1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
11 PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
14 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
16 PP/04 New "dns_use_edns0" global option.
18 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
21 PP/06 Extra paranoia around STARTTLS-with-data-in-buffer.
27 NM/01 Workround for PCRE version dependency in version reporting
30 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
31 This fixes portability to compilers other than gcc, notably
32 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
34 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
35 makefiles for portability to HP-UX and POSIX correctness.
37 PP/01 Permit LOOKUP_foo enabling on the make command-line.
38 Also via indented variable definition in the Makefile.
39 (Debugging by Oliver Heesakkers).
41 PP/02 Restore caching of spamd results with expanded spamd_address.
42 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
44 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
45 Improves build reliability. Fix from: Frank Elsner
47 NM/02 Fix wide character breakage in the rfc2047 coding
48 Fixes bug 1064. Patch from Andrey N. Oktyabrski
50 NM/03 Allow underscore in dnslist lookups
51 Fixes bug 1026. Patch from Graeme Fowler
53 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
54 Code patches from Adam Ciarcinski of NetBSD.
56 NM/04 Fixed exiqgrep to cope with mailq missing size issue
59 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
60 is logged, to avoid truncation. Patch from John Horne.
62 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
63 Patch from Jakob Hirsch.
65 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
66 SQL string expansion failure details.
67 Patch from Andrey Oktyabrski.
69 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
70 Patch from Simon Arlott.
72 PP/09 New lookups functionality failed to compile on old gcc which rejects
73 extern declarations in function scope.
74 Patch from Oliver Fleischmann
76 PP/10 Use sig_atomic_t for flags set from signal handlers.
77 Check getgroups() return and improve debugging.
78 Fixed developed for diagnosis in bug 927 (which turned out to be
81 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
82 Patch from Mark Zealey.
84 PP/12 Bugzilla 1056: Improved spamd server selection.
85 Patch from Mark Zealey.
87 PP/13 Bugzilla 1086: Deal with maildir quota file races.
88 Based on patch from Heiko Schlittermann.
90 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
91 Patch from Uwe Doering, sign-off by Michael Haardt.
93 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
94 variable declaration deep within a block. Bug and patch from
97 PP/15 lookups-Makefile IRIX compatibilty coercion.
99 PP/16 Make DISABLE_DKIM build knob functional.
101 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
102 Patch by Simon Arlott
104 TF/03 Fix valgrind.h portability to C89 compilers that do not support
105 variable argument macros. Our copy now differs from upstream.
111 TF/01 Failure to get a lock on a hints database can have serious
112 consequences so log it to the panic log.
114 TF/02 Log LMTP confirmation messages in the same way as SMTP,
115 controlled using the smtp_confirmation log selector.
117 TF/03 Include the error message when we fail to unlink a spool file.
119 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
120 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
121 for maintaining out-of-tree patches for some time.
123 PP/01 Bugzilla 139: Documentation and portability issues.
124 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
125 Handle per-OS dynamic-module compilation flags.
127 PP/02 Let /dev/null have normal permissions.
128 The 4.73 fixes were a little too stringent and complained about the
129 permissions on /dev/null. Exempt it from some checks.
130 Reported by Andreas M. Kirchwitz.
132 PP/03 Report version information for many libraries, including
133 Exim version information for dynamically loaded libraries. Created
134 version.h, now support a version extension string for distributors
135 who patch heavily. Dynamic module ABI change.
137 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
138 privilege escalation vulnerability whereby the Exim run-time user
139 can cause root to append content of the attacker's choosing to
142 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
145 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
146 If dropping privileges for untrusted macros, we disabled normal logging
147 on the basis that it would fail; for the Exim run-time user, this is not
148 the case, and it resulted in successful deliveries going unlogged.
149 Fixed. Reported by Andreas Metzler.
155 PP/01 Date: & Message-Id: revert to normally being appended to a message,
156 only prepend for the Resent-* case. Fixes regression introduced in
157 Exim 4.70 by NM/22 for Bugzilla 607.
159 PP/02 Include check_rfc2047_length in configure.default because we're seeing
160 increasing numbers of administrators be bitten by this.
162 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
164 PP/03 Bugzilla 994: added openssl_options main configuration option.
166 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
168 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
170 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
172 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
173 without a peer certificate, leading to a segfault because of an
174 assumption that peers always have certificates. Be a little more
175 paranoid. Problem reported by Martin Tscholak.
177 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
178 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
179 NB: ClamAV planning to remove STREAM in "middle of 2010".
180 CL also introduces -bmalware, various -d+acl logging additions and
181 more caution in buffer sizes.
183 PP/09 Implemented reverse_ip expansion operator.
185 PP/10 Bugzilla 937: provide a "debug" ACL control.
187 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
189 PP/12 Bugzilla 973: Implement --version.
191 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
193 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
195 PP/15 Bugzilla 816: support multiple condition rules on Routers.
197 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
198 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
199 ignore trailing whitespace.
201 JJ/02 prevent non-panic DKIM error from being sent to paniclog
203 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
206 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
207 Notification from Dr Andrew Aitchison.
209 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
210 ExtendedDetectionInfo response format.
211 Notification from John Horne.
213 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
216 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
217 XSL and documented dependency on system catalogs, with examples of how
220 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
223 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
224 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
225 configuration file which is writeable by the Exim user or group.
227 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
228 of configuration files to cover files specified with the -C option if
229 they are going to be used with root privileges, not just the default
232 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
233 option (effectively making it always true).
235 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
236 files to be used while preserving root privileges.
238 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
239 that rogue child processes cannot use them.
241 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
242 run-time user, instead of root.
244 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
245 Exim run-time user without dropping privileges.
247 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
248 result string, instead of calling string_vformat() twice with the same
251 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
252 for other users. Others should always drop root privileges if they use
253 -C on the command line, even for a whitelisted configure file.
255 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
257 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
263 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
264 $data_path, and $header_path variables; fixed documentation bugs and
267 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
268 exipick to access non-standard spools, including the "frozen" queue
271 NM/01 Bugzilla 965: Support mysql stored procedures.
272 Patch from Alain Williams
274 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
276 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
277 Patch from Andreas Metzler
279 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
280 Patch from Kirill Miazine
282 NM/05 Bugzilla 671: Added umask to procmail example.
284 JJ/03 installed exipick 20100323.0, fixing doc bug
286 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
287 directory. Notification and patch from Dan Rosenberg.
289 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
291 TK/02 Improve log output when DKIM signing operation fails.
293 MH/01 Treat the transport option dkim_domain as a colon separated
294 list, not as a single string, and sign the message with each element,
295 omitting multiple occurences of the same signer.
297 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
298 Bugzilla 985, 986. Patch by Simon Arlott
300 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
301 Patch by Simon Arlott
303 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
304 MBX locking. Notification from Dan Rosenberg.
310 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
312 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
314 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
316 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
318 NM/04 Bugzilla 915: Flag broken perl installation during build.
324 TK/01 Added patch by Johannes Berg that expands the main option
325 "spamd_address" if it starts with a dollar sign.
327 TK/02 Write list of recipients to X-Envelope-Sender header when building
328 the mbox-format spool file for content scanning (suggested by Jakob
331 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
332 (http://www.dcc-servers.net/) support via dccifd. Activated by
333 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
335 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
336 by Mark Daniel Reidel <mr@df.eu>.
338 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
339 When building exim an external PCRE library is now needed -
340 PCRE is a system library on the majority of modern systems.
341 See entry on PCRE_LIBS in EDITME file.
343 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
344 conversation. Added nologin parameter to request.
345 Patch contributed by Kirill Miazine.
347 TF/01 Do not log submission mode rewrites if they do not change the address.
349 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
351 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
352 log files in place. Contributed by Roberto Lima.
354 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
356 TF/03 Bugzilla 615: When checking the local_parts router precondition
357 after a local_part_suffix or local_part_prefix option, Exim now
358 does not use the address's named list lookup cache, since this
359 contains cached lookups for the whole local part.
361 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
362 Robert Millan. Documentation is in experimental-spec.txt.
364 TF/04 Bugzilla 668: Fix parallel build (make -j).
366 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
368 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
369 Patch provided by Jan Srzednicki.
371 TF/05 Leading white space used to be stripped from $spam_report which
372 wrecked the formatting. Now it is preserved.
374 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
375 that they are available at delivery time.
377 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
379 TF/08 TLS error reporting now respects the incoming_interface and
380 incoming_port log selectors.
382 TF/09 Produce a more useful error message if an SMTP transport's hosts
383 setting expands to an empty string.
385 NM/06 Bugzilla 744: EXPN did not work under TLS.
386 Patch provided by Phil Pennock.
388 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
389 Patch provided by Richard Godbee.
391 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
392 acl_smtp_notquit, added index entry.
394 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
395 Patch provided by Eugene Bujak.
397 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
398 accept(). Patch provided by Maxim Dounin.
400 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
401 Patch provided by Phil Pennock.
403 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
405 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
406 Patch provided by Brad "anomie" Jorsch.
408 NM/14 Bugzilla 622: Added timeout setting to transport filter.
409 Patch provided by Dean Brooks.
411 TK/05 Add native DKIM support (does not depend on external libraries).
413 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
414 Patch provided by Graeme Fowler.
416 NM/16 Bugzilla 851: Documentation example syntax fix.
418 NM/17 Changed NOTICE file to remove references to embedded PCRE.
420 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
423 NM/19 Bugzilla 745: TLS version reporting.
424 Patch provided by Phil Pennock.
426 NM/20 Bugzilla 167: bool: condition support.
427 Patch provided by Phil Pennock.
429 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
430 clients. Patch provided by Phil Pennock.
432 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
433 Patch provided by Brad "anomie" Jorsch.
435 NM/23 Bugzilla 687: Fix misparses in eximstats.
436 Patch provided by Heiko Schlittermann.
438 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
439 Patch provided by Heiko Schlittermann.
441 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
442 plus update to original patch.
444 NM/26 Bugzilla 799: Documentation correction for ratelimit.
446 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
447 Patch provided by David Brownlee.
449 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
451 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
453 NM/30 Bugzilla 888: TLS documentation bugfixes.
455 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
457 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
458 Unlike the original bugzilla I have changed all shell scripts in src tree.
460 NM/33 Bugzilla 898: Transport filter timeout fix.
461 Patch by Todd Rinaldo.
463 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
464 Patch by Serge Demonchaux.
466 NM/35 Bugzilla 39: Base64 decode bug fixes.
467 Patch by Jakob Hirsch.
469 NM/36 Bugzilla 909: Correct connect() call in dcc code.
471 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
473 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
475 NM/39 Bugzilla 911: Fixed MakeLinks build script.
481 TK/01 Add preliminary DKIM support. Currently requires a forked version of
482 ALT-N's libdkim that I have put here:
483 http://duncanthrax.net/exim-experimental/
485 Note to Michael Haardt: I had to rename some vars in sieve.c. They
486 were called 'true' and it seems that C99 defines that as a reserved
487 keyword to be used with 'bool' variable types. That means you could
488 not include C99-style headers which use bools without triggering
489 build errors in sieve.c.
491 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
492 as mailq or other aliases. Changed the --help handling significantly
493 to do whats expected. exim_usage() emits usage/help information.
495 SC/01 Added the -bylocaldomain option to eximstats.
497 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
499 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
501 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
507 PH/01 Another patch from the Sieve maintainer.
509 PH/02 When an IPv6 address is converted to a string for single-key lookup
510 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
511 dots are used instead of colons so that keys in lsearch files need not
512 contain colons. This was done some time before quoting was made available
513 in lsearch files. However, iplsearch files do require colons in IPv6 keys
514 (notated using the quote facility) so as to distinguish them from IPv4
515 keys. This meant that lookups for IP addresses in host lists did not work
516 for iplsearch lookups.
518 This has been fixed by arranging for IPv6 addresses to be expressed with
519 colons if the lookup type is iplsearch. This is not incompatible, because
520 previously such lookups could never work.
522 The situation is now rather anomolous, since one *can* have colons in
523 ordinary lsearch keys. However, making the change in all cases is
524 incompatible and would probably break a number of configurations.
526 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
529 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
530 conversion specification without a maximum field width, thereby enabling
531 a rogue spamd server to cause a buffer overflow. While nobody in their
532 right mind would setup Exim to query an untrusted spamd server, an
533 attacker that gains access to a server running spamd could potentially
534 exploit this vulnerability to run arbitrary code as the Exim user.
536 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
537 $primary_hostname instead of what libspf2 thinks the hosts name is.
539 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
540 a directory entry by the name of the lookup key. Previously, if a
541 symlink pointed to a non-existing file or a file in a directory that
542 Exim lacked permissions to read, a lookup for a key matching that
543 symlink would fail. Now it is enough that a matching directory entry
544 exists, symlink or not. (Bugzilla 503.)
546 PH/03 The body_linecount and body_zerocount variables are now exported in the
549 PH/04 Added the $dnslist_matched variable.
551 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
552 This means they are set thereafter only if the connection becomes
555 PH/06 Added the client_condition to authenticators so that some can be skipped
556 by clients under certain conditions.
558 PH/07 The error message for a badly-placed control=no_multiline_responses left
559 "_responses" off the end of the name.
561 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
563 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
564 (without spaces) instead of just copying the configuration text.
566 PH/10 Added the /noupdate option to the ratelimit ACL condition.
568 PH/11 Added $max_received_linelength.
570 PH/12 Added +ignore_defer and +include_defer to host lists.
572 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
573 way in which PCRE > 7.0 is built.
575 PH/14 Implemented queue_only_load_latch.
577 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
578 MAIL command. The effect was to mangle the value on 64-bit systems.
580 PH/16 Another patch from the Sieve maintainer.
582 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
584 PH/18 If a system quota error occurred while trying to create the file for
585 a maildir delivery, the message "Mailbox is full" was not appended to the
586 bounce if the delivery eventually timed out. Change 4.67/27 below applied
587 only to a quota excession during the actual writing of the file.
589 PH/19 It seems that peer DN values may contain newlines (and other non-printing
590 characters?) which causes problems in log lines. The DN values are now
591 passed through string_printing() before being added to log lines.
593 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
594 and InterBase are left for another time.)
596 PH/21 Added message_body_newlines option.
598 PH/22 Guard against possible overflow in moan_check_errorcopy().
600 PH/23 POSIX allows open() to be a macro; guard against that.
602 PH/24 If the recipient of an error message contained an @ in the local part
603 (suitably quoted, of course), incorrect values were put in $domain and
604 $local_part during the evaluation of errors_copy.
610 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
611 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
614 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
615 issue a MAIL command.
617 PH/02 In an ACL statement such as
619 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
621 if a client was not listed at all, or was listed with a value other than
622 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
623 the condition was not true (as it should be), so access was not denied.
624 The bug was that the ! inversion was incorrectly passed on to the second
625 item. This has been fixed.
627 PH/03 Added additional dnslists conditions == and =& which are different from
628 = and & when the dns lookup returns more than one IP address.
630 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
631 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
633 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
634 FSYNC, which compiles an option called disable_fsync that allows for
635 bypassing fsync(). The documentation is heavily laced with warnings.
637 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
639 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
640 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
641 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
642 including adding "make clean"; (3) Added -fPIC when compiling the test
643 dynamically loaded module, to get rid of a warning.
645 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
646 message fails, move_frozen_messages = true and ignore_bounce_errors_after
647 = 0s. The bug is otherwise harmless.
649 PH/07 There was a bug in the dovecot authenticator such that the value of
650 $auth1 could be overwritten, and so not correctly preserved, after a
651 successful authentication. This usually meant that the value preserved by
652 the server_setid option was incorrect.
654 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
656 PH/09 Installed PCRE release 7.0.
658 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
659 run for batched SMTP input. It is now run at the start of every message
660 in the batch. While fixing this I discovered that the process information
661 (output by running exiwhat) was not always getting set for -bs and -bS
662 input. This is fixed, and it now also says "batched" for BSMTP.
664 PH/11 Added control=no_pipelining.
666 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
667 patch, slightly modified), and move the expansion of helo_data till after
668 the connection is made in the smtp transport (so it can use these
671 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
673 PH/14 Added log_selector = +pid.
675 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
677 PH/16 Add ${if forany and ${if forall.
679 PH/17 Added dsn_from option to vary the From: line in DSNs.
681 PH/18 Flush SMTP output before performing a callout, unless control =
682 no_callout_flush is set.
684 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
685 was true (the default) a successful delivery failed to delete the retry
686 item, thus causing premature timeout of the address. The bug is now
689 PH/20 Added hosts_avoid_pipelining to the smtp transport.
691 PH/21 Long custom messages for fakedefer and fakereject are now split up
692 into multiline reponses in the same way that messages for "deny" and
693 other ACL rejections are.
695 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
696 with slight modification.
698 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
699 draft, changing the syntax and factoring some duplicate code".
701 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
702 for deliveries of the second and subsequent messages over the same SMTP
705 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
706 ${reduce, with only minor "tidies".
708 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
710 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
711 expansion side effects.
713 PH/27 When a message times out after an over-quota error from an Exim-imposed
714 quota, the bounce message says "mailbox is full". This message was not
715 being given when it was a system quota that was exceeded. It now should
718 MH/03 Made $recipients available in local_scan(). local_scan() already has
719 better access to the recipient list through recipients_list[], but
720 $recipients can be useful in postmaster-provided expansion strings.
722 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
723 in the case of a MAIL command with additional options following the
724 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
725 were accidentally chopped off.
727 PH/29 SMTP synchronization checks are implemented when a command is read -
728 there is a check that no more input is waiting when there shouldn't be
729 any. However, for some commands, a delay in an ACL can mean that it is
730 some time before the response is written. In this time, more input might
731 arrive, invalidly. So now there are extra checks after an ACL has run for
732 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
733 pipelining has not been advertised.
735 PH/30 MH's patch to allow iscntrl() characters to be list separators.
737 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
738 returned in the SMTP response when smtp_return_error_details was false.
741 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
742 instead of the C I/O that was originally supplied, because problems were
745 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
746 Exim which did not show up earlier: it was assuming that a call to
747 SSL_CTX_set_info_callback() might give an error value. In fact, there is
748 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
749 was a macro that became an assignment, so it seemed to work. This has
750 changed to a proper function call with a void return, hence the compile
751 error. Exim's code has been fixed.
753 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
756 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
758 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
760 PH/37 If a message is not accepted after it has had an id assigned (e.g.
761 because it turns out to be too big or there is a timeout) there is no
762 "Completed" line in the log. When some messages of this type were
763 selected by exigrep, they were listed as "not completed". Others were
764 picked up by some special patterns. I have improved the selection
765 criteria to be more general.
767 PH/38 The host_find_failed option in the manualroute router can now be set
768 to "ignore", to completely ignore a host whose IP address cannot be
769 found. If all hosts are ignored, the behaviour is controlled by the new
770 host_all_ignored option.
772 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
773 homing or because of multiple MX records with /mx) generated more than
774 one IP address, and the following item turned out to be the local host,
775 all the secondary addresses of the first item were incorrectly removed
776 from the list, along with the local host and any following hosts (which
777 is what is supposed to happen).
779 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
780 whoever called Exim into the -H file. In the case of the daemon it was
781 behaving confusingly. When first started, it used values for whoever
782 started the daemon, but after a SIGHUP it used the Exim user (because it
783 calls itself on a restart). I have changed the code so that it now always
786 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
787 message are rejected with the same error (e.g. no authentication or bad
788 sender address), and a DATA command is nevertheless sent (as can happen
789 with PIPELINING or a stupid MUA), the error message that was given to the
790 RCPT commands is included in the rejection of the DATA command. This is
791 intended to be helpful for MUAs that show only the final error to their
794 PH/42 Another patch from the Sieve maintainer.
796 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
797 Eximstats - Fixed some broken HTML links and added missing column headers
799 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
800 columns for Rejects, Temp Rejects, Ham, and Spam rows.
802 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
804 PH/43 Yet another patch from the Sieve maintainer.
806 PH/44 I found a way to check for a TCP/IP connection going away before sending
807 the response to the final '.' that terminates a message, but only in the
808 case where the client has not sent further data following the '.'
809 (unfortunately, this is allowed). However, in many cases there won't be
810 any further data because there won't be any more messages to send. A call
811 to select() can be used: if it shows that the input is "ready", there is
812 either input waiting, or the socket has been closed. An attempt to read
813 the next input character can distinguish the two cases. Previously, Exim
814 would have sent an OK response which the client would never have see.
815 This could lead to message repetition. This fix should cure that, at
816 least in a lot of common cases.
818 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
819 advertised in response to EHLO.
825 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
826 fixed by 4.65/MH/01 (is this a record?) are fixed:
828 (i) An empty string was always treated as zero by the numeric comparison
829 operators. This behaviour has been restored.
831 (ii) It is documented that the numeric comparison operators always treat
832 their arguments as decimal numbers. This was broken in that numbers
833 starting with 0 were being interpreted as octal.
835 While fixing these problems I realized that there was another issue that
836 hadn't been noticed. Values of message_size_limit (both the global option
837 and the transport option) were treated as octal if they started with 0.
838 The documentation was vague. These values are now always treated as
839 decimal, and I will make that clear in the documentation.
845 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
846 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
849 MH/01 Don't check that the operands of numeric comparison operators are
850 integers when their expansion is in "skipping" mode (fixes bug
851 introduced by 4.64-PH/07).
853 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
854 child addresses, Exim now panics and dies. Previously, because the count
855 is held in a short int, deliveries were likely to be lost. As such a
856 large number of recipients for a single message is ridiculous
857 (performance will be very, very poor), I have chosen to impose a limit
858 rather than extend the field.
864 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
865 leftover -K file (the existence of which was triggered by #402).
866 While we were at it, introduced process PID as part of the -K
867 filename. This should rule out race conditions when creating
870 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
871 processing considerably. Previous code took too long for large mails,
872 triggering a timeout which in turn triggers #401.
874 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
875 in the DK code in transports.c. sendfile() is not really portable,
876 hence the _LINUX specificness.
878 TF/01 In the add_headers option to the mail command in an Exim filter,
879 there was a bug that Exim would claim a syntax error in any
880 header after the first one which had an odd number of characters
883 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
884 callout verification, Exim cached a "reject" for the entire domain. This
885 is correct for most verifications, but it is not correct for a recipient
886 verification with use_sender or use_postmaster set, because in that case
887 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
888 case of MAIL FROM:<> rejection from other early rejections (e.g.
889 rejection of HELO). When verifying a recipient using a non-null MAIL
890 address, the cache is ignored if it shows MAIL FROM:<> rejection.
891 Whatever the result of the callout, the value of the domain cache is
892 left unchanged (for any other kind of callout, getting as far as trying
893 RCPT means that the domain itself is ok).
895 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
898 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
899 manifest itself as EPIPE rather than ECONNECT. When tidying away a
900 session, the daemon ignores ECONNECT errors and logs others; it now
901 ignores EPIPE as well.
903 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
904 (quoted-printable decoding).
906 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
907 later the small subsequent patch to fix an introduced bug.
909 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
911 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
913 PH/08 An error is now given if message_size_limit is specified negative.
915 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
916 to be given (somewhat) arbitrary names.
918 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
921 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
922 miscellaneous code fixes
924 PH/10 Added the log_reject_target ACL modifier to specify where to log
927 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
928 hostname. This is wrong, because it relates to the incoming message (and
929 probably the interface on which it is arriving) and not to the outgoing
930 callout (which could be using a different interface). This has been
931 changed to use the value of the helo_data option from the smtp transport
932 instead - this is what is used when a message is actually being sent. If
933 there is no remote transport (possible with a router that sets up host
934 addresses), $smtp_active_hostname is used.
936 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
937 tweaks were necessary in order to get it to work (see also 21 below):
938 (a) The code assumed that strncpy() returns a negative number on buffer
939 overflow, which isn't the case. Replaced with Exim's string_format()
941 (b) There were several signed/unsigned issues. I just did the minimum
942 hacking in of casts. There is scope for a larger refactoring.
943 (c) The code used strcasecmp() which is not a standard C function.
944 Replaced with Exim's strcmpic() function.
945 (d) The code set only $1; it now sets $auth1 as well.
946 (e) A simple test gave the error "authentication client didn't specify
947 service in request". It would seem that Dovecot has changed its
948 interface. Fortunately there's a specification; I followed it and
949 changed what the client sends and it appears to be working now.
951 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
954 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
955 address A is aliased to B and C, where B exists and C does not. Without
956 -v the output is "A verified" because verification stops after a
957 successful redirection if more than one address is generated. However,
958 with -v the child addresses are also verified. Exim was outputting "A
959 failed to verify" and then showing the successful verification for C,
960 with its parentage. It now outputs "B failed to verify", showing B's
961 parentage before showing the successful verification of C.
963 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
964 look up a TXT record in a specific list after matching in a combined
967 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
968 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
969 they consult the DNS. I had assumed they would set it the way they
970 wanted; and indeed my experiments on Linux seem to show that in some
971 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
972 To be on the safe side, however, I have now made the interface to
973 host_find_byname() similar to host_find_bydns(), with an argument
974 containing the DNS resolver options. The host_find_byname() function now
975 sets these options at its start, just as host_find_bydns() does. The smtp
976 transport options dns_qualify_single and dns_search_parents are passed to
977 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
978 of host_find_byname() use the default settings of RES_DEFNAMES
979 (qualify_single) but not RES_DNSRCH (search_parents).
981 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
982 spool_read_header() do less string testing, by means of a preliminary
983 switch on the second character of optional "-foo" lines. (This is
984 overdue, caused by the large number of possibilities that now exist.
985 Originally there were few.) While I was there, I also converted the
986 str(n)cmp tests so they don't re-test the leading "-" and the first
987 character, in the hope this might squeeze out yet more improvement.
989 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
990 flag allowing group syntax was set by the header_syntax check but not
991 turned off, possible causing trouble later; (2) The flag was not being
992 set at all for the header_verify test, causing "group"-style headers to
993 be rejected. I have now set it in this case, and also caused header_
994 verify to ignore an empty address taken from a group. While doing this, I
995 came across some other cases where the code for allowing group syntax
996 while scanning a header line wasn't quite right (mostly, not resetting
997 the flag correctly in the right place). These bugs could have caused
998 trouble for malformed header lines. I hope it is now all correct.
1000 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1001 with the "reply" argument non-NULL. The code, however (which originally
1002 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1003 but it didn't always do it. This confused somebody who was copying the
1004 code for some other use. I have removed all the tests.
1006 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1007 feature that was used to support insecure browsers during the U.S. crypto
1008 embargo. It requires special client support, and Exim is probably the
1009 only MTA that supported it -- and would never use it because real RSA is
1010 always available. This code has been removed, because it had the bad
1011 effect of slowing Exim down by computing (never used) parameters for the
1012 RSA_EXPORT functionality.
1014 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1015 authenticator to fail if there's a tab character in the incoming data
1016 (there should never be unless someone is messing about, as it's supposed
1017 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1018 if the connection is using TLS or if the remote IP is the same as the
1019 local IP, and the "valid-client-cert option" if a client certificate has
1022 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1023 authenticators. This can be used for authorization after authentication
1024 succeeds. (In the case of plaintext, it servers for both authentication
1027 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1028 if any retry times were supplied.
1030 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1031 connection, where there is no client IP address to check. In this
1032 situation, the verify now always succeeds.
1034 PH/25 Applied John Jetmore's -Mset patch.
1036 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1038 PH/27 In a string expansion for a processed (not raw) header when multiple
1039 headers of the same name were present, leading whitespace was being
1040 removed from all of them, but trailing whitespace was being removed only
1041 from the last one. Now trailing whitespace is removed from each header
1042 before concatenation. Completely empty headers in a concatenation (as
1043 before) are ignored.
1045 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1046 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1048 PH/29 [Removed. This was a change that I later backed out, and forgot to
1049 correct the ChangeLog entry (that I had efficiently created) before
1050 committing the later change.]
1052 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1053 address errors (4xx response to RCPT) over the same connection as other
1054 messages routed to the same hosts. Such deliveries are always "forced",
1055 so retry times are not inspected. This resulted in far too many retries
1056 for the affected addresses. The effect occurred only when there were more
1057 hosts than the hosts_max_try setting in the smtp transport when it had
1058 the 4xx errors. Those hosts that it had tried were not added to the list
1059 of hosts for which the message was waiting, so if all were tried, there
1060 was no problem. Two fixes have been applied:
1062 (i) If there are any address or message errors in an SMTP delivery, none
1063 of the hosts (tried or untried) are now added to the list of hosts
1064 for which the message is waiting, so the message should not be a
1065 candidate for sending over the same connection that was used for a
1066 successful delivery of some other message. This seems entirely
1067 reasonable: after all the message is NOT "waiting for some host".
1068 This is so "obvious" that I'm not sure why it wasn't done
1069 previously. Hope I haven't missed anything, but it can't do any
1070 harm, as the worst effect is to miss an optimization.
1072 (ii) If, despite (i), such a delivery is accidentally attempted, the
1073 routing retry time is respected, so at least it doesn't keep
1074 hammering the server.
1076 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1077 in ${readsocket because some servers need this prod.
1079 PH/32 Added some extra debug output when updating a wait-xxx database.
1081 PH/33 The hint "could be header name not terminated by colon", which has been
1082 given for certain expansion errors for a long time, was not being given
1083 for the ${if def:h_colon_omitted{... case.
1085 PH/34 The spec says: "With one important exception, whenever a domain list is
1086 being scanned, $domain contains the subject domain." There was at least
1087 one case where this was not true.
1089 PH/35 The error "getsockname() failed: connection reset by peer" was being
1090 written to the panic log as well as the main log, but it isn't really
1091 panic-worthy as it just means the connection died rather early on. I have
1092 removed the panic log writing for the ECONNRESET error when getsockname()
1095 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1096 runs only) independently of the message's sender address. This meant
1097 that, if the 4xx error was in fact related to the sender, a different
1098 message to the same recipient with a different sender could confuse
1099 things. In particualar, this can happen when sending to a greylisting
1100 server, but other circumstances could also provoke similar problems.
1101 I have changed the default so that the retry time for these errors is now
1102 based a combination of the sender and recipient addresses. This change
1103 can be overridden by setting address_retry_include_sender=false in the
1106 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1107 remote server are returned as part of bounce messages. This was not
1108 happening for LMTP over a pipe (the lmtp transport), but now it is the
1109 same for both kinds of LMTP.
1111 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1112 in header lines that were in fact CNAMEs. This is no longer the case.
1114 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1115 and queue runs started by the daemon processed all messages. This has
1116 been fixed so that -R and -S can now usefully be given with -q<time>.
1118 PH/40 Import PCRE release 6.7 (fixes some bugs).
1120 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1122 PH/42 Give an error if -q is specified more than once.
1124 PH/43 Renamed the variables $interface_address and $interface_port as
1125 $received_ip_address and $received_port, to make it clear that these
1126 values apply to message reception, and not to the outgoing interface when
1127 a message is delivered. (The old names remain recognized, of course.)
1129 PH/44 There was no timeout on the connect() call when using a Unix domain
1130 socket in the ${readsocket expansion. There now is.
1132 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1133 be meaningful with "accept".
1135 SC/01 Eximstats V1.43
1136 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1138 SC/02 Eximstats V1.44
1139 Use a glob alias rather than an array ref in the generated
1140 parser. This improves both readability and performance.
1142 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1143 Collect SpamAssassin and rejection statistics.
1144 Don't display local sender or destination tables unless
1145 there is data to show.
1146 Added average volumes into the top table text output.
1148 SC/04 Eximstats V1.46
1149 Collect data on the number of addresses (recipients)
1150 as well as the number of messages.
1152 SC/05 Eximstats V1.47
1153 Added 'Message too big' to the list of mail rejection
1154 reasons (thanks to Marco Gaiarin).
1156 SC/06 Eximstats V1.48
1157 Mainlog lines which have GMT offsets and are too short to
1158 have a flag are now skipped.
1160 SC/07 Eximstats V1.49 (Alain Williams)
1161 Added the -emptyok flag.
1163 SC/08 Eximstats V1.50
1164 Fixes for obtaining the IP address from reject messages.
1166 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1167 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1168 whitesspace changes from 4.64-PH/27
1170 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1173 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1174 are found, allow negative numbers in numeric criteria)
1176 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1178 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1179 to match changes made in 4.64-PH/43
1181 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1183 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1184 "domain" is actually an IP address, and force a failure. This locks out
1185 those revolvers/nameservers that support "A-for-A" lookups, in
1186 contravention of the specifications.
1188 PH/48 When a host name was looked up from an IP address, and the subsequent
1189 forward lookup of the name timed out, the host name was left in
1190 $sender_host_name, contrary to the specification.
1192 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1193 restricted to single-key lookups, Exim was not diagnosing an error if
1194 * or *@ was used with a query-style lookup.
1196 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1198 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1199 long ago, but noone interested enough thought of it. Let's just say that
1200 the "1.1" means that there are some new functions that weren't there at
1201 some point in the past.
1203 PH/51 Error processing for expansion failure of helo_data from an smtp
1204 transport during callout processing was broken.
1206 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1207 tested/used via the -bh/-bhc/-bs options.
1209 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1210 bug, fixed in subsequent PCRE releases).
1212 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1213 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1215 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1221 SC/01 Use a glob alias rather than an array ref in eximstats generated
1222 parser. This improves both readability and performance.
1224 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1225 Don't display local sender or destination tables in eximstats unless
1226 there is data to show.
1227 Added average volumes into the eximstats top table text output.
1229 SC/03 Collect data on the number of addresses (recipients) as well
1230 as the number of messages in eximstats.
1232 TF/01 Correct an error in the documentation for the redirect router. Exim
1233 does (usually) call initgroups() when daemonizing.
1235 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1236 with consistent privilege compared to when running as a daemon.
1238 TF/03 Note in the spec that $authenticated_id is not set for local
1239 submissions from trusted users.
1241 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1242 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1244 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1245 by adding some example configuration directives to the default
1246 configuration file. A little bit of work is required to uncomment the
1247 directives and define how usernames and passwords are checked, but
1248 there is now a framework to start from.
1250 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1251 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1252 without this. I don't know how relevant this is to other LDAP libraries.
1254 PH/02 Add the verb name to the "unknown ACL verb" error.
1256 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1258 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1260 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1261 directory not expanded when it should be if an expanded home directory
1262 was set for the address (which is overridden by the transport).
1264 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1267 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1268 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1269 because it is too late at that time, and has no effect.
1271 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1272 security issue with \' (bugzilla #107). I could not use the
1273 PQescapeStringConn() function, because it needs a PGconn value as one of
1276 PH/08 When testing addresses using -bt, indicate those final addresses that
1277 are duplicates that would not cause an additional delivery. At least one
1278 person was confused, thinking that -bt output corresponded to deliveries.
1279 (Suppressing duplicates isn't a good idea as you lose the information
1280 about possibly different redirections that led to the duplicates.)
1282 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1283 systems where poll() doesn't work, in particular OS X.
1285 PH/10 Added more information to debugging output for retry time not reached.
1287 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1288 operations in malware.c.
1290 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1293 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1294 syslog_duplication set false, log lines that would normally be written
1295 both the the main log and to the reject log were not written to syslog at
1298 PH/14 In the default configuration, change the use of "message" in ACL warn
1299 statements to "add_header".
1301 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1302 not followed by a command (e.g. "seen endif").
1304 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1305 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1308 PH/17 Added extra conditions to the default value of delay_warning_condition
1312 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1313 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1314 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1317 The Auto-Submitted: and various List- headers are standardised, whereas I
1318 don't think Precedence: ever was.
1320 PH/18 Refactored debugging code in route_finduser() to show more information,
1321 in particular, the error code if getpwnam() issues one.
1323 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1324 This is apparently needed in addition to the PH/07 change above to avoid
1325 any possible encoding problems.
1327 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1328 but not after initializing Perl.
1330 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1331 output them only if debugging. By default they are written stderr,
1332 apparently, which is not desirable.
1334 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1337 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1340 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1342 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1343 authenticated or an ident call has been made. Suppress the default
1344 values for $authenticated_id and $authenticated_sender (but permit -oMai
1345 and -oMas) when testing with -bh.
1347 PH/24 Re-jigged the order of the tests in the default configuration so that the
1348 tests for valid domains and recipients precede the DNS black list and CSA
1349 tests, on the grounds that those ones are more expensive.
1351 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1352 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1353 This bug exists in every version of Exim that I still have, right back to
1356 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1357 However, an attempt to turn on case-sensitivity in a regex key by
1358 including (?-i) didn't work because the subject string was already
1359 lowercased, and the effects were non-intuitive. It turns out that a
1360 one-line patch can be used to allow (?-i) to work as expected.
1366 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1367 other effects) broke the use of negated acl sub-conditions.
1369 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1372 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1373 "Deny" causes Exim to reject the incoming connection with a 554 error.
1374 Unfortunately, if there is a major crisis, such as a disk failure,
1375 tcp-wrappers gives "deny", whereas what one would like would be some
1376 kind of temporary error. A kludge has been added to help with this.
1377 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1378 554 error is used if errno is still zero or contains ENOENT (which occurs
1379 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1382 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1384 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1385 errors. Otherwise a message that provokes a temporary error (when other
1386 messages do not) can cause a whole host to time out.
1388 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1389 addresses were routed directly to files or pipes from a redirect router.
1390 File deliveries just didn't batch; pipe deliveries might have suffered
1393 PH/06 A failure to get a lock for a hints database would erroneously always say
1394 "Failed to get write lock", even when it was really a read lock.
1396 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1397 of 0600. This has been changed to use the value of the lockfile_mode
1398 option (which defaults to 0600).
1400 PH/08 Applied small patch from the Sieve maintainer.
1402 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1403 folder from quota calculations, a direct delivery into this folder messed
1404 up the contents of the maildirsize file. This was because the regex was
1405 used only to exclude .Trash (or whatever) when the size of the mailbox
1406 was calculated. There was no check that a delivery was happening into an
1407 excluded directory. This bug has been fixed by ignoring all quota
1408 processing for deliveries into excluded directories.
1410 PH/10 Added the maildirfolder_create_regex option to appendfile.
1416 PH/01 The code for finding all the local interface addresses on a FreeBSD
1417 system running IPv6 was broken. This may well have applied to all BSD
1418 systems, as well as to others that have similar system calls. The broken
1419 code found IPv4 interfaces correctly, but gave incorrect values for the
1420 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1421 that it would not match correctly against @[] and not recognize the IPv6
1424 PH/02 The ipliteral router was not recognizing addresses of the form user@
1425 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1427 PH/03 Added disable_ipv6.
1429 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1430 decoded form, because it is most often used to construct To: headers
1431 lines in autoreplies, and the decoded form may well be syntactically
1432 invalid. However, $reply_address has leading white space removed, and all
1433 newlines turned into spaces so that the autoreply transport does not
1436 PH/05 If group was specified without a user on a router, and no group or user
1437 was specified on a transport, the group from the router was ignored.
1439 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1440 for visible compile-time settings that can be used to change these
1441 numbers, for those that want even more. Backwards compatibility with old
1442 spool files has been maintained. However, going back to a previous Exim
1443 release will lost any variables that are in spool files.
1445 PH/07 Two small changes when running in the test harness: increase delay when
1446 passing a TCP/IP connection to a new process, in case the original
1447 process has to generate a bounce, and remove special handling of
1448 127.0.0.2 (sic), which is no longer necessary.
1450 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1451 be the same on different OS.
1453 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1456 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1457 whether --show-vars was specified or not
1459 JJ/02 exipick: Added support for new ACL variable spool format introduced
1462 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1463 syntactically invalid From: or Reply-to: line, and a filter used this to
1464 generate an autoreply, and therefore failed to obtain an address for the
1465 autoreply, Exim could try to deliver to a non-existent relative file
1466 name, causing unrelated and misleading errors. What now happens is that
1467 it logs this as a hard delivery error, but does not attempt to create a
1470 PH/11 The exinext utility has a -C option for testing purposes, but although
1471 the given file was scanned by exinext itself; it wasn't being passed on
1472 when Exim was called.
1474 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1475 an end-of-file indication when reading a command response.
1477 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1478 compiled. In many other places in Exim, IPv6 addresses are always
1479 recognized, so I have changed this. It also means that IPv4 domain
1480 literals of the form [IPV4:n.n.n.n] are now always recognized.
1482 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1483 used if the router is not running as root, for example, when verifying at
1484 ACL time, or when using -bh. The debugging output from this situation was
1485 non-existent - all you got was a failure to exec. I have made two
1488 (a) Failures to set uid/gid, the current directory, or a process leader
1489 in a subprocess such as that created by queryprogram now generate
1490 suitable debugging ouput when -d is set.
1492 (b) The queryprogram router detects when it is not running as root,
1493 outputs suitable debugging information if -d is set, and then runs
1494 the subprocess without attempting to change uid/gid.
1496 PH/15 Minor change to Makefile for building test_host (undocumented testing
1499 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1500 additional section of a DNS packet that returns MX or SRV records.
1501 Instead, it always explicitly searches for A/AAAA records. This avoids
1502 major problems that occur when a DNS server includes only records of one
1503 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1504 fixed another bug: if SRV records were looked up and the corresponding
1505 address records were *not* found in the additional section, the port
1506 values from the SRV records were lost.
1508 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1509 using the correct key (the original address) when searching the retry
1510 rules in order to find which one to use for generating the retry hint.
1512 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1513 adding the default one. Similarly, if it contains a Reply-To: header, the
1514 errors_reply_to option, if set, is not used.
1516 PH/19 When calculating a retry time, Exim used to measure the "time since
1517 failure" by looking at the "first failed" field in the retry record. Now
1518 it does not use this if it is later than than the arrival time of the
1519 message. Instead it uses the arrival time. This makes for better
1520 behaviour in cases where some deliveries succeed, thus re-setting the
1521 "first failed" field. An example is a quota failure for a huge message
1522 when small messages continue to be delivered. Without this change, the
1523 "time since failure" will always be short, possible causing more frequent
1524 delivery attempts for the huge message than are intended.
1525 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1527 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1528 $1, $2, $3) because the numerical variables can be reset during some
1529 expansion items (e.g. "match"), thereby losing the authentication data.
1531 PH/21 Make -bV show the size of off_t variables so that the test suite can
1532 decide whether to run tests for quotas > 2G.
1534 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1535 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1536 filecount value is greater than 2G or if a quota value is greater than 2G
1537 on a system where the size of off_t is not greater than 4, a panic error
1540 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1541 never match. The debug and -bh output now contains an explicit error
1542 message indicating a malformed IPv4 address or mask.
1544 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1545 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1546 PH/23 above applies.
1548 PH/25 Do not write to syslog when running in the test harness. The only
1549 occasion when this arises is a failure to open the main or panic logs
1550 (for which there is an explicit test).
1552 PH/26 Added the /no_tell option to "control=freeze".
1554 PH/27 If a host name lookup failed very early in a connection, for example, if
1555 the IP address matched host_lookup and the reverse lookup yielded a name
1556 that did not have a forward lookup, an error message of the form "no IP
1557 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1558 could be logged. Now it outputs the IP address instead of "NULL".
1560 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1561 allows the sender and the authenticated sender to be set when
1562 submitting a message from within Exim. Since child_open_exim() is
1563 documented for local_scan(), the new function should be too.
1565 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1566 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1567 results in an empty string is now treated as unset.
1569 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1571 PH/31 Added sender_verify_fail logging option.
1573 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1574 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1575 tidied the source and removed it altogether.
1577 PH/33 When a queue run was abandoned because the load average was too high, a
1578 log line was always written; now it is written only if the queue_run log
1579 selector is set. In addition, the log line for abandonment now contains
1580 information about the queue run such as the pid. This is always present
1581 in "start" and "stop" lines but was omitted from the "abandon" line.
1583 PH/34 Omit spaces between a header name and the colon in the error message that
1584 is given when verify = headers_syntax fails (if there are lots of them,
1585 the message gets confusing).
1587 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1588 names, as there are now some PTR records that contain slashes. This check
1589 is only to protect against broken name servers that fall over on strange
1590 characters, so the fact that it applies to all lookups doesn't matter.
1592 PH/36 Now that the new test suite is complete, we can remove some of the
1593 special code in Exim that was needed for the old test suite. For example,
1594 sorting DNS records because real resolvers return them in an arbitrary
1595 order. The new test suite's fake resolver always returns records in the
1598 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1599 bounces) except when queue_only is set, to avoid logging races between
1600 the different processes.
1602 PH/38 Panic-die if .include specifies a non-absolute path.
1604 PH/39 A tweak to the "H" retry rule from its user.
1606 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1607 a label. They prevented compilation on older perls.
1609 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1610 a warning to be raised on newish perls.
1612 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1613 on queue. Changes to match documented behaviour of showing count of
1614 messages matching specified criteria.
1616 PH/40 Changed the default ident timeout from 30s to 5s.
1618 PH/41 Added support for the use of login_cap features, on those BSD systems
1619 that have them, for controlling the resources used by pipe deliveries.
1621 PH/42 The content-scanning code uses fopen() to create files in which to put
1622 message data. Previously it was not paying any attention to the mode of
1623 the files. Exim runs with umask(0) because the rest of the code creates
1624 files with open(), and sets the required mode explicitly. Thus, these
1625 files were ending up world-writeable. This was not a big issue, because,
1626 being within the spool directory, they were not world-accessible. I have
1627 created a function called modefopen, which takes an additional mode
1628 argument. It sets umask(777), creates the file, chmods it to the required
1629 mode, then resets the umask. All the relevant calls to fopen() in the
1630 content scanning code have been changed to use this function.
1632 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1633 to 24 hours. This avoids potential overflow problems when processing G
1634 and H retry rules. I suspect nobody ever tinkers with this value.
1636 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1638 PH/45 When the plaintext authenticator is running as a client, the server's
1639 challenges are checked to ensure they are valid base64 strings. By
1640 default, the authentication attempt is cancelled if an invalid string is
1641 received. Setting client_ignore_invalid_base64 true ignores these errors.
1642 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1643 they are received. Thus, the responses can be made to depend on the
1644 challenges. If an invalid string is ignored, an empty string is placed in
1647 PH/46 Messages that are created by the autoreply transport now contains a
1648 References: header, in accordance with RFCs 2822 and 3834.
1650 PH/47 Added authenticated_sender_force to the smtp transport.
1652 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1654 PH/49 Installed latest patch from the Sieve maintainer.
1656 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1657 was also set, the appendfile transport was unnecessarily scanning a
1658 directory of message files (e.g. for maildir delivery) to find the count
1659 of files (along with the size), even though it did not need this
1660 information. It now does the scan only if it needs to find either the
1661 size of the count of files.
1663 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1665 PH/52 Two bugs concerned with error handling when the smtp transport is
1668 (i) Exim was not creating retry information for temporary errors given
1669 for individual recipients after the DATA command when the smtp transport
1670 was used in LMTP mode. This meant that they could be retried too
1671 frequently, and not timed out correctly.
1673 (ii) Exim was setting the flag that allows error details to be returned
1674 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1675 recipients that were returned after the DATA command.
1677 PH/53 This is related to PH/52, but is more general: for any failing address,
1678 when detailed error information was permitted to be returned to the
1679 sender, but the error was temporary, then after the final timeout, only
1680 "retry timeout exceeded" was returned. Now it returns the full error as
1681 well as "retry timeout exceeded".
1683 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1684 do this, and (what is worse) MTAs that accept it.
1686 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1687 will now be deprecated.
1689 PH/56 New os.c-cygwin from the Cygwin maintainer.
1691 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1692 formats (previously only available in exim formats via -bpr, -bpru,
1693 and -bpra. Now also available in native and exiqgrep formats)
1695 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1696 with very large, slow to parse queues
1698 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1700 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1702 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1703 responses to authentication challenges, though it was showing the
1704 challenges; (ii) I've removed the CR characters from the debug output for
1707 PH/58 Allow for the insertion of a newline as well as a space when a string
1708 is turned into more than one encoded-word during RFC 2047 encoding. The
1709 Sieve code now uses this.
1711 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1712 data_4xx, lost_connection, tls_required.
1714 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1715 message was being sent as an SMTP response.
1717 PH/61 Add -l and -k options to exicyclog.
1719 PH/62 When verifying, if an address was redirected to one new address, so that
1720 verification continued, and the new address failed or deferred after
1721 having set something in $address_data, the value of $address_data was not
1722 passed back to the ACL. This was different to the case when no
1723 redirection occurred. The value is now passed back in both cases.
1725 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1726 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1727 use login_cap.h, so on its own it isn't the distinguishing feature. The
1728 new name refers directly to the setclassresources() function.
1730 PH/65 Added configuration files for NetBSD3.
1732 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1734 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1735 is preferred over IPv4.
1737 PH/68 The bounce_return_message and bounce_return_body options were not being
1738 honoured for bounces generated during the reception of non-SMTP messages.
1739 In particular, this applied to messages rejected by the ACL. This bug has
1740 been fixed. However, if bounce_return_message is true and bounce_return_
1741 body is false, the headers that are returned for a non-SMTP message
1742 include only those that have been read before the error was detected.
1743 (In the case of an ACL rejection, they have all been read.)
1745 PH/69 The HTML version of the specification is now built in a directory called
1746 spec_html instead of spec.html, because the latter looks like a path with
1747 a MIME-type, and this confuses some software.
1749 PH/70 Catch two compiler warnings in sieve.c.
1751 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1752 function verify_get_ident() calls ip_connect() to connect a socket, but
1753 if the "connect()" function timed out, ip_connect() used to close the
1754 socket. However, verify_get_ident() also closes the socket later, and in
1755 between Exim writes to the log, which may get opened at this point. When
1756 the socket was closed in ip_connect(), the log could get the same file
1757 descriptor number as the socket. This naturally causes chaos. The fix is
1758 not to close the socket in ip_connect(); the socket should be closed by
1759 the function that creates it. There was only one place in the code where
1760 this was missing, in the iplookup router, which I don't think anybody now
1761 uses, but I've fixed it anyway.
1763 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1764 well as to direct DNS lookups. Otherwise the handling of names in host
1765 lists is inconsistent and therefore confusing.
1771 PH/01 Two changes to the default runtime configuration:
1773 (1) Move the checks for relay_from_hosts and authenticated clients from
1774 after to before the (commented out) DNS black list checks.
1776 (2) Add control=submission to the relay_from_hosts and authenticated
1777 clients checks, on the grounds that messages accepted by these
1778 statements are most likely to be submissions.
1780 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1782 (1) Generate an error if the third argument for the ${prvs expansion is
1785 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1788 (3) Reset the variables that are obtained from the first argument of
1789 ${prvscheck and used in the second argument before leaving the code,
1790 because their memory is reclaimed, so using them afterwards may do
1793 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1794 one (it's much easier than Tom thought :-).
1796 (5) Because of (4), we can now allow for the use of $prvscheck_result
1797 inside the third argument.
1799 PH/03 For some reason, the default setting of PATH when running a command from
1800 a pipe transport was just "/usr/bin". I have changed it to
1803 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1804 anything to be listed in the output from -bV.
1806 PH/05 When a filter generated an autoreply, the entire To: header line was
1807 quoted in the delivery log line, like this:
1809 => >A.N.Other <ano@some.domain> <original@ddress> ...
1811 This has been changed so that it extracts the operative address. There
1812 may be more than one such address. If so, they are comma-separated, like
1815 => >ano@some.domain,ona@other.domain <original@ddress> ...
1817 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1818 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1819 not being looked up in the rDNS to get a host name, Exim was showing the
1820 IP address twice in Received: lines, even though the IP addresses were
1821 identical. For example:
1823 Received: from [1.2.3.4] (helo=[1.2.3.4])
1825 However, if the real host name was known, it was omitting the HELO data
1826 if it matched the actual IP address. This has been tidied up so that it
1827 doesn't show the same IP address twice.
1829 PH/07 When both +timestamp and +memory debugging was on, the value given by
1830 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1831 called by the debug printing, thereby overwriting the timestamp buffer.
1832 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1834 PH/08 When the original message was included in an autoreply transport, it
1835 always said "this is a copy of the message, including all the headers",
1836 even if body_only or headers_only was set. It now gives an appropriate
1839 PH/09 Applied a patch from the Sieve maintainer which:
1841 o fixes some comments
1842 o adds the (disabled) notify extension core
1843 o adds some debug output for the result of if/elsif tests
1844 o points to the current vacation draft in the documentation
1845 and documents the missing references header update
1849 o fixes a bug in processing the envelope test (when testing
1850 multiple envelope elements, the last element determinted the
1853 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1854 Electronic Mail") by including:
1856 Auto-submitted: auto-generated
1858 in the messages that it generates (bounce messages and others, such as
1859 warnings). In the case of bounce messages for non-SMTP mesages, there was
1860 also a typo: it was using "Auto_submitted" (underscore instead of
1861 hyphen). Since every message generated by Exim is necessarily in response
1862 to another message, thes have all been changed to:
1864 Auto-Submitted: auto-replied
1866 in accordance with these statements in the RFC:
1868 The auto-replied keyword:
1870 - SHOULD be used on messages sent in direct response to another
1871 message by an automatic process,
1873 - MUST NOT be used on manually-generated messages,
1875 - MAY be used on Delivery Status Notifications (DSNs) and Message
1876 Disposition Notifications (MDNs),
1878 - MUST NOT be used on messages generated by automatic or periodic
1879 processes, except for messages which are automatic responses to
1882 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1883 to the default Received: header definition.
1885 PH/12 Added log selector acl_warn_skipped (default on).
1887 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1888 variables because (a) they are in the wrong storage pool and (b) even if
1889 they were copied, it wouldn't work properly because of the caching.
1891 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1892 checking when decoding. Apparently there are clients that generate
1893 overlong encoded strings. Why am I not surprised?
1895 PH/15 If the first argument of "${if match_address" was not empty, but did not
1896 contain an "@" character, Exim crashed. Now it writes a panic log message
1897 and treats the condition as false.
1899 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1901 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1902 extension "envelope-auth". The code is finished and in agreement with
1903 other implementations, but there is no documentation so far and in fact,
1904 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1905 not changing the active code.
1907 Print executed "if" and "elsif" statements when debugging is used. This
1908 helps a great deal to understand what a filter does.
1910 Document more things not specified clearly in RFC3028. I had all this
1911 sorted out, when out of a sudden new issues came to my mind. Oops."
1913 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1916 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1917 canonical form (as documented). However, after a host name lookup from
1918 the IP address, check_host() was doing a simple string comparison with
1919 addresses acquired from the DNS when checking that the found name did
1920 have the original IP as one of its addresses. Since any found IPv6
1921 addresses are likely to be in abbreviated form, the comparison could
1922 fail. Luckily, there already exists a function for doing the comparison
1923 by converting both addresses to binary, so now that is used instead of
1924 the text comparison.
1926 PH/20 There was another similar case to PH/19, when a complete host name was
1927 given in a host list; looking up its IP address could give an abbreviated
1928 form, whereas the current host's name might or might not be abbreviated.
1929 The same fix has been applied.
1935 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1936 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1939 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1940 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1942 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1944 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1945 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1946 Also, for "Auto-Submitted", treat anything other than "no" as
1947 non-personal, in accordance with RFC 3834. (Previously it treated
1948 anything starting "auto-" as non-personal.)
1950 TF/01 The control=submission/name=... option had a problem with syntax
1951 errors if the name included a slash character. The /name= option
1952 now slurps the rest of the string, so it can include any characters
1953 but it must come last in the list of options (after /sender_retain
1956 PH/05 Some modifications to the interface to the fake nameserver for the new
1964 TK/01 Added the "success_on_redirect" address verification option. See
1965 NewStuff for rationale and an example.
1967 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1969 PH/02 Patch to exigrep to allow it to work on syslog lines.
1971 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1972 fread() to skip over the body file's header line, because in Cygwin the
1973 header line is locked and is inaccessible.
1975 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1976 co-exist for some time) to make it clear that it is the Exim ID that is
1977 referenced, not the Message-ID: header line.
1979 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1980 string_format() function, because snprintf() does not exist on all
1983 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1984 ${stat: expansion item.
1986 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1987 protocol synchronization error", to keep the pedants happy.
1989 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1990 well as for IRIX systems, when gcc is being used. See the host.c source
1993 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1995 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1998 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1999 transport and to the smtp transport in LMTP mode.
2001 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2003 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2004 run a filter in a subprocess. This could lead to confusion in subsequent
2005 lookups in the parent process. There should also be a search_tidyup() at
2006 the end of the subprocess.
2008 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2009 only if the host matched helo_try_verify_hosts, which caused the
2010 verification to occur when the EHLO/HELO command was issued. The ACL just
2011 tested the remembered result. Now, if a previous verification attempt has
2012 not happened, "verify = helo" does it there and then.
2014 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2016 TK/03 Fix log output including CR from clamd.
2018 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2019 exist provoked a memory error which could cause a segfault.
2021 PH/15 Installed PCRE 6.2
2023 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2025 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2026 of the problem. Specifically, suggested +O2 rather than +O1 for the
2029 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2031 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2032 record, the port was not correctly logged when the outgoing_port log
2033 selector was set (it logged the transort's default port).
2035 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2036 fallback_hosts, and "hosts" in the smtp transport.
2038 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2039 host errors such as "Connection refused".
2041 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2042 authentication with radiusclient 0.4.9:
2044 - Error returned from rc_read_config was caught wrongly
2045 - Username/password not passed on to radius server due to wrong length.
2047 The presumption is that some radiusclient API changes for 4.51/PH/17
2048 were not taken care of correctly. The code is still untested by me (my
2049 Linux distribution still has 0.3.2 of radiusclient), but it was
2050 contributed by a Radius user.
2052 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2053 expanding the "port" option of the smtp transport.
2055 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2056 while reading a MIME header. Thanks to Tom Hughes for a patch.
2058 PH/24 Include config.h inside local_scan.h so that configuration settings are
2061 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2062 that in an ACL for RCPT (for example), you can examine exactly what was
2065 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2066 commands, but it was not correctly comparing the address with the actual
2067 client host address. Thus, it would show the EHLO address in Received:
2068 header lines when this was not necessary.
2070 PH/27 Added the % operator to ${eval:}.
2072 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2073 it should be ignoring failures (because with -C, for example, it has lost
2074 privilege). It wasn't ignoring creation failures other than "already
2077 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2080 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2081 because an input line was too long, either on its own, or by virtue of
2082 too many continuations, the temporary file was not being removed, and the
2083 return code was incorrect.
2085 PH/31 Missing "BOOL" in function definition in filtertest.c.
2087 PH/32 Applied Sieve patches from the maintainer.
2089 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2091 PH/33 Added "verify = not_blind".
2093 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2094 Local/Makefile (with some defaults set). These are used in built scripts
2095 such as exicyclog, but they have never been used in the exim_install
2096 script (though there are many overriding facilities there). I have
2097 arranged that the exim_install script now takes note of these two
2100 PH/35 Installed configuration files for Dragonfly.
2102 PH/36 When a locally submitted message by a trusted user did not contain a
2103 From: header, and the sender address was obtained from -f or from an SMTP
2104 MAIL command, and the trusted user did not use -F to supply a sender
2105 name, $originator_name was incorrectly used when constructing a From:
2106 header. Furthermore, $originator_name was used for submission mode
2107 messages from external hosts without From: headers in a similar way,
2108 which is clearly wrong.
2110 PH/37 Added control=suppress_local_fixups.
2112 PH/38 When log_selector = +received_sender was set, and the addition of the
2113 sender made the log line's construction buffer exactly full, or one byte
2114 less than full, an overflow happened when the terminating "\n" was
2117 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2118 when the result of a list match is failure because a DNS lookup failed.
2120 PH/40 RM_COMMAND is now used in the building process.
2122 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2123 the "build-* directories that it finds.
2125 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2126 address in a domain literal was a prefix of an interface address.
2128 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2129 when verifying a sender address, unless rewrite_headers is false.
2131 PH/44 Wrote a long comment about why errors_to addresses are verified as
2132 recipients, not senders.
2134 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2135 the ratelimit ACL was added.
2137 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2139 PH/46 Added extra information about PostgreSQL errors to the error string.
2141 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2142 avoiding the need to install special zones in a real server. This is
2143 backwards compatible; if it can't find the fake resolver, it drops back.
2144 Thus, both old and new test suites can be run.
2146 TF/02 Added util/ratelimit.pl
2148 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2149 clock is set back in time.
2151 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2152 Candler <B.Candler@pobox.com>.
2154 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2155 OK for addresses that are the result of redirections.
2157 PH/48 A number of further additions for the benefit of the new test suite,
2158 including a fake gethostbyname() that interfaces to the fake DNS resolver
2161 TF/06 The fix for widen_domains has also been applied to qualify_single and
2162 search_parents which are the other dnslookup options that can cause
2165 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2168 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2170 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2171 service" if it encountered an MX record with an empty target hostname.
2172 The message is now "an MX or SRV record indicated no SMTP service".
2174 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2175 if the DNS of the sending site is misconfigured. This is quite a
2176 common situation. This change restores the behaviour of treating a
2177 helo verification defer as a failure.
2179 PH/51 If self=fail was set on a router, the bounce message did not include the
2180 actual error message.
2186 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2188 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2189 command itself ended in error, the underlying message about the transport
2190 filter timeout was being overwritten with the pipe command error. Now the
2191 underlying error message should be appended to the second error message.
2193 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2195 PH/02 Reduce the amount of output that "make" produces by default. Full output
2196 can still be requested.
2198 PH/03 The warning log line about a condition test deferring for a "warn" verb
2199 was being output only once per connection, rather than after each
2200 occurrence (because it was using the same function as for successful
2201 "warn" verbs). This seems wrong, so I have changed it.
2203 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2204 it should not have, which might have caused a crash in the right
2205 circumstances, but probably never did.
2207 PH/04 Installed a modified version of Tony Finch's patch to make submission
2208 mode fix the return path as well as the Sender: header line, and to
2209 add a /name= option so that you can make the user's friendly name appear
2212 TF/03 Added the control = fakedefer ACL modifier.
2214 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2215 Mark Lowes for thorough testing.
2217 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2219 TK/03 Merged latest SRS patch from Miles Wilton.
2221 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2222 with the definition in sysexits.h (which is #included earlier).
2223 Fortunately, Exim does not actually use EX_OK. The code used to try to
2224 preserve the sysexits.h value, by assumimg that macro definitions were
2225 scanned for macro replacements. I have been disabused of this notion,
2226 so now the code just undefines EX_OK before #including unistd.h.
2228 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2229 in the smtp transport. When a block could not be written in a single
2230 write() function, the timeout was being re-applied to each part-write.
2231 This seems wrong - if the receiver was accepting one byte at a time it
2232 would take for ever. The timeout is now adjusted when this happens. It
2233 doesn't have to be particularly precise.
2235 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2236 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2238 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2239 without a domain if the check to <postmaster@domain> fails.
2241 SC/01 Eximstats: added -xls and the ability to specify output files
2242 (patch written by Frank Heydlauf).
2244 SC/02 Eximstats: use FileHandles for outputing results.
2246 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2248 SC/04 Eximstats: fixed display of large numbers with -nvr option
2250 SC/05 Eximstats: fixed merging of reports with empty tables.
2252 SC/06 Eximstats: added the -include_original_destination flag
2254 SC/07 Eximstats: removed tabs and trailing whitespace.
2256 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2258 TK/06 MBOX spool code: Add real "From " MBOX separator line
2259 so the .eml file is really in mbox format (even though
2260 most programs do not really care). Patch from Alex Miller.
2262 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2263 The latter is generated from $received_to and is only set if the
2264 message has one envelope recipient. SA can use these headers,
2265 obviously out-of-the-box. Patch from Alex Miller.
2267 PH/08 The ${def test on a variable was returning false if the variable's
2268 value was "0", contrary to what the specification has always said!
2269 The result should be true unless the variable is empty.
2271 PH/09 The syntax error of a character other than { following "${if
2272 def:variable_name" (after optional whitespace) was not being diagnosed.
2273 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2274 accidental colon was present, for example, could give incorrect results.
2276 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2277 result is used (not including appendfile, where other changes are about
2280 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2281 This involved changing a lot of size variables from int to off_t. It
2282 should work with maildirs and everything.
2284 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2285 spamd dying while we are connected to it.
2287 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2290 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2291 function for BDB 4.3.
2293 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2295 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2296 the sources for PCRE have been re-arranged and more files are now
2299 PH/15 The code I had for printing potentially long long variables in PH/11
2300 above was not the best (it lost precision). The length of off_t variables
2301 is now inspected at build time, and an appropriate printing format (%ld
2302 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2303 to be "long long int" or "long int". This is needed for the internal
2304 formatting function string_vformat().
2306 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2307 the configuration file to be ":syslog", then the script "guesses" where
2308 the logs files are, rather than using the compiled in default. In our
2309 case the guess is not the same as the compiled default, so the script
2310 suddenly stopped working when I started to use syslog. The patch checks
2311 to see if log_file_path is "". If so, it attempts to read it from exim
2312 with no configuration file to get the compiled in version, before it
2313 falls back to the previous guessing code."
2315 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2316 implementing BATV in an Exim configuration. See NewStuff for the gory
2319 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2320 Makefile that are specific to HP-UX.
2322 PH/18 If the "use_postmaster" option was set for a recipient callout together
2323 with the "random" option, the postmaster address was used as the MAIL
2324 FROM address for the random test, but not for the subsequent recipient
2325 test. It is now used for both.
2327 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2328 patch removes a few documentation additions to RFC 3028, because the
2329 latest draft now contains them. It adds the new en;ascii-case comparator
2330 and a new error check for 8bit text in MIME parts. Comparator and
2331 require names are now matched exactly. I enabled the subaddress
2332 extension, but it is not well tested yet (read: it works for me)."
2334 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2335 rework some of the code of TK/09 above to avoid the hardwired use of
2336 "%lld" and "long long". Replaced the call to snprintf() with a call to
2339 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2340 records point to non-existent hosts", "retry timeout exceeded", and
2341 "retry time not reached for any host after a long failure period".
2343 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2344 experimental DomainKeys support:
2346 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2347 (2) On an error such as an illegally used "control", the wrong name for
2348 the control was given.
2350 These problems did NOT occur unless DomainKeys support was compiled.
2352 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2354 PH/24 Added ${if match_ip condition.
2356 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2357 databases so that it will be absolutely obvious if a crash occurs in the
2358 DB library. This is a regular occurrence (often caused by mis-matched
2361 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2362 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2363 picked up on a user's system that detects such things. There doesn't seem
2364 to be a gcc warning option for this - only an attribute that has to be
2365 put on the function's prototype. It seems that in Fedora Core 4 they have
2366 set this on a number of new functions. No doubt there will be more in due
2369 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2370 specify a transport. However, if an address that was verified by such a
2371 router was the subject of a callout, Exim crashed because it tried to
2372 read the rcpt_include_affixes from the non-existent transport. Now it
2373 just assumes that the setting of that option is false. This bug was
2374 introduced by 4.51/PH/31.
2376 PH/28 Changed -d+all to exclude +memory, because that information is very
2377 rarely of interest, but it makes the output a lot bigger. People tend to
2378 do -d+all out of habit.
2380 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2381 code in os-type was giving problems when libc.so lives in lib64, like on
2384 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2385 aren't the modern standard, and it seems that some systems' include files
2386 don't always have them. Exim was already checking for some of the newer
2387 ones like T_AAAA, and defining it itself. I've added checks for all the
2388 record types that Exim uses.
2390 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2391 not automatically generating a new one, as it is supposed to. This
2392 prevented TLS from working. If the file did exist, but contained invalid
2393 data, a new version was generated, as expected. It was only the case of a
2394 non-existent file that was broken.
2396 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2397 with a change in libdomainkeys > 0.64.
2399 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2400 from DNS. If the selector record carries the flag, it now has
2401 precedence over the domain-wide flag.
2403 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2405 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2406 the use of a port name that isn't defined in /etc/services) occurred, the
2407 message was deferred as in a normal delivery, and thus remained on the
2408 spool, instead of being failed because of the mua_wrapper setting. This
2409 is now fixed, and I tidied up some of the mua_wrapper messages at the
2412 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2413 the messages in a hash of arrays rather than using individual hashes.
2414 This is a bit cleaner and results in dramatic memory savings, albeit
2415 at a slight CPU cost.
2417 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2418 as requested by Marc Sherman.
2420 SC/10 Eximstats: added histograms for user specified patterns as requested
2423 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2425 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2426 fopen() in the content-scanning modules that did not already have it.
2432 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2433 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2435 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2437 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2439 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2440 to test Sieve filters that use "vacation".
2442 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2443 that changes the way the GnuTLS parameters are stored in the cache file.
2444 The new format can be generated externally. For backward compatibility,
2445 if the data in the cache doesn't make sense, Exim assumes it has read an
2446 old-format file, and it generates new data and writes a new file. This
2447 means that you can't go back to an older release without removing the
2450 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2451 work if there are any delivery delays because "one_time" forces the
2452 parent to be marked "delivered", so its unseen clone is never tried
2453 again. For this reason, Exim now forbids the simultaneous setting of
2456 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2457 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2458 entry if you want to know the details. The fix, however, neglected to
2459 consider the case when local delivery batching is involved. The test for
2460 "previously delivered" was not happening when checking to see if an
2461 address could be batched with a previous (undelivered) one; under
2462 certain circumstances this could lead to multiple deliveries to the same
2465 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2466 in its include files, and this causes problems building Exim.
2468 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2469 header_syntax) but Exim was just ignoring anything given after a slash.
2470 In particular, this caused confusion with an attempt to use "verify =
2471 reverse_host_lookup/defer_ok". An error is now given when options are
2472 supplied for verify items that do not have them. (Maybe reverse_host_
2473 lookup should have a defer_ok option, but that's a different point.)
2475 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2476 defined by RFC 821) to 2048, because there were problems with some AUTH
2477 commands, and RFC 1869 says the size should be increased for extended
2478 SMTP commands that take arguments.
2480 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2483 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2484 "unknown" error; now it says that the functionality isn't in the binary.
2486 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2487 an address' error message when a string expansion fails (syntax or
2488 whatever). Otherwise the password may appear in the log. Following change
2489 PH/42 below, there is no longer a chance of it appearing in a bounce
2492 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2494 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2495 of its addresses was ever tried. (Bugzilla bug #2.)
2497 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2498 the result incorrectly in the debug output. (It correctly added a newline
2499 to what was transported.)
2501 TF/01 Added $received_time.
2503 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2504 commented out examples of how to interface to a virus scanner and to
2505 SpamAssassin. Also added commented examples of av_scanner and
2506 spamd_address settings.
2508 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2509 and controls are allowed in which ACLs. There were a couple of minor
2510 errors. Some of the entries in the conditions table (which is a table of
2511 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2512 negation of where the condition IS allowed.
2514 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2516 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2517 header file does not have a version number, so I've had to invent a new
2518 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2519 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2520 radiusclient), but it was contributed by a Radius user.
2522 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2523 files or directories, for OpenSSL.
2525 PH/19 When an Exim process that is running as root has to create an Exim log
2526 file, it does so in a subprocess that runs as exim:exim so as to get the
2527 ownership right at creation (otherwise, other Exim processes might see
2528 the file with the wrong ownership). There was no test for failure of this
2529 fork() call, which would lead to the process getting stuck as it waited
2530 for a non-existent subprocess. Forks do occasionally fail when resources
2531 run out. I reviewed all the other calls to fork(); they all seem to check
2534 PH/20 When checking for unexpected SMTP input at connect time (before writing
2535 the banner), Exim was not dealing correctly with a non-positive return
2536 from the read() function. If the client had disconnected by this time,
2537 the result was a log entry for a synchronization error with an empty
2538 string after "input=" when read() returned zero. If read() returned -1
2539 (an event I could not check), uninitialized data bytes were printed.
2540 There were reports of junk text (parts of files, etc) appearing after
2543 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2545 PH/22 Added support for macro redefinition, and (re)definition in between
2546 driver and ACL definitions.
2548 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2549 forgetting to use the resulting value; it was using the unexpanded value.
2551 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2552 hadn't been configured. The fix is from Juergen Kreileder, who
2553 understands it better than I do:
2555 "Here's what I see happening with three configured cyrus_sasl
2556 authenticators configured (plain, login, cram-md5):
2558 On startup auth_cyrus_sasl_init() gets called for each of these.
2559 This means three calls to sasl_listmech() without a specified mech_list.
2560 => SASL tests which mechs of all available mechs actually work
2561 => three warnings about OTP not working
2562 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2564 With the patch, sasl_listmech() also gets called three times. But now
2565 SASL's mech_list option is set to the server_mech specified in the the
2566 authenticator. Or in other words, the answer from sasl_listmech()
2567 gets limited to just the mech you're testing for (which is different
2569 => the return list contains just 'plain' or 'login', 'cram-md5' or
2570 nothing depending on the value of ob->server_mech.
2572 I've just tested the patch: Authentication still works fine,
2573 unavailable mechs specified in the exim configuration are still
2574 caught, and the auth.log warnings about OTP are gone."
2576 PH/25 When debugging is enabled, the contents of the command line are added
2577 to the debugging output, even when log_selector=+arguments is not
2580 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2581 answer is "GNU", and only if the return is "GNU/something" is the answer
2584 PH/27 $acl_verify_message is now set immediately after the failure of a
2585 verification in an ACL, and so is available in subsequent modifiers. In
2586 particular, the message can be preserved by coding like this:
2588 warn !verify = sender
2589 set acl_m0 = $acl_verify_message
2591 Previously, $acl_verify_message was set only while expanding "message"
2592 and "log_message" when a very denied access.
2594 PH/28 Modified OS/os.c-Linux with
2596 -#ifndef OS_LOAD_AVERAGE
2597 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2599 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2600 nomenclature these days.)
2602 PH/29 Installed patch from the Sieve maintainer that adds the options
2603 sieve_useraddress and sieve_subaddress to the redirect router.
2605 PH/30 In these circumstances:
2606 . Two addresses routed to the same list of hosts;
2607 . First host does not offer TLS;
2608 . First host accepts first address;
2609 . First host gives temporary error to second address;
2610 . Second host offers TLS and a TLS session is established;
2611 . Second host accepts second address.
2612 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2613 and peerdn, if requested) that were in fact used only for the second
2616 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2617 attention to any local part prefix or suffix that was matched by the
2618 router that accepted the address. It now behaves in the same way as it
2619 does for delivery: the affixes are removed from the local part unless
2620 rcpt_include_affixes is set on the transport.
2622 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2623 timeout during the DATA phase of an incoming message.
2625 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2626 applied a patch sanctioned by the Sieve maintainer.
2628 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2629 the uid or gid is negative. A case of a negative gid caused this to be
2630 noticed. The fix allows for either to be negative.
2632 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2633 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2636 PH/36 The change PH/12 above was broken. Fixed it.
2638 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2639 the grounds that routing the same address twice would always produce the
2640 same answer. This might have been true once, but it is certainly no
2641 longer true now. Routing a child address may depend on the previous
2642 routing that produced that child. Some complicated redirection strategies
2643 went wrong when messages had multiple recipients, and made Exim's
2644 behaviour dependent on the order in which the addresses were given.
2646 I have moved the duplicate checking until after the routing is complete.
2647 Exim scans the addresses that are assigned to local and remote
2648 transports, and removes any duplicates. This means that more work will be
2649 done, as duplicates will always all be routed, but duplicates are
2650 presumably rare, so I don't expect this is of any significance.
2652 For deliveries to pipes, files, and autoreplies, the duplicate checking
2653 still happens during the routing process, since they are not going to be
2656 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2657 It corrects a timeout issue with spamd. This is Ian's comment: "The
2658 background is that sometimes spamd either never reads data from a
2659 connection it has accepted, or it never writes response data. The exiscan
2660 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2661 blindly assumes that writes won't block so it may never time out."
2663 PH/39 Allow G after quota size as well as K and M.
2665 PH/40 The value set for $authenticated_id in an authenticator may not contain
2666 binary zeroes or newlines because the value is written to log lines and
2667 to spool files. There was no check on this. Now the value is run through
2668 the string_printing() function so that such characters are converted to
2669 printable escape sequences.
2671 PH/41 $message_linecount is a new variable that contains the total number of
2672 lines in the message. Compare $body_linecount, which is the count for the
2675 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2676 bounce and delay warning messages, except in certain special cases, which
2679 (a) An SMTP error message from a remote host;
2680 (b) A message specified in a :fail: redirection;
2681 (c) A message specified in a "fail" command in a system filter;
2682 (d) A message specified in a FAIL return from the queryprogram router;
2683 (e) A message specified by the cannot_route_message router option.
2685 In these cases only, Exim does include the error details in bounce and
2686 warning messages. There are also a few cases where bland messages such
2687 as "unrouteable address" or "local delivery error" are given.
2689 PH/43 $value is now also set for the "else" part of a ${run expansion.
2691 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2692 being worked on, but at least Exim now implements the latest version to
2695 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2696 process to complete was treated as a delivery failure, a timeout while
2697 writing the message to the pipe was logged, but erroneously treated as a
2698 successful delivery. Such timeouts include transport filter timeouts. For
2699 consistency with the overall process timeout, these timeouts are now
2700 treated as errors, giving rise to delivery failures by default. However,
2701 there is now a new Boolean option for the pipe transport called
2702 timeout_defer, which, if set TRUE, converts the failures into defers for
2703 both kinds of timeout. A transport filter timeout is now identified in
2706 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2707 systems where "make" and "gmake" are different, calling "gmake" at top
2708 level broke things. I've arranged for the value of $(MAKE) to be passed
2709 from the Makefile to this script so that it can call the same version of
2713 A note about Exim versions 4.44 and 4.50
2714 ----------------------------------------
2716 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2717 changes of various kinds. As a consequence, a big documentation update was
2718 needed. This delayed the release for rather longer than seemed good, especially
2719 in the light of a couple of (minor) security issues. Therefore, the changes
2720 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2721 release. So 4.44 and 4.50 are in effect two different branches that both start
2724 I have left the 4.50 change log unchanged; it contains all the changes since
2725 4.43. The change log for 4.44 is below; many of its items are identical to
2726 those for 4.50. This seems to be the most sensible way to preserve the
2727 historical information.
2733 1. Minor wording change to the doc/README.SIEVE file.
2735 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2736 computation of the current number of files was incorrect.
2738 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2739 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2742 4. Give more explanation in the error message when the command for a transport
2743 filter fails to execute.
2745 5. There are several places where Exim runs a non-Exim command in a
2746 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2747 was being done only for the command run by the queryprogram router. It is
2748 now done for all such subprocesses. The other cases are: ${run, transport
2749 filters, and the commands run by the lmtp and pipe transports.
2751 6. Added CONFIGURE_GROUP build-time option.
2753 7. Some older OS have a limit of 256 on the maximum number of file
2754 descriptors. Exim was using setrlimit() to set 1000 as a large value
2755 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2756 systems. I've change it so that if it can't get 1000, it tries for 256.
2758 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2759 was an oversight, and furthermore, ever since the addition of extra
2760 controls (e.g. 4.43/32), the checks on when to allow different forms of
2761 "control" were broken. There should now be diagnostics for all cases when a
2762 control that does not make sense is encountered.
2764 9. Added the /retain_sender option to "control=submission".
2766 10. $recipients is now available in the predata ACL (oversight).
2768 11. Tidy the search cache before the fork to do a delivery from a message
2769 received from the command line. Otherwise the child will trigger a lookup
2770 failure and thereby defer the delivery if it tries to use (for example) a
2771 cached ldap connection that the parent has called unbind on.
2773 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2774 of $address_data from the recipient verification was clobbered by the
2775 sender verification.
2777 13. The value of address_data from a sender verification is now available in
2778 $sender_address_data in subsequent conditions in the ACL statement.
2780 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2782 15. Added a new option "connect=<time>" to callout options, to set a different
2785 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2786 was its contents. (It was OK if the option was not defined at all.)
2788 17. A "Completed" log line is now written for messages that are removed from
2789 the spool by the -Mrm option.
2791 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2792 information about exactly what failed.
2794 19. Added -dd to debug only the daemon process.
2796 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2797 handles timeouts, both on the server side and network timeouts. Renamed the
2798 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2800 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2801 It is now set to "smtps".
2803 22. $host_address is now set to the target address during the checking of
2804 ignore_target_hosts.
2806 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2807 being passed; this would have caused $sender_host_name to have been used if
2808 matching the list had actually called for a host name (not very likely,
2809 since this list is usually IP addresses). A host name is now passed as
2812 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2813 code that specifies a non-restarting handler (typically sigaction() in
2814 modern systems) in an attempt to fix a rare and obscure crash bug.
2816 25. Narrowed the window for a race in the daemon that could cause it to ignore
2817 SIGCHLD signals. This is not a major problem, because they are used only to
2818 wake it up if nothing else does.
2820 26. A malformed maildirsize file could cause Exim to calculate negative values
2821 for the mailbox size or file count. Odd effects could occur as a result.
2822 The maildirsize information is now recalculated if the size or filecount
2825 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2826 support for a long time. Removed HAVE_SYS_VFS_H.
2828 28. Installed the latest version of exipick from John Jetmore.
2830 29. In an address list, if the pattern was not a regular expression, an empty
2831 subject address (from a bounce message) matched only if the pattern was an
2832 empty string. Non-empty patterns were not even tested. This was the wrong
2833 because it is perfectly reasonable to use an empty address as part of a
2834 database query. An empty address is now tested by patterns that are
2835 lookups. However, all the other forms of pattern expect the subject to
2836 contain a local part and a domain, and therefore, for them, an empty
2837 address still always fails if the pattern is not itself empty.
2839 30. Exim went into a mad DNS loop when attempting to do a callout where the
2840 host was specified on an smtp transport, and looking it up yielded more
2841 than one IP address.
2843 31. Re-factored the code for checking spool and log partition space into a
2844 function that finds that data and another that does the check. The former
2845 is then used to implement four new variables: $spool_space, $log_space,
2846 $spool_inodes, and $log_inodes.
2848 32. The RFC2047 encoding function was originally intended for short strings
2849 such as real names; it was not keeping to the 75-character limit for
2850 encoded words that the RFC imposes. It now respects the limit, and
2851 generates multiple encoded words if necessary. To be on the safe side, I
2852 have increased the buffer size for the ${rfc2047: expansion operator from
2855 33. It is now permitted to omit both strings after an "if" condition; if the
2856 condition is true, the result is "true". As before, when the second string
2857 is omitted, a false condition yields an empty string. This makes it less
2858 cumbersome to write custom ACL and router conditions.
2860 34. Failure to deliver a bounce message always caused it to be frozen, even if
2861 there was an errors_to setting on the router. The errors_to setting is now
2864 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2865 canonical form (fully expanded) before being placed in
2866 $sender_host_address.
2868 36. The table in the code that translates DNS record types into text (T_A to
2869 "A" for instance) was missing entries for NS and CNAME. It is just possible
2870 that this could have caused confusion if both these types were looked up
2871 for the same domain, because the text type is used as part of Exim's
2872 per-process caching. But the chance of anyone hitting this buglet seems
2875 37. The dnsdb lookup has been extended in a number of ways.
2877 (1) There is a new type, "zns", which walks up the domain tree until it
2878 finds some nameserver records. It should be used with care.
2880 (2) There is a new type, "mxh", which is like "mx" except that it returns
2881 just the host names, not the priorities.
2883 (3) It is now possible to give a list of domains (or IP addresses) to be
2884 looked up. The behaviour when one of the lookups defers can be
2885 controlled by a keyword.
2887 (4) It is now possible to specify the separator character for use when
2888 multiple records are returned.
2890 38. The dnslists ACL condition has been extended: it is now possible to supply
2891 a list of IP addresses and/or domains to be looked up in a particular DNS
2894 39. Added log_selector=+queue_time_overall.
2896 40. When running the queue in the test harness, wait just a tad after forking a
2897 delivery process, to get repeatability of debugging output.
2899 41. Include certificate and key file names in error message when GnuTLS fails
2900 to set them up, because the GnuTLS error message doesn't include the name
2901 of the failing file when there is a problem reading it.
2903 42. Allow both -bf and -bF in the same test run.
2905 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2907 44. The "Exiscan patch" is now merged into the mainline Exim source.
2909 45. Sometimes the final signoff response after QUIT could fail to get
2910 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2911 < 0 before doing a fflush(). This bug looks as though it goes back to the
2912 introduction of TLS in release 3.20, but "sometimes" must have been rare
2913 because the tests only now provoked it.
2915 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2916 (this can affect the format of dates).
2918 47. exim_tidydb, when checking for the continued existence of a message for
2919 which it has found a message-specific retry record, was not finding
2920 messages that were in split spool directories. Consequently, it was
2921 deleting retry records that should have stayed in existence.
2923 48. Steve fixed some bugs in eximstats.
2925 49. The SPA authentication driver was not abandoning authentication and moving
2926 on to the next authenticator when an expansion was forced to fail,
2927 contradicting the general specification for all authenticators. Instead it
2928 was generating a temporary error. It now behaves as specified.
2930 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2931 (the order specifies the preference for clients). The order is now AES256,
2932 AES128, 3DES, ARCFOUR128.
2934 51. Small patch to Sieve code - explicitly set From: when generating an
2937 52. Exim crashed if a remote delivery caused a very long error message to be
2938 recorded - for instance if somebody sent an entire SpamAssassin report back
2939 as a large number of 550 error lines. This bug was coincidentally fixed by
2940 increasing the size of one of Exim's internal buffers (big_buffer) that
2941 happened as part of the Exiscan merge. However, to be on the safe side, I
2942 have made the code more robust (and fixed the comments that describe what
2945 53. Now that there can be additional text after "Completed" in log lines (if
2946 the queue_time_overall log selector is set), a one-byte patch to exigrep
2947 was needed to allow it to recognize "Completed" as not the last thing in
2950 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2951 patch that reportedly fixes this has been added. I am not expert enough to
2952 create a test for it. This is what the patch creator wrote:
2954 "I found a little strange behaviour of ldap code when working with
2955 Windows 2003 AD Domain, where users was placed in more than one
2956 Organization Units. When I tried to give exim partial DN, the exit code
2957 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2958 But simultaneously result of request was absolutely normal ldap result,
2959 so I produce this patch..."
2961 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2962 REFERENCE, so I have modified the code to exclude the patch when that macro
2965 55. Some experimental protocols are using DNS PTR records for new purposes. The
2966 keys for these records are domain names, not reversed IP addresses. The
2967 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2968 leaves it alone. Component reversal etc. now happens only for IP addresses.
2971 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2973 57. Double the size of the debug message buffer (to 2048) so that more of very
2974 long debug lines gets shown.
2976 58. The exicyclog utility now does better if the number of log files to keep
2977 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2979 59. Two changes related to the smtp_active_hostname option:
2981 (1) $smtp_active_hostname is now available as a variable.
2982 (2) The default for smtp_banner uses $smtp_active_hostname instead
2983 of $primary_hostname.
2985 60. The host_aton() function is supposed to be passed a string that is known
2986 to be a valid IP address. However, in the case of IPv6 addresses, it was
2987 not checking this. This is a hostage to fortune. Exim now panics and dies
2988 if the condition is not met. A case was found where this could be provoked
2989 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2990 components; fortuitously, this particular loophole had already been fixed
2991 by change 4.50/55 above.
2993 If there are any other similar loopholes, the new check in host_aton()
2994 itself should stop them being exploited. The report I received stated that
2995 data on the command line could provoke the exploit when Exim was running as
2996 exim, but did not say which command line option was involved. All I could
2997 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2998 running as the user.
3001 61. There was a buffer overflow vulnerability in the SPA authentication code
3002 (which came originally from the Samba project). I have added a test to the
3003 spa_base64_to_bits() function which I hope fixes it.
3006 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3007 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3009 63. The daemon start-up calls getloadavg() while still root for those OS that
3010 need the first call to be done as root, but it missed one case: when
3011 deliver_queue_load_max is set with deliver_drop_privilege. This is
3012 necessary for the benefit of the queue runner, because there is no re-exec
3013 when deliver_drop_privilege is set.
3015 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3016 This has been fixed.
3018 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3019 was in use, was not putting the data itself into the right store pool;
3020 consequently, it could be overwritten for a subsequent message in the same
3021 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3024 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3026 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3027 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3028 the function were treating the return as a boolean value, which happened to
3029 work because 0=false and not-0=true, but is not correct code.
3031 68. The host_aton() function was not handling scoped IPv6 addresses (those
3032 with, for example, "%eth0" on the end) correctly.
3034 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3035 negated items (that is, ~something) in unsigned ints. Some compilers
3036 apparently mutter when there is no cast.
3038 70. If an address verification called from an ACL failed, and did not produce a
3039 user-specific message (i.e. there was only a "system" message), nothing was
3040 put in $acl_verify_message. In this situation, it now puts the system
3043 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3044 session; change 4.31/43 added the unwanted input to the log line - except
3045 that it did not do this in the start of session case. It now does.
3047 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3048 This is wrong and can cause the other end to generate a synchronization
3049 error if it is another Exim or anything else that does the synchronization
3050 check. A QUIT command is no longer sent after a timeout.
3052 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3053 during host lookups.
3055 74. The defer_ok option of callout verification was not working if it was used
3056 when verifying addresses in header lines, that is, for this case:
3058 verify = header_sender/callout=defer_ok
3060 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3061 those file descriptors could be used for SMTP connections. If anything
3062 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3063 could be sent to the SMTP client, causing chaos. The daemon now opens
3064 stdin, stdout, and stderr to /dev/null when it puts itself into the
3067 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3068 log by default. The user can override this with suitable Perl magic.
3070 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3071 the log message when discard triggers, was not working for the DATA ACL or
3072 for the non-SMTP ACL.
3074 78. Error message wording change in sieve.c.
3076 79. If smtp_accept_max_per_host was set, the number of connections could be
3077 restricted to fewer than expected, because the daemon was trying to set up
3078 a new connection before checking whether the processes handling previous
3079 connections had finished. The check for completed processes is now done
3080 earlier. On busy systems, this bug wouldn't be noticed because something
3081 else would have woken the daemon, and it would have reaped the completed
3084 80. If a message was submitted locally by a user whose login name contained one
3085 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3086 It caused a spool format error. I have fixed the spool reading code. A
3087 related problem was that the "from" clause in the Received: line became
3088 illegal because of the space(s). It is now covered by ${quote_local_part.
3090 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3093 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3095 83. Patch from Sieve maintainer to fix unterminated string problem in
3096 "vacation" handling.
3098 84. Some minor changes to the Linux configuration files to help with other
3099 OS variants using glibc.
3101 85. One more patch for Sieve to update vacation handling to latest spec.
3104 ----------------------------------------------------
3105 See the note above about the 4.44 and 4.50 releases.
3106 ----------------------------------------------------
3112 1. Change 4.43/35 introduced a bug that caused file counts to be
3113 incorrectly computed when quota_filecount was set in an appendfile
3116 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3117 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3120 3. Give more explanation in the error message when the command for a transport
3121 filter fails to execute.
3123 4. There are several places where Exim runs a non-Exim command in a
3124 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3125 was being done only for the command run by the queryprogram router. It is
3126 now done for all such subprocesses. The other cases are: ${run, transport
3127 filters, and the commands run by the lmtp and pipe transports.
3129 5. Some older OS have a limit of 256 on the maximum number of file
3130 descriptors. Exim was using setrlimit() to set 1000 as a large value
3131 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3132 systems. I've change it so that if it can't get 1000, it tries for 256.
3134 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3135 was an oversight, and furthermore, ever since the addition of extra
3136 controls (e.g. 4.43/32), the checks on when to allow different forms of
3137 "control" were broken. There should now be diagnostics for all cases when a
3138 control that does not make sense is encountered.
3140 7. $recipients is now available in the predata ACL (oversight).
3142 8. Tidy the search cache before the fork to do a delivery from a message
3143 received from the command line. Otherwise the child will trigger a lookup
3144 failure and thereby defer the delivery if it tries to use (for example) a
3145 cached ldap connection that the parent has called unbind on.
3147 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3148 of $address_data from the recipient verification was clobbered by the
3149 sender verification.
3151 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3152 was its contents. (It was OK if the option was not defined at all.)
3154 11. A "Completed" log line is now written for messages that are removed from
3155 the spool by the -Mrm option.
3157 12. $host_address is now set to the target address during the checking of
3158 ignore_target_hosts.
3160 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3161 being passed; this would have caused $sender_host_name to have been used if
3162 matching the list had actually called for a host name (not very likely,
3163 since this list is usually IP addresses). A host name is now passed as
3166 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3167 code that specifies a non-restarting handler (typically sigaction() in
3168 modern systems) in an attempt to fix a rare and obscure crash bug.
3170 15. Narrowed the window for a race in the daemon that could cause it to ignore
3171 SIGCHLD signals. This is not a major problem, because they are used only to
3172 wake it up if nothing else does.
3174 16. A malformed maildirsize file could cause Exim to calculate negative values
3175 for the mailbox size or file count. Odd effects could occur as a result.
3176 The maildirsize information is now recalculated if the size or filecount
3179 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3180 support for a long time. Removed HAVE_SYS_VFS_H.
3182 18. Updated exipick to current release from John Jetmore.
3184 19. Allow an empty sender to be matched against a lookup in an address list.
3185 Previously the only cases considered were a regular expression, or an
3188 20. Exim went into a mad DNS lookup loop when doing a callout where the
3189 host was specified on the transport, if the DNS lookup yielded more than
3192 21. The RFC2047 encoding function was originally intended for short strings
3193 such as real names; it was not keeping to the 75-character limit for
3194 encoded words that the RFC imposes. It now respects the limit, and
3195 generates multiple encoded words if necessary. To be on the safe side, I
3196 have increased the buffer size for the ${rfc2047: expansion operator from
3199 22. Failure to deliver a bounce message always caused it to be frozen, even if
3200 there was an errors_to setting on the router. The errors_to setting is now
3203 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3204 canonical form (fully expanded) before being placed in
3205 $sender_host_address.
3207 24. Updated eximstats to version 1.33
3209 25. Include certificate and key file names in error message when GnuTLS fails
3210 to set them up, because the GnuTLS error message doesn't include the name
3211 of the failing file when there is a problem reading it.
3213 26. Expand error message when OpenSSL has problems setting up cert/key files.
3216 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3217 (this can affect the format of dates).
3219 28. exim_tidydb, when checking for the continued existence of a message for
3220 which it has found a message-specific retry record, was not finding
3221 messages that were in split spool directories. Consequently, it was
3222 deleting retry records that should have stayed in existence.
3224 29. eximstats updated to version 1.35
3225 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3226 1.35 - bugfix such that pie charts by volume are generated correctly
3228 30. The SPA authentication driver was not abandoning authentication and moving
3229 on to the next authenticator when an expansion was forced to fail,
3230 contradicting the general specification for all authenticators. Instead it
3231 was generating a temporary error. It now behaves as specified.
3233 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3234 (the order specifies the preference for clients). The order is now AES256,
3235 AES128, 3DES, ARCFOUR128.
3237 31. Small patch to Sieve code - explicitly set From: when generating an
3240 32. Exim crashed if a remote delivery caused a very long error message to be
3241 recorded - for instance if somebody sent an entire SpamAssassin report back
3242 as a large number of 550 error lines. This bug was coincidentally fixed by
3243 increasing the size of one of Exim's internal buffers (big_buffer) that
3244 happened as part of the Exiscan merge. However, to be on the safe side, I
3245 have made the code more robust (and fixed the comments that describe what
3248 33. Some experimental protocols are using DNS PTR records for new purposes. The
3249 keys for these records are domain names, not reversed IP addresses. The
3250 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3251 leaves it alone. Component reversal etc. now happens only for IP addresses.
3254 34. The host_aton() function is supposed to be passed a string that is known
3255 to be a valid IP address. However, in the case of IPv6 addresses, it was
3256 not checking this. This is a hostage to fortune. Exim now panics and dies
3257 if the condition is not met. A case was found where this could be provoked
3258 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3259 components; fortuitously, this particular loophole had already been fixed
3260 by change 4.50/55 or 4.44/33 above.
3262 If there are any other similar loopholes, the new check in host_aton()
3263 itself should stop them being exploited. The report I received stated that
3264 data on the command line could provoke the exploit when Exim was running as
3265 exim, but did not say which command line option was involved. All I could
3266 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3267 running as the user.
3270 35. There was a buffer overflow vulnerability in the SPA authentication code
3271 (which came originally from the Samba project). I have added a test to the
3272 spa_base64_to_bits() function which I hope fixes it.
3275 36. The daemon start-up calls getloadavg() while still root for those OS that
3276 need the first call to be done as root, but it missed one case: when
3277 deliver_queue_load_max is set with deliver_drop_privilege. This is
3278 necessary for the benefit of the queue runner, because there is no re-exec
3279 when deliver_drop_privilege is set.
3281 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3282 was in use, was not putting the data itself into the right store pool;
3283 consequently, it could be overwritten for a subsequent message in the same
3284 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3287 38. Sometimes the final signoff response after QUIT could fail to get
3288 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3289 < 0 before doing a fflush(). This bug looks as though it goes back to the
3290 introduction of TLS in release 3.20, but "sometimes" must have been rare
3291 because the tests only now provoked it.
3297 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3298 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3299 Now it returns an integer. A number of calls were still expecting a T/F
3300 return. Fortuitously, in all cases, the tests worked in OK situations,
3301 which is the norm. However, things would have gone wrong on any write
3302 failures on the smtp file descriptor. This function is used when sending
3303 messages over SMTP and also when doing verify callouts.
3305 2. When Exim is called to do synchronous delivery of a locally submitted
3306 message (the -odf or -odi options), it no longer closes stderr before doing
3309 3. Implemented the mua_wrapper option.
3311 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3313 5. Implemented the functions header_remove(), header_testname(),
3314 header_add_at_position(), and receive_remove_recipient(), and exported them
3317 6. If an ACL "warn" statement specified the addition of headers, Exim already
3318 inserted X-ACL-Warn: at the start if there was no header name. However, it
3319 was not making this test for the second and subsequent header lines if
3320 there were newlines in the string. This meant that an invalid header could
3321 be inserted if Exim was badly configured.
3323 7. Allow an ACL "warn" statement to add header lines at the start or after all
3324 the Received: headers, as well as at the end.
3326 8. Added the rcpt_4xx retry error code.
3328 9. Added postmaster_mailfrom=xxx to callout verification option.
3330 10. Added mailfrom=xxxx to the callout verification option, for verify=
3333 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3334 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3336 12. Inserted some casts to stop certain compilers warning when using pointer
3337 differences as field lengths or precisions in printf-type calls (mostly
3338 affecting debugging statements).
3340 13. Added optional readline() support for -be (dynamically loaded).
3342 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3343 same clock tick as a message's arrival, so that its received time was the
3344 same as the "first fail" time on the retry record, and that message
3345 remained on the queue past the ultimate address timeout, every queue runner
3346 would try a delivery (because it was past the ultimate address timeout) but
3347 after another failure, the ultimate address timeout, which should have then
3348 bounced the address, did not kick in. This was a "< instead of <=" error;
3349 in most cases the first failure would have been in the next clock tick
3350 after the received time, and all would be well.
3352 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3353 being recognized when the domain list was tested by the match_domain
3354 condition in an expansion string.
3356 16. Added the ${str2b64: operator.
3358 17. Exim was always calling setrlimit() to set a large limit for the number of
3359 processes, without checking whether the existing limit was already
3360 adequate. (It did check for the limit on file descriptors.) Furthermore,
3361 errors from getrlimit() and setrlimit() were being ignored. Now they are
3362 logged to the main and panic logs, but Exim does carry on, to try to do its
3363 job under whatever limits there are.
3365 18. Imported PCRE 5.0.
3367 19. Trivial typo in log message " temporarily refused connection" (the leading
3370 20. If the log selector return_path_on_delivery was set and an address was
3371 redirected to /dev/null, the delivery process crashed because it assumed
3372 that a return path would always be set for a "successful" delivery. In this
3373 case, the whole delivery is bypassed as an optimization, and therefore no
3376 21. Internal re-arrangement: the function for sending a challenge and reading
3377 a response while authentication was assuming a zero-terminated challenge
3378 string. It's now changed to take a pointer and a length, to allow for
3379 binary data in such strings.
3381 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3383 23. Exim was not respecting finduser_retries when seeking the login of the
3384 uid under which it was called; it was always trying 10 times. (The default
3385 setting of finduser_retries is zero.) Also, it was sleeping after the final
3386 failure, which is pointless.
3388 24. Implemented tls_on_connect_ports.
3390 25. Implemented acl_smtp_predata.
3392 26. If the domain in control=submission is set empty, Exim assumes that the
3393 authenticated id is a complete email address when it generates From: or
3394 Sender: header lines.
3396 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3397 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3398 chown and chgrp in /bin and hostname in /usr/bin.
3400 28. Exim was keeping the "process log" file open after each use, just as it
3401 does for the main log. This opens the possibility of it remaining open for
3402 long periods when the USR1 signal hits a daemon. Occasional processlog
3403 errors were reported, that could have been caused by this. Anyway, it seems
3404 much more sensible not to leave this file open at all, so that is what now
3407 29. The long-running daemon process does not normally write to the log once it
3408 has entered its main loop, and it closes the log before doing so. This is
3409 so that log files can straightforwardly be renamed and moved. However,
3410 there are a couple of unusual error situations where the daemon does write
3411 log entries, and I had neglected to close the log afterwards.
3413 30. The text of an SMTP error response that was received during a remote
3414 delivery was being truncated at 512 bytes. This is too short for some of
3415 the long messages that one sometimes sees. I've increased the limit to
3418 31. It is now possible to make retry rules that apply only when a message has a
3419 specific sender, in particular, an empty sender.
3421 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3422 it possible to be selective about when SMTP synchronization is enforced.
3424 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3426 32. Implemented hosts_connection_nolog.
3428 33. Added an ACL for QUIT.
3430 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3433 35. Added mailbox_size and mailbox_filecount to appendfile.
3435 36. Added control = no_multiline_responses to ACLs.
3437 37. There was a bug in the logic of the code that waits for the clock to tick
3438 in the case where the clock went backwards by a substantial amount such
3439 that the microsecond fraction of "now" was more than the microsecond
3440 fraction of "then" (but the whole seconds number was less).
3442 38. Added support for the libradius Radius client library this is found on
3443 FreeBSD (previously only the radiusclient library was supported).
3449 1. When certain lookups returned multiple values in the form name=value, the
3450 quoting of the values was not always being done properly. Specifically:
3451 (a) If the value started with a double quote, but contained no whitespace,
3453 (b) If the value contained whitespace other than a space character (i.e.
3454 tabs or newlines or carriage returns) it was not quoted.
3455 This fix has been applied to the mysql and pgsql lookups by writing a
3456 separate quoting function and calling it from the lookup code. The fix
3457 should probably also be applied to nisplus, ibase and oracle lookups, but
3458 since I cannot test any of those, I have not disturbed their existing code.
3460 2. A hit in the callout cache for a specific address caused a log line with no
3461 reason for rejecting RCPT. Now it says "Previous (cached) callout
3462 verification failure".
3464 3. There was an off-by-one bug in the queryprogram router. An over-long
3465 return line was truncated at 256 instead of 255 characters, thereby
3466 overflowing its buffer with the terminating zero. As well as fixing this, I
3467 have increased the buffer size to 1024 (and made a note to document this).
3469 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3470 when Exim is waiting for an SMTP response from a remote server, Exim
3471 restarts its select() call on the socket, thereby resetting its timeout.
3472 This is not a problem when such interrupts are rare. Somebody set up a cron
3473 job to run exiwhat every 2 minutes, which is less than the normal select()
3474 timeout (5 or 10 minutes). This meant that the select() timeout never
3475 kicked in because it was always reset. I have fixed this by comparing the
3476 time when an interrupt arrives with the time at the start of the first call
3477 to select(). If more time than the timeout has elapsed, the interrupt is
3478 treated as a timeout.
3480 5. Some internal re-factoring in preparation for the addition of Sieve
3481 extensions (by MH). In particular, the "personal" test is moved to a
3482 separate function, and given an option for scanning Cc: and Bcc: (which is
3483 not set for Exim filters).
3485 6. When Exim created an email address using the login of the caller as the
3486 local part (e.g. when creating a From: or Sender: header line), it was not
3487 quoting the local part when it contained special characters such as @.
3489 7. Installed new OpenBSD configuration files.
3491 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3492 try to make them clearer.
3494 9. Callout options, other than the timeout value, were being ignored when
3495 verifying sender addresses in header lines. For example, when using
3497 verify = header_sender/callout=no_cache
3499 the cache was (incorrectly) being used.
3501 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3502 only the Cygwin environment.
3504 11. When return_path_on_delivery was set as a log selector, if different remote
3505 addresses in the same message used different return paths and parallel
3506 remote delivery occurred, the wrong values would sometimes be logged.
3507 (Whenever a remote delivery process finished, the return path value from
3508 the most recently started remote delivery process was logged.)
3510 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3511 lines when AUTH and/or TLS are in use. This is the "received protocol"
3512 field. Exim used to use "asmtp" for authenticated SMTP, without any
3513 indication (in the protocol name) for TLS use. Now it follows the RFC and
3514 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3515 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3516 names appear in log lines as well as in Received: header lines.
3518 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3519 extensions, and comparison tests, and to fix some bugs.
3521 14. Changes to the "personal" filter test:
3523 (1) The test was buggy in that it was just doing the equivalent of
3524 "contains" tests on header lines. For example, if a user's address was
3525 anne@some.where, the "personal" test would incorrectly be true for
3527 To: susanne@some.where
3529 This test is now done by extracting each address from the header in turn,
3530 and checking the entire address. Other tests that are part of "personal"
3531 are now done using regular expressions (for example, to check local parts
3532 of addresses in From: header lines).
3534 (2) The list of non-personal local parts in From: addresses has been
3535 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3536 taken from the Sieve specification recommendations.
3538 (3) If the message contains any header line starting with "List-" it is
3539 treated as non-personal.
3541 (4) The test for "circular" in the Subject: header line has been removed
3542 because it now seems ill-conceived.
3544 15. Minor typos in src/EDITME comments corrected.
3546 16. Installed latest exipick from John Jetmore.
3548 17. If headers_add on a router specified a text string that was too long for
3549 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3550 of string_sprintf() is now avoided.
3552 18. $message_body_size was not set (it was always zero) when running the DATA
3553 ACL and the local_scan() function.
3555 19. For the "mail" command in an Exim filter, no default was being set for
3556 the once_repeat time, causing a random time value to be used if "once" was
3557 specified. (If the value happened to be <= 0, no repeat happened.) The
3558 default is now 0s, meaning "never repeat". The "vacation" command was OK
3559 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3560 (I found it when inspecting the code).
3562 20. There is now an overall timeout for performing a callout verification. It
3563 defaults to 4 times the callout timeout, which applies to individual SMTP
3564 commands during the callout. The overall timeout applies when there is more
3565 than one host that can be tried. The timeout is checked before trying the
3566 next host. This prevents very long delays if there are a large number of
3567 hosts and all are timing out (e.g. when the network connections are timing
3568 out). The value of the overall timeout can be changed by specifying an
3569 additional sub-option for "callout", called "maxwait". For example:
3571 verify = sender/callout=5s,maxwait=20s
3573 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3574 to the end before writing, but this should make it even safer).
3576 22. Exim was forgetting that it had advertised PIPELINING for the second and
3577 subsequent messages on an SMTP connection. It was also not resetting its
3578 memory on STARTTLS and an internal HELO.
3580 23. When Exim logs an SMTP synchronization error within a session, it now
3581 records whether PIPELINING has been advertised or not.
3583 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3584 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3585 rather than long int.
3587 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3589 26. Added the never_mail option to autoreply.
3595 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3596 crash if the getsockname() call failed; this can happen if a connection is
3597 closed very soon after it is established. The problem was simply in the
3598 order in which certain operations were done, causing Exim to try to write
3599 to the SMTP stream before it had set up the file descriptor. The bug has
3600 been fixed by making things happen in the correct order.
3606 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3607 before the connection was closed, thus losing the rejection response.
3609 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3610 some early Solaris releases, but causes trouble in current releases where
3611 socklen_t is defined.
3613 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3616 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3619 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3620 that releasing the top of it at the end releases what was used for sub-
3621 expansions (unless the block got too big). However, discard this block if
3622 the first thing is a variable or header, so that we can use its block when
3623 it is dynamic (useful for very large $message_headers, for example).
3625 6. Lookups now cache *every* query, not just the most recent. A new, separate
3626 store pool is used for this. It can be recovered when all lookup caches are
3627 flushed. Lookups now release memory at the end of their result strings.
3628 This has involved some general refactoring of the lookup sources.
3630 7. Some code has been added to the store_xxx() functions to reduce the amount
3631 of flapping under certain conditions.
3633 8. log_incoming_interface used to affect only the <= reception log lines. Now
3634 it causes the local interface and port to be added to several more SMTP log
3635 lines, for example "SMTP connection from", and rejection lines.
3637 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3639 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3641 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3642 could be overwritten at the end of the current message (or the start of a
3643 new message if it was set in a HELO ACL). The value is now preserved for
3644 the duration of the SMTP connection.
3646 12. If a transport had a headers_rewrite setting, and a matching header line
3647 contained an unqualified address, that address was qualified, even if it
3648 did not match any rewriting rules. The underlying bug was that the values
3649 of the flags that permit the existence of unqualified sender and recipient
3650 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3651 non-local messages, and by -bnq for local messages) were not being
3652 preserved with the message after it was received.
3654 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3655 "next input=" as part of the text comprising the host identity instead of
3656 the correct text. The code was using the same buffer for two different
3657 strings. However, depending on which order the printing function evaluated
3658 its arguments, the bug did not always show up. Under Linux, for example, my
3659 test suite worked just fine.
3661 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3662 doesn't work with some older versions of Perl. It has been changed to "my",
3663 which in any case is probably the better facility to use.
3665 15. A really picky compiler found some instances of statements for creating
3666 error messages that either had too many or two few arguments for the format
3669 16. The size of the buffer for calls to the DNS resolver has been increased
3670 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3671 for addresses that have a lot of PTR records. This alleviates a problem; it
3672 does not fully solve it.
3674 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3675 buffer now truncates the list and logs the incident, which is the same
3676 action as happens when Exim is looking up a host name and its aliases.
3677 Previously in this situation something unpredictable would happen;
3678 sometimes it was "internal error: store_reset failed".
3680 18. If a server dropped the connection unexpectedly when an Exim client was
3681 using GnuTLS and trying to read a response, the client delivery process
3682 crashed while trying to generate an error log message.
3684 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3685 string, for example:
3687 warn message = H1: something\nH2: something
3689 the text was added as a single header line from Exim's point of view
3690 though it ended up OK in the delivered message. However, searching for the
3691 second and subsequent header lines using $h_h2: did not work. This has been
3692 fixed. Similarly, if a system filter added multiple headers in this way,
3693 the routers could not see them.
3695 20. Expanded the error message when iplsearch is called with an invalid key to
3696 suggest using net-iplsearch in a host list.
3698 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3699 ACLs are no longer actually imposed (and a message to that effect is
3702 22. If a "gecos" field in a passwd entry contained escaped characters, in
3703 particular, if it contained a \" sequence, Exim got it wrong when building
3704 a From: or a Sender: header from that name. A second bug also caused
3705 incorrect handling when an unquoted " was present following a character
3706 that needed quoting.
3708 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3709 was not being matched caselessly.
3711 24. Arranged for all hyphens in the exim.8 source to be escaped with
3714 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3715 properly. Recipient callout cache records were still being keyed to include
3716 the sender, even when use_sender was set false. This led to far more
3717 callouts that were necessary. The sender is no longer included in the key
3718 when use_sender is false.
3720 26. Added "control = submission" modifier to ACLs.
3722 27. Added the ${base62d: operator to decode base 62 numbers.
3724 28. dnsdb lookups can now access SRV records.
3726 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3727 the configuration file.
3729 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3730 (-v) mode. This makes the output for a verbose queue run more intelligible.
3732 31. Added a use_postmaster feature to recipient callouts.
3734 32. Added the $body_zerocount variable, containing the number of binary zero
3735 bytes in the message body.
3737 33. The time of last modification of the "new" subdirectory is now used as the
3738 "mailbox time last read" when there is a quota error for a maildir
3741 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3743 35. Added +ignore_unknown as a special item in host lists.
3745 36. Code for decoding IPv6 addresses in host lists is now included, even if
3746 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3747 address was recognized as an IP address, but was then not correctly decoded
3748 into binary, causing unexpected and incorrect effects when compared with
3755 1. Very minor rewording of debugging text in manualroute to say "list of
3756 hosts" instead of "hostlist".
3758 2. If verify=header_syntax was set, and a header line with an unqualified
3759 address (no domain) and a large number of spaces between the end of the
3760 name and the colon was received, the reception process suffered a buffer
3761 overflow, and (when I tested it) crashed. This was caused by some obsolete
3762 code that should have been removed. The fix is to remove it!
3764 3. When running in the test harness, delay a bit after writing a bounce
3765 message to get a bit more predictability in the log output.
3767 4. Added a call to search_tidyup() just before forking a reception process. In
3768 theory, someone could use a lookup in the expansion of smtp_accept_max_
3769 per_host which, without the tidyup, could leave open a database connection.
3771 5. Added the variables $recipient_data and $sender_data which get set from a
3772 lookup success in an ACL "recipients" or "senders" condition, or a router
3773 "senders" option, similar to $domain_data and $local_part_data.
3775 6. Moved the writing of debug_print from before to after the "senders" test
3778 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3779 problems for message scanning, either using a data ACL, or using
3780 local_scan() because the Received: header was not generated till after they
3781 were called (in order to set the time as the time of reception completion).
3782 I have revised the way this works. The header is now generated after the
3783 body is received, but before the ACL or local_scan() are called. After they
3784 are run, the timestamp in the header is updated.
3790 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3791 before starting a queue runner without re-exec. This happened only when
3792 deliver_drop_privilege was set or when the Exim user was set to root. The
3793 effect of the bug was that timeouts during subsequent deliveries caused
3794 crashes instead of being properly handled. The handler is now left at its
3795 default (and expected) setting.
3797 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3798 message, again when deliver_drop_privilege is set or Exim is run as root.
3799 The bug described in (1) was not present in this case, but the tidying up
3800 of the other signals was missing. I have made the two cases consistent.
3802 3. The ignore_target_hosts setting on a manualroute router was being ignored
3803 for hosts that were looked up using the /MX notation.
3805 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3808 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3809 operated on the sender address. After changing the $sender_address to <>
3810 for the sender address verify, Exim was re-instated it as the original
3811 (before rewriting) address, but remembering that it had rewritten it, so it
3812 wasn't rewriting it again. This bug also had the effect of breaking the
3813 sender address verification caching when the sender address was rewritten.
3815 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3816 This has been changed so that if the ip literal address matches
3817 ignore_target_hosts, the router declines.
3819 7. Added expansion conditions match_domain, match_address, and match_local_
3820 part (NOT match_host).
3822 8. The placeholder for the Received: header didn't have a length field set.
3824 9. Added code to Exim itself and to exim_lock to test for a specific race
3825 condition that could lead to file corruption when using MBX delivery. The
3826 issue is with the lockfile that is created in /tmp. If this file is removed
3827 after a process has opened it but before that process has acquired a lock,
3828 there is the potential for a second process to recreate the file and also
3829 acquire a lock. This could lead to two Exim processes writing to the file
3830 at the same time. The added code performs the same test as UW imapd; it
3831 checks after acquiring the lock that its file descriptor still refers to
3832 the same named file.
3834 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3835 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3838 11. Added the smtp_active_hostname option. If used, this will typically be made
3839 to depend on the incoming interface address. Because $interface_address is
3840 not set up until the daemon has forked a reception process, error responses
3841 that can happen earlier (such as "too many connections") no longer contain
3844 12. If an expansion in a condition on a "warn" statement fails because a lookup
3845 defers, the "warn" statement is abandoned, and the next ACL statement is
3846 processed. Previously this caused the whole ACL to be aborted.
3848 13. Added the iplsearch lookup type.
3850 14. Added ident_timeout as a log selector.
3852 15. Added tls_certificate_verified as a log selector.
3854 16. Added a global option tls_require_ciphers (compare the smtp transport
3855 option of the same name). This controls incoming TLS connections.
3857 17. I finally figured out how to make tls_require_ciphers do a similar thing
3858 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3859 before starting the TLS session.
3861 18. Tabs are now shown as \t in -bP output.
3863 19. If the log selector return_path_on_delivery was set, Exim crashed when
3864 bouncing a message because it had too many Received: header lines.
3866 20. If two routers both had headers_remove settings, and the first one included
3867 a superfluous trailing colon, the final name in the first list and the
3868 first name in the second list were incorrectly joined into one item (with a
3869 colon in the middle).
3875 1. Added -C and -D options to the exinext utility, mainly to make it easier
3876 to include in the automated testing, but these could be helpful when
3877 multiple configurations are in use.
3879 2. The exinext utility was not formatting the output nicely when there was
3880 an alternate port involved in the retry record key, nor when there was a
3881 message id as well (for retries that were specific to a specific message
3882 and a specific host). It was also confused by IPv6 addresses, because of
3883 the additional colons they contain. I have fixed the IPv4 problem, and
3884 patched it up to do a reasonable job for IPv6.
3886 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3887 delivery, the log line now contains "pipelined" if PIPELINING was used.
3889 4. An SMTP transport process used to panic and die if the bind() call to set
3890 an explicit outgoing interface failed. This has been changed; it is now
3891 treated in the same way as a connect() failure.
3893 5. A reference to $sender_host_name in the part of a conditional expansion
3894 that was being skipped was still causing a DNS lookup. This no longer
3897 6. The def: expansion condition was not recognizing references to header lines
3898 that used bh_ and bheader_.
3900 7. Added the _cache feature to named lists.
3902 8. The code for checking quota_filecount in the appendfile transport was
3903 allowing one more file than it should have been.
3905 9. For compatibility with Sendmail, the command line option
3913 and sets the incoming protocol and host name (for trusted callers). The
3914 host name and its colon can be omitted when only the protocol is to be set.
3915 Note the Exim already has two private options, -pd and -ps, that refer to
3916 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3917 "s", but I don't think that's a major issue.
3919 10. A number of refactoring changes to the code, none of which should affect
3922 (a) The number of logging options was getting close to filling up the
3923 32-bit word that was used as a bit map. I have split them into two classes:
3924 those that are passed in the argument to log_write(), and those that are
3925 only ever tested independently outside of that function. These are now in
3926 separate 32-bit words, so there is plenty of room for expansion again.
3927 There is no change in the user interface or the logging behaviour.
3929 (b) When building, for example, log lines, the code previously used a
3930 macro that called string_cat() twice, in order to add two strings. This is
3931 not really sufficiently general. Furthermore, there was one instance where
3932 it was actually wrong because one of the argument was used twice, and in
3933 one call a function was used. (As it happened, calling the function twice
3934 did not affect the overall behaviour.) The macro has been replaced by a
3935 function that can join an arbitrary number of extra strings onto a growing
3938 (c) The code for expansion conditions now uses a table and a binary chop
3939 instead of a serial search (which was left over from when there were very
3940 few conditions). Also, it now recognizes conditions like "pam" even when
3941 the relevant support is not compiled in: a suitably worded error message is
3942 given if an attempt is made to use such a condition.
3944 11. Added ${time_interval:xxxxx}.
3946 12. A bug was causing one of the ddress fields not to be passed back correctly
3947 from remote delivery subprocesses. The field in question was not being
3948 subsequently used, so this caused to problems in practice.
3950 13. Added new log selectors queue_time and deliver_time.
3952 14. Might have fixed a bug in maildirsizefile handling that threw up
3953 "unexpected character" debug warnings, and recalculated the data
3954 unnecessarily. In any case, I expanded the warning message to give more
3957 15. Added the message "Restricted characters in address" to the statements in
3958 the default ACL that block characters like @ and % in local parts.
3960 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3961 Three changes have been made:
3963 (a) There was a serious bug; a negative response to MAIL caused the whole
3964 recipient domain to be cached as invalid, thereby blocking all messages
3965 to all local parts at the same domain, from all senders. This bug has
3966 been fixed. The domain is no longer cached after a negative response to
3967 MAIL if the sender used is not empty.
3969 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3972 (c) A new callout option, "use_sender" has been added for people who want
3973 the modified behaviour.
3979 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3982 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3983 indeed breaks things for older releases.
3985 3. Added additional logging to the case where there is a problem reading data
3986 from a filter that is running in a subprocess using a pipe, in order to
3987 try to track down a specific problem.
3989 4. Testing facility fudge: when running in the test harness and attempting
3990 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3991 getting "No route to host". Convert this to a timeout.
3993 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3996 6. Some OS don't have socklen_t but use size_t instead. This affects the
3997 fifth argument of getsockopt() amongst other things. This is now
3998 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3999 can be set for individual OS. I have set it for SunOS5, OSF1, and
4000 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4001 some earlier ones do not.
4003 7. Change 4.30/15 was not doing the test caselessly.
4005 8. The standard form for an IPv6 address literal was being rejected by address
4006 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4007 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4008 this, as well as the form without the "IPv6" on the front (but only when
4009 address literals are enabled, of course).
4011 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4013 10. Exim crashed if a message with an empty sender address specified by -f
4014 encountered a router with an errors_to setting. This could be provoked only
4015 by a command such as
4019 where an empty string was supplied; "<>" did not hit this bug.
4021 11. Installed PCRE release 4.5.
4023 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4024 remained set. It is now erased.
4026 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4027 times from message ids (which are base 36 rather than the normal 62).
4029 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4030 were being counted as actual protocol errors, and logged if the log
4031 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4032 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4033 and DATA following a set of rejected RCPTs do not count as protocol errors.
4034 In other words, Exim assumes they were pipelined, though this may not
4035 actually be the case. Of course, in all cases the client gets an
4036 appropriate error code.
4038 15. If a lookup fails in an ACL condition, a message about the failure may
4039 be available; it is used if testing the ACL cannot continue, because most
4040 such messages specify what the cause of the deferral is. However, some
4041 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4042 that caused an old message to be retained and used if a later statement
4043 caused a defer, replacing the real cause of the deferral.
4045 16. If an IP address had so many PTR records that the DNS lookup buffer
4046 was not large enough to hold them, Exim could crash while trying to process
4047 the truncated data. It now detects and logs this case.
4049 17. Further to 4.21/58, another change has been made: if (and only if) the
4050 first line of a message (the first header line) ends with CRLF, a bare LF
4051 in a subsequent header line has a space inserted after it, so as not to
4052 terminate the header.
4054 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4055 unnecessarily, used atoi() instead of strtol(), and didn't check the
4056 termination when getting file sizes from file names by regex.
4058 19. Completely re-implemented the support for maildirsize files, in the light
4059 of a number of problems with the previous contributed implementation
4060 (4.30/29). In particular:
4062 . If the quota is zero, the maildirsize file is maintained, but no quota is
4065 . If the maildir directory does not exist, it is created before any attempt
4066 to write a maildirsize file.
4068 . The quota value in the file is just a cache; if the quota is changed in
4069 the transport, the new value overrides.
4071 . A regular expression is available for excluding directories from the
4074 20. The autoreply transport checks the characters in options that define the
4075 message's headers; it allows continued headers, but it was checking with
4076 isspace() after an embedded newline instead of explicitly looking for a
4079 21. If all the "regular" hosts to which an address was routed had passed their
4080 expiry times, and had not reached their retry times, the address was
4081 bounced, even if fallback hosts were defined. Now Exim should go on to try
4084 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4085 equivalent code in the SMTP transport. Some hosts send humungous responses
4086 to HELO/EHLO, more than 1024 it seems.
4088 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4089 gives compiler warnings in some environments. I've now done it "properly",
4092 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4093 (because of problems with the built-in one) was declared to return uschar *
4094 instead of char *, causing compiler failure.
4096 25. Fixed a file descriptor leak when processing alias/forward files.
4098 26. Fixed a minor format string issue in dbfn.c.
4100 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4102 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4103 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4106 29. When a sender address is verified, it is cached, to save repeating the test
4107 when there is more than one recipient in a message. However, when the
4108 verification involves a callout, it is possible for different callout
4109 options to be set for different recipients. It is too complicated to keep
4110 track of this in the cache, so now Exim always runs a verification when a
4111 callout is required, relying on the callout cache for the optimization.
4112 The overhead is duplication of the address routing, but this should not be
4115 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4116 to be verified with callout=postmaster, and the main callout worked but the
4117 postmaster check failed, the verification correctly failed. However, if a
4118 subsequent RCPT command asked for sender verification *without* the
4119 postmaster check, incorrect caching caused this verification also to fail,
4122 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4123 it was not caching the DNS options (qualify_single, search_parents) that
4124 were used when the lookup failed. A subsequent lookup with different
4125 options therefore always gave the same answer, though there were cases
4126 where it should not have. (Example: a "domains = !$mx_any" option on a
4127 dnslookup router: the "domains" option is always processed without any
4128 widening, but the router might have qualify_single set.) Now Exim uses the
4129 cached value only when the same options are set.
4131 32. Added John Jetmore's "exipick" utility to the distribution.
4133 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4134 than a timeout (e.g. a certificate is required, and is not provided), an
4135 Exim server now closes the connection immediately. Previously it waited for
4136 the client to close - but if the client is SSL, it seems that they each
4137 wait for each other, leading to a delay before one of them times out.
4139 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4140 maintained 0.8.x compatibility because I don't think many are using it, and
4141 it is clearly obsolete.
4143 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4146 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4147 client certificate was expired. A simple patch fixes this, though I don't
4148 understand the full logic of why the verify callback is called multiple
4151 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4152 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4153 which causes problems with some clients (such as the Certicom SSL Plus
4154 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4155 disables the coutermeasure allowing Eudora to connect."
4157 38. Exim was not checking that a write() to a log file succeeded. This could
4158 lead to Bad Things if a log got too big, in particular if it hit a file
4159 size limit. Exim now panics and dies if it cannot write to a log file, just
4160 as it does if it cannot open a log file.
4162 39. Modified OS/Makefile-Linux so that it now contains
4164 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4166 The two -D definitions ensure that Exim is compiled with large file
4167 support, which makes it possible to handle log files that are bigger than
4170 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4171 instance) a domain was checked against a named list that involved a lookup,
4172 causing $domain_data to be set, then another domain was checked against the
4173 same list, then the first domain was re-checked, the value of $domain_data
4174 after the final check could be wrong. In particular, if the second check
4175 failed, it could be set empty. This bug probably also applied to
4178 41. The strip_trailing_dot option was not being applied to the address given
4179 with the -f command-line option.
4181 42. The code for reading a message's header from the spool was incrementing
4182 $received_count, but never initializing it. This meant that the value was
4183 incorrect (doubled) while delivering a message in the same process in which
4184 it was received. In the most common configuration of Exim, this never
4185 happens - a fresh exec is done - but it can happen when
4186 deliver_drop_privilege is set.
4188 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4189 it now includes up to 150 characters of the unexpected data in the log
4192 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4193 and building data strings. The size of both of these buffers was 10 000
4194 bytes - far larger than anybody would *ever* want, thought I. Needless to
4195 say, somebody hit the limit. I have increased the maximum line length to
4196 20 000 and the maximum data length of concatenated lines to 100 000. I have
4197 also fixed two bugs, because there was no checking on these buffers. Tsk,
4198 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4199 buffer is too small.
4201 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4202 lsearch lookups. Now it does.
4204 46. When parsing a route_list item in a manualroute router, a fixed-length
4205 buffer was used for the list of hosts. I made this 1024 bytes long,
4206 thinking that nobody would ever have a list of hosts that long. Wrong.
4207 Somebody had a whole pile of complicated expansion conditions, and the
4208 string was silently truncated, leading to an expansion error. It turns out
4209 that it is easier to change to an unlimited length (owing to other changes
4210 that have happened since this code was originally written) than to build
4211 structure for giving a limitation error. The length of the item that
4212 expands into the list of hosts is now unlimited.
4214 47. The lsearch lookup could not handle data where the length of text line was
4215 more than 4095 characters. Such lines were truncated, leading to shortened
4216 data being returned. It should now handle lines of any length.
4218 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4219 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4222 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4223 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4224 no longer happens in comment lines. A list of replacements is now placed
4225 at the head of all of the source files, except those whose only change is
4226 to replace PERL_COMMAND in the very first #! line.
4228 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4229 messages on the queue, with a bottom-up merge sort, using code contributed
4230 by Michael Haardt. This should make operations like -bp somewhat faster on
4231 large queues. It won't affect queue runners, except when queue_run_in_order
4234 51. Installed eximstats 1.31 in the distribution.
4236 52. Added support for SRV lookups to the dnslookup router.
4238 53. If an ACL referred to $message_body or $message_body_end, the value was not
4239 reset for any messages that followed in the same SMTP session.
4241 54. The store-handling optimization for building very long strings was not
4242 differentiating between the different store pools. I don't think this
4243 actually made any difference in practice, but I've tidied it.
4245 55. While running the routers to verify a sender address, $sender_address
4246 was still set to the sender address. This is wrong, because when routing to
4247 send a bounce to the sender, it would be empty. Therefore, I have changed
4248 it so that, while verifying a sender address, $sender_address is set to <>.
4249 (There is no change to what happens when verifying a recipient address.)
4251 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4252 target A or AAAA records (if not already returned) without resetting the
4253 qualify_single or search_parents options of the DNS resolver. These are
4254 inappropriate in this case because the targets of MX and SRV records must
4255 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4256 target that, when qualified, matched something in the local domain. These
4257 two options are now turned off when doing these lookups.
4259 57. It seems that at least some releases of Reiserfs (which does not have the
4260 concept of a fixed number of inodes) returns zero and not -1 for the
4261 number of available inodes. This interacted badly with check_spool_inodes,
4262 which assumed that -1 was the "no such thing" setting. What I have done is
4263 to check that the total number of inodes is greater than zero before doing
4264 the test of how many are available.
4266 58. When a "warn" ACL statement has a log_message modifier, the message is
4267 remembered, and not repeated. This is to avoid a lot of repetition when a
4268 message has many recipients that cause the same warning to be written.
4269 Howewer, Exim was preserving the list of already written lines for an
4270 entire SMTP session, which doesn't seem right. The memory is now reset if a
4271 new message is started.
4273 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4274 debugging output unless log_rewrite was also set.
4276 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4277 of (int)(handle) when we know that handle contains (void *)(-1).
4279 61. The Exim daemon panic-logs an error return when it closes the incoming
4280 connection. However "connection reset by peer" seems to be common, and
4281 isn't really an error worthy of noting specially, so that particular error
4284 62. When Exim is trying to find all the local interfaces, it used to panic and
4285 die if the ioctl to get the interface flags failed. However, it seems that
4286 on at least one OS (Solaris 9) it is possible to have an interface that is
4287 included in the list of interfaces, but for which you get a failure error
4288 for this call. This happens when the interface is not "plumbed" into a
4289 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4290 failure of the "get flags" call assumes that the interface is down.
4292 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4293 makes life easier for people who are doing arithmetic on fields extracted
4294 from dates, where you often get leading zeros that should not be
4295 interpreted as octal.
4297 64. Added qualify_domain to the redirect router, to override the global
4300 65. If a pathologically long header line contained very many addresses (the
4301 report of this problem mentioned 10 000) and each of them was rewritten,
4302 Exim could use up a very large amount of memory. (It kept on making new
4303 copies of the header line as it rewrote, and never released the old ones.)
4304 At the expense of a bit more processing, the header rewriting function has
4305 been changed so that it no longer eats memory in this way.
4307 66. The generation of the Received: header has been moved from the time that a
4308 message starts to be received, to the time that it finishes. The timestamp
4309 in the Received: header should now be very close to that of the <= log
4310 line. There are two side-effects of this change:
4312 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4313 logged header lines no longer include the local Received: line, because
4314 it has not yet been created. The same applies to a copy of the message
4315 that is returned to a non-SMTP sender when a message is rejected.
4317 (b) When a filter file is tested using -bf, no additional Received: header
4318 is added to the test message. After some thought, I decided that this
4321 This change does not affect the value of $received_for. It is still set
4322 after address rewriting, but before local_scan() is called.
4324 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4326 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4327 gave an unhelpful panic error message, and a defer error. I have managed to
4328 change this behaviour so that it now rejects any supplied certificate,
4329 which seems right, as the list of acceptable certificates is empty.
4331 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4332 gave an unhelpful defer error. I have not managed to make this reject any
4333 supplied certificates, but the error message it gives is "no certificate
4334 supplied", which is not helpful.
4336 70. exigrep's output now also includes lines that are not associated with any
4337 message, but which match the given pattern. Implemented by a patch from
4338 Martin Sluka, which also tidied up the Perl a bit.
4340 71. Recipient callout verification, like sender verification, was using <> in
4341 the MAIL FROM command. This isn't really the right thing, since the actual
4342 sender may affect whether the remote host accepts the recipient or not. I
4343 have changed it to use the actual sender in the callout; this means that
4344 the cache record is now keyed on a recipient/sender pair, not just the
4345 recipient address. There doesn't seem to be a real danger of callout loops,
4346 since a callout by the remote host to check the sender would use <>.
4347 [SEE ABOVE: changed after hitting problems.]
4349 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4350 temporary errors. However, in the case of such a code being given after
4351 the end of a data transmission (i.e. after ".") Exim was failing to write
4352 a retry record for the message. (Yes, there was some broken host that was
4353 actually sending 8xx at this point.)
4355 73. An unknown lookup type in a host list could cause Exim to panic-die when
4356 the list was checked. (An example that provoked this was putting <; in the
4357 middle of a list instead of at the start.) If this happened during a DATA
4358 ACL check, a -D file could be left lying around. This kind of configuration
4359 error no longer causes Exim to die; instead it causes a defer errror. The
4360 incident is still logged to the main and panic logs.
4362 74. Buglet left over from Exim 3 conversion. The message "too many messages
4363 in one connection" was written to the rejectlog but not the mainlog, except
4364 when address rewriting (yes!) was being logged.
4366 75. Added write_rejectlog option.
4368 76. When a system filter was run not as root (that is, when system_filter_user
4369 was set), the values of the $n variables were not being returned to the
4370 main process; thus, they were not subsequently available in the $sn
4373 77. Added +return_path_on_delivery log selector.
4375 78. A connection timeout was being treated differently from recipients deferred
4376 when testing hosts_max_try with a message that was older than the host's
4377 retry timeout. (The host should not be counted, thus allowing all hosts to
4378 be tried at least once before bouncing.) This may have been the cause of an
4379 occasionally reported bug whereby a message would remain on the queue
4380 longer than the retry timeout, but would be bounced if a delivery was
4381 forced. I say "may" because I never totally pinned down the problem;
4382 setting up timeout/retry tests is difficult. See also the next item.
4384 79. The ultimate address timeout was not being applied to errors that involved
4385 a combination of host plus message (for example, a timeout on a MAIL
4386 command). When an address resolved to a number of possible hosts, and they
4387 were not all tried for each delivery (e.g. because of hosts_max_try), a
4388 message could remain on the queue longer than the retry timeout.
4390 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4393 81. Fixed an obscure SMTP outgoing bug which required at least the following
4394 conditions: (a) there was another message waiting for the same server;
4395 (b) the server returned 5xx to all RCPT commands in the first message so
4396 that the message was not completed; (c) the server dropped the connection
4397 or gave a negative response to the RSET that Exim sends to abort the
4398 transaction. The observed case was a dropped connection after DATA that had
4399 been sent in pipelining mode. That is, the server had advertised PIPELINING
4400 but was not implementing it correctly. The effect of the bug was incorrect
4401 behaviour, such as trying another host, and this could lead to a crash.
4407 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4408 and daemon.c were passed as pointers to ints; they should have been
4409 pointers to socklen_t variables (which are typically unsigned ints).
4411 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4414 3. Fixed a really odd bug that affected only the testing scheme; patching a
4415 certain fixed string in the binary changed the value of another string that
4416 happened to be identical to the end of the original first string.
4418 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4419 name", it returns that address as the IP address. On some operating
4420 systems (e.g. Solaris), it also passes back the IP address string as the
4421 "host name". However, on others (e.g. Linux), it passes back an empty
4422 string. Exim wasn't checking for this, and was changing the host name to an
4423 empty string, assuming it had been canonicized.
4425 5. Although rare, it is permitted to have more than one PTR record for a given
4426 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4427 all the names associated with an address, because they do in Solaris.
4428 However, it seems that they do not in Linux for data that comes from the
4429 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4430 I found this out when I moved to a new Linux workstation and tried to run
4431 the Exim test suite.
4433 To get round this problem I have changed the code so that it now does its
4434 own call to the DNS to look up PTR records when searching for a host name.
4435 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4436 addresses that are only in /etc/hosts are still found.
4438 This behaviour is, however, controlled by an option called host_lookup_
4439 order, which defaults to "bydns:byaddr". If people want to use the other
4440 order, or indeed, just use one or the other means of lookup, they can
4441 specify it in this variable.
4443 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4444 some operating systems, this comes back from gethostbyaddr() as an empty
4445 string, and this is what Exim used to test for. However, it seems that in
4446 other systems, "." is yielded. Exim now tests for this case too.
4448 7. The values of check_spool_space and check_log_space are now held internally
4449 as a number of kilobytes instead of an absolute number of bytes. If a
4450 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4451 kilobyte. This means that much larger values can be stored.
4453 8. Exim monitor: an attempt to get the action menu when not actually pointing
4454 at a message produces an empty menu entitled "No message selected". This
4455 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4456 no entries in it ("Shell widget menu has zero width and/or height"). So I
4457 have added a single, blank menu entry in this case.
4459 9. Added ${quote_local_part.
4461 10. MIME decoding is now applied to the contents of Subject: header lines when
4464 11. Now that a reference to $sender_host_address automatically causes a reverse
4465 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4466 host lookup before query-style lookups in lists that might use this
4467 variable. This has therefore been abolished, and the "net-" prefix is no
4468 longer necessary for query-style lookups.
4470 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4471 have been a typo for LFLAGS, so it has been changed.
4473 13. The install script calls Exim with "-C /dev/null" in order to find the
4474 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4475 to be output. Howeve, since Exim outputs its version number before the
4476 error, it didn't break the script. It just looked ugly. I fixed this by
4477 always allowing "-C /dev/null" if the caller is root.
4479 14. Ignore overlarge ACL variable number when reading spool file - insurance
4480 against a later release with more variables having written the file.
4482 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4483 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4484 this, as well as the form without the "IPv6" on the front.
4486 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4487 OS/Makefile-Darwin file.
4489 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4490 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4492 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4493 variable. However, some OS use ints for inodes. I've added cast to long int
4494 to get rid of the compiler warning.
4496 19. I had forgotten to lock out "/../" in configuration file names when
4497 ALT_CONFIG_PREFIX was set.
4499 20. Routers used for verification do not need to specify transports. However,
4500 if such a router generated a host list, and callout was configured, Exim
4501 crashed, because it could not find a port number from the (non-existent)
4502 transport. It now assumes port 25 in this circumstance.
4504 21. Added the -t option to exigrep.
4506 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4507 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4508 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4509 them is set, LOOKUP_LSEARCH is forced.
4511 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4513 24. Added sender and host information to the "rejected by local_scan()" log
4514 line; previously there was no indication of these.
4516 25. Added .include_if_exists.
4518 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4519 Linux. It turns out that not all file systems support this. Apparently some
4520 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4521 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4522 not supported on the file descriptor, is now ignored when Exim is trying to
4523 sync a directory. This applies only to Linux.
4525 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4527 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4528 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4531 29. Added support for maildirsize files from supplied patch (modified a bit).
4533 30. The use of :fail: followed by an empty string could lead Exim to respond to
4534 sender verification failures with (e.g.):
4536 550 Verification failed for <xxx>
4537 550 Sender verify failed
4539 where the first response line was missing the '-' that indicates it is not
4540 the final line of the response.
4542 31. The loop for finding the name of the user that called Exim had a hardwired
4543 limit of 10; it now uses the value of finduser_retries, which is used for
4544 all other user lookups.
4546 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4549 33. Exim was neglecting to zero errno before one call of strtol() when
4550 expanding a string and expecting an integer value. On some systems this
4551 resulted in spurious "integer overflow" errors. Also, it was casting the
4552 result into an int without checking.
4554 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4555 did not work. The code looks as if it has *never* worked, though it appears
4556 to have been documented since at least releast 1.62. I have made it work.
4558 35. The "timeout_DNS" error in retry rules, also documented since at least
4559 1.62, also never worked. As it isn't clear exactly what this means, and
4560 clearly it isn't a major issue, I have abolished the feature by treating it
4561 as "timeout", and writing a warning to the main and panic logs.
4563 36. The display of retry rules for -brt wasn't always showing the error code
4566 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4567 timeout_connect_A, timeout_connect_MX.
4569 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4570 to the empty sender.
4572 39. The daemon was not analysing the content of -oX till after it had closed
4573 stderr and disconnected from the controlling terminal. This meant that any
4574 syntax errors were only noted on the panic log, and the return code from
4575 the command was 0. By re-arranging the code a little, I've made the
4576 decoding happen first, so such errors now appear on stderr, and the return
4577 code is 1. However, the actual setting up of the sockets still happens in
4578 the disconnected process, so errors there are still only recorded on the
4581 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4582 connections (as happens on some IP stacks) was logged at start up time as
4583 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4584 it from "IPv6 and IPv4", which means that two separate sockets are being
4587 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4588 says whether AF_INET or AF_INET6 was passed as an argument.
4590 42. Exiwhat output was messed up when time zones were included in log
4593 43. Exiwhat now gives more information about the daemon's listening ports,
4594 and whether -tls-on-connect was used.
4596 44. The "port" option of the smtp transport is now expanded.
4598 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4599 silently ignored. Now an error message is written to the main and panic
4602 46. There's a new ACL modifier called "logwrite" which writes to a log file
4603 as soon as it is encountered.
4605 47. Added $local_user_uid and $local_user_gid at routing time.
4607 48. Exim crashed when trying to verify a sender address that was being
4610 49. Exim was recognizing only a space character after ".include". It now also
4611 recognizes a tab character.
4613 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4614 extracting the relevant information from the specification. The man page no
4615 longer contains scrambled data for the -d option, and I've added a section
4616 at the front about calling Exim under different names.
4618 51. Added "extra_headers" argument to the "mail" command in filter files.
4620 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4623 53. Installed eximstats 1.29.
4625 54. Added transport_filter_timeout as a generic transport option.
4627 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4628 Cc: header lines. This was required by RFC 822, but it not required by RFC
4631 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4632 incoming messages that did not have them. Now it does so only if the
4633 message originates locally, that is, if there is no associated remote host
4634 address. When Resent- header lines are present, this applies to the Resent-
4635 lines rather than the non-Resent- lines.
4637 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4638 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4640 58. Messages for configuration errors now include the name of the main
4641 configuration file - useful now that there may be more than one file in a
4642 list (.included file names were always shown).
4644 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4645 for those rare installations that do not start the daemon as root or run it
4646 setuid root. I've cut out the call to initgroups() if the daemon is not
4649 60. The Exim user and group can now be bound into the binary as text strings
4650 that are looked up at the start of Exim's processing.
4652 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4654 62. Added $mailstore_basename variable.
4656 63. Installed patch to sieve.c from Michael Haardt.
4658 64. When Exim failed to open the panic log after failing to open the main log,
4659 the original message it was trying to log was written to stderr and debug
4660 output, but if they were not available (the usual case in production), it
4661 was lost. Now it is written to syslog before the two lines that record the
4662 failures to open the logs.
4664 65. Users' Exim filters run in subprocesses under the user's uid. It is
4665 possible for a "deliver" command or an alias in a "personal" command to
4666 provoke an address rewrite. If logging of address rewriting is configured,
4667 this fails because the process is not running as root or exim. There may be
4668 a better way of dealing with this, but for the moment (because 4.30 needs
4669 to be released), I have disabled address rewrite logging when running a
4670 filter in a non-root, non-exim process.
4676 1. The buildconfig auxiliary program wasn't quoting the value set for
4677 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4678 not defined. This bug was masked in 4.22 by the effect that was fixed in
4681 2. Some messages that were rejected after a message id was allocated were
4682 shown as "incomplete" by exigrep. It no longer does this for messages that
4683 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4685 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4686 have allow_domain_literals set, the ID did not get logged in the <= line.
4687 Domain literals are now always recognized in Message-ID: header lines.
4689 4. The first argument for a ${extract expansion item is the key name or field
4690 number. Leading and trailing spaces in this item were not being ignored,
4691 causing some misleading effects.
4693 5. When deliver_drop_privilege was set, single queue runner processes started
4694 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4695 same command in the process it spins off) were not dropping privilege.
4697 6. When the daemon running as "exim" started a queue runner, it always
4698 re-executed Exim in the spun-off process. This is a waste of effort when
4699 deliver_drop_privilege is set. The new process now just calls the
4700 queue-runner function directly.
4706 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4709 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4710 ignored. Though the use of -oP was forcing the writing of a pid file, it
4711 was always written to the default place.
4713 3. If the message "no IP address found for host xxxx" is generated during
4714 incoming verification, it is now followed by identification of the incoming
4715 connection (so you can more easily find what provoked it).
4717 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4719 5. Added some features to "harden" Exim a bit more against certain attacks:
4721 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4722 be put in Local/Makefile. This is like the never_users runtime option,
4723 but it cannot be overridden. The default setting is "root".
4725 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4726 prefix string with which any file named in a -C command line option
4729 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4730 is retained for -C and -D only if the caller of Exim is root. Without
4731 it, the exim user may also use -C and -D and retain privilege.
4733 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4734 command line option is disabled.
4736 6. Macro names set by the -D option must start with an upper case letter, just
4737 like macro names defined in the configuration file.
4739 7. Added "dereference=" facility to LDAP.
4741 8. Two instances of the typo "uknown" in the source files are fixed.
4743 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4744 the Configure-Makefile script screwed up while processing it.
4746 10. Incorporated PCRE 4.4.
4748 11. The SMTP synchronization check was not operating right at the start of an
4749 SMTP session. For example, it could not catch a HELO sent before the client
4750 waited for the greeting. There is now a check for outstanding input at the
4751 point when the greeting is written. Because of the duplex, asynchronous
4752 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4753 way, but not yet received, when the check is performed.
4755 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4756 on TCP/IP sockets, because this apparently causes some broken clients to
4759 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4760 unchanged) from the Cygwin maintainer.
4762 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4763 was supported instead of testing for mbx. Effectively a typo.
4765 15. The spa authenticator server code was not checking that the input it
4766 received was valid base64.
4768 16. The debug output line for the "set" modifier in ACLs was not showing the
4769 name of the variable that was being set.
4771 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4773 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4774 lookup on its own. Something else was needed to trigger the lookup. For
4775 example, a match in host_lookup or the need for a host name in a host list.
4776 Now, if $sender_host_name is referenced and the host name has not yet been
4777 looked up, a lookup is performed. If the lookup fails, the variable remains
4778 empty, and $host_lookup_failed is set to "1".
4780 19. Added "eqi" as a case-independent comparison operator.
4782 20. The saslauthd authentication condition could segfault if neither service
4783 nor realm was specified.
4785 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4786 error message that was logged was misleading, and incoming SMTP
4787 connections were dropped. The message is now more accurate, and temporary
4788 errors are given to SMTP connections.
4790 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4791 (except RSET) with a 421 error, until QUIT is received. However, it was
4792 failing to send a response to QUIT.
4794 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4795 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4796 ACL is now run after the helo_try_verify_hosts code.
4798 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4799 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4800 case-independent, so other case variants are also recognized). Apparently
4801 some systems use these upper case variants.
4803 25. If more than two messages were waiting for the same host, and a transport
4804 filter was specified for the transport, Exim sent two messages over the
4805 same TCP/IP connection, and then failed with "socket operation on non-
4806 socket" when it tried to send the third.
4808 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4810 27. The extern definition of crypt16() in expand.c was not being excluded when
4811 the OS had its own crypt16() function.
4813 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4814 as a preferred synonym for return_size_limit, both as an option and as an
4817 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4819 30. Changed the default configuration ACL to relax the local part checking rule
4820 for addresses that are not in any local domains. For these addresses,
4821 slashes and pipe symbols are allowed within local parts, but the sequence
4822 /../ is explicitly forbidden.
4824 31. SPA server authentication was not clearing the challenge buffer before
4827 32. log_message in a "warn" ACL statement was writing to the reject log as
4828 well as to the main log, which contradicts the documentation and doesn't
4829 seem right (because no rejection is happening). So I have stopped it.
4831 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4832 However, I am unable to do any testing of this.
4834 34. Fixed an infelicity in the appendfile transport. When checking directories
4835 for a mailbox, to see if any needed to be created, it was accidentally
4836 using path names with one or more superfluous leading slashes; tracing
4837 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4839 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4840 contents are added to the log line that is written for every discarded
4841 recipient. (Previously a log_message setting was ignored.)
4843 36. The ${quote: operator now quotes the string if it is empty.
4845 37. The install script runs exim in order to find its version number. If for
4846 some reason other than non-existence or emptiness, which it checks, it
4847 could not run './exim', it was installing it with an empty version number,
4848 i.e. as "exim-". This error state is now caught, and the installation is
4851 38. An argument was missing from the function that creates an error message
4852 when Exim fails to connect to the socket for saslauthd authentication.
4853 This could cause Exim to crash, or give a corrupted message.
4855 39. Added isip, isip4, and isip6 to ${if conditions.
4857 40. The ACL variables $acl_xx are now saved with the message, and can be
4858 accessed later in routers, transports, and filters.
4860 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4861 strings in the file are not string-expanded.
4863 42. If a MAIL command specified a SIZE value that was too large to fit into an
4864 int variable, the check against message_size_limit failed. Such values are
4865 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4866 day this will have to be increased, but I don't think I want to be around
4867 when emails are that large.
4874 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4875 iconv() is not standard in FreeBSD.
4877 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4878 IPv6 enabled. The observed symptom was a segmentation fault on return from
4879 the function os_common_find_running_interfaces() in src/os.c.
4881 3. In the check_special_case() function in daemon.c I had used "errno" as an
4882 argument name, which causes warnings on some systems. This was basically a
4883 typo, since it was named "eno" in the comments!
4885 4. The code that waits for the clock to tick (at a resolution of some fraction
4886 of a second) so as to ensure message-id uniqueness was always waiting for
4887 at least one whole tick, when it could have waited for less. [This is
4888 almost certainly not relevant at current processor speeds, where it is
4889 unlikely to ever wait at all. But we try to future-proof.]
4891 5. The function that sleeps for a time interval that includes fractions of a
4892 second contained a race. It did not block SIGALRM between setting the
4893 timer, and suspending (a couple of lines later). If the interval was short
4894 and the sigsuspend() was delayed until after it had expired, the suspension
4895 never ended. On busy systems this could lead to processes getting stuck for
4898 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4899 process, before it forks any delivery processes. The open lookup caching
4900 mechanism meant that the open file or database connection was passed into
4901 the delivery process. The problem was that delivery processes always tidy
4902 up cached lookup data. This could cause a problem for the next delivery
4903 process started by the queue runner, because the external queue runner
4904 process does not know about the closure. So the next delivery process
4905 still has data in the lookup cache. In the case of a file lookup, there was
4906 no problem because closing a file descriptor in a subprocess doesn't affect
4907 the parent. However, if the lookup was caching a connection to a database,
4908 the connection was closed, and the second delivery process was likely to
4909 see errors such as "PGSQL: query failed: server closed the connection
4910 unexpectedly". The problem has been fixed by closing all cached lookups
4911 in a queue runner before running a delivery process.
4913 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4914 seem to have the "const" qualifier which it has on other OS. I've
4917 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4918 *of the same type* (client or server) with the same public name that an
4919 error should be diagnosed.
4921 9. When Exim looked up a host name for an IP address, but failed to find the
4922 original IP address when looking up the host name (a safety check), it
4923 output the message "<ip address> does not match any IP for NULL", which was
4924 confusing, to say the least. The bug was that the host name should have
4925 appeared instead of "NULL".
4927 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4928 user that is built into the binary, and the -C or -D options is used, root
4929 privilege is dropped before the configuration file is read. In addition,
4930 logging is switched to stderr instead of the normal log files. If the
4931 configuration then re-defines the Exim user, the unprivileged environment
4932 is probably not what is expected, so Exim logs a panic warning message (but
4935 However, if deliver_drop_privilege is set, the unprivileged state may well
4936 be exactly what is intended, so the warning has been cut out in that case,
4937 and Exim is allowed to try to write to its normal log files.
4943 1. smtp_return_error_details was not giving details for temporary sender
4944 or receiver verification errors.
4946 2. Diagnose a configuration error if two authenticators have the same public
4949 3. Exim used not to create the message log file for a message until the first
4950 delivery attempt. This could be confusing when incoming messages were held
4951 for policy or load reasons. The message log file is now created at the time
4952 the message is received, and an initial "Received" line is written to it.
4954 4. The automatically generated man page for command line options had a minor
4955 bug that caused no ill effects; however, a more serious problem was that
4956 the procedure for building the man page automatically didn't always
4957 operate. Consequently, release 4.20 contains an out-of-date version. This
4958 shouldn't happen again.
4960 5. When building Exim with embedded Perl support, the script that builds the
4961 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4962 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4964 6. The freeze_tell option was not being used for messages that were frozen on
4965 arrival, either by an ACL or by local_scan().
4967 7. Added the smtp_incomplete_transaction log selector.
4969 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4970 was accepting AUTH without a new EHLO.
4972 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4973 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4976 10. Logging of TCP/IP connections (when configured) now happens in the main
4977 daemon process instead of the child process, so that the TCP/IP connection
4978 count is more accurate (but it can never be perfect).
4980 11. The use of "drop" in a nested ACL was not being handled correctly in the
4981 outer ACL. Now, if condition failure induced by the nested "drop" causes
4982 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4983 or "require"), the connection is dropped.
4985 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4986 that yield "discard" can now be used with an "accept" or a "discard" verb,
4987 but an error is generated for any others (because I can't see a useful way
4988 to define what should happen).
4990 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4991 are now processed in the same way as lines in the Exim configuration file.
4992 In particular, continuation lines are supported.
4994 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4996 15. Added -ti meaning -t -i.
4998 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4999 domains, and warn by logging if others are found.
5001 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5002 returned from GIFCONF when Exim is trying to find the list of interfaces on
5003 a host. The code in os.c has been modified to copy each ifreq to an aligned
5004 structure in all cases.
5006 Also, in some cases, the returned ifreq's were being copied to a 'struct
5007 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5008 means the last couple of bytes of an IPv6 address could be chopped if the
5009 ifreq contained only a normal sockaddr (14 bytes storage).
5011 18. Named domain lists were not supported in the hosts_treat_as_local option.
5012 An entry such as +xxxx was not recognized, and was treated as a literal
5015 19. Ensure that header lines added by a DATA ACL are included in the reject log
5016 if the ACL subsequently rejects the message.
5018 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5019 MD5 (which is deprecated).
5021 21. When testing a filter file using -bf, Exim was writing a message when it
5022 took the sender from a "From " line in the message, but it was not doing so
5023 when it took $return_path from a Return-Path: header line. It now does.
5025 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5026 with a valid header line field name (a series of printing characters
5027 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5029 23. Changed "disc" in the source to "disk" to conform to the documentation and
5030 the book and for uniformity.
5032 24. Ignore Sendmail's -Ooption=value command line item.
5034 25. When execve() failed while trying to run a command in a pipe transport,
5035 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5036 could be confused with a return value of 69 from the command itself. This
5037 has been changed to 127, the value the shell returns if it is asked to run
5038 a non-existent command. The wording for the related log line suggests a
5039 non-existent command as the problem.
5041 26. If received_header_text expands to an empty string, do not add a Received:
5042 header line to the message. (Well, it adds a token one on the spool, but
5043 marks it "old" so that it doesn't get used or transmitted.)
5045 27. Installed eximstats 1.28 (addition of -nt option).
5047 28. There was no check for failure on the call to getsockname() in the daemon
5048 code. This can fail if there is a shortage of resources on the system, with
5049 ENOMEM, for example. A temporary error is now given on failure.
5051 29. Contrary to the C standard, it seems that in some environments, the
5052 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5053 program. Exim now does this explicitly; it affects the formatting of
5054 timestamps using strftime().
5056 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5057 complaints. I've now initialized all the variables, to avoid this.
5059 32. Header lines added by a system filter were not being "seen" during
5060 transport-time rewrites.
5062 33. The info_callback() function passed to OpenSSL is set up with type void
5063 (*)(SSL *, int, int), as described somewhere. However, when calling the
5064 function (actually a macro) that sets it up, the type void(*)() is
5065 expected. I've put in a cast to prevent warnings from picky compilers.
5067 34. If a DNS black list lookup found a CNAME record, but there were no A
5068 records associated with the domain it pointed at, Exim crashed.
5070 35. If a DNS black list lookup returned more than one A record, Exim ignored
5071 all but the first. It now scans all returned addresses if a particular IP
5072 value is being sought. In this situation, the contents of the
5073 $dnslist_value variable are a list of all the addresses, separated by a
5076 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5077 to accept a host name and all its aliases if the forward lookup for any of
5078 them yielded the IP address of the incoming connection. Now it accepts only
5079 those names whose forward lookup yields the correct IP address. Any other
5080 names are discarded. This closes a loophole whereby a rogue DNS
5081 administrator could create reverse DNS records to break through a
5082 wildcarded host restriction in an ACL.
5084 37. If a user filter or a system filter that ran in a subprocess used any of
5085 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5086 the wrong values were passed to the pipe command ($thisaddress had the
5087 value of $0, $0 had the value of $1, etc). This bug was introduced by
5088 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5090 38. Improved the line breaking for long SMTP error messages from ACLs.
5091 Previously, if there was no break point between 40 and 75 characters, Exim
5092 left the rest of the message alone. Two changes have been made: (a) I've
5093 reduced the minimum length to 35 characters; (b) if it can't find a break
5094 point between 35 and 75 characters, it looks ahead and uses the first one
5095 that it finds. This may give the occasional overlong line, but at least the
5096 remaining text gets split now.
5098 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5099 file descriptors might be low, and that setting 1000 would always raise it.
5100 It turns out that in some environments, the limit is already over 1000 and
5101 that lowering it causes trouble. So now Exim takes care not to decrease it.
5103 40. When delivering a message, the value of $return_path is set to $sender_
5104 address at the start of routing (routers may change the value). By an
5105 oversight, this default was not being set up when an address was tested by
5106 -bt or -bv, which affected the outcome if any router or filter referred to
5109 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5110 log only when "message" or "log_message" are set. However, if one of the
5111 conditions was an address verification, or a call to a nested ACL, the
5112 messages generated by the underlying test were being passed through. This
5113 no longer happens. The underlying message is available in $acl_verify_
5114 message for both "message" and "log_message" expansions, so it can be
5115 passed through if needed.
5117 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5118 new expansion $bh_ to give the encoded byte string without charset
5119 translation. Translation happens only if iconv() is available; HAVE_ICONV
5120 indicates this at build time. HEADERS_CHARSET gives the charset to
5121 translate to; headers_charset can change it in the configuration, and
5122 "headers charset" can change it in an individual filter file.
5124 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5125 that creates RFC 2047 encoded "words" labels them as that charset instead
5126 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5127 expansion operator; (ii) when Exim creates a From: line for a local
5128 message; (iii) when a header line is rewritten to include a "phrase" part.
5130 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5131 buggy, causing it to skip the first lines of messages whose message ID
5132 ended in 'D'. This would not have bitten before Exim release 4.14, because
5133 message IDs were unlikely to end in 'D' before then. The effect was to have
5134 incorrect size information for certain domains.
5136 45. #include "config.h" was missing at the start of the crypt16.c module. This
5137 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5140 46. If there was a timeout during a "random" callout check, Exim treated it as
5141 a failure of the random address, and carried on sending RSET and the real
5142 address. If the delay was just some slowness somewhere, the response to the
5143 original RCPT would be taken as a response to RSET and so on, causing
5144 mayhem of various kinds.
5146 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5147 when I implemented it. It didn't allow for the fact that some option values
5148 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5149 the right test for positive values.
5151 48. Domain names in DNS records are case-independent. Exim always looks them up
5152 in lower case. Some resolvers return domain names in exactly the case they
5153 appear in the zone file, that is, they may contain uppercase letters. Not
5154 all resolvers do this - some return always lower case. Exim was treating a
5155 change of case by a resolver as a change of domain, similar to a widening
5156 of a domain abbreviation. This triggered its re-routing code and so it was
5157 trying to route what was effectively the same domain again. This normally
5158 caused routing to fail (because the router wouldn't handle the domain
5159 twice). Now Exim checks for this case specially, and just changes the
5160 casing of the domain that it ultimately uses when it transmits the message
5163 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5166 50. If a filter generated a file delivery with a non-absolute name (possible if
5167 no home directory exists for the router), the forbid_file option was not
5170 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5171 the existing equality matching.
5173 52. Exim was using ints instead of ino_t variables in some places where it was
5174 dealing with inode numbers.
5176 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5177 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5178 called TMPDIR, and if it finds it is different, it changes its value.
5180 54. The smtp_printf() function is now made available to local_scan() so
5181 additional output lines can be written before returning. There is also an
5182 smtp_fflush() function to enable the detection of a dropped connection.
5183 The variables smtp_input and smtp_batched_input are exported to
5186 55. Changed the default runtime configuration: the message "Unknown user"
5187 has been removed from the ACL, and instead placed on the localuser router,
5188 using the cannot_route_message feature. This means that any verification
5189 failures that generate their own messages won't get overridden. Similarly,
5190 the "Unrouteable address" message that was in the ACL for unverifiable
5191 relay addresses has also been removed.
5193 56. Added hosts_avoid_esmtp to the smtp transport.
5195 57. The exicyclog script was not checking for the esoteric option
5196 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5197 will work only if exicyclog is run under the appropriate euid.
5199 58. Following a discussion on the list, the rules by which Exim recognises line
5200 endings on incoming messages have been changed. The -dropcr and drop_cr
5201 options are now no-ops, retained only for backwards compatibility. The
5202 following line terminators are recognized: LF CRLF CR. However, special
5203 processing applies to CR:
5205 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5206 nor a local message in the state where . is a terminator.
5208 (ii) If a bare CR is encountered in a header line, an extra space is added
5209 after the line terminator so as not to end the header. The reasoning
5210 behind this is that bare CRs in header lines are most likely either
5211 to be mistakes, or people trying to play silly games.
5213 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5214 was being incorrectly given as 18 bytes larger than it should have been.
5215 This is a VOB (very old bug).
5217 60. This may never have affected anything current, but just in case it has:
5218 When the local host is found other than at the start of a list of hosts,
5219 the local host, those with the same MX, and any that follow, are discarded.
5220 When the list in question was part of a longer list of hosts, the following
5221 hosts (not currently being processed) were also being discarded. This no
5222 longer happens. I'm not sure if this situation could ever has previously
5225 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5228 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5229 header. This is something that is recommended in a new Internet Draft, and
5230 is something that is documented as being done by Sendmail. There are two
5231 possible values. For messages generated by the autoreply transport, Exim
5234 Auto-Submitted: auto-replied
5236 whereas for all other generated messages (e.g. bounces) it adds
5238 Auto-Submitted: auto-generated
5240 63. The "personal" condition in filters now includes a test for the
5241 Auto-Submitted: header. If it contains the string "auto-" the message it
5242 not considered personal.
5244 64. Added rcpt_include_affixes as a generic transport option.
5246 65. Added queue_only_override (default true).
5248 66. Added the syslog_duplication option.
5250 67. If what should have been the first header line of a message consisted of
5251 a space followed by a colon, Exim was mis-interpreting it as a header line.
5252 It isn't of course - it is syntactically invalid and should therefore be
5253 treated as the start of the message body. The misbehaviour could have
5254 caused a number of strange effects, including loss of data in subsequent
5255 header lines, and spool format errors.
5257 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5258 client host had authenticated. This control can now be exercised by an ACL
5259 for more flexibility.
5261 69. By default, callouts do not happen when testing with -bh. There is now a
5262 variant, -bhc, which does actually run the callout code, including
5263 consulting and updating the callout cache.
5265 70. Added support for saslauthd authentication, courtesy of Alexander
5268 71. If statvfs() failed on the spool or log directories while checking their
5269 size for availability, Exim confusingly gave the error "space shortage".
5270 Furthermore, in debugging mode it crashed with a floating point exception.
5271 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5272 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5273 serious problem, Exim now writes to the main and panic logs when this
5274 happens, with details of the failure. It then refuses to accept the
5275 incoming message, giving the message "spool directory problem" or "log
5276 directory problem" with a 421 code for SMTP messages.
5278 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5279 0, 1, and 2 exist, because some OS complain for execs without them (see
5280 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5281 descriptors. However, the code omitted to check that the open succeeded,
5282 causing mysterious errors if for some reason the permissions on /dev/null
5283 got screwed. Now Exim writes a message to the main and panic logs, and
5284 bombs out if it can't open /dev/null.
5286 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5287 interact so that it is all more flexible. It is supposed to remain
5288 backwards compatible. Also added extra_local_interfaces.
5290 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5291 to bomb out with an assertion failure - to the client this appears as a
5292 connection drop. This problem occurs in the part of the code that was taken
5293 from the Samba project. Fortunately, the assertion is in a very simple
5294 function, so I have fixed this by reproducing the function inline in the
5295 one place where it is called, and arranging for authentication to fail
5296 instead of killing the process with assert().
5298 75. The SPA client code was not working when the server requested OEM rather
5299 than Unicode encoding.
5301 76. Added code to make require_files with a specific uid setting more usable in
5302 the case where statting the file as root fails - usually a non-root-mounted
5303 NFS file system. When this happens and the failure is EACCES, Exim now
5304 forks a subprocess and does the per-uid checking as the relevant uid.
5306 77. Added process_log_path.
5308 78. If log_file_path was not explicitly set, a setting of check_log_space or
5309 check_log_inodes was ignored.
5311 79. If a space check for the spool or log partitions fails, the incident is now
5312 logged. Of course, in the latter case the data may get lost...
5314 80. Added the %p formatting code to string_format() so that it can be used to
5315 print addresses in debug_print(). Adjusted all the address printing in the
5316 debugging in store.c to use %p rather than %d.
5318 81. There was a concern that a line of code in smtp_in.c could overflow a
5319 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5320 initially expressed, the concern was not well-founded, because trailing
5321 spaces are removed early. However, if the trailing spaces were followed by
5322 a NULL, they did not get removed, so the overflow was possible. Two fixes
5325 (a) I re-wrote the offending code in a cleaner fashion.
5326 (b) If an incoming SMTP command contains a NULL character, it is rejected
5329 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5330 runs initgroups(), so that if the Exim user is in any additional groups,
5331 they will be used during message reception.
5337 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob