087. Patch from Heiko Schlittermann.
202 PP/07 Make maildir_use_size_file an _expandable_ boolean.
203 Bugzilla 1089. Patch from Heiko Schlittermann.
205 PP/08 Handle ${run} returning more data than OS pipe buffer size.
206 Bugzilla 1131. Patch from Holger Weiß.
208 PP/09 Handle IPv6 addresses with SPF.
209 Bugzilla 860. Patch from Wolfgang Breyha.
211 PP/10 GnuTLS: support TLS 1.2 & 1.1.
213 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
216 PP/11 match_* no longer expand right-hand-side by default.
217 New compile-time build option, EXPAND_LISTMATCH_RHS.
218 New expansion conditions, "inlist", "inlisti".
220 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
222 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
224 PP/14 fix log_write() format string regression from TF/03.
225 Bugzilla 1152. Patch from Dmitry Isaikin.
231 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
233 PP/02 Harmonised TLS library version reporting; only show if debugging.
234 Layout now matches that introduced for other libraries in 4.74 PP/03.
236 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
238 PP/04 New "dns_use_edns0" global option.
240 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
243 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
244 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
246 TK/01 Updated PolarSSL code to 0.14.2.
247 Bugzilla 1097. Patch from Andreas Metzler.
249 PP/07 Catch divide-by-zero in ${eval:...}.
252 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
255 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
256 format-string attack -- SECURITY: remote arbitrary code execution.
258 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
259 time unintentionally subject to list matching rules, letting the header
260 cause arbitrary Exim lookups (of items which can occur in lists, *not*
261 arbitrary string expansion). This allowed for information disclosure.
263 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
264 INT_MIN/-1 -- value coerced to INT_MAX.
270 NM/01 Workround for PCRE version dependency in version reporting
273 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
274 This fixes portability to compilers other than gcc, notably
275 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
277 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
278 makefiles for portability to HP-UX and POSIX correctness.
280 PP/01 Permit LOOKUP_foo enabling on the make command-line.
281 Also via indented variable definition in the Makefile.
282 (Debugging by Oliver Heesakkers).
284 PP/02 Restore caching of spamd results with expanded spamd_address.
285 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
287 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
288 Improves build reliability. Fix from: Frank Elsner
290 NM/02 Fix wide character breakage in the rfc2047 coding
291 Fixes bug 1064. Patch from Andrey N. Oktyabrski
293 NM/03 Allow underscore in dnslist lookups
294 Fixes bug 1026. Patch from Graeme Fowler
296 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
297 Code patches from Adam Ciarcinski of NetBSD.
299 NM/04 Fixed exiqgrep to cope with mailq missing size issue
302 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
303 is logged, to avoid truncation. Patch from John Horne.
305 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
306 Patch from Jakob Hirsch.
308 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
309 SQL string expansion failure details.
310 Patch from Andrey Oktyabrski.
312 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
313 Patch from Simon Arlott.
315 PP/09 New lookups functionality failed to compile on old gcc which rejects
316 extern declarations in function scope.
317 Patch from Oliver Fleischmann
319 PP/10 Use sig_atomic_t for flags set from signal handlers.
320 Check getgroups() return and improve debugging.
321 Fixed developed for diagnosis in bug 927 (which turned out to be
324 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
325 Patch from Mark Zealey.
327 PP/12 Bugzilla 1056: Improved spamd server selection.
328 Patch from Mark Zealey.
330 PP/13 Bugzilla 1086: Deal with maildir quota file races.
331 Based on patch from Heiko Schlittermann.
333 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
334 Patch from Uwe Doering, sign-off by Michael Haardt.
336 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
337 variable declaration deep within a block. Bug and patch from
340 PP/15 lookups-Makefile IRIX compatibilty coercion.
342 PP/16 Make DISABLE_DKIM build knob functional.
344 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
345 Patch by Simon Arlott
347 TF/03 Fix valgrind.h portability to C89 compilers that do not support
348 variable argument macros. Our copy now differs from upstream.
354 TF/01 Failure to get a lock on a hints database can have serious
355 consequences so log it to the panic log.
357 TF/02 Log LMTP confirmation messages in the same way as SMTP,
358 controlled using the smtp_confirmation log selector.
360 TF/03 Include the error message when we fail to unlink a spool file.
362 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
363 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
364 for maintaining out-of-tree patches for some time.
366 PP/01 Bugzilla 139: Documentation and portability issues.
367 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
368 Handle per-OS dynamic-module compilation flags.
370 PP/02 Let /dev/null have normal permissions.
371 The 4.73 fixes were a little too stringent and complained about the
372 permissions on /dev/null. Exempt it from some checks.
373 Reported by Andreas M. Kirchwitz.
375 PP/03 Report version information for many libraries, including
376 Exim version information for dynamically loaded libraries. Created
377 version.h, now support a version extension string for distributors
378 who patch heavily. Dynamic module ABI change.
380 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
381 privilege escalation vulnerability whereby the Exim run-time user
382 can cause root to append content of the attacker's choosing to
385 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
388 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
389 If dropping privileges for untrusted macros, we disabled normal logging
390 on the basis that it would fail; for the Exim run-time user, this is not
391 the case, and it resulted in successful deliveries going unlogged.
392 Fixed. Reported by Andreas Metzler.
398 PP/01 Date: & Message-Id: revert to normally being appended to a message,
399 only prepend for the Resent-* case. Fixes regression introduced in
400 Exim 4.70 by NM/22 for Bugzilla 607.
402 PP/02 Include check_rfc2047_length in configure.default because we're seeing
403 increasing numbers of administrators be bitten by this.
405 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
407 PP/03 Bugzilla 994: added openssl_options main configuration option.
409 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
411 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
413 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
415 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
416 without a peer certificate, leading to a segfault because of an
417 assumption that peers always have certificates. Be a little more
418 paranoid. Problem reported by Martin Tscholak.
420 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
421 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
422 NB: ClamAV planning to remove STREAM in "middle of 2010".
423 CL also introduces -bmalware, various -d+acl logging additions and
424 more caution in buffer sizes.
426 PP/09 Implemented reverse_ip expansion operator.
428 PP/10 Bugzilla 937: provide a "debug" ACL control.
430 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
432 PP/12 Bugzilla 973: Implement --version.
434 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
436 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
438 PP/15 Bugzilla 816: support multiple condition rules on Routers.
440 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
441 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
442 ignore trailing whitespace.
444 JJ/02 prevent non-panic DKIM error from being sent to paniclog
446 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
449 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
450 Notification from Dr Andrew Aitchison.
452 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
453 ExtendedDetectionInfo response format.
454 Notification from John Horne.
456 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
459 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
460 XSL and documented dependency on system catalogs, with examples of how
463 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
466 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
467 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
468 configuration file which is writeable by the Exim user or group.
470 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
471 of configuration files to cover files specified with the -C option if
472 they are going to be used with root privileges, not just the default
475 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
476 option (effectively making it always true).
478 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
479 files to be used while preserving root privileges.
481 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
482 that rogue child processes cannot use them.
484 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
485 run-time user, instead of root.
487 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
488 Exim run-time user without dropping privileges.
490 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
491 result string, instead of calling string_vformat() twice with the same
494 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
495 for other users. Others should always drop root privileges if they use
496 -C on the command line, even for a whitelisted configure file.
498 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
500 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
506 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
507 $data_path, and $header_path variables; fixed documentation bugs and
510 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
511 exipick to access non-standard spools, including the "frozen" queue
514 NM/01 Bugzilla 965: Support mysql stored procedures.
515 Patch from Alain Williams
517 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
519 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
520 Patch from Andreas Metzler
522 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
523 Patch from Kirill Miazine
525 NM/05 Bugzilla 671: Added umask to procmail example.
527 JJ/03 installed exipick 20100323.0, fixing doc bug
529 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
530 directory. Notification and patch from Dan Rosenberg.
532 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
534 TK/02 Improve log output when DKIM signing operation fails.
536 MH/01 Treat the transport option dkim_domain as a colon separated
537 list, not as a single string, and sign the message with each element,
538 omitting multiple occurences of the same signer.
540 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
541 Bugzilla 985, 986. Patch by Simon Arlott
543 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
544 Patch by Simon Arlott
546 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
547 MBX locking. Notification from Dan Rosenberg.
553 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
555 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
557 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
559 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
561 NM/04 Bugzilla 915: Flag broken perl installation during build.
567 TK/01 Added patch by Johannes Berg that expands the main option
568 "spamd_address" if it starts with a dollar sign.
570 TK/02 Write list of recipients to X-Envelope-Sender header when building
571 the mbox-format spool file for content scanning (suggested by Jakob
574 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
575 (http://www.dcc-servers.net/) support via dccifd. Activated by
576 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
578 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
579 by Mark Daniel Reidel <mr@df.eu>.
581 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
582 When building exim an external PCRE library is now needed -
583 PCRE is a system library on the majority of modern systems.
584 See entry on PCRE_LIBS in EDITME file.
586 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
587 conversation. Added nologin parameter to request.
588 Patch contributed by Kirill Miazine.
590 TF/01 Do not log submission mode rewrites if they do not change the address.
592 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
594 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
595 log files in place. Contributed by Roberto Lima.
597 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
599 TF/03 Bugzilla 615: When checking the local_parts router precondition
600 after a local_part_suffix or local_part_prefix option, Exim now
601 does not use the address's named list lookup cache, since this
602 contains cached lookups for the whole local part.
604 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
605 Robert Millan. Documentation is in experimental-spec.txt.
607 TF/04 Bugzilla 668: Fix parallel build (make -j).
609 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
611 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
612 Patch provided by Jan Srzednicki.
614 TF/05 Leading white space used to be stripped from $spam_report which
615 wrecked the formatting. Now it is preserved.
617 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
618 that they are available at delivery time.
620 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
622 TF/08 TLS error reporting now respects the incoming_interface and
623 incoming_port log selectors.
625 TF/09 Produce a more useful error message if an SMTP transport's hosts
626 setting expands to an empty string.
628 NM/06 Bugzilla 744: EXPN did not work under TLS.
629 Patch provided by Phil Pennock.
631 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
632 Patch provided by Richard Godbee.
634 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
635 acl_smtp_notquit, added index entry.
637 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
638 Patch provided by Eugene Bujak.
640 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
641 accept(). Patch provided by Maxim Dounin.
643 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
644 Patch provided by Phil Pennock.
646 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
648 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
649 Patch provided by Brad "anomie" Jorsch.
651 NM/14 Bugzilla 622: Added timeout setting to transport filter.
652 Patch provided by Dean Brooks.
654 TK/05 Add native DKIM support (does not depend on external libraries).
656 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
657 Patch provided by Graeme Fowler.
659 NM/16 Bugzilla 851: Documentation example syntax fix.
661 NM/17 Changed NOTICE file to remove references to embedded PCRE.
663 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
666 NM/19 Bugzilla 745: TLS version reporting.
667 Patch provided by Phil Pennock.
669 NM/20 Bugzilla 167: bool: condition support.
670 Patch provided by Phil Pennock.
672 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
673 clients. Patch provided by Phil Pennock.
675 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
676 Patch provided by Brad "anomie" Jorsch.
678 NM/23 Bugzilla 687: Fix misparses in eximstats.
679 Patch provided by Heiko Schlittermann.
681 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
682 Patch provided by Heiko Schlittermann.
684 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
685 plus update to original patch.
687 NM/26 Bugzilla 799: Documentation correction for ratelimit.
689 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
690 Patch provided by David Brownlee.
692 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
694 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
696 NM/30 Bugzilla 888: TLS documentation bugfixes.
698 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
700 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
701 Unlike the original bugzilla I have changed all shell scripts in src tree.
703 NM/33 Bugzilla 898: Transport filter timeout fix.
704 Patch by Todd Rinaldo.
706 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
707 Patch by Serge Demonchaux.
709 NM/35 Bugzilla 39: Base64 decode bug fixes.
710 Patch by Jakob Hirsch.
712 NM/36 Bugzilla 909: Correct connect() call in dcc code.
714 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
716 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
718 NM/39 Bugzilla 911: Fixed MakeLinks build script.
724 TK/01 Add preliminary DKIM support. Currently requires a forked version of
725 ALT-N's libdkim that I have put here:
726 http://duncanthrax.net/exim-experimental/
728 Note to Michael Haardt: I had to rename some vars in sieve.c. They
729 were called 'true' and it seems that C99 defines that as a reserved
730 keyword to be used with 'bool' variable types. That means you could
731 not include C99-style headers which use bools without triggering
732 build errors in sieve.c.
734 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
735 as mailq or other aliases. Changed the --help handling significantly
736 to do whats expected. exim_usage() emits usage/help information.
738 SC/01 Added the -bylocaldomain option to eximstats.
740 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
742 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
744 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
750 PH/01 Another patch from the Sieve maintainer.
752 PH/02 When an IPv6 address is converted to a string for single-key lookup
753 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
754 dots are used instead of colons so that keys in lsearch files need not
755 contain colons. This was done some time before quoting was made available
756 in lsearch files. However, iplsearch files do require colons in IPv6 keys
757 (notated using the quote facility) so as to distinguish them from IPv4
758 keys. This meant that lookups for IP addresses in host lists did not work
759 for iplsearch lookups.
761 This has been fixed by arranging for IPv6 addresses to be expressed with
762 colons if the lookup type is iplsearch. This is not incompatible, because
763 previously such lookups could never work.
765 The situation is now rather anomolous, since one *can* have colons in
766 ordinary lsearch keys. However, making the change in all cases is
767 incompatible and would probably break a number of configurations.
769 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
772 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
773 conversion specification without a maximum field width, thereby enabling
774 a rogue spamd server to cause a buffer overflow. While nobody in their
775 right mind would setup Exim to query an untrusted spamd server, an
776 attacker that gains access to a server running spamd could potentially
777 exploit this vulnerability to run arbitrary code as the Exim user.
779 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
780 $primary_hostname instead of what libspf2 thinks the hosts name is.
782 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
783 a directory entry by the name of the lookup key. Previously, if a
784 symlink pointed to a non-existing file or a file in a directory that
785 Exim lacked permissions to read, a lookup for a key matching that
786 symlink would fail. Now it is enough that a matching directory entry
787 exists, symlink or not. (Bugzilla 503.)
789 PH/03 The body_linecount and body_zerocount variables are now exported in the
792 PH/04 Added the $dnslist_matched variable.
794 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
795 This means they are set thereafter only if the connection becomes
798 PH/06 Added the client_condition to authenticators so that some can be skipped
799 by clients under certain conditions.
801 PH/07 The error message for a badly-placed control=no_multiline_responses left
802 "_responses" off the end of the name.
804 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
806 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
807 (without spaces) instead of just copying the configuration text.
809 PH/10 Added the /noupdate option to the ratelimit ACL condition.
811 PH/11 Added $max_received_linelength.
813 PH/12 Added +ignore_defer and +include_defer to host lists.
815 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
816 way in which PCRE > 7.0 is built.
818 PH/14 Implemented queue_only_load_latch.
820 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
821 MAIL command. The effect was to mangle the value on 64-bit systems.
823 PH/16 Another patch from the Sieve maintainer.
825 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
827 PH/18 If a system quota error occurred while trying to create the file for
828 a maildir delivery, the message "Mailbox is full" was not appended to the
829 bounce if the delivery eventually timed out. Change 4.67/27 below applied
830 only to a quota excession during the actual writing of the file.
832 PH/19 It seems that peer DN values may contain newlines (and other non-printing
833 characters?) which causes problems in log lines. The DN values are now
834 passed through string_printing() before being added to log lines.
836 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
837 and InterBase are left for another time.)
839 PH/21 Added message_body_newlines option.
841 PH/22 Guard against possible overflow in moan_check_errorcopy().
843 PH/23 POSIX allows open() to be a macro; guard against that.
845 PH/24 If the recipient of an error message contained an @ in the local part
846 (suitably quoted, of course), incorrect values were put in $domain and
847 $local_part during the evaluation of errors_copy.
853 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
854 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
857 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
858 issue a MAIL command.
860 PH/02 In an ACL statement such as
862 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
864 if a client was not listed at all, or was listed with a value other than
865 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
866 the condition was not true (as it should be), so access was not denied.
867 The bug was that the ! inversion was incorrectly passed on to the second
868 item. This has been fixed.
870 PH/03 Added additional dnslists conditions == and =& which are different from
871 = and & when the dns lookup returns more than one IP address.
873 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
874 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
876 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
877 FSYNC, which compiles an option called disable_fsync that allows for
878 bypassing fsync(). The documentation is heavily laced with warnings.
880 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
882 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
883 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
884 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
885 including adding "make clean"; (3) Added -fPIC when compiling the test
886 dynamically loaded module, to get rid of a warning.
888 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
889 message fails, move_frozen_messages = true and ignore_bounce_errors_after
890 = 0s. The bug is otherwise harmless.
892 PH/07 There was a bug in the dovecot authenticator such that the value of
893 $auth1 could be overwritten, and so not correctly preserved, after a
894 successful authentication. This usually meant that the value preserved by
895 the server_setid option was incorrect.
897 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
899 PH/09 Installed PCRE release 7.0.
901 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
902 run for batched SMTP input. It is now run at the start of every message
903 in the batch. While fixing this I discovered that the process information
904 (output by running exiwhat) was not always getting set for -bs and -bS
905 input. This is fixed, and it now also says "batched" for BSMTP.
907 PH/11 Added control=no_pipelining.
909 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
910 patch, slightly modified), and move the expansion of helo_data till after
911 the connection is made in the smtp transport (so it can use these
914 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
916 PH/14 Added log_selector = +pid.
918 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
920 PH/16 Add ${if forany and ${if forall.
922 PH/17 Added dsn_from option to vary the From: line in DSNs.
924 PH/18 Flush SMTP output before performing a callout, unless control =
925 no_callout_flush is set.
927 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
928 was true (the default) a successful delivery failed to delete the retry
929 item, thus causing premature timeout of the address. The bug is now
932 PH/20 Added hosts_avoid_pipelining to the smtp transport.
934 PH/21 Long custom messages for fakedefer and fakereject are now split up
935 into multiline reponses in the same way that messages for "deny" and
936 other ACL rejections are.
938 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
939 with slight modification.
941 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
942 draft, changing the syntax and factoring some duplicate code".
944 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
945 for deliveries of the second and subsequent messages over the same SMTP
948 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
949 ${reduce, with only minor "tidies".
951 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
953 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
954 expansion side effects.
956 PH/27 When a message times out after an over-quota error from an Exim-imposed
957 quota, the bounce message says "mailbox is full". This message was not
958 being given when it was a system quota that was exceeded. It now should
961 MH/03 Made $recipients available in local_scan(). local_scan() already has
962 better access to the recipient list through recipients_list[], but
963 $recipients can be useful in postmaster-provided expansion strings.
965 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
966 in the case of a MAIL command with additional options following the
967 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
968 were accidentally chopped off.
970 PH/29 SMTP synchronization checks are implemented when a command is read -
971 there is a check that no more input is waiting when there shouldn't be
972 any. However, for some commands, a delay in an ACL can mean that it is
973 some time before the response is written. In this time, more input might
974 arrive, invalidly. So now there are extra checks after an ACL has run for
975 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
976 pipelining has not been advertised.
978 PH/30 MH's patch to allow iscntrl() characters to be list separators.
980 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
981 returned in the SMTP response when smtp_return_error_details was false.
984 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
985 instead of the C I/O that was originally supplied, because problems were
988 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
989 Exim which did not show up earlier: it was assuming that a call to
990 SSL_CTX_set_info_callback() might give an error value. In fact, there is
991 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
992 was a macro that became an assignment, so it seemed to work. This has
993 changed to a proper function call with a void return, hence the compile
994 error. Exim's code has been fixed.
996 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
999 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1001 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1003 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1004 because it turns out to be too big or there is a timeout) there is no
1005 "Completed" line in the log. When some messages of this type were
1006 selected by exigrep, they were listed as "not completed". Others were
1007 picked up by some special patterns. I have improved the selection
1008 criteria to be more general.
1010 PH/38 The host_find_failed option in the manualroute router can now be set
1011 to "ignore", to completely ignore a host whose IP address cannot be
1012 found. If all hosts are ignored, the behaviour is controlled by the new
1013 host_all_ignored option.
1015 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1016 homing or because of multiple MX records with /mx) generated more than
1017 one IP address, and the following item turned out to be the local host,
1018 all the secondary addresses of the first item were incorrectly removed
1019 from the list, along with the local host and any following hosts (which
1020 is what is supposed to happen).
1022 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1023 whoever called Exim into the -H file. In the case of the daemon it was
1024 behaving confusingly. When first started, it used values for whoever
1025 started the daemon, but after a SIGHUP it used the Exim user (because it
1026 calls itself on a restart). I have changed the code so that it now always
1029 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1030 message are rejected with the same error (e.g. no authentication or bad
1031 sender address), and a DATA command is nevertheless sent (as can happen
1032 with PIPELINING or a stupid MUA), the error message that was given to the
1033 RCPT commands is included in the rejection of the DATA command. This is
1034 intended to be helpful for MUAs that show only the final error to their
1037 PH/42 Another patch from the Sieve maintainer.
1039 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1040 Eximstats - Fixed some broken HTML links and added missing column headers
1042 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1043 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1045 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1047 PH/43 Yet another patch from the Sieve maintainer.
1049 PH/44 I found a way to check for a TCP/IP connection going away before sending
1050 the response to the final '.' that terminates a message, but only in the
1051 case where the client has not sent further data following the '.'
1052 (unfortunately, this is allowed). However, in many cases there won't be
1053 any further data because there won't be any more messages to send. A call
1054 to select() can be used: if it shows that the input is "ready", there is
1055 either input waiting, or the socket has been closed. An attempt to read
1056 the next input character can distinguish the two cases. Previously, Exim
1057 would have sent an OK response which the client would never have see.
1058 This could lead to message repetition. This fix should cure that, at
1059 least in a lot of common cases.
1061 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1062 advertised in response to EHLO.
1068 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1069 fixed by 4.65/MH/01 (is this a record?) are fixed:
1071 (i) An empty string was always treated as zero by the numeric comparison
1072 operators. This behaviour has been restored.
1074 (ii) It is documented that the numeric comparison operators always treat
1075 their arguments as decimal numbers. This was broken in that numbers
1076 starting with 0 were being interpreted as octal.
1078 While fixing these problems I realized that there was another issue that
1079 hadn't been noticed. Values of message_size_limit (both the global option
1080 and the transport option) were treated as octal if they started with 0.
1081 The documentation was vague. These values are now always treated as
1082 decimal, and I will make that clear in the documentation.
1088 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1089 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1092 MH/01 Don't check that the operands of numeric comparison operators are
1093 integers when their expansion is in "skipping" mode (fixes bug
1094 introduced by 4.64-PH/07).
1096 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1097 child addresses, Exim now panics and dies. Previously, because the count
1098 is held in a short int, deliveries were likely to be lost. As such a
1099 large number of recipients for a single message is ridiculous
1100 (performance will be very, very poor), I have chosen to impose a limit
1101 rather than extend the field.
1107 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1108 leftover -K file (the existence of which was triggered by #402).
1109 While we were at it, introduced process PID as part of the -K
1110 filename. This should rule out race conditions when creating
1113 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1114 processing considerably. Previous code took too long for large mails,
1115 triggering a timeout which in turn triggers #401.
1117 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1118 in the DK code in transports.c. sendfile() is not really portable,
1119 hence the _LINUX specificness.
1121 TF/01 In the add_headers option to the mail command in an Exim filter,
1122 there was a bug that Exim would claim a syntax error in any
1123 header after the first one which had an odd number of characters
1126 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1127 callout verification, Exim cached a "reject" for the entire domain. This
1128 is correct for most verifications, but it is not correct for a recipient
1129 verification with use_sender or use_postmaster set, because in that case
1130 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1131 case of MAIL FROM:<> rejection from other early rejections (e.g.
1132 rejection of HELO). When verifying a recipient using a non-null MAIL
1133 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1134 Whatever the result of the callout, the value of the domain cache is
1135 left unchanged (for any other kind of callout, getting as far as trying
1136 RCPT means that the domain itself is ok).
1138 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1141 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1142 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1143 session, the daemon ignores ECONNECT errors and logs others; it now
1144 ignores EPIPE as well.
1146 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1147 (quoted-printable decoding).
1149 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1150 later the small subsequent patch to fix an introduced bug.
1152 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1154 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1156 PH/08 An error is now given if message_size_limit is specified negative.
1158 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1159 to be given (somewhat) arbitrary names.
1161 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1164 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1165 miscellaneous code fixes
1167 PH/10 Added the log_reject_target ACL modifier to specify where to log
1170 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1171 hostname. This is wrong, because it relates to the incoming message (and
1172 probably the interface on which it is arriving) and not to the outgoing
1173 callout (which could be using a different interface). This has been
1174 changed to use the value of the helo_data option from the smtp transport
1175 instead - this is what is used when a message is actually being sent. If
1176 there is no remote transport (possible with a router that sets up host
1177 addresses), $smtp_active_hostname is used.
1179 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1180 tweaks were necessary in order to get it to work (see also 21 below):
1181 (a) The code assumed that strncpy() returns a negative number on buffer
1182 overflow, which isn't the case. Replaced with Exim's string_format()
1184 (b) There were several signed/unsigned issues. I just did the minimum
1185 hacking in of casts. There is scope for a larger refactoring.
1186 (c) The code used strcasecmp() which is not a standard C function.
1187 Replaced with Exim's strcmpic() function.
1188 (d) The code set only $1; it now sets $auth1 as well.
1189 (e) A simple test gave the error "authentication client didn't specify
1190 service in request". It would seem that Dovecot has changed its
1191 interface. Fortunately there's a specification; I followed it and
1192 changed what the client sends and it appears to be working now.
1194 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1197 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1198 address A is aliased to B and C, where B exists and C does not. Without
1199 -v the output is "A verified" because verification stops after a
1200 successful redirection if more than one address is generated. However,
1201 with -v the child addresses are also verified. Exim was outputting "A
1202 failed to verify" and then showing the successful verification for C,
1203 with its parentage. It now outputs "B failed to verify", showing B's
1204 parentage before showing the successful verification of C.
1206 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1207 look up a TXT record in a specific list after matching in a combined
1210 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1211 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1212 they consult the DNS. I had assumed they would set it the way they
1213 wanted; and indeed my experiments on Linux seem to show that in some
1214 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1215 To be on the safe side, however, I have now made the interface to
1216 host_find_byname() similar to host_find_bydns(), with an argument
1217 containing the DNS resolver options. The host_find_byname() function now
1218 sets these options at its start, just as host_find_bydns() does. The smtp
1219 transport options dns_qualify_single and dns_search_parents are passed to
1220 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1221 of host_find_byname() use the default settings of RES_DEFNAMES
1222 (qualify_single) but not RES_DNSRCH (search_parents).
1224 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1225 spool_read_header() do less string testing, by means of a preliminary
1226 switch on the second character of optional "-foo" lines. (This is
1227 overdue, caused by the large number of possibilities that now exist.
1228 Originally there were few.) While I was there, I also converted the
1229 str(n)cmp tests so they don't re-test the leading "-" and the first
1230 character, in the hope this might squeeze out yet more improvement.
1232 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1233 flag allowing group syntax was set by the header_syntax check but not
1234 turned off, possible causing trouble later; (2) The flag was not being
1235 set at all for the header_verify test, causing "group"-style headers to
1236 be rejected. I have now set it in this case, and also caused header_
1237 verify to ignore an empty address taken from a group. While doing this, I
1238 came across some other cases where the code for allowing group syntax
1239 while scanning a header line wasn't quite right (mostly, not resetting
1240 the flag correctly in the right place). These bugs could have caused
1241 trouble for malformed header lines. I hope it is now all correct.
1243 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1244 with the "reply" argument non-NULL. The code, however (which originally
1245 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1246 but it didn't always do it. This confused somebody who was copying the
1247 code for some other use. I have removed all the tests.
1249 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1250 feature that was used to support insecure browsers during the U.S. crypto
1251 embargo. It requires special client support, and Exim is probably the
1252 only MTA that supported it -- and would never use it because real RSA is
1253 always available. This code has been removed, because it had the bad
1254 effect of slowing Exim down by computing (never used) parameters for the
1255 RSA_EXPORT functionality.
1257 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1258 authenticator to fail if there's a tab character in the incoming data
1259 (there should never be unless someone is messing about, as it's supposed
1260 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1261 if the connection is using TLS or if the remote IP is the same as the
1262 local IP, and the "valid-client-cert option" if a client certificate has
1265 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1266 authenticators. This can be used for authorization after authentication
1267 succeeds. (In the case of plaintext, it servers for both authentication
1270 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1271 if any retry times were supplied.
1273 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1274 connection, where there is no client IP address to check. In this
1275 situation, the verify now always succeeds.
1277 PH/25 Applied John Jetmore's -Mset patch.
1279 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1281 PH/27 In a string expansion for a processed (not raw) header when multiple
1282 headers of the same name were present, leading whitespace was being
1283 removed from all of them, but trailing whitespace was being removed only
1284 from the last one. Now trailing whitespace is removed from each header
1285 before concatenation. Completely empty headers in a concatenation (as
1286 before) are ignored.
1288 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1289 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1291 PH/29 [Removed. This was a change that I later backed out, and forgot to
1292 correct the ChangeLog entry (that I had efficiently created) before
1293 committing the later change.]
1295 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1296 address errors (4xx response to RCPT) over the same connection as other
1297 messages routed to the same hosts. Such deliveries are always "forced",
1298 so retry times are not inspected. This resulted in far too many retries
1299 for the affected addresses. The effect occurred only when there were more
1300 hosts than the hosts_max_try setting in the smtp transport when it had
1301 the 4xx errors. Those hosts that it had tried were not added to the list
1302 of hosts for which the message was waiting, so if all were tried, there
1303 was no problem. Two fixes have been applied:
1305 (i) If there are any address or message errors in an SMTP delivery, none
1306 of the hosts (tried or untried) are now added to the list of hosts
1307 for which the message is waiting, so the message should not be a
1308 candidate for sending over the same connection that was used for a
1309 successful delivery of some other message. This seems entirely
1310 reasonable: after all the message is NOT "waiting for some host".
1311 This is so "obvious" that I'm not sure why it wasn't done
1312 previously. Hope I haven't missed anything, but it can't do any
1313 harm, as the worst effect is to miss an optimization.
1315 (ii) If, despite (i), such a delivery is accidentally attempted, the
1316 routing retry time is respected, so at least it doesn't keep
1317 hammering the server.
1319 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1320 in ${readsocket because some servers need this prod.
1322 PH/32 Added some extra debug output when updating a wait-xxx database.
1324 PH/33 The hint "could be header name not terminated by colon", which has been
1325 given for certain expansion errors for a long time, was not being given
1326 for the ${if def:h_colon_omitted{... case.
1328 PH/34 The spec says: "With one important exception, whenever a domain list is
1329 being scanned, $domain contains the subject domain." There was at least
1330 one case where this was not true.
1332 PH/35 The error "getsockname() failed: connection reset by peer" was being
1333 written to the panic log as well as the main log, but it isn't really
1334 panic-worthy as it just means the connection died rather early on. I have
1335 removed the panic log writing for the ECONNRESET error when getsockname()
1338 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1339 runs only) independently of the message's sender address. This meant
1340 that, if the 4xx error was in fact related to the sender, a different
1341 message to the same recipient with a different sender could confuse
1342 things. In particualar, this can happen when sending to a greylisting
1343 server, but other circumstances could also provoke similar problems.
1344 I have changed the default so that the retry time for these errors is now
1345 based a combination of the sender and recipient addresses. This change
1346 can be overridden by setting address_retry_include_sender=false in the
1349 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1350 remote server are returned as part of bounce messages. This was not
1351 happening for LMTP over a pipe (the lmtp transport), but now it is the
1352 same for both kinds of LMTP.
1354 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1355 in header lines that were in fact CNAMEs. This is no longer the case.
1357 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1358 and queue runs started by the daemon processed all messages. This has
1359 been fixed so that -R and -S can now usefully be given with -q<time>.
1361 PH/40 Import PCRE release 6.7 (fixes some bugs).
1363 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1365 PH/42 Give an error if -q is specified more than once.
1367 PH/43 Renamed the variables $interface_address and $interface_port as
1368 $received_ip_address and $received_port, to make it clear that these
1369 values apply to message reception, and not to the outgoing interface when
1370 a message is delivered. (The old names remain recognized, of course.)
1372 PH/44 There was no timeout on the connect() call when using a Unix domain
1373 socket in the ${readsocket expansion. There now is.
1375 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1376 be meaningful with "accept".
1378 SC/01 Eximstats V1.43
1379 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1381 SC/02 Eximstats V1.44
1382 Use a glob alias rather than an array ref in the generated
1383 parser. This improves both readability and performance.
1385 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1386 Collect SpamAssassin and rejection statistics.
1387 Don't display local sender or destination tables unless
1388 there is data to show.
1389 Added average volumes into the top table text output.
1391 SC/04 Eximstats V1.46
1392 Collect data on the number of addresses (recipients)
1393 as well as the number of messages.
1395 SC/05 Eximstats V1.47
1396 Added 'Message too big' to the list of mail rejection
1397 reasons (thanks to Marco Gaiarin).
1399 SC/06 Eximstats V1.48
1400 Mainlog lines which have GMT offsets and are too short to
1401 have a flag are now skipped.
1403 SC/07 Eximstats V1.49 (Alain Williams)
1404 Added the -emptyok flag.
1406 SC/08 Eximstats V1.50
1407 Fixes for obtaining the IP address from reject messages.
1409 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1410 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1411 whitesspace changes from 4.64-PH/27
1413 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1416 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1417 are found, allow negative numbers in numeric criteria)
1419 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1421 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1422 to match changes made in 4.64-PH/43
1424 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1426 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1427 "domain" is actually an IP address, and force a failure. This locks out
1428 those revolvers/nameservers that support "A-for-A" lookups, in
1429 contravention of the specifications.
1431 PH/48 When a host name was looked up from an IP address, and the subsequent
1432 forward lookup of the name timed out, the host name was left in
1433 $sender_host_name, contrary to the specification.
1435 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1436 restricted to single-key lookups, Exim was not diagnosing an error if
1437 * or *@ was used with a query-style lookup.
1439 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1441 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1442 long ago, but noone interested enough thought of it. Let's just say that
1443 the "1.1" means that there are some new functions that weren't there at
1444 some point in the past.
1446 PH/51 Error processing for expansion failure of helo_data from an smtp
1447 transport during callout processing was broken.
1449 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1450 tested/used via the -bh/-bhc/-bs options.
1452 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1453 bug, fixed in subsequent PCRE releases).
1455 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1456 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1458 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1464 SC/01 Use a glob alias rather than an array ref in eximstats generated
1465 parser. This improves both readability and performance.
1467 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1468 Don't display local sender or destination tables in eximstats unless
1469 there is data to show.
1470 Added average volumes into the eximstats top table text output.
1472 SC/03 Collect data on the number of addresses (recipients) as well
1473 as the number of messages in eximstats.
1475 TF/01 Correct an error in the documentation for the redirect router. Exim
1476 does (usually) call initgroups() when daemonizing.
1478 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1479 with consistent privilege compared to when running as a daemon.
1481 TF/03 Note in the spec that $authenticated_id is not set for local
1482 submissions from trusted users.
1484 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1485 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1487 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1488 by adding some example configuration directives to the default
1489 configuration file. A little bit of work is required to uncomment the
1490 directives and define how usernames and passwords are checked, but
1491 there is now a framework to start from.
1493 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1494 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1495 without this. I don't know how relevant this is to other LDAP libraries.
1497 PH/02 Add the verb name to the "unknown ACL verb" error.
1499 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1501 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1503 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1504 directory not expanded when it should be if an expanded home directory
1505 was set for the address (which is overridden by the transport).
1507 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1510 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1511 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1512 because it is too late at that time, and has no effect.
1514 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1515 security issue with \' (bugzilla #107). I could not use the
1516 PQescapeStringConn() function, because it needs a PGconn value as one of
1519 PH/08 When testing addresses using -bt, indicate those final addresses that
1520 are duplicates that would not cause an additional delivery. At least one
1521 person was confused, thinking that -bt output corresponded to deliveries.
1522 (Suppressing duplicates isn't a good idea as you lose the information
1523 about possibly different redirections that led to the duplicates.)
1525 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1526 systems where poll() doesn't work, in particular OS X.
1528 PH/10 Added more information to debugging output for retry time not reached.
1530 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1531 operations in malware.c.
1533 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1536 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1537 syslog_duplication set false, log lines that would normally be written
1538 both the the main log and to the reject log were not written to syslog at
1541 PH/14 In the default configuration, change the use of "message" in ACL warn
1542 statements to "add_header".
1544 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1545 not followed by a command (e.g. "seen endif").
1547 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1548 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1551 PH/17 Added extra conditions to the default value of delay_warning_condition
1555 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1556 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1557 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1560 The Auto-Submitted: and various List- headers are standardised, whereas I
1561 don't think Precedence: ever was.
1563 PH/18 Refactored debugging code in route_finduser() to show more information,
1564 in particular, the error code if getpwnam() issues one.
1566 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1567 This is apparently needed in addition to the PH/07 change above to avoid
1568 any possible encoding problems.
1570 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1571 but not after initializing Perl.
1573 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1574 output them only if debugging. By default they are written stderr,
1575 apparently, which is not desirable.
1577 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1580 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1583 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1585 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1586 authenticated or an ident call has been made. Suppress the default
1587 values for $authenticated_id and $authenticated_sender (but permit -oMai
1588 and -oMas) when testing with -bh.
1590 PH/24 Re-jigged the order of the tests in the default configuration so that the
1591 tests for valid domains and recipients precede the DNS black list and CSA
1592 tests, on the grounds that those ones are more expensive.
1594 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1595 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1596 This bug exists in every version of Exim that I still have, right back to
1599 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1600 However, an attempt to turn on case-sensitivity in a regex key by
1601 including (?-i) didn't work because the subject string was already
1602 lowercased, and the effects were non-intuitive. It turns out that a
1603 one-line patch can be used to allow (?-i) to work as expected.
1609 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1610 other effects) broke the use of negated acl sub-conditions.
1612 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1615 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1616 "Deny" causes Exim to reject the incoming connection with a 554 error.
1617 Unfortunately, if there is a major crisis, such as a disk failure,
1618 tcp-wrappers gives "deny", whereas what one would like would be some
1619 kind of temporary error. A kludge has been added to help with this.
1620 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1621 554 error is used if errno is still zero or contains ENOENT (which occurs
1622 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1625 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1627 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1628 errors. Otherwise a message that provokes a temporary error (when other
1629 messages do not) can cause a whole host to time out.
1631 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1632 addresses were routed directly to files or pipes from a redirect router.
1633 File deliveries just didn't batch; pipe deliveries might have suffered
1636 PH/06 A failure to get a lock for a hints database would erroneously always say
1637 "Failed to get write lock", even when it was really a read lock.
1639 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1640 of 0600. This has been changed to use the value of the lockfile_mode
1641 option (which defaults to 0600).
1643 PH/08 Applied small patch from the Sieve maintainer.
1645 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1646 folder from quota calculations, a direct delivery into this folder messed
1647 up the contents of the maildirsize file. This was because the regex was
1648 used only to exclude .Trash (or whatever) when the size of the mailbox
1649 was calculated. There was no check that a delivery was happening into an
1650 excluded directory. This bug has been fixed by ignoring all quota
1651 processing for deliveries into excluded directories.
1653 PH/10 Added the maildirfolder_create_regex option to appendfile.
1659 PH/01 The code for finding all the local interface addresses on a FreeBSD
1660 system running IPv6 was broken. This may well have applied to all BSD
1661 systems, as well as to others that have similar system calls. The broken
1662 code found IPv4 interfaces correctly, but gave incorrect values for the
1663 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1664 that it would not match correctly against @[] and not recognize the IPv6
1667 PH/02 The ipliteral router was not recognizing addresses of the form user@
1668 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1670 PH/03 Added disable_ipv6.
1672 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1673 decoded form, because it is most often used to construct To: headers
1674 lines in autoreplies, and the decoded form may well be syntactically
1675 invalid. However, $reply_address has leading white space removed, and all
1676 newlines turned into spaces so that the autoreply transport does not
1679 PH/05 If group was specified without a user on a router, and no group or user
1680 was specified on a transport, the group from the router was ignored.
1682 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1683 for visible compile-time settings that can be used to change these
1684 numbers, for those that want even more. Backwards compatibility with old
1685 spool files has been maintained. However, going back to a previous Exim
1686 release will lost any variables that are in spool files.
1688 PH/07 Two small changes when running in the test harness: increase delay when
1689 passing a TCP/IP connection to a new process, in case the original
1690 process has to generate a bounce, and remove special handling of
1691 127.0.0.2 (sic), which is no longer necessary.
1693 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1694 be the same on different OS.
1696 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1699 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1700 whether --show-vars was specified or not
1702 JJ/02 exipick: Added support for new ACL variable spool format introduced
1705 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1706 syntactically invalid From: or Reply-to: line, and a filter used this to
1707 generate an autoreply, and therefore failed to obtain an address for the
1708 autoreply, Exim could try to deliver to a non-existent relative file
1709 name, causing unrelated and misleading errors. What now happens is that
1710 it logs this as a hard delivery error, but does not attempt to create a
1713 PH/11 The exinext utility has a -C option for testing purposes, but although
1714 the given file was scanned by exinext itself; it wasn't being passed on
1715 when Exim was called.
1717 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1718 an end-of-file indication when reading a command response.
1720 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1721 compiled. In many other places in Exim, IPv6 addresses are always
1722 recognized, so I have changed this. It also means that IPv4 domain
1723 literals of the form [IPV4:n.n.n.n] are now always recognized.
1725 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1726 used if the router is not running as root, for example, when verifying at
1727 ACL time, or when using -bh. The debugging output from this situation was
1728 non-existent - all you got was a failure to exec. I have made two
1731 (a) Failures to set uid/gid, the current directory, or a process leader
1732 in a subprocess such as that created by queryprogram now generate
1733 suitable debugging ouput when -d is set.
1735 (b) The queryprogram router detects when it is not running as root,
1736 outputs suitable debugging information if -d is set, and then runs
1737 the subprocess without attempting to change uid/gid.
1739 PH/15 Minor change to Makefile for building test_host (undocumented testing
1742 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1743 additional section of a DNS packet that returns MX or SRV records.
1744 Instead, it always explicitly searches for A/AAAA records. This avoids
1745 major problems that occur when a DNS server includes only records of one
1746 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1747 fixed another bug: if SRV records were looked up and the corresponding
1748 address records were *not* found in the additional section, the port
1749 values from the SRV records were lost.
1751 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1752 using the correct key (the original address) when searching the retry
1753 rules in order to find which one to use for generating the retry hint.
1755 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1756 adding the default one. Similarly, if it contains a Reply-To: header, the
1757 errors_reply_to option, if set, is not used.
1759 PH/19 When calculating a retry time, Exim used to measure the "time since
1760 failure" by looking at the "first failed" field in the retry record. Now
1761 it does not use this if it is later than than the arrival time of the
1762 message. Instead it uses the arrival time. This makes for better
1763 behaviour in cases where some deliveries succeed, thus re-setting the
1764 "first failed" field. An example is a quota failure for a huge message
1765 when small messages continue to be delivered. Without this change, the
1766 "time since failure" will always be short, possible causing more frequent
1767 delivery attempts for the huge message than are intended.
1768 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1770 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1771 $1, $2, $3) because the numerical variables can be reset during some
1772 expansion items (e.g. "match"), thereby losing the authentication data.
1774 PH/21 Make -bV show the size of off_t variables so that the test suite can
1775 decide whether to run tests for quotas > 2G.
1777 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1778 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1779 filecount value is greater than 2G or if a quota value is greater than 2G
1780 on a system where the size of off_t is not greater than 4, a panic error
1783 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1784 never match. The debug and -bh output now contains an explicit error
1785 message indicating a malformed IPv4 address or mask.
1787 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1788 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1789 PH/23 above applies.
1791 PH/25 Do not write to syslog when running in the test harness. The only
1792 occasion when this arises is a failure to open the main or panic logs
1793 (for which there is an explicit test).
1795 PH/26 Added the /no_tell option to "control=freeze".
1797 PH/27 If a host name lookup failed very early in a connection, for example, if
1798 the IP address matched host_lookup and the reverse lookup yielded a name
1799 that did not have a forward lookup, an error message of the form "no IP
1800 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1801 could be logged. Now it outputs the IP address instead of "NULL".
1803 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1804 allows the sender and the authenticated sender to be set when
1805 submitting a message from within Exim. Since child_open_exim() is
1806 documented for local_scan(), the new function should be too.
1808 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1809 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1810 results in an empty string is now treated as unset.
1812 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1814 PH/31 Added sender_verify_fail logging option.
1816 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1817 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1818 tidied the source and removed it altogether.
1820 PH/33 When a queue run was abandoned because the load average was too high, a
1821 log line was always written; now it is written only if the queue_run log
1822 selector is set. In addition, the log line for abandonment now contains
1823 information about the queue run such as the pid. This is always present
1824 in "start" and "stop" lines but was omitted from the "abandon" line.
1826 PH/34 Omit spaces between a header name and the colon in the error message that
1827 is given when verify = headers_syntax fails (if there are lots of them,
1828 the message gets confusing).
1830 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1831 names, as there are now some PTR records that contain slashes. This check
1832 is only to protect against broken name servers that fall over on strange
1833 characters, so the fact that it applies to all lookups doesn't matter.
1835 PH/36 Now that the new test suite is complete, we can remove some of the
1836 special code in Exim that was needed for the old test suite. For example,
1837 sorting DNS records because real resolvers return them in an arbitrary
1838 order. The new test suite's fake resolver always returns records in the
1841 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1842 bounces) except when queue_only is set, to avoid logging races between
1843 the different processes.
1845 PH/38 Panic-die if .include specifies a non-absolute path.
1847 PH/39 A tweak to the "H" retry rule from its user.
1849 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1850 a label. They prevented compilation on older perls.
1852 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1853 a warning to be raised on newish perls.
1855 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1856 on queue. Changes to match documented behaviour of showing count of
1857 messages matching specified criteria.
1859 PH/40 Changed the default ident timeout from 30s to 5s.
1861 PH/41 Added support for the use of login_cap features, on those BSD systems
1862 that have them, for controlling the resources used by pipe deliveries.
1864 PH/42 The content-scanning code uses fopen() to create files in which to put
1865 message data. Previously it was not paying any attention to the mode of
1866 the files. Exim runs with umask(0) because the rest of the code creates
1867 files with open(), and sets the required mode explicitly. Thus, these
1868 files were ending up world-writeable. This was not a big issue, because,
1869 being within the spool directory, they were not world-accessible. I have
1870 created a function called modefopen, which takes an additional mode
1871 argument. It sets umask(777), creates the file, chmods it to the required
1872 mode, then resets the umask. All the relevant calls to fopen() in the
1873 content scanning code have been changed to use this function.
1875 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1876 to 24 hours. This avoids potential overflow problems when processing G
1877 and H retry rules. I suspect nobody ever tinkers with this value.
1879 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1881 PH/45 When the plaintext authenticator is running as a client, the server's
1882 challenges are checked to ensure they are valid base64 strings. By
1883 default, the authentication attempt is cancelled if an invalid string is
1884 received. Setting client_ignore_invalid_base64 true ignores these errors.
1885 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1886 they are received. Thus, the responses can be made to depend on the
1887 challenges. If an invalid string is ignored, an empty string is placed in
1890 PH/46 Messages that are created by the autoreply transport now contains a
1891 References: header, in accordance with RFCs 2822 and 3834.
1893 PH/47 Added authenticated_sender_force to the smtp transport.
1895 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1897 PH/49 Installed latest patch from the Sieve maintainer.
1899 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1900 was also set, the appendfile transport was unnecessarily scanning a
1901 directory of message files (e.g. for maildir delivery) to find the count
1902 of files (along with the size), even though it did not need this
1903 information. It now does the scan only if it needs to find either the
1904 size of the count of files.
1906 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1908 PH/52 Two bugs concerned with error handling when the smtp transport is
1911 (i) Exim was not creating retry information for temporary errors given
1912 for individual recipients after the DATA command when the smtp transport
1913 was used in LMTP mode. This meant that they could be retried too
1914 frequently, and not timed out correctly.
1916 (ii) Exim was setting the flag that allows error details to be returned
1917 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1918 recipients that were returned after the DATA command.
1920 PH/53 This is related to PH/52, but is more general: for any failing address,
1921 when detailed error information was permitted to be returned to the
1922 sender, but the error was temporary, then after the final timeout, only
1923 "retry timeout exceeded" was returned. Now it returns the full error as
1924 well as "retry timeout exceeded".
1926 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1927 do this, and (what is worse) MTAs that accept it.
1929 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1930 will now be deprecated.
1932 PH/56 New os.c-cygwin from the Cygwin maintainer.
1934 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1935 formats (previously only available in exim formats via -bpr, -bpru,
1936 and -bpra. Now also available in native and exiqgrep formats)
1938 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1939 with very large, slow to parse queues
1941 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1943 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1945 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1946 responses to authentication challenges, though it was showing the
1947 challenges; (ii) I've removed the CR characters from the debug output for
1950 PH/58 Allow for the insertion of a newline as well as a space when a string
1951 is turned into more than one encoded-word during RFC 2047 encoding. The
1952 Sieve code now uses this.
1954 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1955 data_4xx, lost_connection, tls_required.
1957 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1958 message was being sent as an SMTP response.
1960 PH/61 Add -l and -k options to exicyclog.
1962 PH/62 When verifying, if an address was redirected to one new address, so that
1963 verification continued, and the new address failed or deferred after
1964 having set something in $address_data, the value of $address_data was not
1965 passed back to the ACL. This was different to the case when no
1966 redirection occurred. The value is now passed back in both cases.
1968 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1969 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1970 use login_cap.h, so on its own it isn't the distinguishing feature. The
1971 new name refers directly to the setclassresources() function.
1973 PH/65 Added configuration files for NetBSD3.
1975 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1977 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1978 is preferred over IPv4.
1980 PH/68 The bounce_return_message and bounce_return_body options were not being
1981 honoured for bounces generated during the reception of non-SMTP messages.
1982 In particular, this applied to messages rejected by the ACL. This bug has
1983 been fixed. However, if bounce_return_message is true and bounce_return_
1984 body is false, the headers that are returned for a non-SMTP message
1985 include only those that have been read before the error was detected.
1986 (In the case of an ACL rejection, they have all been read.)
1988 PH/69 The HTML version of the specification is now built in a directory called
1989 spec_html instead of spec.html, because the latter looks like a path with
1990 a MIME-type, and this confuses some software.
1992 PH/70 Catch two compiler warnings in sieve.c.
1994 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1995 function verify_get_ident() calls ip_connect() to connect a socket, but
1996 if the "connect()" function timed out, ip_connect() used to close the
1997 socket. However, verify_get_ident() also closes the socket later, and in
1998 between Exim writes to the log, which may get opened at this point. When
1999 the socket was closed in ip_connect(), the log could get the same file
2000 descriptor number as the socket. This naturally causes chaos. The fix is
2001 not to close the socket in ip_connect(); the socket should be closed by
2002 the function that creates it. There was only one place in the code where
2003 this was missing, in the iplookup router, which I don't think anybody now
2004 uses, but I've fixed it anyway.
2006 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2007 well as to direct DNS lookups. Otherwise the handling of names in host
2008 lists is inconsistent and therefore confusing.
2014 PH/01 Two changes to the default runtime configuration:
2016 (1) Move the checks for relay_from_hosts and authenticated clients from
2017 after to before the (commented out) DNS black list checks.
2019 (2) Add control=submission to the relay_from_hosts and authenticated
2020 clients checks, on the grounds that messages accepted by these
2021 statements are most likely to be submissions.
2023 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2025 (1) Generate an error if the third argument for the ${prvs expansion is
2028 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2031 (3) Reset the variables that are obtained from the first argument of
2032 ${prvscheck and used in the second argument before leaving the code,
2033 because their memory is reclaimed, so using them afterwards may do
2036 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2037 one (it's much easier than Tom thought :-).
2039 (5) Because of (4), we can now allow for the use of $prvscheck_result
2040 inside the third argument.
2042 PH/03 For some reason, the default setting of PATH when running a command from
2043 a pipe transport was just "/usr/bin". I have changed it to
2046 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2047 anything to be listed in the output from -bV.
2049 PH/05 When a filter generated an autoreply, the entire To: header line was
2050 quoted in the delivery log line, like this:
2052 => >A.N.Other <ano@some.domain> <original@ddress> ...
2054 This has been changed so that it extracts the operative address. There
2055 may be more than one such address. If so, they are comma-separated, like
2058 => >ano@some.domain,ona@other.domain <original@ddress> ...
2060 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2061 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2062 not being looked up in the rDNS to get a host name, Exim was showing the
2063 IP address twice in Received: lines, even though the IP addresses were
2064 identical. For example:
2066 Received: from [1.2.3.4] (helo=[1.2.3.4])
2068 However, if the real host name was known, it was omitting the HELO data
2069 if it matched the actual IP address. This has been tidied up so that it
2070 doesn't show the same IP address twice.
2072 PH/07 When both +timestamp and +memory debugging was on, the value given by
2073 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2074 called by the debug printing, thereby overwriting the timestamp buffer.
2075 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2077 PH/08 When the original message was included in an autoreply transport, it
2078 always said "this is a copy of the message, including all the headers",
2079 even if body_only or headers_only was set. It now gives an appropriate
2082 PH/09 Applied a patch from the Sieve maintainer which:
2084 o fixes some comments
2085 o adds the (disabled) notify extension core
2086 o adds some debug output for the result of if/elsif tests
2087 o points to the current vacation draft in the documentation
2088 and documents the missing references header update
2092 o fixes a bug in processing the envelope test (when testing
2093 multiple envelope elements, the last element determinted the
2096 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2097 Electronic Mail") by including:
2099 Auto-submitted: auto-generated
2101 in the messages that it generates (bounce messages and others, such as
2102 warnings). In the case of bounce messages for non-SMTP mesages, there was
2103 also a typo: it was using "Auto_submitted" (underscore instead of
2104 hyphen). Since every message generated by Exim is necessarily in response
2105 to another message, thes have all been changed to:
2107 Auto-Submitted: auto-replied
2109 in accordance with these statements in the RFC:
2111 The auto-replied keyword:
2113 - SHOULD be used on messages sent in direct response to another
2114 message by an automatic process,
2116 - MUST NOT be used on manually-generated messages,
2118 - MAY be used on Delivery Status Notifications (DSNs) and Message
2119 Disposition Notifications (MDNs),
2121 - MUST NOT be used on messages generated by automatic or periodic
2122 processes, except for messages which are automatic responses to
2125 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2126 to the default Received: header definition.
2128 PH/12 Added log selector acl_warn_skipped (default on).
2130 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2131 variables because (a) they are in the wrong storage pool and (b) even if
2132 they were copied, it wouldn't work properly because of the caching.
2134 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2135 checking when decoding. Apparently there are clients that generate
2136 overlong encoded strings. Why am I not surprised?
2138 PH/15 If the first argument of "${if match_address" was not empty, but did not
2139 contain an "@" character, Exim crashed. Now it writes a panic log message
2140 and treats the condition as false.
2142 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2144 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2145 extension "envelope-auth". The code is finished and in agreement with
2146 other implementations, but there is no documentation so far and in fact,
2147 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2148 not changing the active code.
2150 Print executed "if" and "elsif" statements when debugging is used. This
2151 helps a great deal to understand what a filter does.
2153 Document more things not specified clearly in RFC3028. I had all this
2154 sorted out, when out of a sudden new issues came to my mind. Oops."
2156 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2159 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2160 canonical form (as documented). However, after a host name lookup from
2161 the IP address, check_host() was doing a simple string comparison with
2162 addresses acquired from the DNS when checking that the found name did
2163 have the original IP as one of its addresses. Since any found IPv6
2164 addresses are likely to be in abbreviated form, the comparison could
2165 fail. Luckily, there already exists a function for doing the comparison
2166 by converting both addresses to binary, so now that is used instead of
2167 the text comparison.
2169 PH/20 There was another similar case to PH/19, when a complete host name was
2170 given in a host list; looking up its IP address could give an abbreviated
2171 form, whereas the current host's name might or might not be abbreviated.
2172 The same fix has been applied.
2178 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2179 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2182 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2183 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2185 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2187 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2188 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2189 Also, for "Auto-Submitted", treat anything other than "no" as
2190 non-personal, in accordance with RFC 3834. (Previously it treated
2191 anything starting "auto-" as non-personal.)
2193 TF/01 The control=submission/name=... option had a problem with syntax
2194 errors if the name included a slash character. The /name= option
2195 now slurps the rest of the string, so it can include any characters
2196 but it must come last in the list of options (after /sender_retain
2199 PH/05 Some modifications to the interface to the fake nameserver for the new
2207 TK/01 Added the "success_on_redirect" address verification option. See
2208 NewStuff for rationale and an example.
2210 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2212 PH/02 Patch to exigrep to allow it to work on syslog lines.
2214 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2215 fread() to skip over the body file's header line, because in Cygwin the
2216 header line is locked and is inaccessible.
2218 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2219 co-exist for some time) to make it clear that it is the Exim ID that is
2220 referenced, not the Message-ID: header line.
2222 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2223 string_format() function, because snprintf() does not exist on all
2226 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2227 ${stat: expansion item.
2229 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2230 protocol synchronization error", to keep the pedants happy.
2232 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2233 well as for IRIX systems, when gcc is being used. See the host.c source
2236 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2238 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2241 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2242 transport and to the smtp transport in LMTP mode.
2244 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2246 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2247 run a filter in a subprocess. This could lead to confusion in subsequent
2248 lookups in the parent process. There should also be a search_tidyup() at
2249 the end of the subprocess.
2251 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2252 only if the host matched helo_try_verify_hosts, which caused the
2253 verification to occur when the EHLO/HELO command was issued. The ACL just
2254 tested the remembered result. Now, if a previous verification attempt has
2255 not happened, "verify = helo" does it there and then.
2257 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2259 TK/03 Fix log output including CR from clamd.
2261 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2262 exist provoked a memory error which could cause a segfault.
2264 PH/15 Installed PCRE 6.2
2266 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2268 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2269 of the problem. Specifically, suggested +O2 rather than +O1 for the
2272 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2274 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2275 record, the port was not correctly logged when the outgoing_port log
2276 selector was set (it logged the transort's default port).
2278 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2279 fallback_hosts, and "hosts" in the smtp transport.
2281 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2282 host errors such as "Connection refused".
2284 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2285 authentication with radiusclient 0.4.9:
2287 - Error returned from rc_read_config was caught wrongly
2288 - Username/password not passed on to radius server due to wrong length.
2290 The presumption is that some radiusclient API changes for 4.51/PH/17
2291 were not taken care of correctly. The code is still untested by me (my
2292 Linux distribution still has 0.3.2 of radiusclient), but it was
2293 contributed by a Radius user.
2295 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2296 expanding the "port" option of the smtp transport.
2298 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2299 while reading a MIME header. Thanks to Tom Hughes for a patch.
2301 PH/24 Include config.h inside local_scan.h so that configuration settings are
2304 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2305 that in an ACL for RCPT (for example), you can examine exactly what was
2308 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2309 commands, but it was not correctly comparing the address with the actual
2310 client host address. Thus, it would show the EHLO address in Received:
2311 header lines when this was not necessary.
2313 PH/27 Added the % operator to ${eval:}.
2315 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2316 it should be ignoring failures (because with -C, for example, it has lost
2317 privilege). It wasn't ignoring creation failures other than "already
2320 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2323 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2324 because an input line was too long, either on its own, or by virtue of
2325 too many continuations, the temporary file was not being removed, and the
2326 return code was incorrect.
2328 PH/31 Missing "BOOL" in function definition in filtertest.c.
2330 PH/32 Applied Sieve patches from the maintainer.
2332 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2334 PH/33 Added "verify = not_blind".
2336 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2337 Local/Makefile (with some defaults set). These are used in built scripts
2338 such as exicyclog, but they have never been used in the exim_install
2339 script (though there are many overriding facilities there). I have
2340 arranged that the exim_install script now takes note of these two
2343 PH/35 Installed configuration files for Dragonfly.
2345 PH/36 When a locally submitted message by a trusted user did not contain a
2346 From: header, and the sender address was obtained from -f or from an SMTP
2347 MAIL command, and the trusted user did not use -F to supply a sender
2348 name, $originator_name was incorrectly used when constructing a From:
2349 header. Furthermore, $originator_name was used for submission mode
2350 messages from external hosts without From: headers in a similar way,
2351 which is clearly wrong.
2353 PH/37 Added control=suppress_local_fixups.
2355 PH/38 When log_selector = +received_sender was set, and the addition of the
2356 sender made the log line's construction buffer exactly full, or one byte
2357 less than full, an overflow happened when the terminating "\n" was
2360 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2361 when the result of a list match is failure because a DNS lookup failed.
2363 PH/40 RM_COMMAND is now used in the building process.
2365 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2366 the "build-* directories that it finds.
2368 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2369 address in a domain literal was a prefix of an interface address.
2371 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2372 when verifying a sender address, unless rewrite_headers is false.
2374 PH/44 Wrote a long comment about why errors_to addresses are verified as
2375 recipients, not senders.
2377 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2378 the ratelimit ACL was added.
2380 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2382 PH/46 Added extra information about PostgreSQL errors to the error string.
2384 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2385 avoiding the need to install special zones in a real server. This is
2386 backwards compatible; if it can't find the fake resolver, it drops back.
2387 Thus, both old and new test suites can be run.
2389 TF/02 Added util/ratelimit.pl
2391 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2392 clock is set back in time.
2394 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2395 Candler <B.Candler@pobox.com>.
2397 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2398 OK for addresses that are the result of redirections.
2400 PH/48 A number of further additions for the benefit of the new test suite,
2401 including a fake gethostbyname() that interfaces to the fake DNS resolver
2404 TF/06 The fix for widen_domains has also been applied to qualify_single and
2405 search_parents which are the other dnslookup options that can cause
2408 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2411 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2413 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2414 service" if it encountered an MX record with an empty target hostname.
2415 The message is now "an MX or SRV record indicated no SMTP service".
2417 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2418 if the DNS of the sending site is misconfigured. This is quite a
2419 common situation. This change restores the behaviour of treating a
2420 helo verification defer as a failure.
2422 PH/51 If self=fail was set on a router, the bounce message did not include the
2423 actual error message.
2429 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2431 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2432 command itself ended in error, the underlying message about the transport
2433 filter timeout was being overwritten with the pipe command error. Now the
2434 underlying error message should be appended to the second error message.
2436 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2438 PH/02 Reduce the amount of output that "make" produces by default. Full output
2439 can still be requested.
2441 PH/03 The warning log line about a condition test deferring for a "warn" verb
2442 was being output only once per connection, rather than after each
2443 occurrence (because it was using the same function as for successful
2444 "warn" verbs). This seems wrong, so I have changed it.
2446 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2447 it should not have, which might have caused a crash in the right
2448 circumstances, but probably never did.
2450 PH/04 Installed a modified version of Tony Finch's patch to make submission
2451 mode fix the return path as well as the Sender: header line, and to
2452 add a /name= option so that you can make the user's friendly name appear
2455 TF/03 Added the control = fakedefer ACL modifier.
2457 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2458 Mark Lowes for thorough testing.
2460 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2462 TK/03 Merged latest SRS patch from Miles Wilton.
2464 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2465 with the definition in sysexits.h (which is #included earlier).
2466 Fortunately, Exim does not actually use EX_OK. The code used to try to
2467 preserve the sysexits.h value, by assumimg that macro definitions were
2468 scanned for macro replacements. I have been disabused of this notion,
2469 so now the code just undefines EX_OK before #including unistd.h.
2471 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2472 in the smtp transport. When a block could not be written in a single
2473 write() function, the timeout was being re-applied to each part-write.
2474 This seems wrong - if the receiver was accepting one byte at a time it
2475 would take for ever. The timeout is now adjusted when this happens. It
2476 doesn't have to be particularly precise.
2478 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2479 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2481 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2482 without a domain if the check to <postmaster@domain> fails.
2484 SC/01 Eximstats: added -xls and the ability to specify output files
2485 (patch written by Frank Heydlauf).
2487 SC/02 Eximstats: use FileHandles for outputing results.
2489 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2491 SC/04 Eximstats: fixed display of large numbers with -nvr option
2493 SC/05 Eximstats: fixed merging of reports with empty tables.
2495 SC/06 Eximstats: added the -include_original_destination flag
2497 SC/07 Eximstats: removed tabs and trailing whitespace.
2499 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2501 TK/06 MBOX spool code: Add real "From " MBOX separator line
2502 so the .eml file is really in mbox format (even though
2503 most programs do not really care). Patch from Alex Miller.
2505 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2506 The latter is generated from $received_to and is only set if the
2507 message has one envelope recipient. SA can use these headers,
2508 obviously out-of-the-box. Patch from Alex Miller.
2510 PH/08 The ${def test on a variable was returning false if the variable's
2511 value was "0", contrary to what the specification has always said!
2512 The result should be true unless the variable is empty.
2514 PH/09 The syntax error of a character other than { following "${if
2515 def:variable_name" (after optional whitespace) was not being diagnosed.
2516 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2517 accidental colon was present, for example, could give incorrect results.
2519 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2520 result is used (not including appendfile, where other changes are about
2523 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2524 This involved changing a lot of size variables from int to off_t. It
2525 should work with maildirs and everything.
2527 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2528 spamd dying while we are connected to it.
2530 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2533 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2534 function for BDB 4.3.
2536 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2538 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2539 the sources for PCRE have been re-arranged and more files are now
2542 PH/15 The code I had for printing potentially long long variables in PH/11
2543 above was not the best (it lost precision). The length of off_t variables
2544 is now inspected at build time, and an appropriate printing format (%ld
2545 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2546 to be "long long int" or "long int". This is needed for the internal
2547 formatting function string_vformat().
2549 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2550 the configuration file to be ":syslog", then the script "guesses" where
2551 the logs files are, rather than using the compiled in default. In our
2552 case the guess is not the same as the compiled default, so the script
2553 suddenly stopped working when I started to use syslog. The patch checks
2554 to see if log_file_path is "". If so, it attempts to read it from exim
2555 with no configuration file to get the compiled in version, before it
2556 falls back to the previous guessing code."
2558 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2559 implementing BATV in an Exim configuration. See NewStuff for the gory
2562 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2563 Makefile that are specific to HP-UX.
2565 PH/18 If the "use_postmaster" option was set for a recipient callout together
2566 with the "random" option, the postmaster address was used as the MAIL
2567 FROM address for the random test, but not for the subsequent recipient
2568 test. It is now used for both.
2570 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2571 patch removes a few documentation additions to RFC 3028, because the
2572 latest draft now contains them. It adds the new en;ascii-case comparator
2573 and a new error check for 8bit text in MIME parts. Comparator and
2574 require names are now matched exactly. I enabled the subaddress
2575 extension, but it is not well tested yet (read: it works for me)."
2577 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2578 rework some of the code of TK/09 above to avoid the hardwired use of
2579 "%lld" and "long long". Replaced the call to snprintf() with a call to
2582 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2583 records point to non-existent hosts", "retry timeout exceeded", and
2584 "retry time not reached for any host after a long failure period".
2586 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2587 experimental DomainKeys support:
2589 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2590 (2) On an error such as an illegally used "control", the wrong name for
2591 the control was given.
2593 These problems did NOT occur unless DomainKeys support was compiled.
2595 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2597 PH/24 Added ${if match_ip condition.
2599 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2600 databases so that it will be absolutely obvious if a crash occurs in the
2601 DB library. This is a regular occurrence (often caused by mis-matched
2604 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2605 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2606 picked up on a user's system that detects such things. There doesn't seem
2607 to be a gcc warning option for this - only an attribute that has to be
2608 put on the function's prototype. It seems that in Fedora Core 4 they have
2609 set this on a number of new functions. No doubt there will be more in due
2612 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2613 specify a transport. However, if an address that was verified by such a
2614 router was the subject of a callout, Exim crashed because it tried to
2615 read the rcpt_include_affixes from the non-existent transport. Now it
2616 just assumes that the setting of that option is false. This bug was
2617 introduced by 4.51/PH/31.
2619 PH/28 Changed -d+all to exclude +memory, because that information is very
2620 rarely of interest, but it makes the output a lot bigger. People tend to
2621 do -d+all out of habit.
2623 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2624 code in os-type was giving problems when libc.so lives in lib64, like on
2627 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2628 aren't the modern standard, and it seems that some systems' include files
2629 don't always have them. Exim was already checking for some of the newer
2630 ones like T_AAAA, and defining it itself. I've added checks for all the
2631 record types that Exim uses.
2633 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2634 not automatically generating a new one, as it is supposed to. This
2635 prevented TLS from working. If the file did exist, but contained invalid
2636 data, a new version was generated, as expected. It was only the case of a
2637 non-existent file that was broken.
2639 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2640 with a change in libdomainkeys > 0.64.
2642 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2643 from DNS. If the selector record carries the flag, it now has
2644 precedence over the domain-wide flag.
2646 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2648 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2649 the use of a port name that isn't defined in /etc/services) occurred, the
2650 message was deferred as in a normal delivery, and thus remained on the
2651 spool, instead of being failed because of the mua_wrapper setting. This
2652 is now fixed, and I tidied up some of the mua_wrapper messages at the
2655 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2656 the messages in a hash of arrays rather than using individual hashes.
2657 This is a bit cleaner and results in dramatic memory savings, albeit
2658 at a slight CPU cost.
2660 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2661 as requested by Marc Sherman.
2663 SC/10 Eximstats: added histograms for user specified patterns as requested
2666 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2668 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2669 fopen() in the content-scanning modules that did not already have it.
2675 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2676 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2678 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2680 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2682 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2683 to test Sieve filters that use "vacation".
2685 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2686 that changes the way the GnuTLS parameters are stored in the cache file.
2687 The new format can be generated externally. For backward compatibility,
2688 if the data in the cache doesn't make sense, Exim assumes it has read an
2689 old-format file, and it generates new data and writes a new file. This
2690 means that you can't go back to an older release without removing the
2693 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2694 work if there are any delivery delays because "one_time" forces the
2695 parent to be marked "delivered", so its unseen clone is never tried
2696 again. For this reason, Exim now forbids the simultaneous setting of
2699 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2700 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2701 entry if you want to know the details. The fix, however, neglected to
2702 consider the case when local delivery batching is involved. The test for
2703 "previously delivered" was not happening when checking to see if an
2704 address could be batched with a previous (undelivered) one; under
2705 certain circumstances this could lead to multiple deliveries to the same
2708 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2709 in its include files, and this causes problems building Exim.
2711 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2712 header_syntax) but Exim was just ignoring anything given after a slash.
2713 In particular, this caused confusion with an attempt to use "verify =
2714 reverse_host_lookup/defer_ok". An error is now given when options are
2715 supplied for verify items that do not have them. (Maybe reverse_host_
2716 lookup should have a defer_ok option, but that's a different point.)
2718 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2719 defined by RFC 821) to 2048, because there were problems with some AUTH
2720 commands, and RFC 1869 says the size should be increased for extended
2721 SMTP commands that take arguments.
2723 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2726 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2727 "unknown" error; now it says that the functionality isn't in the binary.
2729 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2730 an address' error message when a string expansion fails (syntax or
2731 whatever). Otherwise the password may appear in the log. Following change
2732 PH/42 below, there is no longer a chance of it appearing in a bounce
2735 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2737 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2738 of its addresses was ever tried. (Bugzilla bug #2.)
2740 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2741 the result incorrectly in the debug output. (It correctly added a newline
2742 to what was transported.)
2744 TF/01 Added $received_time.
2746 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2747 commented out examples of how to interface to a virus scanner and to
2748 SpamAssassin. Also added commented examples of av_scanner and
2749 spamd_address settings.
2751 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2752 and controls are allowed in which ACLs. There were a couple of minor
2753 errors. Some of the entries in the conditions table (which is a table of
2754 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2755 negation of where the condition IS allowed.
2757 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2759 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2760 header file does not have a version number, so I've had to invent a new
2761 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2762 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2763 radiusclient), but it was contributed by a Radius user.
2765 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2766 files or directories, for OpenSSL.
2768 PH/19 When an Exim process that is running as root has to create an Exim log
2769 file, it does so in a subprocess that runs as exim:exim so as to get the
2770 ownership right at creation (otherwise, other Exim processes might see
2771 the file with the wrong ownership). There was no test for failure of this
2772 fork() call, which would lead to the process getting stuck as it waited
2773 for a non-existent subprocess. Forks do occasionally fail when resources
2774 run out. I reviewed all the other calls to fork(); they all seem to check
2777 PH/20 When checking for unexpected SMTP input at connect time (before writing
2778 the banner), Exim was not dealing correctly with a non-positive return
2779 from the read() function. If the client had disconnected by this time,
2780 the result was a log entry for a synchronization error with an empty
2781 string after "input=" when read() returned zero. If read() returned -1
2782 (an event I could not check), uninitialized data bytes were printed.
2783 There were reports of junk text (parts of files, etc) appearing after
2786 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2788 PH/22 Added support for macro redefinition, and (re)definition in between
2789 driver and ACL definitions.
2791 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2792 forgetting to use the resulting value; it was using the unexpanded value.
2794 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2795 hadn't been configured. The fix is from Juergen Kreileder, who
2796 understands it better than I do:
2798 "Here's what I see happening with three configured cyrus_sasl
2799 authenticators configured (plain, login, cram-md5):
2801 On startup auth_cyrus_sasl_init() gets called for each of these.
2802 This means three calls to sasl_listmech() without a specified mech_list.
2803 => SASL tests which mechs of all available mechs actually work
2804 => three warnings about OTP not working
2805 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2807 With the patch, sasl_listmech() also gets called three times. But now
2808 SASL's mech_list option is set to the server_mech specified in the the
2809 authenticator. Or in other words, the answer from sasl_listmech()
2810 gets limited to just the mech you're testing for (which is different
2812 => the return list contains just 'plain' or 'login', 'cram-md5' or
2813 nothing depending on the value of ob->server_mech.
2815 I've just tested the patch: Authentication still works fine,
2816 unavailable mechs specified in the exim configuration are still
2817 caught, and the auth.log warnings about OTP are gone."
2819 PH/25 When debugging is enabled, the contents of the command line are added
2820 to the debugging output, even when log_selector=+arguments is not
2823 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2824 answer is "GNU", and only if the return is "GNU/something" is the answer
2827 PH/27 $acl_verify_message is now set immediately after the failure of a
2828 verification in an ACL, and so is available in subsequent modifiers. In
2829 particular, the message can be preserved by coding like this:
2831 warn !verify = sender
2832 set acl_m0 = $acl_verify_message
2834 Previously, $acl_verify_message was set only while expanding "message"
2835 and "log_message" when a very denied access.
2837 PH/28 Modified OS/os.c-Linux with
2839 -#ifndef OS_LOAD_AVERAGE
2840 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2842 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2843 nomenclature these days.)
2845 PH/29 Installed patch from the Sieve maintainer that adds the options
2846 sieve_useraddress and sieve_subaddress to the redirect router.
2848 PH/30 In these circumstances:
2849 . Two addresses routed to the same list of hosts;
2850 . First host does not offer TLS;
2851 . First host accepts first address;
2852 . First host gives temporary error to second address;
2853 . Second host offers TLS and a TLS session is established;
2854 . Second host accepts second address.
2855 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2856 and peerdn, if requested) that were in fact used only for the second
2859 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2860 attention to any local part prefix or suffix that was matched by the
2861 router that accepted the address. It now behaves in the same way as it
2862 does for delivery: the affixes are removed from the local part unless
2863 rcpt_include_affixes is set on the transport.
2865 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2866 timeout during the DATA phase of an incoming message.
2868 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2869 applied a patch sanctioned by the Sieve maintainer.
2871 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2872 the uid or gid is negative. A case of a negative gid caused this to be
2873 noticed. The fix allows for either to be negative.
2875 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2876 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2879 PH/36 The change PH/12 above was broken. Fixed it.
2881 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2882 the grounds that routing the same address twice would always produce the
2883 same answer. This might have been true once, but it is certainly no
2884 longer true now. Routing a child address may depend on the previous
2885 routing that produced that child. Some complicated redirection strategies
2886 went wrong when messages had multiple recipients, and made Exim's
2887 behaviour dependent on the order in which the addresses were given.
2889 I have moved the duplicate checking until after the routing is complete.
2890 Exim scans the addresses that are assigned to local and remote
2891 transports, and removes any duplicates. This means that more work will be
2892 done, as duplicates will always all be routed, but duplicates are
2893 presumably rare, so I don't expect this is of any significance.
2895 For deliveries to pipes, files, and autoreplies, the duplicate checking
2896 still happens during the routing process, since they are not going to be
2899 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2900 It corrects a timeout issue with spamd. This is Ian's comment: "The
2901 background is that sometimes spamd either never reads data from a
2902 connection it has accepted, or it never writes response data. The exiscan
2903 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2904 blindly assumes that writes won't block so it may never time out."
2906 PH/39 Allow G after quota size as well as K and M.
2908 PH/40 The value set for $authenticated_id in an authenticator may not contain
2909 binary zeroes or newlines because the value is written to log lines and
2910 to spool files. There was no check on this. Now the value is run through
2911 the string_printing() function so that such characters are converted to
2912 printable escape sequences.
2914 PH/41 $message_linecount is a new variable that contains the total number of
2915 lines in the message. Compare $body_linecount, which is the count for the
2918 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2919 bounce and delay warning messages, except in certain special cases, which
2922 (a) An SMTP error message from a remote host;
2923 (b) A message specified in a :fail: redirection;
2924 (c) A message specified in a "fail" command in a system filter;
2925 (d) A message specified in a FAIL return from the queryprogram router;
2926 (e) A message specified by the cannot_route_message router option.
2928 In these cases only, Exim does include the error details in bounce and
2929 warning messages. There are also a few cases where bland messages such
2930 as "unrouteable address" or "local delivery error" are given.
2932 PH/43 $value is now also set for the "else" part of a ${run expansion.
2934 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2935 being worked on, but at least Exim now implements the latest version to
2938 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2939 process to complete was treated as a delivery failure, a timeout while
2940 writing the message to the pipe was logged, but erroneously treated as a
2941 successful delivery. Such timeouts include transport filter timeouts. For
2942 consistency with the overall process timeout, these timeouts are now
2943 treated as errors, giving rise to delivery failures by default. However,
2944 there is now a new Boolean option for the pipe transport called
2945 timeout_defer, which, if set TRUE, converts the failures into defers for
2946 both kinds of timeout. A transport filter timeout is now identified in
2949 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2950 systems where "make" and "gmake" are different, calling "gmake" at top
2951 level broke things. I've arranged for the value of $(MAKE) to be passed
2952 from the Makefile to this script so that it can call the same version of
2956 A note about Exim versions 4.44 and 4.50
2957 ----------------------------------------
2959 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2960 changes of various kinds. As a consequence, a big documentation update was
2961 needed. This delayed the release for rather longer than seemed good, especially
2962 in the light of a couple of (minor) security issues. Therefore, the changes
2963 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2964 release. So 4.44 and 4.50 are in effect two different branches that both start
2967 I have left the 4.50 change log unchanged; it contains all the changes since
2968 4.43. The change log for 4.44 is below; many of its items are identical to
2969 those for 4.50. This seems to be the most sensible way to preserve the
2970 historical information.
2976 1. Minor wording change to the doc/README.SIEVE file.
2978 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2979 computation of the current number of files was incorrect.
2981 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2982 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2985 4. Give more explanation in the error message when the command for a transport
2986 filter fails to execute.
2988 5. There are several places where Exim runs a non-Exim command in a
2989 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2990 was being done only for the command run by the queryprogram router. It is
2991 now done for all such subprocesses. The other cases are: ${run, transport
2992 filters, and the commands run by the lmtp and pipe transports.
2994 6. Added CONFIGURE_GROUP build-time option.
2996 7. Some older OS have a limit of 256 on the maximum number of file
2997 descriptors. Exim was using setrlimit() to set 1000 as a large value
2998 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2999 systems. I've change it so that if it can't get 1000, it tries for 256.
3001 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3002 was an oversight, and furthermore, ever since the addition of extra
3003 controls (e.g. 4.43/32), the checks on when to allow different forms of
3004 "control" were broken. There should now be diagnostics for all cases when a
3005 control that does not make sense is encountered.
3007 9. Added the /retain_sender option to "control=submission".
3009 10. $recipients is now available in the predata ACL (oversight).
3011 11. Tidy the search cache before the fork to do a delivery from a message
3012 received from the command line. Otherwise the child will trigger a lookup
3013 failure and thereby defer the delivery if it tries to use (for example) a
3014 cached ldap connection that the parent has called unbind on.
3016 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3017 of $address_data from the recipient verification was clobbered by the
3018 sender verification.
3020 13. The value of address_data from a sender verification is now available in
3021 $sender_address_data in subsequent conditions in the ACL statement.
3023 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3025 15. Added a new option "connect=<time>" to callout options, to set a different
3028 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3029 was its contents. (It was OK if the option was not defined at all.)
3031 17. A "Completed" log line is now written for messages that are removed from
3032 the spool by the -Mrm option.
3034 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3035 information about exactly what failed.
3037 19. Added -dd to debug only the daemon process.
3039 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3040 handles timeouts, both on the server side and network timeouts. Renamed the
3041 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3043 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3044 It is now set to "smtps".
3046 22. $host_address is now set to the target address during the checking of
3047 ignore_target_hosts.
3049 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3050 being passed; this would have caused $sender_host_name to have been used if
3051 matching the list had actually called for a host name (not very likely,
3052 since this list is usually IP addresses). A host name is now passed as
3055 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3056 code that specifies a non-restarting handler (typically sigaction() in
3057 modern systems) in an attempt to fix a rare and obscure crash bug.
3059 25. Narrowed the window for a race in the daemon that could cause it to ignore
3060 SIGCHLD signals. This is not a major problem, because they are used only to
3061 wake it up if nothing else does.
3063 26. A malformed maildirsize file could cause Exim to calculate negative values
3064 for the mailbox size or file count. Odd effects could occur as a result.
3065 The maildirsize information is now recalculated if the size or filecount
3068 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3069 support for a long time. Removed HAVE_SYS_VFS_H.
3071 28. Installed the latest version of exipick from John Jetmore.
3073 29. In an address list, if the pattern was not a regular expression, an empty
3074 subject address (from a bounce message) matched only if the pattern was an
3075 empty string. Non-empty patterns were not even tested. This was the wrong
3076 because it is perfectly reasonable to use an empty address as part of a
3077 database query. An empty address is now tested by patterns that are
3078 lookups. However, all the other forms of pattern expect the subject to
3079 contain a local part and a domain, and therefore, for them, an empty
3080 address still always fails if the pattern is not itself empty.
3082 30. Exim went into a mad DNS loop when attempting to do a callout where the
3083 host was specified on an smtp transport, and looking it up yielded more
3084 than one IP address.
3086 31. Re-factored the code for checking spool and log partition space into a
3087 function that finds that data and another that does the check. The former
3088 is then used to implement four new variables: $spool_space, $log_space,
3089 $spool_inodes, and $log_inodes.
3091 32. The RFC2047 encoding function was originally intended for short strings
3092 such as real names; it was not keeping to the 75-character limit for
3093 encoded words that the RFC imposes. It now respects the limit, and
3094 generates multiple encoded words if necessary. To be on the safe side, I
3095 have increased the buffer size for the ${rfc2047: expansion operator from
3098 33. It is now permitted to omit both strings after an "if" condition; if the
3099 condition is true, the result is "true". As before, when the second string
3100 is omitted, a false condition yields an empty string. This makes it less
3101 cumbersome to write custom ACL and router conditions.
3103 34. Failure to deliver a bounce message always caused it to be frozen, even if
3104 there was an errors_to setting on the router. The errors_to setting is now
3107 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3108 canonical form (fully expanded) before being placed in
3109 $sender_host_address.
3111 36. The table in the code that translates DNS record types into text (T_A to
3112 "A" for instance) was missing entries for NS and CNAME. It is just possible
3113 that this could have caused confusion if both these types were looked up
3114 for the same domain, because the text type is used as part of Exim's
3115 per-process caching. But the chance of anyone hitting this buglet seems
3118 37. The dnsdb lookup has been extended in a number of ways.
3120 (1) There is a new type, "zns", which walks up the domain tree until it
3121 finds some nameserver records. It should be used with care.
3123 (2) There is a new type, "mxh", which is like "mx" except that it returns
3124 just the host names, not the priorities.
3126 (3) It is now possible to give a list of domains (or IP addresses) to be
3127 looked up. The behaviour when one of the lookups defers can be
3128 controlled by a keyword.
3130 (4) It is now possible to specify the separator character for use when
3131 multiple records are returned.
3133 38. The dnslists ACL condition has been extended: it is now possible to supply
3134 a list of IP addresses and/or domains to be looked up in a particular DNS
3137 39. Added log_selector=+queue_time_overall.
3139 40. When running the queue in the test harness, wait just a tad after forking a
3140 delivery process, to get repeatability of debugging output.
3142 41. Include certificate and key file names in error message when GnuTLS fails
3143 to set them up, because the GnuTLS error message doesn't include the name
3144 of the failing file when there is a problem reading it.
3146 42. Allow both -bf and -bF in the same test run.
3148 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3150 44. The "Exiscan patch" is now merged into the mainline Exim source.
3152 45. Sometimes the final signoff response after QUIT could fail to get
3153 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3154 < 0 before doing a fflush(). This bug looks as though it goes back to the
3155 introduction of TLS in release 3.20, but "sometimes" must have been rare
3156 because the tests only now provoked it.
3158 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3159 (this can affect the format of dates).
3161 47. exim_tidydb, when checking for the continued existence of a message for
3162 which it has found a message-specific retry record, was not finding
3163 messages that were in split spool directories. Consequently, it was
3164 deleting retry records that should have stayed in existence.
3166 48. Steve fixed some bugs in eximstats.
3168 49. The SPA authentication driver was not abandoning authentication and moving
3169 on to the next authenticator when an expansion was forced to fail,
3170 contradicting the general specification for all authenticators. Instead it
3171 was generating a temporary error. It now behaves as specified.
3173 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3174 (the order specifies the preference for clients). The order is now AES256,
3175 AES128, 3DES, ARCFOUR128.
3177 51. Small patch to Sieve code - explicitly set From: when generating an
3180 52. Exim crashed if a remote delivery caused a very long error message to be
3181 recorded - for instance if somebody sent an entire SpamAssassin report back
3182 as a large number of 550 error lines. This bug was coincidentally fixed by
3183 increasing the size of one of Exim's internal buffers (big_buffer) that
3184 happened as part of the Exiscan merge. However, to be on the safe side, I
3185 have made the code more robust (and fixed the comments that describe what
3188 53. Now that there can be additional text after "Completed" in log lines (if
3189 the queue_time_overall log selector is set), a one-byte patch to exigrep
3190 was needed to allow it to recognize "Completed" as not the last thing in
3193 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3194 patch that reportedly fixes this has been added. I am not expert enough to
3195 create a test for it. This is what the patch creator wrote:
3197 "I found a little strange behaviour of ldap code when working with
3198 Windows 2003 AD Domain, where users was placed in more than one
3199 Organization Units. When I tried to give exim partial DN, the exit code
3200 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3201 But simultaneously result of request was absolutely normal ldap result,
3202 so I produce this patch..."
3204 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3205 REFERENCE, so I have modified the code to exclude the patch when that macro
3208 55. Some experimental protocols are using DNS PTR records for new purposes. The
3209 keys for these records are domain names, not reversed IP addresses. The
3210 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3211 leaves it alone. Component reversal etc. now happens only for IP addresses.
3214 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3216 57. Double the size of the debug message buffer (to 2048) so that more of very
3217 long debug lines gets shown.
3219 58. The exicyclog utility now does better if the number of log files to keep
3220 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3222 59. Two changes related to the smtp_active_hostname option:
3224 (1) $smtp_active_hostname is now available as a variable.
3225 (2) The default for smtp_banner uses $smtp_active_hostname instead
3226 of $primary_hostname.
3228 60. The host_aton() function is supposed to be passed a string that is known
3229 to be a valid IP address. However, in the case of IPv6 addresses, it was
3230 not checking this. This is a hostage to fortune. Exim now panics and dies
3231 if the condition is not met. A case was found where this could be provoked
3232 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3233 components; fortuitously, this particular loophole had already been fixed
3234 by change 4.50/55 above.
3236 If there are any other similar loopholes, the new check in host_aton()
3237 itself should stop them being exploited. The report I received stated that
3238 data on the command line could provoke the exploit when Exim was running as
3239 exim, but did not say which command line option was involved. All I could
3240 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3241 running as the user.
3244 61. There was a buffer overflow vulnerability in the SPA authentication code
3245 (which came originally from the Samba project). I have added a test to the
3246 spa_base64_to_bits() function which I hope fixes it.
3249 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3250 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3252 63. The daemon start-up calls getloadavg() while still root for those OS that
3253 need the first call to be done as root, but it missed one case: when
3254 deliver_queue_load_max is set with deliver_drop_privilege. This is
3255 necessary for the benefit of the queue runner, because there is no re-exec
3256 when deliver_drop_privilege is set.
3258 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3259 This has been fixed.
3261 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3262 was in use, was not putting the data itself into the right store pool;
3263 consequently, it could be overwritten for a subsequent message in the same
3264 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3267 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3269 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3270 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3271 the function were treating the return as a boolean value, which happened to
3272 work because 0=false and not-0=true, but is not correct code.
3274 68. The host_aton() function was not handling scoped IPv6 addresses (those
3275 with, for example, "%eth0" on the end) correctly.
3277 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3278 negated items (that is, ~something) in unsigned ints. Some compilers
3279 apparently mutter when there is no cast.
3281 70. If an address verification called from an ACL failed, and did not produce a
3282 user-specific message (i.e. there was only a "system" message), nothing was
3283 put in $acl_verify_message. In this situation, it now puts the system
3286 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3287 session; change 4.31/43 added the unwanted input to the log line - except
3288 that it did not do this in the start of session case. It now does.
3290 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3291 This is wrong and can cause the other end to generate a synchronization
3292 error if it is another Exim or anything else that does the synchronization
3293 check. A QUIT command is no longer sent after a timeout.
3295 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3296 during host lookups.
3298 74. The defer_ok option of callout verification was not working if it was used
3299 when verifying addresses in header lines, that is, for this case:
3301 verify = header_sender/callout=defer_ok
3303 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3304 those file descriptors could be used for SMTP connections. If anything
3305 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3306 could be sent to the SMTP client, causing chaos. The daemon now opens
3307 stdin, stdout, and stderr to /dev/null when it puts itself into the
3310 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3311 log by default. The user can override this with suitable Perl magic.
3313 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3314 the log message when discard triggers, was not working for the DATA ACL or
3315 for the non-SMTP ACL.
3317 78. Error message wording change in sieve.c.
3319 79. If smtp_accept_max_per_host was set, the number of connections could be
3320 restricted to fewer than expected, because the daemon was trying to set up
3321 a new connection before checking whether the processes handling previous
3322 connections had finished. The check for completed processes is now done
3323 earlier. On busy systems, this bug wouldn't be noticed because something
3324 else would have woken the daemon, and it would have reaped the completed
3327 80. If a message was submitted locally by a user whose login name contained one
3328 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3329 It caused a spool format error. I have fixed the spool reading code. A
3330 related problem was that the "from" clause in the Received: line became
3331 illegal because of the space(s). It is now covered by ${quote_local_part.
3333 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3336 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3338 83. Patch from Sieve maintainer to fix unterminated string problem in
3339 "vacation" handling.
3341 84. Some minor changes to the Linux configuration files to help with other
3342 OS variants using glibc.
3344 85. One more patch for Sieve to update vacation handling to latest spec.
3347 ----------------------------------------------------
3348 See the note above about the 4.44 and 4.50 releases.
3349 ----------------------------------------------------
3355 1. Change 4.43/35 introduced a bug that caused file counts to be
3356 incorrectly computed when quota_filecount was set in an appendfile
3359 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3360 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3363 3. Give more explanation in the error message when the command for a transport
3364 filter fails to execute.
3366 4. There are several places where Exim runs a non-Exim command in a
3367 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3368 was being done only for the command run by the queryprogram router. It is
3369 now done for all such subprocesses. The other cases are: ${run, transport
3370 filters, and the commands run by the lmtp and pipe transports.
3372 5. Some older OS have a limit of 256 on the maximum number of file
3373 descriptors. Exim was using setrlimit() to set 1000 as a large value
3374 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3375 systems. I've change it so that if it can't get 1000, it tries for 256.
3377 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3378 was an oversight, and furthermore, ever since the addition of extra
3379 controls (e.g. 4.43/32), the checks on when to allow different forms of
3380 "control" were broken. There should now be diagnostics for all cases when a
3381 control that does not make sense is encountered.
3383 7. $recipients is now available in the predata ACL (oversight).
3385 8. Tidy the search cache before the fork to do a delivery from a message
3386 received from the command line. Otherwise the child will trigger a lookup
3387 failure and thereby defer the delivery if it tries to use (for example) a
3388 cached ldap connection that the parent has called unbind on.
3390 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3391 of $address_data from the recipient verification was clobbered by the
3392 sender verification.
3394 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3395 was its contents. (It was OK if the option was not defined at all.)
3397 11. A "Completed" log line is now written for messages that are removed from
3398 the spool by the -Mrm option.
3400 12. $host_address is now set to the target address during the checking of
3401 ignore_target_hosts.
3403 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3404 being passed; this would have caused $sender_host_name to have been used if
3405 matching the list had actually called for a host name (not very likely,
3406 since this list is usually IP addresses). A host name is now passed as
3409 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3410 code that specifies a non-restarting handler (typically sigaction() in
3411 modern systems) in an attempt to fix a rare and obscure crash bug.
3413 15. Narrowed the window for a race in the daemon that could cause it to ignore
3414 SIGCHLD signals. This is not a major problem, because they are used only to
3415 wake it up if nothing else does.
3417 16. A malformed maildirsize file could cause Exim to calculate negative values
3418 for the mailbox size or file count. Odd effects could occur as a result.
3419 The maildirsize information is now recalculated if the size or filecount
3422 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3423 support for a long time. Removed HAVE_SYS_VFS_H.
3425 18. Updated exipick to current release from John Jetmore.
3427 19. Allow an empty sender to be matched against a lookup in an address list.
3428 Previously the only cases considered were a regular expression, or an
3431 20. Exim went into a mad DNS lookup loop when doing a callout where the
3432 host was specified on the transport, if the DNS lookup yielded more than
3435 21. The RFC2047 encoding function was originally intended for short strings
3436 such as real names; it was not keeping to the 75-character limit for
3437 encoded words that the RFC imposes. It now respects the limit, and
3438 generates multiple encoded words if necessary. To be on the safe side, I
3439 have increased the buffer size for the ${rfc2047: expansion operator from
3442 22. Failure to deliver a bounce message always caused it to be frozen, even if
3443 there was an errors_to setting on the router. The errors_to setting is now
3446 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3447 canonical form (fully expanded) before being placed in
3448 $sender_host_address.
3450 24. Updated eximstats to version 1.33
3452 25. Include certificate and key file names in error message when GnuTLS fails
3453 to set them up, because the GnuTLS error message doesn't include the name
3454 of the failing file when there is a problem reading it.
3456 26. Expand error message when OpenSSL has problems setting up cert/key files.
3459 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3460 (this can affect the format of dates).
3462 28. exim_tidydb, when checking for the continued existence of a message for
3463 which it has found a message-specific retry record, was not finding
3464 messages that were in split spool directories. Consequently, it was
3465 deleting retry records that should have stayed in existence.
3467 29. eximstats updated to version 1.35
3468 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3469 1.35 - bugfix such that pie charts by volume are generated correctly
3471 30. The SPA authentication driver was not abandoning authentication and moving
3472 on to the next authenticator when an expansion was forced to fail,
3473 contradicting the general specification for all authenticators. Instead it
3474 was generating a temporary error. It now behaves as specified.
3476 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3477 (the order specifies the preference for clients). The order is now AES256,
3478 AES128, 3DES, ARCFOUR128.
3480 31. Small patch to Sieve code - explicitly set From: when generating an
3483 32. Exim crashed if a remote delivery caused a very long error message to be
3484 recorded - for instance if somebody sent an entire SpamAssassin report back
3485 as a large number of 550 error lines. This bug was coincidentally fixed by
3486 increasing the size of one of Exim's internal buffers (big_buffer) that
3487 happened as part of the Exiscan merge. However, to be on the safe side, I
3488 have made the code more robust (and fixed the comments that describe what
3491 33. Some experimental protocols are using DNS PTR records for new purposes. The
3492 keys for these records are domain names, not reversed IP addresses. The
3493 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3494 leaves it alone. Component reversal etc. now happens only for IP addresses.
3497 34. The host_aton() function is supposed to be passed a string that is known
3498 to be a valid IP address. However, in the case of IPv6 addresses, it was
3499 not checking this. This is a hostage to fortune. Exim now panics and dies
3500 if the condition is not met. A case was found where this could be provoked
3501 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3502 components; fortuitously, this particular loophole had already been fixed
3503 by change 4.50/55 or 4.44/33 above.
3505 If there are any other similar loopholes, the new check in host_aton()
3506 itself should stop them being exploited. The report I received stated that
3507 data on the command line could provoke the exploit when Exim was running as
3508 exim, but did not say which command line option was involved. All I could
3509 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3510 running as the user.
3513 35. There was a buffer overflow vulnerability in the SPA authentication code
3514 (which came originally from the Samba project). I have added a test to the
3515 spa_base64_to_bits() function which I hope fixes it.
3518 36. The daemon start-up calls getloadavg() while still root for those OS that
3519 need the first call to be done as root, but it missed one case: when
3520 deliver_queue_load_max is set with deliver_drop_privilege. This is
3521 necessary for the benefit of the queue runner, because there is no re-exec
3522 when deliver_drop_privilege is set.
3524 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3525 was in use, was not putting the data itself into the right store pool;
3526 consequently, it could be overwritten for a subsequent message in the same
3527 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3530 38. Sometimes the final signoff response after QUIT could fail to get
3531 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3532 < 0 before doing a fflush(). This bug looks as though it goes back to the
3533 introduction of TLS in release 3.20, but "sometimes" must have been rare
3534 because the tests only now provoked it.
3540 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3541 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3542 Now it returns an integer. A number of calls were still expecting a T/F
3543 return. Fortuitously, in all cases, the tests worked in OK situations,
3544 which is the norm. However, things would have gone wrong on any write
3545 failures on the smtp file descriptor. This function is used when sending
3546 messages over SMTP and also when doing verify callouts.
3548 2. When Exim is called to do synchronous delivery of a locally submitted
3549 message (the -odf or -odi options), it no longer closes stderr before doing
3552 3. Implemented the mua_wrapper option.
3554 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3556 5. Implemented the functions header_remove(), header_testname(),
3557 header_add_at_position(), and receive_remove_recipient(), and exported them
3560 6. If an ACL "warn" statement specified the addition of headers, Exim already
3561 inserted X-ACL-Warn: at the start if there was no header name. However, it
3562 was not making this test for the second and subsequent header lines if
3563 there were newlines in the string. This meant that an invalid header could
3564 be inserted if Exim was badly configured.
3566 7. Allow an ACL "warn" statement to add header lines at the start or after all
3567 the Received: headers, as well as at the end.
3569 8. Added the rcpt_4xx retry error code.
3571 9. Added postmaster_mailfrom=xxx to callout verification option.
3573 10. Added mailfrom=xxxx to the callout verification option, for verify=
3576 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3577 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3579 12. Inserted some casts to stop certain compilers warning when using pointer
3580 differences as field lengths or precisions in printf-type calls (mostly
3581 affecting debugging statements).
3583 13. Added optional readline() support for -be (dynamically loaded).
3585 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3586 same clock tick as a message's arrival, so that its received time was the
3587 same as the "first fail" time on the retry record, and that message
3588 remained on the queue past the ultimate address timeout, every queue runner
3589 would try a delivery (because it was past the ultimate address timeout) but
3590 after another failure, the ultimate address timeout, which should have then
3591 bounced the address, did not kick in. This was a "< instead of <=" error;
3592 in most cases the first failure would have been in the next clock tick
3593 after the received time, and all would be well.
3595 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3596 being recognized when the domain list was tested by the match_domain
3597 condition in an expansion string.
3599 16. Added the ${str2b64: operator.
3601 17. Exim was always calling setrlimit() to set a large limit for the number of
3602 processes, without checking whether the existing limit was already
3603 adequate. (It did check for the limit on file descriptors.) Furthermore,
3604 errors from getrlimit() and setrlimit() were being ignored. Now they are
3605 logged to the main and panic logs, but Exim does carry on, to try to do its
3606 job under whatever limits there are.
3608 18. Imported PCRE 5.0.
3610 19. Trivial typo in log message " temporarily refused connection" (the leading
3613 20. If the log selector return_path_on_delivery was set and an address was
3614 redirected to /dev/null, the delivery process crashed because it assumed
3615 that a return path would always be set for a "successful" delivery. In this
3616 case, the whole delivery is bypassed as an optimization, and therefore no
3619 21. Internal re-arrangement: the function for sending a challenge and reading
3620 a response while authentication was assuming a zero-terminated challenge
3621 string. It's now changed to take a pointer and a length, to allow for
3622 binary data in such strings.
3624 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3626 23. Exim was not respecting finduser_retries when seeking the login of the
3627 uid under which it was called; it was always trying 10 times. (The default
3628 setting of finduser_retries is zero.) Also, it was sleeping after the final
3629 failure, which is pointless.
3631 24. Implemented tls_on_connect_ports.
3633 25. Implemented acl_smtp_predata.
3635 26. If the domain in control=submission is set empty, Exim assumes that the
3636 authenticated id is a complete email address when it generates From: or
3637 Sender: header lines.
3639 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3640 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3641 chown and chgrp in /bin and hostname in /usr/bin.
3643 28. Exim was keeping the "process log" file open after each use, just as it
3644 does for the main log. This opens the possibility of it remaining open for
3645 long periods when the USR1 signal hits a daemon. Occasional processlog
3646 errors were reported, that could have been caused by this. Anyway, it seems
3647 much more sensible not to leave this file open at all, so that is what now
3650 29. The long-running daemon process does not normally write to the log once it
3651 has entered its main loop, and it closes the log before doing so. This is
3652 so that log files can straightforwardly be renamed and moved. However,
3653 there are a couple of unusual error situations where the daemon does write
3654 log entries, and I had neglected to close the log afterwards.
3656 30. The text of an SMTP error response that was received during a remote
3657 delivery was being truncated at 512 bytes. This is too short for some of
3658 the long messages that one sometimes sees. I've increased the limit to
3661 31. It is now possible to make retry rules that apply only when a message has a
3662 specific sender, in particular, an empty sender.
3664 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3665 it possible to be selective about when SMTP synchronization is enforced.
3667 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3669 32. Implemented hosts_connection_nolog.
3671 33. Added an ACL for QUIT.
3673 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3676 35. Added mailbox_size and mailbox_filecount to appendfile.
3678 36. Added control = no_multiline_responses to ACLs.
3680 37. There was a bug in the logic of the code that waits for the clock to tick
3681 in the case where the clock went backwards by a substantial amount such
3682 that the microsecond fraction of "now" was more than the microsecond
3683 fraction of "then" (but the whole seconds number was less).
3685 38. Added support for the libradius Radius client library this is found on
3686 FreeBSD (previously only the radiusclient library was supported).
3692 1. When certain lookups returned multiple values in the form name=value, the
3693 quoting of the values was not always being done properly. Specifically:
3694 (a) If the value started with a double quote, but contained no whitespace,
3696 (b) If the value contained whitespace other than a space character (i.e.
3697 tabs or newlines or carriage returns) it was not quoted.
3698 This fix has been applied to the mysql and pgsql lookups by writing a
3699 separate quoting function and calling it from the lookup code. The fix
3700 should probably also be applied to nisplus, ibase and oracle lookups, but
3701 since I cannot test any of those, I have not disturbed their existing code.
3703 2. A hit in the callout cache for a specific address caused a log line with no
3704 reason for rejecting RCPT. Now it says "Previous (cached) callout
3705 verification failure".
3707 3. There was an off-by-one bug in the queryprogram router. An over-long
3708 return line was truncated at 256 instead of 255 characters, thereby
3709 overflowing its buffer with the terminating zero. As well as fixing this, I
3710 have increased the buffer size to 1024 (and made a note to document this).
3712 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3713 when Exim is waiting for an SMTP response from a remote server, Exim
3714 restarts its select() call on the socket, thereby resetting its timeout.
3715 This is not a problem when such interrupts are rare. Somebody set up a cron
3716 job to run exiwhat every 2 minutes, which is less than the normal select()
3717 timeout (5 or 10 minutes). This meant that the select() timeout never
3718 kicked in because it was always reset. I have fixed this by comparing the
3719 time when an interrupt arrives with the time at the start of the first call
3720 to select(). If more time than the timeout has elapsed, the interrupt is
3721 treated as a timeout.
3723 5. Some internal re-factoring in preparation for the addition of Sieve
3724 extensions (by MH). In particular, the "personal" test is moved to a
3725 separate function, and given an option for scanning Cc: and Bcc: (which is
3726 not set for Exim filters).
3728 6. When Exim created an email address using the login of the caller as the
3729 local part (e.g. when creating a From: or Sender: header line), it was not
3730 quoting the local part when it contained special characters such as @.
3732 7. Installed new OpenBSD configuration files.
3734 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3735 try to make them clearer.
3737 9. Callout options, other than the timeout value, were being ignored when
3738 verifying sender addresses in header lines. For example, when using
3740 verify = header_sender/callout=no_cache
3742 the cache was (incorrectly) being used.
3744 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3745 only the Cygwin environment.
3747 11. When return_path_on_delivery was set as a log selector, if different remote
3748 addresses in the same message used different return paths and parallel
3749 remote delivery occurred, the wrong values would sometimes be logged.
3750 (Whenever a remote delivery process finished, the return path value from
3751 the most recently started remote delivery process was logged.)
3753 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3754 lines when AUTH and/or TLS are in use. This is the "received protocol"
3755 field. Exim used to use "asmtp" for authenticated SMTP, without any
3756 indication (in the protocol name) for TLS use. Now it follows the RFC and
3757 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3758 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3759 names appear in log lines as well as in Received: header lines.
3761 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3762 extensions, and comparison tests, and to fix some bugs.
3764 14. Changes to the "personal" filter test:
3766 (1) The test was buggy in that it was just doing the equivalent of
3767 "contains" tests on header lines. For example, if a user's address was
3768 anne@some.where, the "personal" test would incorrectly be true for
3770 To: susanne@some.where
3772 This test is now done by extracting each address from the header in turn,
3773 and checking the entire address. Other tests that are part of "personal"
3774 are now done using regular expressions (for example, to check local parts
3775 of addresses in From: header lines).
3777 (2) The list of non-personal local parts in From: addresses has been
3778 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3779 taken from the Sieve specification recommendations.
3781 (3) If the message contains any header line starting with "List-" it is
3782 treated as non-personal.
3784 (4) The test for "circular" in the Subject: header line has been removed
3785 because it now seems ill-conceived.
3787 15. Minor typos in src/EDITME comments corrected.
3789 16. Installed latest exipick from John Jetmore.
3791 17. If headers_add on a router specified a text string that was too long for
3792 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3793 of string_sprintf() is now avoided.
3795 18. $message_body_size was not set (it was always zero) when running the DATA
3796 ACL and the local_scan() function.
3798 19. For the "mail" command in an Exim filter, no default was being set for
3799 the once_repeat time, causing a random time value to be used if "once" was
3800 specified. (If the value happened to be <= 0, no repeat happened.) The
3801 default is now 0s, meaning "never repeat". The "vacation" command was OK
3802 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3803 (I found it when inspecting the code).
3805 20. There is now an overall timeout for performing a callout verification. It
3806 defaults to 4 times the callout timeout, which applies to individual SMTP
3807 commands during the callout. The overall timeout applies when there is more
3808 than one host that can be tried. The timeout is checked before trying the
3809 next host. This prevents very long delays if there are a large number of
3810 hosts and all are timing out (e.g. when the network connections are timing
3811 out). The value of the overall timeout can be changed by specifying an
3812 additional sub-option for "callout", called "maxwait". For example:
3814 verify = sender/callout=5s,maxwait=20s
3816 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3817 to the end before writing, but this should make it even safer).
3819 22. Exim was forgetting that it had advertised PIPELINING for the second and
3820 subsequent messages on an SMTP connection. It was also not resetting its
3821 memory on STARTTLS and an internal HELO.
3823 23. When Exim logs an SMTP synchronization error within a session, it now
3824 records whether PIPELINING has been advertised or not.
3826 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3827 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3828 rather than long int.
3830 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3832 26. Added the never_mail option to autoreply.
3838 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3839 crash if the getsockname() call failed; this can happen if a connection is
3840 closed very soon after it is established. The problem was simply in the
3841 order in which certain operations were done, causing Exim to try to write
3842 to the SMTP stream before it had set up the file descriptor. The bug has
3843 been fixed by making things happen in the correct order.
3849 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3850 before the connection was closed, thus losing the rejection response.
3852 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3853 some early Solaris releases, but causes trouble in current releases where
3854 socklen_t is defined.
3856 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3859 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3862 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3863 that releasing the top of it at the end releases what was used for sub-
3864 expansions (unless the block got too big). However, discard this block if
3865 the first thing is a variable or header, so that we can use its block when
3866 it is dynamic (useful for very large $message_headers, for example).
3868 6. Lookups now cache *every* query, not just the most recent. A new, separate
3869 store pool is used for this. It can be recovered when all lookup caches are
3870 flushed. Lookups now release memory at the end of their result strings.
3871 This has involved some general refactoring of the lookup sources.
3873 7. Some code has been added to the store_xxx() functions to reduce the amount
3874 of flapping under certain conditions.
3876 8. log_incoming_interface used to affect only the <= reception log lines. Now
3877 it causes the local interface and port to be added to several more SMTP log
3878 lines, for example "SMTP connection from", and rejection lines.
3880 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3882 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3884 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3885 could be overwritten at the end of the current message (or the start of a
3886 new message if it was set in a HELO ACL). The value is now preserved for
3887 the duration of the SMTP connection.
3889 12. If a transport had a headers_rewrite setting, and a matching header line
3890 contained an unqualified address, that address was qualified, even if it
3891 did not match any rewriting rules. The underlying bug was that the values
3892 of the flags that permit the existence of unqualified sender and recipient
3893 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3894 non-local messages, and by -bnq for local messages) were not being
3895 preserved with the message after it was received.
3897 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3898 "next input=" as part of the text comprising the host identity instead of
3899 the correct text. The code was using the same buffer for two different
3900 strings. However, depending on which order the printing function evaluated
3901 its arguments, the bug did not always show up. Under Linux, for example, my
3902 test suite worked just fine.
3904 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3905 doesn't work with some older versions of Perl. It has been changed to "my",
3906 which in any case is probably the better facility to use.
3908 15. A really picky compiler found some instances of statements for creating
3909 error messages that either had too many or two few arguments for the format
3912 16. The size of the buffer for calls to the DNS resolver has been increased
3913 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3914 for addresses that have a lot of PTR records. This alleviates a problem; it
3915 does not fully solve it.
3917 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3918 buffer now truncates the list and logs the incident, which is the same
3919 action as happens when Exim is looking up a host name and its aliases.
3920 Previously in this situation something unpredictable would happen;
3921 sometimes it was "internal error: store_reset failed".
3923 18. If a server dropped the connection unexpectedly when an Exim client was
3924 using GnuTLS and trying to read a response, the client delivery process
3925 crashed while trying to generate an error log message.
3927 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3928 string, for example:
3930 warn message = H1: something\nH2: something
3932 the text was added as a single header line from Exim's point of view
3933 though it ended up OK in the delivered message. However, searching for the
3934 second and subsequent header lines using $h_h2: did not work. This has been
3935 fixed. Similarly, if a system filter added multiple headers in this way,
3936 the routers could not see them.
3938 20. Expanded the error message when iplsearch is called with an invalid key to
3939 suggest using net-iplsearch in a host list.
3941 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3942 ACLs are no longer actually imposed (and a message to that effect is
3945 22. If a "gecos" field in a passwd entry contained escaped characters, in
3946 particular, if it contained a \" sequence, Exim got it wrong when building
3947 a From: or a Sender: header from that name. A second bug also caused
3948 incorrect handling when an unquoted " was present following a character
3949 that needed quoting.
3951 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3952 was not being matched caselessly.
3954 24. Arranged for all hyphens in the exim.8 source to be escaped with
3957 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3958 properly. Recipient callout cache records were still being keyed to include
3959 the sender, even when use_sender was set false. This led to far more
3960 callouts that were necessary. The sender is no longer included in the key
3961 when use_sender is false.
3963 26. Added "control = submission" modifier to ACLs.
3965 27. Added the ${base62d: operator to decode base 62 numbers.
3967 28. dnsdb lookups can now access SRV records.
3969 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3970 the configuration file.
3972 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3973 (-v) mode. This makes the output for a verbose queue run more intelligible.
3975 31. Added a use_postmaster feature to recipient callouts.
3977 32. Added the $body_zerocount variable, containing the number of binary zero
3978 bytes in the message body.
3980 33. The time of last modification of the "new" subdirectory is now used as the
3981 "mailbox time last read" when there is a quota error for a maildir
3984 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3986 35. Added +ignore_unknown as a special item in host lists.
3988 36. Code for decoding IPv6 addresses in host lists is now included, even if
3989 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3990 address was recognized as an IP address, but was then not correctly decoded
3991 into binary, causing unexpected and incorrect effects when compared with
3998 1. Very minor rewording of debugging text in manualroute to say "list of
3999 hosts" instead of "hostlist".
4001 2. If verify=header_syntax was set, and a header line with an unqualified
4002 address (no domain) and a large number of spaces between the end of the
4003 name and the colon was received, the reception process suffered a buffer
4004 overflow, and (when I tested it) crashed. This was caused by some obsolete
4005 code that should have been removed. The fix is to remove it!
4007 3. When running in the test harness, delay a bit after writing a bounce
4008 message to get a bit more predictability in the log output.
4010 4. Added a call to search_tidyup() just before forking a reception process. In
4011 theory, someone could use a lookup in the expansion of smtp_accept_max_
4012 per_host which, without the tidyup, could leave open a database connection.
4014 5. Added the variables $recipient_data and $sender_data which get set from a
4015 lookup success in an ACL "recipients" or "senders" condition, or a router
4016 "senders" option, similar to $domain_data and $local_part_data.
4018 6. Moved the writing of debug_print from before to after the "senders" test
4021 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4022 problems for message scanning, either using a data ACL, or using
4023 local_scan() because the Received: header was not generated till after they
4024 were called (in order to set the time as the time of reception completion).
4025 I have revised the way this works. The header is now generated after the
4026 body is received, but before the ACL or local_scan() are called. After they
4027 are run, the timestamp in the header is updated.
4033 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4034 before starting a queue runner without re-exec. This happened only when
4035 deliver_drop_privilege was set or when the Exim user was set to root. The
4036 effect of the bug was that timeouts during subsequent deliveries caused
4037 crashes instead of being properly handled. The handler is now left at its
4038 default (and expected) setting.
4040 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4041 message, again when deliver_drop_privilege is set or Exim is run as root.
4042 The bug described in (1) was not present in this case, but the tidying up
4043 of the other signals was missing. I have made the two cases consistent.
4045 3. The ignore_target_hosts setting on a manualroute router was being ignored
4046 for hosts that were looked up using the /MX notation.
4048 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4051 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4052 operated on the sender address. After changing the $sender_address to <>
4053 for the sender address verify, Exim was re-instated it as the original
4054 (before rewriting) address, but remembering that it had rewritten it, so it
4055 wasn't rewriting it again. This bug also had the effect of breaking the
4056 sender address verification caching when the sender address was rewritten.
4058 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4059 This has been changed so that if the ip literal address matches
4060 ignore_target_hosts, the router declines.
4062 7. Added expansion conditions match_domain, match_address, and match_local_
4063 part (NOT match_host).
4065 8. The placeholder for the Received: header didn't have a length field set.
4067 9. Added code to Exim itself and to exim_lock to test for a specific race
4068 condition that could lead to file corruption when using MBX delivery. The
4069 issue is with the lockfile that is created in /tmp. If this file is removed
4070 after a process has opened it but before that process has acquired a lock,
4071 there is the potential for a second process to recreate the file and also
4072 acquire a lock. This could lead to two Exim processes writing to the file
4073 at the same time. The added code performs the same test as UW imapd; it
4074 checks after acquiring the lock that its file descriptor still refers to
4075 the same named file.
4077 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4078 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4081 11. Added the smtp_active_hostname option. If used, this will typically be made
4082 to depend on the incoming interface address. Because $interface_address is
4083 not set up until the daemon has forked a reception process, error responses
4084 that can happen earlier (such as "too many connections") no longer contain
4087 12. If an expansion in a condition on a "warn" statement fails because a lookup
4088 defers, the "warn" statement is abandoned, and the next ACL statement is
4089 processed. Previously this caused the whole ACL to be aborted.
4091 13. Added the iplsearch lookup type.
4093 14. Added ident_timeout as a log selector.
4095 15. Added tls_certificate_verified as a log selector.
4097 16. Added a global option tls_require_ciphers (compare the smtp transport
4098 option of the same name). This controls incoming TLS connections.
4100 17. I finally figured out how to make tls_require_ciphers do a similar thing
4101 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4102 before starting the TLS session.
4104 18. Tabs are now shown as \t in -bP output.
4106 19. If the log selector return_path_on_delivery was set, Exim crashed when
4107 bouncing a message because it had too many Received: header lines.
4109 20. If two routers both had headers_remove settings, and the first one included
4110 a superfluous trailing colon, the final name in the first list and the
4111 first name in the second list were incorrectly joined into one item (with a
4112 colon in the middle).
4118 1. Added -C and -D options to the exinext utility, mainly to make it easier
4119 to include in the automated testing, but these could be helpful when
4120 multiple configurations are in use.
4122 2. The exinext utility was not formatting the output nicely when there was
4123 an alternate port involved in the retry record key, nor when there was a
4124 message id as well (for retries that were specific to a specific message
4125 and a specific host). It was also confused by IPv6 addresses, because of
4126 the additional colons they contain. I have fixed the IPv4 problem, and
4127 patched it up to do a reasonable job for IPv6.
4129 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4130 delivery, the log line now contains "pipelined" if PIPELINING was used.
4132 4. An SMTP transport process used to panic and die if the bind() call to set
4133 an explicit outgoing interface failed. This has been changed; it is now
4134 treated in the same way as a connect() failure.
4136 5. A reference to $sender_host_name in the part of a conditional expansion
4137 that was being skipped was still causing a DNS lookup. This no longer
4140 6. The def: expansion condition was not recognizing references to header lines
4141 that used bh_ and bheader_.
4143 7. Added the _cache feature to named lists.
4145 8. The code for checking quota_filecount in the appendfile transport was
4146 allowing one more file than it should have been.
4148 9. For compatibility with Sendmail, the command line option
4156 and sets the incoming protocol and host name (for trusted callers). The
4157 host name and its colon can be omitted when only the protocol is to be set.
4158 Note the Exim already has two private options, -pd and -ps, that refer to
4159 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4160 "s", but I don't think that's a major issue.
4162 10. A number of refactoring changes to the code, none of which should affect
4165 (a) The number of logging options was getting close to filling up the
4166 32-bit word that was used as a bit map. I have split them into two classes:
4167 those that are passed in the argument to log_write(), and those that are
4168 only ever tested independently outside of that function. These are now in
4169 separate 32-bit words, so there is plenty of room for expansion again.
4170 There is no change in the user interface or the logging behaviour.
4172 (b) When building, for example, log lines, the code previously used a
4173 macro that called string_cat() twice, in order to add two strings. This is
4174 not really sufficiently general. Furthermore, there was one instance where
4175 it was actually wrong because one of the argument was used twice, and in
4176 one call a function was used. (As it happened, calling the function twice
4177 did not affect the overall behaviour.) The macro has been replaced by a
4178 function that can join an arbitrary number of extra strings onto a growing
4181 (c) The code for expansion conditions now uses a table and a binary chop
4182 instead of a serial search (which was left over from when there were very
4183 few conditions). Also, it now recognizes conditions like "pam" even when
4184 the relevant support is not compiled in: a suitably worded error message is
4185 given if an attempt is made to use such a condition.
4187 11. Added ${time_interval:xxxxx}.
4189 12. A bug was causing one of the ddress fields not to be passed back correctly
4190 from remote delivery subprocesses. The field in question was not being
4191 subsequently used, so this caused to problems in practice.
4193 13. Added new log selectors queue_time and deliver_time.
4195 14. Might have fixed a bug in maildirsizefile handling that threw up
4196 "unexpected character" debug warnings, and recalculated the data
4197 unnecessarily. In any case, I expanded the warning message to give more
4200 15. Added the message "Restricted characters in address" to the statements in
4201 the default ACL that block characters like @ and % in local parts.
4203 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4204 Three changes have been made:
4206 (a) There was a serious bug; a negative response to MAIL caused the whole
4207 recipient domain to be cached as invalid, thereby blocking all messages
4208 to all local parts at the same domain, from all senders. This bug has
4209 been fixed. The domain is no longer cached after a negative response to
4210 MAIL if the sender used is not empty.
4212 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4215 (c) A new callout option, "use_sender" has been added for people who want
4216 the modified behaviour.
4222 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4225 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4226 indeed breaks things for older releases.
4228 3. Added additional logging to the case where there is a problem reading data
4229 from a filter that is running in a subprocess using a pipe, in order to
4230 try to track down a specific problem.
4232 4. Testing facility fudge: when running in the test harness and attempting
4233 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4234 getting "No route to host". Convert this to a timeout.
4236 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4239 6. Some OS don't have socklen_t but use size_t instead. This affects the
4240 fifth argument of getsockopt() amongst other things. This is now
4241 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4242 can be set for individual OS. I have set it for SunOS5, OSF1, and
4243 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4244 some earlier ones do not.
4246 7. Change 4.30/15 was not doing the test caselessly.
4248 8. The standard form for an IPv6 address literal was being rejected by address
4249 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4250 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4251 this, as well as the form without the "IPv6" on the front (but only when
4252 address literals are enabled, of course).
4254 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4256 10. Exim crashed if a message with an empty sender address specified by -f
4257 encountered a router with an errors_to setting. This could be provoked only
4258 by a command such as
4262 where an empty string was supplied; "<>" did not hit this bug.
4264 11. Installed PCRE release 4.5.
4266 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4267 remained set. It is now erased.
4269 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4270 times from message ids (which are base 36 rather than the normal 62).
4272 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4273 were being counted as actual protocol errors, and logged if the log
4274 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4275 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4276 and DATA following a set of rejected RCPTs do not count as protocol errors.
4277 In other words, Exim assumes they were pipelined, though this may not
4278 actually be the case. Of course, in all cases the client gets an
4279 appropriate error code.
4281 15. If a lookup fails in an ACL condition, a message about the failure may
4282 be available; it is used if testing the ACL cannot continue, because most
4283 such messages specify what the cause of the deferral is. However, some
4284 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4285 that caused an old message to be retained and used if a later statement
4286 caused a defer, replacing the real cause of the deferral.
4288 16. If an IP address had so many PTR records that the DNS lookup buffer
4289 was not large enough to hold them, Exim could crash while trying to process
4290 the truncated data. It now detects and logs this case.
4292 17. Further to 4.21/58, another change has been made: if (and only if) the
4293 first line of a message (the first header line) ends with CRLF, a bare LF
4294 in a subsequent header line has a space inserted after it, so as not to
4295 terminate the header.
4297 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4298 unnecessarily, used atoi() instead of strtol(), and didn't check the
4299 termination when getting file sizes from file names by regex.
4301 19. Completely re-implemented the support for maildirsize files, in the light
4302 of a number of problems with the previous contributed implementation
4303 (4.30/29). In particular:
4305 . If the quota is zero, the maildirsize file is maintained, but no quota is
4308 . If the maildir directory does not exist, it is created before any attempt
4309 to write a maildirsize file.
4311 . The quota value in the file is just a cache; if the quota is changed in
4312 the transport, the new value overrides.
4314 . A regular expression is available for excluding directories from the
4317 20. The autoreply transport checks the characters in options that define the
4318 message's headers; it allows continued headers, but it was checking with
4319 isspace() after an embedded newline instead of explicitly looking for a
4322 21. If all the "regular" hosts to which an address was routed had passed their
4323 expiry times, and had not reached their retry times, the address was
4324 bounced, even if fallback hosts were defined. Now Exim should go on to try
4327 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4328 equivalent code in the SMTP transport. Some hosts send humungous responses
4329 to HELO/EHLO, more than 1024 it seems.
4331 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4332 gives compiler warnings in some environments. I've now done it "properly",
4335 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4336 (because of problems with the built-in one) was declared to return uschar *
4337 instead of char *, causing compiler failure.
4339 25. Fixed a file descriptor leak when processing alias/forward files.
4341 26. Fixed a minor format string issue in dbfn.c.
4343 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4345 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4346 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4349 29. When a sender address is verified, it is cached, to save repeating the test
4350 when there is more than one recipient in a message. However, when the
4351 verification involves a callout, it is possible for different callout
4352 options to be set for different recipients. It is too complicated to keep
4353 track of this in the cache, so now Exim always runs a verification when a
4354 callout is required, relying on the callout cache for the optimization.
4355 The overhead is duplication of the address routing, but this should not be
4358 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4359 to be verified with callout=postmaster, and the main callout worked but the
4360 postmaster check failed, the verification correctly failed. However, if a
4361 subsequent RCPT command asked for sender verification *without* the
4362 postmaster check, incorrect caching caused this verification also to fail,
4365 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4366 it was not caching the DNS options (qualify_single, search_parents) that
4367 were used when the lookup failed. A subsequent lookup with different
4368 options therefore always gave the same answer, though there were cases
4369 where it should not have. (Example: a "domains = !$mx_any" option on a
4370 dnslookup router: the "domains" option is always processed without any
4371 widening, but the router might have qualify_single set.) Now Exim uses the
4372 cached value only when the same options are set.
4374 32. Added John Jetmore's "exipick" utility to the distribution.
4376 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4377 than a timeout (e.g. a certificate is required, and is not provided), an
4378 Exim server now closes the connection immediately. Previously it waited for
4379 the client to close - but if the client is SSL, it seems that they each
4380 wait for each other, leading to a delay before one of them times out.
4382 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4383 maintained 0.8.x compatibility because I don't think many are using it, and
4384 it is clearly obsolete.
4386 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4389 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4390 client certificate was expired. A simple patch fixes this, though I don't
4391 understand the full logic of why the verify callback is called multiple
4394 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4395 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4396 which causes problems with some clients (such as the Certicom SSL Plus
4397 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4398 disables the coutermeasure allowing Eudora to connect."
4400 38. Exim was not checking that a write() to a log file succeeded. This could
4401 lead to Bad Things if a log got too big, in particular if it hit a file
4402 size limit. Exim now panics and dies if it cannot write to a log file, just
4403 as it does if it cannot open a log file.
4405 39. Modified OS/Makefile-Linux so that it now contains
4407 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4409 The two -D definitions ensure that Exim is compiled with large file
4410 support, which makes it possible to handle log files that are bigger than
4413 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4414 instance) a domain was checked against a named list that involved a lookup,
4415 causing $domain_data to be set, then another domain was checked against the
4416 same list, then the first domain was re-checked, the value of $domain_data
4417 after the final check could be wrong. In particular, if the second check
4418 failed, it could be set empty. This bug probably also applied to
4421 41. The strip_trailing_dot option was not being applied to the address given
4422 with the -f command-line option.
4424 42. The code for reading a message's header from the spool was incrementing
4425 $received_count, but never initializing it. This meant that the value was
4426 incorrect (doubled) while delivering a message in the same process in which
4427 it was received. In the most common configuration of Exim, this never
4428 happens - a fresh exec is done - but it can happen when
4429 deliver_drop_privilege is set.
4431 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4432 it now includes up to 150 characters of the unexpected data in the log
4435 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4436 and building data strings. The size of both of these buffers was 10 000
4437 bytes - far larger than anybody would *ever* want, thought I. Needless to
4438 say, somebody hit the limit. I have increased the maximum line length to
4439 20 000 and the maximum data length of concatenated lines to 100 000. I have
4440 also fixed two bugs, because there was no checking on these buffers. Tsk,
4441 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4442 buffer is too small.
4444 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4445 lsearch lookups. Now it does.
4447 46. When parsing a route_list item in a manualroute router, a fixed-length
4448 buffer was used for the list of hosts. I made this 1024 bytes long,
4449 thinking that nobody would ever have a list of hosts that long. Wrong.
4450 Somebody had a whole pile of complicated expansion conditions, and the
4451 string was silently truncated, leading to an expansion error. It turns out
4452 that it is easier to change to an unlimited length (owing to other changes
4453 that have happened since this code was originally written) than to build
4454 structure for giving a limitation error. The length of the item that
4455 expands into the list of hosts is now unlimited.
4457 47. The lsearch lookup could not handle data where the length of text line was
4458 more than 4095 characters. Such lines were truncated, leading to shortened
4459 data being returned. It should now handle lines of any length.
4461 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4462 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4465 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4466 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4467 no longer happens in comment lines. A list of replacements is now placed
4468 at the head of all of the source files, except those whose only change is
4469 to replace PERL_COMMAND in the very first #! line.
4471 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4472 messages on the queue, with a bottom-up merge sort, using code contributed
4473 by Michael Haardt. This should make operations like -bp somewhat faster on
4474 large queues. It won't affect queue runners, except when queue_run_in_order
4477 51. Installed eximstats 1.31 in the distribution.
4479 52. Added support for SRV lookups to the dnslookup router.
4481 53. If an ACL referred to $message_body or $message_body_end, the value was not
4482 reset for any messages that followed in the same SMTP session.
4484 54. The store-handling optimization for building very long strings was not
4485 differentiating between the different store pools. I don't think this
4486 actually made any difference in practice, but I've tidied it.
4488 55. While running the routers to verify a sender address, $sender_address
4489 was still set to the sender address. This is wrong, because when routing to
4490 send a bounce to the sender, it would be empty. Therefore, I have changed
4491 it so that, while verifying a sender address, $sender_address is set to <>.
4492 (There is no change to what happens when verifying a recipient address.)
4494 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4495 target A or AAAA records (if not already returned) without resetting the
4496 qualify_single or search_parents options of the DNS resolver. These are
4497 inappropriate in this case because the targets of MX and SRV records must
4498 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4499 target that, when qualified, matched something in the local domain. These
4500 two options are now turned off when doing these lookups.
4502 57. It seems that at least some releases of Reiserfs (which does not have the
4503 concept of a fixed number of inodes) returns zero and not -1 for the
4504 number of available inodes. This interacted badly with check_spool_inodes,
4505 which assumed that -1 was the "no such thing" setting. What I have done is
4506 to check that the total number of inodes is greater than zero before doing
4507 the test of how many are available.
4509 58. When a "warn" ACL statement has a log_message modifier, the message is
4510 remembered, and not repeated. This is to avoid a lot of repetition when a
4511 message has many recipients that cause the same warning to be written.
4512 Howewer, Exim was preserving the list of already written lines for an
4513 entire SMTP session, which doesn't seem right. The memory is now reset if a
4514 new message is started.
4516 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4517 debugging output unless log_rewrite was also set.
4519 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4520 of (int)(handle) when we know that handle contains (void *)(-1).
4522 61. The Exim daemon panic-logs an error return when it closes the incoming
4523 connection. However "connection reset by peer" seems to be common, and
4524 isn't really an error worthy of noting specially, so that particular error
4527 62. When Exim is trying to find all the local interfaces, it used to panic and
4528 die if the ioctl to get the interface flags failed. However, it seems that
4529 on at least one OS (Solaris 9) it is possible to have an interface that is
4530 included in the list of interfaces, but for which you get a failure error
4531 for this call. This happens when the interface is not "plumbed" into a
4532 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4533 failure of the "get flags" call assumes that the interface is down.
4535 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4536 makes life easier for people who are doing arithmetic on fields extracted
4537 from dates, where you often get leading zeros that should not be
4538 interpreted as octal.
4540 64. Added qualify_domain to the redirect router, to override the global
4543 65. If a pathologically long header line contained very many addresses (the
4544 report of this problem mentioned 10 000) and each of them was rewritten,
4545 Exim could use up a very large amount of memory. (It kept on making new
4546 copies of the header line as it rewrote, and never released the old ones.)
4547 At the expense of a bit more processing, the header rewriting function has
4548 been changed so that it no longer eats memory in this way.
4550 66. The generation of the Received: header has been moved from the time that a
4551 message starts to be received, to the time that it finishes. The timestamp
4552 in the Received: header should now be very close to that of the <= log
4553 line. There are two side-effects of this change:
4555 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4556 logged header lines no longer include the local Received: line, because
4557 it has not yet been created. The same applies to a copy of the message
4558 that is returned to a non-SMTP sender when a message is rejected.
4560 (b) When a filter file is tested using -bf, no additional Received: header
4561 is added to the test message. After some thought, I decided that this
4564 This change does not affect the value of $received_for. It is still set
4565 after address rewriting, but before local_scan() is called.
4567 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4569 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4570 gave an unhelpful panic error message, and a defer error. I have managed to
4571 change this behaviour so that it now rejects any supplied certificate,
4572 which seems right, as the list of acceptable certificates is empty.
4574 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4575 gave an unhelpful defer error. I have not managed to make this reject any
4576 supplied certificates, but the error message it gives is "no certificate
4577 supplied", which is not helpful.
4579 70. exigrep's output now also includes lines that are not associated with any
4580 message, but which match the given pattern. Implemented by a patch from
4581 Martin Sluka, which also tidied up the Perl a bit.
4583 71. Recipient callout verification, like sender verification, was using <> in
4584 the MAIL FROM command. This isn't really the right thing, since the actual
4585 sender may affect whether the remote host accepts the recipient or not. I
4586 have changed it to use the actual sender in the callout; this means that
4587 the cache record is now keyed on a recipient/sender pair, not just the
4588 recipient address. There doesn't seem to be a real danger of callout loops,
4589 since a callout by the remote host to check the sender would use <>.
4590 [SEE ABOVE: changed after hitting problems.]
4592 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4593 temporary errors. However, in the case of such a code being given after
4594 the end of a data transmission (i.e. after ".") Exim was failing to write
4595 a retry record for the message. (Yes, there was some broken host that was
4596 actually sending 8xx at this point.)
4598 73. An unknown lookup type in a host list could cause Exim to panic-die when
4599 the list was checked. (An example that provoked this was putting <; in the
4600 middle of a list instead of at the start.) If this happened during a DATA
4601 ACL check, a -D file could be left lying around. This kind of configuration
4602 error no longer causes Exim to die; instead it causes a defer errror. The
4603 incident is still logged to the main and panic logs.
4605 74. Buglet left over from Exim 3 conversion. The message "too many messages
4606 in one connection" was written to the rejectlog but not the mainlog, except
4607 when address rewriting (yes!) was being logged.
4609 75. Added write_rejectlog option.
4611 76. When a system filter was run not as root (that is, when system_filter_user
4612 was set), the values of the $n variables were not being returned to the
4613 main process; thus, they were not subsequently available in the $sn
4616 77. Added +return_path_on_delivery log selector.
4618 78. A connection timeout was being treated differently from recipients deferred
4619 when testing hosts_max_try with a message that was older than the host's
4620 retry timeout. (The host should not be counted, thus allowing all hosts to
4621 be tried at least once before bouncing.) This may have been the cause of an
4622 occasionally reported bug whereby a message would remain on the queue
4623 longer than the retry timeout, but would be bounced if a delivery was
4624 forced. I say "may" because I never totally pinned down the problem;
4625 setting up timeout/retry tests is difficult. See also the next item.
4627 79. The ultimate address timeout was not being applied to errors that involved
4628 a combination of host plus message (for example, a timeout on a MAIL
4629 command). When an address resolved to a number of possible hosts, and they
4630 were not all tried for each delivery (e.g. because of hosts_max_try), a
4631 message could remain on the queue longer than the retry timeout.
4633 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4636 81. Fixed an obscure SMTP outgoing bug which required at least the following
4637 conditions: (a) there was another message waiting for the same server;
4638 (b) the server returned 5xx to all RCPT commands in the first message so
4639 that the message was not completed; (c) the server dropped the connection
4640 or gave a negative response to the RSET that Exim sends to abort the
4641 transaction. The observed case was a dropped connection after DATA that had
4642 been sent in pipelining mode. That is, the server had advertised PIPELINING
4643 but was not implementing it correctly. The effect of the bug was incorrect
4644 behaviour, such as trying another host, and this could lead to a crash.
4650 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4651 and daemon.c were passed as pointers to ints; they should have been
4652 pointers to socklen_t variables (which are typically unsigned ints).
4654 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4657 3. Fixed a really odd bug that affected only the testing scheme; patching a
4658 certain fixed string in the binary changed the value of another string that
4659 happened to be identical to the end of the original first string.
4661 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4662 name", it returns that address as the IP address. On some operating
4663 systems (e.g. Solaris), it also passes back the IP address string as the
4664 "host name". However, on others (e.g. Linux), it passes back an empty
4665 string. Exim wasn't checking for this, and was changing the host name to an
4666 empty string, assuming it had been canonicized.
4668 5. Although rare, it is permitted to have more than one PTR record for a given
4669 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4670 all the names associated with an address, because they do in Solaris.
4671 However, it seems that they do not in Linux for data that comes from the
4672 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4673 I found this out when I moved to a new Linux workstation and tried to run
4674 the Exim test suite.
4676 To get round this problem I have changed the code so that it now does its
4677 own call to the DNS to look up PTR records when searching for a host name.
4678 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4679 addresses that are only in /etc/hosts are still found.
4681 This behaviour is, however, controlled by an option called host_lookup_
4682 order, which defaults to "bydns:byaddr". If people want to use the other
4683 order, or indeed, just use one or the other means of lookup, they can
4684 specify it in this variable.
4686 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4687 some operating systems, this comes back from gethostbyaddr() as an empty
4688 string, and this is what Exim used to test for. However, it seems that in
4689 other systems, "." is yielded. Exim now tests for this case too.
4691 7. The values of check_spool_space and check_log_space are now held internally
4692 as a number of kilobytes instead of an absolute number of bytes. If a
4693 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4694 kilobyte. This means that much larger values can be stored.
4696 8. Exim monitor: an attempt to get the action menu when not actually pointing
4697 at a message produces an empty menu entitled "No message selected". This
4698 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4699 no entries in it ("Shell widget menu has zero width and/or height"). So I
4700 have added a single, blank menu entry in this case.
4702 9. Added ${quote_local_part.
4704 10. MIME decoding is now applied to the contents of Subject: header lines when
4707 11. Now that a reference to $sender_host_address automatically causes a reverse
4708 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4709 host lookup before query-style lookups in lists that might use this
4710 variable. This has therefore been abolished, and the "net-" prefix is no
4711 longer necessary for query-style lookups.
4713 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4714 have been a typo for LFLAGS, so it has been changed.
4716 13. The install script calls Exim with "-C /dev/null" in order to find the
4717 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4718 to be output. Howeve, since Exim outputs its version number before the
4719 error, it didn't break the script. It just looked ugly. I fixed this by
4720 always allowing "-C /dev/null" if the caller is root.
4722 14. Ignore overlarge ACL variable number when reading spool file - insurance
4723 against a later release with more variables having written the file.
4725 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4726 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4727 this, as well as the form without the "IPv6" on the front.
4729 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4730 OS/Makefile-Darwin file.
4732 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4733 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4735 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4736 variable. However, some OS use ints for inodes. I've added cast to long int
4737 to get rid of the compiler warning.
4739 19. I had forgotten to lock out "/../" in configuration file names when
4740 ALT_CONFIG_PREFIX was set.
4742 20. Routers used for verification do not need to specify transports. However,
4743 if such a router generated a host list, and callout was configured, Exim
4744 crashed, because it could not find a port number from the (non-existent)
4745 transport. It now assumes port 25 in this circumstance.
4747 21. Added the -t option to exigrep.
4749 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4750 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4751 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4752 them is set, LOOKUP_LSEARCH is forced.
4754 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4756 24. Added sender and host information to the "rejected by local_scan()" log
4757 line; previously there was no indication of these.
4759 25. Added .include_if_exists.
4761 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4762 Linux. It turns out that not all file systems support this. Apparently some
4763 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4764 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4765 not supported on the file descriptor, is now ignored when Exim is trying to
4766 sync a directory. This applies only to Linux.
4768 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4770 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4771 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4774 29. Added support for maildirsize files from supplied patch (modified a bit).
4776 30. The use of :fail: followed by an empty string could lead Exim to respond to
4777 sender verification failures with (e.g.):
4779 550 Verification failed for <xxx>
4780 550 Sender verify failed
4782 where the first response line was missing the '-' that indicates it is not
4783 the final line of the response.
4785 31. The loop for finding the name of the user that called Exim had a hardwired
4786 limit of 10; it now uses the value of finduser_retries, which is used for
4787 all other user lookups.
4789 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4792 33. Exim was neglecting to zero errno before one call of strtol() when
4793 expanding a string and expecting an integer value. On some systems this
4794 resulted in spurious "integer overflow" errors. Also, it was casting the
4795 result into an int without checking.
4797 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4798 did not work. The code looks as if it has *never* worked, though it appears
4799 to have been documented since at least releast 1.62. I have made it work.
4801 35. The "timeout_DNS" error in retry rules, also documented since at least
4802 1.62, also never worked. As it isn't clear exactly what this means, and
4803 clearly it isn't a major issue, I have abolished the feature by treating it
4804 as "timeout", and writing a warning to the main and panic logs.
4806 36. The display of retry rules for -brt wasn't always showing the error code
4809 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4810 timeout_connect_A, timeout_connect_MX.
4812 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4813 to the empty sender.
4815 39. The daemon was not analysing the content of -oX till after it had closed
4816 stderr and disconnected from the controlling terminal. This meant that any
4817 syntax errors were only noted on the panic log, and the return code from
4818 the command was 0. By re-arranging the code a little, I've made the
4819 decoding happen first, so such errors now appear on stderr, and the return
4820 code is 1. However, the actual setting up of the sockets still happens in
4821 the disconnected process, so errors there are still only recorded on the
4824 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4825 connections (as happens on some IP stacks) was logged at start up time as
4826 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4827 it from "IPv6 and IPv4", which means that two separate sockets are being
4830 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4831 says whether AF_INET or AF_INET6 was passed as an argument.
4833 42. Exiwhat output was messed up when time zones were included in log
4836 43. Exiwhat now gives more information about the daemon's listening ports,
4837 and whether -tls-on-connect was used.
4839 44. The "port" option of the smtp transport is now expanded.
4841 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4842 silently ignored. Now an error message is written to the main and panic
4845 46. There's a new ACL modifier called "logwrite" which writes to a log file
4846 as soon as it is encountered.
4848 47. Added $local_user_uid and $local_user_gid at routing time.
4850 48. Exim crashed when trying to verify a sender address that was being
4853 49. Exim was recognizing only a space character after ".include". It now also
4854 recognizes a tab character.
4856 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4857 extracting the relevant information from the specification. The man page no
4858 longer contains scrambled data for the -d option, and I've added a section
4859 at the front about calling Exim under different names.
4861 51. Added "extra_headers" argument to the "mail" command in filter files.
4863 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4866 53. Installed eximstats 1.29.
4868 54. Added transport_filter_timeout as a generic transport option.
4870 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4871 Cc: header lines. This was required by RFC 822, but it not required by RFC
4874 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4875 incoming messages that did not have them. Now it does so only if the
4876 message originates locally, that is, if there is no associated remote host
4877 address. When Resent- header lines are present, this applies to the Resent-
4878 lines rather than the non-Resent- lines.
4880 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4881 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4883 58. Messages for configuration errors now include the name of the main
4884 configuration file - useful now that there may be more than one file in a
4885 list (.included file names were always shown).
4887 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4888 for those rare installations that do not start the daemon as root or run it
4889 setuid root. I've cut out the call to initgroups() if the daemon is not
4892 60. The Exim user and group can now be bound into the binary as text strings
4893 that are looked up at the start of Exim's processing.
4895 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4897 62. Added $mailstore_basename variable.
4899 63. Installed patch to sieve.c from Michael Haardt.
4901 64. When Exim failed to open the panic log after failing to open the main log,
4902 the original message it was trying to log was written to stderr and debug
4903 output, but if they were not available (the usual case in production), it
4904 was lost. Now it is written to syslog before the two lines that record the
4905 failures to open the logs.
4907 65. Users' Exim filters run in subprocesses under the user's uid. It is
4908 possible for a "deliver" command or an alias in a "personal" command to
4909 provoke an address rewrite. If logging of address rewriting is configured,
4910 this fails because the process is not running as root or exim. There may be
4911 a better way of dealing with this, but for the moment (because 4.30 needs
4912 to be released), I have disabled address rewrite logging when running a
4913 filter in a non-root, non-exim process.
4919 1. The buildconfig auxiliary program wasn't quoting the value set for
4920 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4921 not defined. This bug was masked in 4.22 by the effect that was fixed in
4924 2. Some messages that were rejected after a message id was allocated were
4925 shown as "incomplete" by exigrep. It no longer does this for messages that
4926 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4928 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4929 have allow_domain_literals set, the ID did not get logged in the <= line.
4930 Domain literals are now always recognized in Message-ID: header lines.
4932 4. The first argument for a ${extract expansion item is the key name or field
4933 number. Leading and trailing spaces in this item were not being ignored,
4934 causing some misleading effects.
4936 5. When deliver_drop_privilege was set, single queue runner processes started
4937 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4938 same command in the process it spins off) were not dropping privilege.
4940 6. When the daemon running as "exim" started a queue runner, it always
4941 re-executed Exim in the spun-off process. This is a waste of effort when
4942 deliver_drop_privilege is set. The new process now just calls the
4943 queue-runner function directly.
4949 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4952 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4953 ignored. Though the use of -oP was forcing the writing of a pid file, it
4954 was always written to the default place.
4956 3. If the message "no IP address found for host xxxx" is generated during
4957 incoming verification, it is now followed by identification of the incoming
4958 connection (so you can more easily find what provoked it).
4960 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4962 5. Added some features to "harden" Exim a bit more against certain attacks:
4964 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4965 be put in Local/Makefile. This is like the never_users runtime option,
4966 but it cannot be overridden. The default setting is "root".
4968 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4969 prefix string with which any file named in a -C command line option
4972 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4973 is retained for -C and -D only if the caller of Exim is root. Without
4974 it, the exim user may also use -C and -D and retain privilege.
4976 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4977 command line option is disabled.
4979 6. Macro names set by the -D option must start with an upper case letter, just
4980 like macro names defined in the configuration file.
4982 7. Added "dereference=" facility to LDAP.
4984 8. Two instances of the typo "uknown" in the source files are fixed.
4986 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4987 the Configure-Makefile script screwed up while processing it.
4989 10. Incorporated PCRE 4.4.
4991 11. The SMTP synchronization check was not operating right at the start of an
4992 SMTP session. For example, it could not catch a HELO sent before the client
4993 waited for the greeting. There is now a check for outstanding input at the
4994 point when the greeting is written. Because of the duplex, asynchronous
4995 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4996 way, but not yet received, when the check is performed.
4998 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4999 on TCP/IP sockets, because this apparently causes some broken clients to
5002 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5003 unchanged) from the Cygwin maintainer.
5005 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5006 was supported instead of testing for mbx. Effectively a typo.
5008 15. The spa authenticator server code was not checking that the input it
5009 received was valid base64.
5011 16. The debug output line for the "set" modifier in ACLs was not showing the
5012 name of the variable that was being set.
5014 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5016 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5017 lookup on its own. Something else was needed to trigger the lookup. For
5018 example, a match in host_lookup or the need for a host name in a host list.
5019 Now, if $sender_host_name is referenced and the host name has not yet been
5020 looked up, a lookup is performed. If the lookup fails, the variable remains
5021 empty, and $host_lookup_failed is set to "1".
5023 19. Added "eqi" as a case-independent comparison operator.
5025 20. The saslauthd authentication condition could segfault if neither service
5026 nor realm was specified.
5028 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5029 error message that was logged was misleading, and incoming SMTP
5030 connections were dropped. The message is now more accurate, and temporary
5031 errors are given to SMTP connections.
5033 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5034 (except RSET) with a 421 error, until QUIT is received. However, it was
5035 failing to send a response to QUIT.
5037 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5038 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5039 ACL is now run after the helo_try_verify_hosts code.
5041 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5042 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5043 case-independent, so other case variants are also recognized). Apparently
5044 some systems use these upper case variants.
5046 25. If more than two messages were waiting for the same host, and a transport
5047 filter was specified for the transport, Exim sent two messages over the
5048 same TCP/IP connection, and then failed with "socket operation on non-
5049 socket" when it tried to send the third.
5051 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5053 27. The extern definition of crypt16() in expand.c was not being excluded when
5054 the OS had its own crypt16() function.
5056 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5057 as a preferred synonym for return_size_limit, both as an option and as an
5060 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5062 30. Changed the default configuration ACL to relax the local part checking rule
5063 for addresses that are not in any local domains. For these addresses,
5064 slashes and pipe symbols are allowed within local parts, but the sequence
5065 /../ is explicitly forbidden.
5067 31. SPA server authentication was not clearing the challenge buffer before
5070 32. log_message in a "warn" ACL statement was writing to the reject log as
5071 well as to the main log, which contradicts the documentation and doesn't
5072 seem right (because no rejection is happening). So I have stopped it.
5074 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5075 However, I am unable to do any testing of this.
5077 34. Fixed an infelicity in the appendfile transport. When checking directories
5078 for a mailbox, to see if any needed to be created, it was accidentally
5079 using path names with one or more superfluous leading slashes; tracing
5080 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5082 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5083 contents are added to the log line that is written for every discarded
5084 recipient. (Previously a log_message setting was ignored.)
5086 36. The ${quote: operator now quotes the string if it is empty.
5088 37. The install script runs exim in order to find its version number. If for
5089 some reason other than non-existence or emptiness, which it checks, it
5090 could not run './exim', it was installing it with an empty version number,
5091 i.e. as "exim-". This error state is now caught, and the installation is
5094 38. An argument was missing from the function that creates an error message
5095 when Exim fails to connect to the socket for saslauthd authentication.
5096 This could cause Exim to crash, or give a corrupted message.
5098 39. Added isip, isip4, and isip6 to ${if conditions.
5100 40. The ACL variables $acl_xx are now saved with the message, and can be
5101 accessed later in routers, transports, and filters.
5103 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5104 strings in the file are not string-expanded.
5106 42. If a MAIL command specified a SIZE value that was too large to fit into an
5107 int variable, the check against message_size_limit failed. Such values are
5108 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5109 day this will have to be increased, but I don't think I want to be around
5110 when emails are that large.
5117 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5118 iconv() is not standard in FreeBSD.
5120 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5121 IPv6 enabled. The observed symptom was a segmentation fault on return from
5122 the function os_common_find_running_interfaces() in src/os.c.
5124 3. In the check_special_case() function in daemon.c I had used "errno" as an
5125 argument name, which causes warnings on some systems. This was basically a
5126 typo, since it was named "eno" in the comments!
5128 4. The code that waits for the clock to tick (at a resolution of some fraction
5129 of a second) so as to ensure message-id uniqueness was always waiting for
5130 at least one whole tick, when it could have waited for less. [This is
5131 almost certainly not relevant at current processor speeds, where it is
5132 unlikely to ever wait at all. But we try to future-proof.]
5134 5. The function that sleeps for a time interval that includes fractions of a
5135 second contained a race. It did not block SIGALRM between setting the
5136 timer, and suspending (a couple of lines later). If the interval was short
5137 and the sigsuspend() was delayed until after it had expired, the suspension
5138 never ended. On busy systems this could lead to processes getting stuck for
5141 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5142 process, before it forks any delivery processes. The open lookup caching
5143 mechanism meant that the open file or database connection was passed into
5144 the delivery process. The problem was that delivery processes always tidy
5145 up cached lookup data. This could cause a problem for the next delivery
5146 process started by the queue runner, because the external queue runner
5147 process does not know about the closure. So the next delivery process
5148 still has data in the lookup cache. In the case of a file lookup, there was
5149 no problem because closing a file descriptor in a subprocess doesn't affect
5150 the parent. However, if the lookup was caching a connection to a database,
5151 the connection was closed, and the second delivery process was likely to
5152 see errors such as "PGSQL: query failed: server closed the connection
5153 unexpectedly". The problem has been fixed by closing all cached lookups
5154 in a queue runner before running a delivery process.
5156 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5157 seem to have the "const" qualifier which it has on other OS. I've
5160 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5161 *of the same type* (client or server) with the same public name that an
5162 error should be diagnosed.
5164 9. When Exim looked up a host name for an IP address, but failed to find the
5165 original IP address when looking up the host name (a safety check), it
5166 output the message "<ip address> does not match any IP for NULL", which was
5167 confusing, to say the least. The bug was that the host name should have
5168 appeared instead of "NULL".
5170 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5171 user that is built into the binary, and the -C or -D options is used, root
5172 privilege is dropped before the configuration file is read. In addition,
5173 logging is switched to stderr instead of the normal log files. If the
5174 configuration then re-defines the Exim user, the unprivileged environment
5175 is probably not what is expected, so Exim logs a panic warning message (but
5178 However, if deliver_drop_privilege is set, the unprivileged state may well
5179 be exactly what is intended, so the warning has been cut out in that case,
5180 and Exim is allowed to try to write to its normal log files.
5186 1. smtp_return_error_details was not giving details for temporary sender
5187 or receiver verification errors.
5189 2. Diagnose a configuration error if two authenticators have the same public
5192 3. Exim used not to create the message log file for a message until the first
5193 delivery attempt. This could be confusing when incoming messages were held
5194 for policy or load reasons. The message log file is now created at the time
5195 the message is received, and an initial "Received" line is written to it.
5197 4. The automatically generated man page for command line options had a minor
5198 bug that caused no ill effects; however, a more serious problem was that
5199 the procedure for building the man page automatically didn't always
5200 operate. Consequently, release 4.20 contains an out-of-date version. This
5201 shouldn't happen again.
5203 5. When building Exim with embedded Perl support, the script that builds the
5204 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5205 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5207 6. The freeze_tell option was not being used for messages that were frozen on
5208 arrival, either by an ACL or by local_scan().
5210 7. Added the smtp_incomplete_transaction log selector.
5212 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5213 was accepting AUTH without a new EHLO.
5215 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5216 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5219 10. Logging of TCP/IP connections (when configured) now happens in the main
5220 daemon process instead of the child process, so that the TCP/IP connection
5221 count is more accurate (but it can never be perfect).
5223 11. The use of "drop" in a nested ACL was not being handled correctly in the
5224 outer ACL. Now, if condition failure induced by the nested "drop" causes
5225 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5226 or "require"), the connection is dropped.
5228 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5229 that yield "discard" can now be used with an "accept" or a "discard" verb,
5230 but an error is generated for any others (because I can't see a useful way
5231 to define what should happen).
5233 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5234 are now processed in the same way as lines in the Exim configuration file.
5235 In particular, continuation lines are supported.
5237 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5239 15. Added -ti meaning -t -i.
5241 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5242 domains, and warn by logging if others are found.
5244 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5245 returned from GIFCONF when Exim is trying to find the list of interfaces on
5246 a host. The code in os.c has been modified to copy each ifreq to an aligned
5247 structure in all cases.
5249 Also, in some cases, the returned ifreq's were being copied to a 'struct
5250 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5251 means the last couple of bytes of an IPv6 address could be chopped if the
5252 ifreq contained only a normal sockaddr (14 bytes storage).
5254 18. Named domain lists were not supported in the hosts_treat_as_local option.
5255 An entry such as +xxxx was not recognized, and was treated as a literal
5258 19. Ensure that header lines added by a DATA ACL are included in the reject log
5259 if the ACL subsequently rejects the message.
5261 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5262 MD5 (which is deprecated).
5264 21. When testing a filter file using -bf, Exim was writing a message when it
5265 took the sender from a "From " line in the message, but it was not doing so
5266 when it took $return_path from a Return-Path: header line. It now does.
5268 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5269 with a valid header line field name (a series of printing characters
5270 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5272 23. Changed "disc" in the source to "disk" to conform to the documentation and
5273 the book and for uniformity.
5275 24. Ignore Sendmail's -Ooption=value command line item.
5277 25. When execve() failed while trying to run a command in a pipe transport,
5278 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5279 could be confused with a return value of 69 from the command itself. This
5280 has been changed to 127, the value the shell returns if it is asked to run
5281 a non-existent command. The wording for the related log line suggests a
5282 non-existent command as the problem.
5284 26. If received_header_text expands to an empty string, do not add a Received:
5285 header line to the message. (Well, it adds a token one on the spool, but
5286 marks it "old" so that it doesn't get used or transmitted.)
5288 27. Installed eximstats 1.28 (addition of -nt option).
5290 28. There was no check for failure on the call to getsockname() in the daemon
5291 code. This can fail if there is a shortage of resources on the system, with
5292 ENOMEM, for example. A temporary error is now given on failure.
5294 29. Contrary to the C standard, it seems that in some environments, the
5295 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5296 program. Exim now does this explicitly; it affects the formatting of
5297 timestamps using strftime().
5299 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5300 complaints. I've now initialized all the variables, to avoid this.
5302 32. Header lines added by a system filter were not being "seen" during
5303 transport-time rewrites.
5305 33. The info_callback() function passed to OpenSSL is set up with type void
5306 (*)(SSL *, int, int), as described somewhere. However, when calling the
5307 function (actually a macro) that sets it up, the type void(*)() is
5308 expected. I've put in a cast to prevent warnings from picky compilers.
5310 34. If a DNS black list lookup found a CNAME record, but there were no A
5311 records associated with the domain it pointed at, Exim crashed.
5313 35. If a DNS black list lookup returned more than one A record, Exim ignored
5314 all but the first. It now scans all returned addresses if a particular IP
5315 value is being sought. In this situation, the contents of the
5316 $dnslist_value variable are a list of all the addresses, separated by a
5319 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5320 to accept a host name and all its aliases if the forward lookup for any of
5321 them yielded the IP address of the incoming connection. Now it accepts only
5322 those names whose forward lookup yields the correct IP address. Any other
5323 names are discarded. This closes a loophole whereby a rogue DNS
5324 administrator could create reverse DNS records to break through a
5325 wildcarded host restriction in an ACL.
5327 37. If a user filter or a system filter that ran in a subprocess used any of
5328 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5329 the wrong values were passed to the pipe command ($thisaddress had the
5330 value of $0, $0 had the value of $1, etc). This bug was introduced by
5331 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5333 38. Improved the line breaking for long SMTP error messages from ACLs.
5334 Previously, if there was no break point between 40 and 75 characters, Exim
5335 left the rest of the message alone. Two changes have been made: (a) I've
5336 reduced the minimum length to 35 characters; (b) if it can't find a break
5337 point between 35 and 75 characters, it looks ahead and uses the first one
5338 that it finds. This may give the occasional overlong line, but at least the
5339 remaining text gets split now.
5341 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5342 file descriptors might be low, and that setting 1000 would always raise it.
5343 It turns out that in some environments, the limit is already over 1000 and
5344 that lowering it causes trouble. So now Exim takes care not to decrease it.
5346 40. When delivering a message, the value of $return_path is set to $sender_
5347 address at the start of routing (routers may change the value). By an
5348 oversight, this default was not being set up when an address was tested by
5349 -bt or -bv, which affected the outcome if any router or filter referred to
5352 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5353 log only when "message" or "log_message" are set. However, if one of the
5354 conditions was an address verification, or a call to a nested ACL, the
5355 messages generated by the underlying test were being passed through. This
5356 no longer happens. The underlying message is available in $acl_verify_
5357 message for both "message" and "log_message" expansions, so it can be
5358 passed through if needed.
5360 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5361 new expansion $bh_ to give the encoded byte string without charset
5362 translation. Translation happens only if iconv() is available; HAVE_ICONV
5363 indicates this at build time. HEADERS_CHARSET gives the charset to
5364 translate to; headers_charset can change it in the configuration, and
5365 "headers charset" can change it in an individual filter file.
5367 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5368 that creates RFC 2047 encoded "words" labels them as that charset instead
5369 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5370 expansion operator; (ii) when Exim creates a From: line for a local
5371 message; (iii) when a header line is rewritten to include a "phrase" part.
5373 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5374 buggy, causing it to skip the first lines of messages whose message ID
5375 ended in 'D'. This would not have bitten before Exim release 4.14, because
5376 message IDs were unlikely to end in 'D' before then. The effect was to have
5377 incorrect size information for certain domains.
5379 45. #include "config.h" was missing at the start of the crypt16.c module. This
5380 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5383 46. If there was a timeout during a "random" callout check, Exim treated it as
5384 a failure of the random address, and carried on sending RSET and the real
5385 address. If the delay was just some slowness somewhere, the response to the
5386 original RCPT would be taken as a response to RSET and so on, causing
5387 mayhem of various kinds.
5389 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5390 when I implemented it. It didn't allow for the fact that some option values
5391 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5392 the right test for positive values.
5394 48. Domain names in DNS records are case-independent. Exim always looks them up
5395 in lower case. Some resolvers return domain names in exactly the case they
5396 appear in the zone file, that is, they may contain uppercase letters. Not
5397 all resolvers do this - some return always lower case. Exim was treating a
5398 change of case by a resolver as a change of domain, similar to a widening
5399 of a domain abbreviation. This triggered its re-routing code and so it was
5400 trying to route what was effectively the same domain again. This normally
5401 caused routing to fail (because the router wouldn't handle the domain
5402 twice). Now Exim checks for this case specially, and just changes the
5403 casing of the domain that it ultimately uses when it transmits the message
5406 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5409 50. If a filter generated a file delivery with a non-absolute name (possible if
5410 no home directory exists for the router), the forbid_file option was not
5413 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5414 the existing equality matching.
5416 52. Exim was using ints instead of ino_t variables in some places where it was
5417 dealing with inode numbers.
5419 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5420 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5421 called TMPDIR, and if it finds it is different, it changes its value.
5423 54. The smtp_printf() function is now made available to local_scan() so
5424 additional output lines can be written before returning. There is also an
5425 smtp_fflush() function to enable the detection of a dropped connection.
5426 The variables smtp_input and smtp_batched_input are exported to
5429 55. Changed the default runtime configuration: the message "Unknown user"
5430 has been removed from the ACL, and instead placed on the localuser router,
5431 using the cannot_route_message feature. This means that any verification
5432 failures that generate their own messages won't get overridden. Similarly,
5433 the "Unrouteable address" message that was in the ACL for unverifiable
5434 relay addresses has also been removed.
5436 56. Added hosts_avoid_esmtp to the smtp transport.
5438 57. The exicyclog script was not checking for the esoteric option
5439 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5440 will work only if exicyclog is run under the appropriate euid.
5442 58. Following a discussion on the list, the rules by which Exim recognises line
5443 endings on incoming messages have been changed. The -dropcr and drop_cr
5444 options are now no-ops, retained only for backwards compatibility. The
5445 following line terminators are recognized: LF CRLF CR. However, special
5446 processing applies to CR:
5448 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5449 nor a local message in the state where . is a terminator.
5451 (ii) If a bare CR is encountered in a header line, an extra space is added
5452 after the line terminator so as not to end the header. The reasoning
5453 behind this is that bare CRs in header lines are most likely either
5454 to be mistakes, or people trying to play silly games.
5456 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5457 was being incorrectly given as 18 bytes larger than it should have been.
5458 This is a VOB (very old bug).
5460 60. This may never have affected anything current, but just in case it has:
5461 When the local host is found other than at the start of a list of hosts,
5462 the local host, those with the same MX, and any that follow, are discarded.
5463 When the list in question was part of a longer list of hosts, the following
5464 hosts (not currently being processed) were also being discarded. This no
5465 longer happens. I'm not sure if this situation could ever has previously
5468 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5471 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5472 header. This is something that is recommended in a new Internet Draft, and
5473 is something that is documented as being done by Sendmail. There are two
5474 possible values. For messages generated by the autoreply transport, Exim
5477 Auto-Submitted: auto-replied
5479 whereas for all other generated messages (e.g. bounces) it adds
5481 Auto-Submitted: auto-generated
5483 63. The "personal" condition in filters now includes a test for the
5484 Auto-Submitted: header. If it contains the string "auto-" the message it
5485 not considered personal.
5487 64. Added rcpt_include_affixes as a generic transport option.
5489 65. Added queue_only_override (default true).
5491 66. Added the syslog_duplication option.
5493 67. If what should have been the first header line of a message consisted of
5494 a space followed by a colon, Exim was mis-interpreting it as a header line.
5495 It isn't of course - it is syntactically invalid and should therefore be
5496 treated as the start of the message body. The misbehaviour could have
5497 caused a number of strange effects, including loss of data in subsequent
5498 header lines, and spool format errors.
5500 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5501 client host had authenticated. This control can now be exercised by an ACL
5502 for more flexibility.
5504 69. By default, callouts do not happen when testing with -bh. There is now a
5505 variant, -bhc, which does actually run the callout code, including
5506 consulting and updating the callout cache.
5508 70. Added support for saslauthd authentication, courtesy of Alexander
5511 71. If statvfs() failed on the spool or log directories while checking their
5512 size for availability, Exim confusingly gave the error "space shortage".
5513 Furthermore, in debugging mode it crashed with a floating point exception.
5514 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5515 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5516 serious problem, Exim now writes to the main and panic logs when this
5517 happens, with details of the failure. It then refuses to accept the
5518 incoming message, giving the message "spool directory problem" or "log
5519 directory problem" with a 421 code for SMTP messages.
5521 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5522 0, 1, and 2 exist, because some OS complain for execs without them (see
5523 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5524 descriptors. However, the code omitted to check that the open succeeded,
5525 causing mysterious errors if for some reason the permissions on /dev/null
5526 got screwed. Now Exim writes a message to the main and panic logs, and
5527 bombs out if it can't open /dev/null.
5529 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5530 interact so that it is all more flexible. It is supposed to remain
5531 backwards compatible. Also added extra_local_interfaces.
5533 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5534 to bomb out with an assertion failure - to the client this appears as a
5535 connection drop. This problem occurs in the part of the code that was taken
5536 from the Samba project. Fortunately, the assertion is in a very simple
5537 function, so I have fixed this by reproducing the function inline in the
5538 one place where it is called, and arranging for authentication to fail
5539 instead of killing the process with assert().
5541 75. The SPA client code was not working when the server requested OEM rather
5542 than Unicode encoding.
5544 76. Added code to make require_files with a specific uid setting more usable in
5545 the case where statting the file as root fails - usually a non-root-mounted
5546 NFS file system. When this happens and the failure is EACCES, Exim now
5547 forks a subprocess and does the per-uid checking as the relevant uid.
5549 77. Added process_log_path.
5551 78. If log_file_path was not explicitly set, a setting of check_log_space or
5552 check_log_inodes was ignored.
5554 79. If a space check for the spool or log partitions fails, the incident is now
5555 logged. Of course, in the latter case the data may get lost...
5557 80. Added the %p formatting code to string_format() so that it can be used to
5558 print addresses in debug_print(). Adjusted all the address printing in the
5559 debugging in store.c to use %p rather than %d.
5561 81. There was a concern that a line of code in smtp_in.c could overflow a
5562 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5563 initially expressed, the concern was not well-founded, because trailing
5564 spaces are removed early. However, if the trailing spaces were followed by
5565 a NULL, they did not get removed, so the overflow was possible. Two fixes
5568 (a) I re-wrote the offending code in a cleaner fashion.
5569 (b) If an incoming SMTP command contains a NULL character, it is rejected
5572 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5573 runs initgroups(), so that if the Exim user is in any additional groups,
5574 they will be used during message reception.
5580 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob