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 buffer usage at the STARTTLS transition.
22 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
24 TK/01 Updated PolarSSL code to 0.14.2.
25 Bugzilla 1097. Patch from Andreas Metzler.
27 PP/07 Catch divide-by-zero in ${eval:...}.
30 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
33 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
40 NM/01 Workround for PCRE version dependency in version reporting
43 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
44 This fixes portability to compilers other than gcc, notably
45 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
47 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
48 makefiles for portability to HP-UX and POSIX correctness.
50 PP/01 Permit LOOKUP_foo enabling on the make command-line.
51 Also via indented variable definition in the Makefile.
52 (Debugging by Oliver Heesakkers).
54 PP/02 Restore caching of spamd results with expanded spamd_address.
55 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
57 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
58 Improves build reliability. Fix from: Frank Elsner
60 NM/02 Fix wide character breakage in the rfc2047 coding
61 Fixes bug 1064. Patch from Andrey N. Oktyabrski
63 NM/03 Allow underscore in dnslist lookups
64 Fixes bug 1026. Patch from Graeme Fowler
66 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
67 Code patches from Adam Ciarcinski of NetBSD.
69 NM/04 Fixed exiqgrep to cope with mailq missing size issue
72 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
73 is logged, to avoid truncation. Patch from John Horne.
75 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
76 Patch from Jakob Hirsch.
78 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
79 SQL string expansion failure details.
80 Patch from Andrey Oktyabrski.
82 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
83 Patch from Simon Arlott.
85 PP/09 New lookups functionality failed to compile on old gcc which rejects
86 extern declarations in function scope.
87 Patch from Oliver Fleischmann
89 PP/10 Use sig_atomic_t for flags set from signal handlers.
90 Check getgroups() return and improve debugging.
91 Fixed developed for diagnosis in bug 927 (which turned out to be
94 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
95 Patch from Mark Zealey.
97 PP/12 Bugzilla 1056: Improved spamd server selection.
98 Patch from Mark Zealey.
100 PP/13 Bugzilla 1086: Deal with maildir quota file races.
101 Based on patch from Heiko Schlittermann.
103 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
104 Patch from Uwe Doering, sign-off by Michael Haardt.
106 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
107 variable declaration deep within a block. Bug and patch from
110 PP/15 lookups-Makefile IRIX compatibilty coercion.
112 PP/16 Make DISABLE_DKIM build knob functional.
114 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
115 Patch by Simon Arlott
117 TF/03 Fix valgrind.h portability to C89 compilers that do not support
118 variable argument macros. Our copy now differs from upstream.
124 TF/01 Failure to get a lock on a hints database can have serious
125 consequences so log it to the panic log.
127 TF/02 Log LMTP confirmation messages in the same way as SMTP,
128 controlled using the smtp_confirmation log selector.
130 TF/03 Include the error message when we fail to unlink a spool file.
132 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
133 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
134 for maintaining out-of-tree patches for some time.
136 PP/01 Bugzilla 139: Documentation and portability issues.
137 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
138 Handle per-OS dynamic-module compilation flags.
140 PP/02 Let /dev/null have normal permissions.
141 The 4.73 fixes were a little too stringent and complained about the
142 permissions on /dev/null. Exempt it from some checks.
143 Reported by Andreas M. Kirchwitz.
145 PP/03 Report version information for many libraries, including
146 Exim version information for dynamically loaded libraries. Created
147 version.h, now support a version extension string for distributors
148 who patch heavily. Dynamic module ABI change.
150 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
151 privilege escalation vulnerability whereby the Exim run-time user
152 can cause root to append content of the attacker's choosing to
155 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
158 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
159 If dropping privileges for untrusted macros, we disabled normal logging
160 on the basis that it would fail; for the Exim run-time user, this is not
161 the case, and it resulted in successful deliveries going unlogged.
162 Fixed. Reported by Andreas Metzler.
168 PP/01 Date: & Message-Id: revert to normally being appended to a message,
169 only prepend for the Resent-* case. Fixes regression introduced in
170 Exim 4.70 by NM/22 for Bugzilla 607.
172 PP/02 Include check_rfc2047_length in configure.default because we're seeing
173 increasing numbers of administrators be bitten by this.
175 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
177 PP/03 Bugzilla 994: added openssl_options main configuration option.
179 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
181 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
183 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
185 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
186 without a peer certificate, leading to a segfault because of an
187 assumption that peers always have certificates. Be a little more
188 paranoid. Problem reported by Martin Tscholak.
190 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
191 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
192 NB: ClamAV planning to remove STREAM in "middle of 2010".
193 CL also introduces -bmalware, various -d+acl logging additions and
194 more caution in buffer sizes.
196 PP/09 Implemented reverse_ip expansion operator.
198 PP/10 Bugzilla 937: provide a "debug" ACL control.
200 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
202 PP/12 Bugzilla 973: Implement --version.
204 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
206 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
208 PP/15 Bugzilla 816: support multiple condition rules on Routers.
210 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
211 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
212 ignore trailing whitespace.
214 JJ/02 prevent non-panic DKIM error from being sent to paniclog
216 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
219 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
220 Notification from Dr Andrew Aitchison.
222 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
223 ExtendedDetectionInfo response format.
224 Notification from John Horne.
226 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
229 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
230 XSL and documented dependency on system catalogs, with examples of how
233 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
236 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
237 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
238 configuration file which is writeable by the Exim user or group.
240 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
241 of configuration files to cover files specified with the -C option if
242 they are going to be used with root privileges, not just the default
245 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
246 option (effectively making it always true).
248 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
249 files to be used while preserving root privileges.
251 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
252 that rogue child processes cannot use them.
254 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
255 run-time user, instead of root.
257 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
258 Exim run-time user without dropping privileges.
260 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
261 result string, instead of calling string_vformat() twice with the same
264 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
265 for other users. Others should always drop root privileges if they use
266 -C on the command line, even for a whitelisted configure file.
268 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
270 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
276 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
277 $data_path, and $header_path variables; fixed documentation bugs and
280 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
281 exipick to access non-standard spools, including the "frozen" queue
284 NM/01 Bugzilla 965: Support mysql stored procedures.
285 Patch from Alain Williams
287 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
289 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
290 Patch from Andreas Metzler
292 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
293 Patch from Kirill Miazine
295 NM/05 Bugzilla 671: Added umask to procmail example.
297 JJ/03 installed exipick 20100323.0, fixing doc bug
299 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
300 directory. Notification and patch from Dan Rosenberg.
302 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
304 TK/02 Improve log output when DKIM signing operation fails.
306 MH/01 Treat the transport option dkim_domain as a colon separated
307 list, not as a single string, and sign the message with each element,
308 omitting multiple occurences of the same signer.
310 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
311 Bugzilla 985, 986. Patch by Simon Arlott
313 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
314 Patch by Simon Arlott
316 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
317 MBX locking. Notification from Dan Rosenberg.
323 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
325 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
327 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
329 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
331 NM/04 Bugzilla 915: Flag broken perl installation during build.
337 TK/01 Added patch by Johannes Berg that expands the main option
338 "spamd_address" if it starts with a dollar sign.
340 TK/02 Write list of recipients to X-Envelope-Sender header when building
341 the mbox-format spool file for content scanning (suggested by Jakob
344 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
345 (http://www.dcc-servers.net/) support via dccifd. Activated by
346 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
348 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
349 by Mark Daniel Reidel <mr@df.eu>.
351 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
352 When building exim an external PCRE library is now needed -
353 PCRE is a system library on the majority of modern systems.
354 See entry on PCRE_LIBS in EDITME file.
356 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
357 conversation. Added nologin parameter to request.
358 Patch contributed by Kirill Miazine.
360 TF/01 Do not log submission mode rewrites if they do not change the address.
362 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
364 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
365 log files in place. Contributed by Roberto Lima.
367 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
369 TF/03 Bugzilla 615: When checking the local_parts router precondition
370 after a local_part_suffix or local_part_prefix option, Exim now
371 does not use the address's named list lookup cache, since this
372 contains cached lookups for the whole local part.
374 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
375 Robert Millan. Documentation is in experimental-spec.txt.
377 TF/04 Bugzilla 668: Fix parallel build (make -j).
379 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
381 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
382 Patch provided by Jan Srzednicki.
384 TF/05 Leading white space used to be stripped from $spam_report which
385 wrecked the formatting. Now it is preserved.
387 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
388 that they are available at delivery time.
390 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
392 TF/08 TLS error reporting now respects the incoming_interface and
393 incoming_port log selectors.
395 TF/09 Produce a more useful error message if an SMTP transport's hosts
396 setting expands to an empty string.
398 NM/06 Bugzilla 744: EXPN did not work under TLS.
399 Patch provided by Phil Pennock.
401 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
402 Patch provided by Richard Godbee.
404 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
405 acl_smtp_notquit, added index entry.
407 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
408 Patch provided by Eugene Bujak.
410 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
411 accept(). Patch provided by Maxim Dounin.
413 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
414 Patch provided by Phil Pennock.
416 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
418 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
419 Patch provided by Brad "anomie" Jorsch.
421 NM/14 Bugzilla 622: Added timeout setting to transport filter.
422 Patch provided by Dean Brooks.
424 TK/05 Add native DKIM support (does not depend on external libraries).
426 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
427 Patch provided by Graeme Fowler.
429 NM/16 Bugzilla 851: Documentation example syntax fix.
431 NM/17 Changed NOTICE file to remove references to embedded PCRE.
433 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
436 NM/19 Bugzilla 745: TLS version reporting.
437 Patch provided by Phil Pennock.
439 NM/20 Bugzilla 167: bool: condition support.
440 Patch provided by Phil Pennock.
442 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
443 clients. Patch provided by Phil Pennock.
445 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
446 Patch provided by Brad "anomie" Jorsch.
448 NM/23 Bugzilla 687: Fix misparses in eximstats.
449 Patch provided by Heiko Schlittermann.
451 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
452 Patch provided by Heiko Schlittermann.
454 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
455 plus update to original patch.
457 NM/26 Bugzilla 799: Documentation correction for ratelimit.
459 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
460 Patch provided by David Brownlee.
462 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
464 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
466 NM/30 Bugzilla 888: TLS documentation bugfixes.
468 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
470 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
471 Unlike the original bugzilla I have changed all shell scripts in src tree.
473 NM/33 Bugzilla 898: Transport filter timeout fix.
474 Patch by Todd Rinaldo.
476 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
477 Patch by Serge Demonchaux.
479 NM/35 Bugzilla 39: Base64 decode bug fixes.
480 Patch by Jakob Hirsch.
482 NM/36 Bugzilla 909: Correct connect() call in dcc code.
484 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
486 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
488 NM/39 Bugzilla 911: Fixed MakeLinks build script.
494 TK/01 Add preliminary DKIM support. Currently requires a forked version of
495 ALT-N's libdkim that I have put here:
496 http://duncanthrax.net/exim-experimental/
498 Note to Michael Haardt: I had to rename some vars in sieve.c. They
499 were called 'true' and it seems that C99 defines that as a reserved
500 keyword to be used with 'bool' variable types. That means you could
501 not include C99-style headers which use bools without triggering
502 build errors in sieve.c.
504 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
505 as mailq or other aliases. Changed the --help handling significantly
506 to do whats expected. exim_usage() emits usage/help information.
508 SC/01 Added the -bylocaldomain option to eximstats.
510 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
512 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
514 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
520 PH/01 Another patch from the Sieve maintainer.
522 PH/02 When an IPv6 address is converted to a string for single-key lookup
523 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
524 dots are used instead of colons so that keys in lsearch files need not
525 contain colons. This was done some time before quoting was made available
526 in lsearch files. However, iplsearch files do require colons in IPv6 keys
527 (notated using the quote facility) so as to distinguish them from IPv4
528 keys. This meant that lookups for IP addresses in host lists did not work
529 for iplsearch lookups.
531 This has been fixed by arranging for IPv6 addresses to be expressed with
532 colons if the lookup type is iplsearch. This is not incompatible, because
533 previously such lookups could never work.
535 The situation is now rather anomolous, since one *can* have colons in
536 ordinary lsearch keys. However, making the change in all cases is
537 incompatible and would probably break a number of configurations.
539 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
542 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
543 conversion specification without a maximum field width, thereby enabling
544 a rogue spamd server to cause a buffer overflow. While nobody in their
545 right mind would setup Exim to query an untrusted spamd server, an
546 attacker that gains access to a server running spamd could potentially
547 exploit this vulnerability to run arbitrary code as the Exim user.
549 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
550 $primary_hostname instead of what libspf2 thinks the hosts name is.
552 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
553 a directory entry by the name of the lookup key. Previously, if a
554 symlink pointed to a non-existing file or a file in a directory that
555 Exim lacked permissions to read, a lookup for a key matching that
556 symlink would fail. Now it is enough that a matching directory entry
557 exists, symlink or not. (Bugzilla 503.)
559 PH/03 The body_linecount and body_zerocount variables are now exported in the
562 PH/04 Added the $dnslist_matched variable.
564 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
565 This means they are set thereafter only if the connection becomes
568 PH/06 Added the client_condition to authenticators so that some can be skipped
569 by clients under certain conditions.
571 PH/07 The error message for a badly-placed control=no_multiline_responses left
572 "_responses" off the end of the name.
574 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
576 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
577 (without spaces) instead of just copying the configuration text.
579 PH/10 Added the /noupdate option to the ratelimit ACL condition.
581 PH/11 Added $max_received_linelength.
583 PH/12 Added +ignore_defer and +include_defer to host lists.
585 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
586 way in which PCRE > 7.0 is built.
588 PH/14 Implemented queue_only_load_latch.
590 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
591 MAIL command. The effect was to mangle the value on 64-bit systems.
593 PH/16 Another patch from the Sieve maintainer.
595 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
597 PH/18 If a system quota error occurred while trying to create the file for
598 a maildir delivery, the message "Mailbox is full" was not appended to the
599 bounce if the delivery eventually timed out. Change 4.67/27 below applied
600 only to a quota excession during the actual writing of the file.
602 PH/19 It seems that peer DN values may contain newlines (and other non-printing
603 characters?) which causes problems in log lines. The DN values are now
604 passed through string_printing() before being added to log lines.
606 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
607 and InterBase are left for another time.)
609 PH/21 Added message_body_newlines option.
611 PH/22 Guard against possible overflow in moan_check_errorcopy().
613 PH/23 POSIX allows open() to be a macro; guard against that.
615 PH/24 If the recipient of an error message contained an @ in the local part
616 (suitably quoted, of course), incorrect values were put in $domain and
617 $local_part during the evaluation of errors_copy.
623 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
624 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
627 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
628 issue a MAIL command.
630 PH/02 In an ACL statement such as
632 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
634 if a client was not listed at all, or was listed with a value other than
635 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
636 the condition was not true (as it should be), so access was not denied.
637 The bug was that the ! inversion was incorrectly passed on to the second
638 item. This has been fixed.
640 PH/03 Added additional dnslists conditions == and =& which are different from
641 = and & when the dns lookup returns more than one IP address.
643 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
644 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
646 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
647 FSYNC, which compiles an option called disable_fsync that allows for
648 bypassing fsync(). The documentation is heavily laced with warnings.
650 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
652 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
653 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
654 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
655 including adding "make clean"; (3) Added -fPIC when compiling the test
656 dynamically loaded module, to get rid of a warning.
658 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
659 message fails, move_frozen_messages = true and ignore_bounce_errors_after
660 = 0s. The bug is otherwise harmless.
662 PH/07 There was a bug in the dovecot authenticator such that the value of
663 $auth1 could be overwritten, and so not correctly preserved, after a
664 successful authentication. This usually meant that the value preserved by
665 the server_setid option was incorrect.
667 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
669 PH/09 Installed PCRE release 7.0.
671 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
672 run for batched SMTP input. It is now run at the start of every message
673 in the batch. While fixing this I discovered that the process information
674 (output by running exiwhat) was not always getting set for -bs and -bS
675 input. This is fixed, and it now also says "batched" for BSMTP.
677 PH/11 Added control=no_pipelining.
679 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
680 patch, slightly modified), and move the expansion of helo_data till after
681 the connection is made in the smtp transport (so it can use these
684 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
686 PH/14 Added log_selector = +pid.
688 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
690 PH/16 Add ${if forany and ${if forall.
692 PH/17 Added dsn_from option to vary the From: line in DSNs.
694 PH/18 Flush SMTP output before performing a callout, unless control =
695 no_callout_flush is set.
697 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
698 was true (the default) a successful delivery failed to delete the retry
699 item, thus causing premature timeout of the address. The bug is now
702 PH/20 Added hosts_avoid_pipelining to the smtp transport.
704 PH/21 Long custom messages for fakedefer and fakereject are now split up
705 into multiline reponses in the same way that messages for "deny" and
706 other ACL rejections are.
708 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
709 with slight modification.
711 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
712 draft, changing the syntax and factoring some duplicate code".
714 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
715 for deliveries of the second and subsequent messages over the same SMTP
718 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
719 ${reduce, with only minor "tidies".
721 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
723 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
724 expansion side effects.
726 PH/27 When a message times out after an over-quota error from an Exim-imposed
727 quota, the bounce message says "mailbox is full". This message was not
728 being given when it was a system quota that was exceeded. It now should
731 MH/03 Made $recipients available in local_scan(). local_scan() already has
732 better access to the recipient list through recipients_list[], but
733 $recipients can be useful in postmaster-provided expansion strings.
735 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
736 in the case of a MAIL command with additional options following the
737 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
738 were accidentally chopped off.
740 PH/29 SMTP synchronization checks are implemented when a command is read -
741 there is a check that no more input is waiting when there shouldn't be
742 any. However, for some commands, a delay in an ACL can mean that it is
743 some time before the response is written. In this time, more input might
744 arrive, invalidly. So now there are extra checks after an ACL has run for
745 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
746 pipelining has not been advertised.
748 PH/30 MH's patch to allow iscntrl() characters to be list separators.
750 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
751 returned in the SMTP response when smtp_return_error_details was false.
754 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
755 instead of the C I/O that was originally supplied, because problems were
758 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
759 Exim which did not show up earlier: it was assuming that a call to
760 SSL_CTX_set_info_callback() might give an error value. In fact, there is
761 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
762 was a macro that became an assignment, so it seemed to work. This has
763 changed to a proper function call with a void return, hence the compile
764 error. Exim's code has been fixed.
766 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
769 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
771 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
773 PH/37 If a message is not accepted after it has had an id assigned (e.g.
774 because it turns out to be too big or there is a timeout) there is no
775 "Completed" line in the log. When some messages of this type were
776 selected by exigrep, they were listed as "not completed". Others were
777 picked up by some special patterns. I have improved the selection
778 criteria to be more general.
780 PH/38 The host_find_failed option in the manualroute router can now be set
781 to "ignore", to completely ignore a host whose IP address cannot be
782 found. If all hosts are ignored, the behaviour is controlled by the new
783 host_all_ignored option.
785 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
786 homing or because of multiple MX records with /mx) generated more than
787 one IP address, and the following item turned out to be the local host,
788 all the secondary addresses of the first item were incorrectly removed
789 from the list, along with the local host and any following hosts (which
790 is what is supposed to happen).
792 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
793 whoever called Exim into the -H file. In the case of the daemon it was
794 behaving confusingly. When first started, it used values for whoever
795 started the daemon, but after a SIGHUP it used the Exim user (because it
796 calls itself on a restart). I have changed the code so that it now always
799 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
800 message are rejected with the same error (e.g. no authentication or bad
801 sender address), and a DATA command is nevertheless sent (as can happen
802 with PIPELINING or a stupid MUA), the error message that was given to the
803 RCPT commands is included in the rejection of the DATA command. This is
804 intended to be helpful for MUAs that show only the final error to their
807 PH/42 Another patch from the Sieve maintainer.
809 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
810 Eximstats - Fixed some broken HTML links and added missing column headers
812 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
813 columns for Rejects, Temp Rejects, Ham, and Spam rows.
815 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
817 PH/43 Yet another patch from the Sieve maintainer.
819 PH/44 I found a way to check for a TCP/IP connection going away before sending
820 the response to the final '.' that terminates a message, but only in the
821 case where the client has not sent further data following the '.'
822 (unfortunately, this is allowed). However, in many cases there won't be
823 any further data because there won't be any more messages to send. A call
824 to select() can be used: if it shows that the input is "ready", there is
825 either input waiting, or the socket has been closed. An attempt to read
826 the next input character can distinguish the two cases. Previously, Exim
827 would have sent an OK response which the client would never have see.
828 This could lead to message repetition. This fix should cure that, at
829 least in a lot of common cases.
831 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
832 advertised in response to EHLO.
838 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
839 fixed by 4.65/MH/01 (is this a record?) are fixed:
841 (i) An empty string was always treated as zero by the numeric comparison
842 operators. This behaviour has been restored.
844 (ii) It is documented that the numeric comparison operators always treat
845 their arguments as decimal numbers. This was broken in that numbers
846 starting with 0 were being interpreted as octal.
848 While fixing these problems I realized that there was another issue that
849 hadn't been noticed. Values of message_size_limit (both the global option
850 and the transport option) were treated as octal if they started with 0.
851 The documentation was vague. These values are now always treated as
852 decimal, and I will make that clear in the documentation.
858 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
859 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
862 MH/01 Don't check that the operands of numeric comparison operators are
863 integers when their expansion is in "skipping" mode (fixes bug
864 introduced by 4.64-PH/07).
866 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
867 child addresses, Exim now panics and dies. Previously, because the count
868 is held in a short int, deliveries were likely to be lost. As such a
869 large number of recipients for a single message is ridiculous
870 (performance will be very, very poor), I have chosen to impose a limit
871 rather than extend the field.
877 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
878 leftover -K file (the existence of which was triggered by #402).
879 While we were at it, introduced process PID as part of the -K
880 filename. This should rule out race conditions when creating
883 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
884 processing considerably. Previous code took too long for large mails,
885 triggering a timeout which in turn triggers #401.
887 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
888 in the DK code in transports.c. sendfile() is not really portable,
889 hence the _LINUX specificness.
891 TF/01 In the add_headers option to the mail command in an Exim filter,
892 there was a bug that Exim would claim a syntax error in any
893 header after the first one which had an odd number of characters
896 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
897 callout verification, Exim cached a "reject" for the entire domain. This
898 is correct for most verifications, but it is not correct for a recipient
899 verification with use_sender or use_postmaster set, because in that case
900 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
901 case of MAIL FROM:<> rejection from other early rejections (e.g.
902 rejection of HELO). When verifying a recipient using a non-null MAIL
903 address, the cache is ignored if it shows MAIL FROM:<> rejection.
904 Whatever the result of the callout, the value of the domain cache is
905 left unchanged (for any other kind of callout, getting as far as trying
906 RCPT means that the domain itself is ok).
908 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
911 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
912 manifest itself as EPIPE rather than ECONNECT. When tidying away a
913 session, the daemon ignores ECONNECT errors and logs others; it now
914 ignores EPIPE as well.
916 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
917 (quoted-printable decoding).
919 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
920 later the small subsequent patch to fix an introduced bug.
922 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
924 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
926 PH/08 An error is now given if message_size_limit is specified negative.
928 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
929 to be given (somewhat) arbitrary names.
931 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
934 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
935 miscellaneous code fixes
937 PH/10 Added the log_reject_target ACL modifier to specify where to log
940 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
941 hostname. This is wrong, because it relates to the incoming message (and
942 probably the interface on which it is arriving) and not to the outgoing
943 callout (which could be using a different interface). This has been
944 changed to use the value of the helo_data option from the smtp transport
945 instead - this is what is used when a message is actually being sent. If
946 there is no remote transport (possible with a router that sets up host
947 addresses), $smtp_active_hostname is used.
949 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
950 tweaks were necessary in order to get it to work (see also 21 below):
951 (a) The code assumed that strncpy() returns a negative number on buffer
952 overflow, which isn't the case. Replaced with Exim's string_format()
954 (b) There were several signed/unsigned issues. I just did the minimum
955 hacking in of casts. There is scope for a larger refactoring.
956 (c) The code used strcasecmp() which is not a standard C function.
957 Replaced with Exim's strcmpic() function.
958 (d) The code set only $1; it now sets $auth1 as well.
959 (e) A simple test gave the error "authentication client didn't specify
960 service in request". It would seem that Dovecot has changed its
961 interface. Fortunately there's a specification; I followed it and
962 changed what the client sends and it appears to be working now.
964 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
967 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
968 address A is aliased to B and C, where B exists and C does not. Without
969 -v the output is "A verified" because verification stops after a
970 successful redirection if more than one address is generated. However,
971 with -v the child addresses are also verified. Exim was outputting "A
972 failed to verify" and then showing the successful verification for C,
973 with its parentage. It now outputs "B failed to verify", showing B's
974 parentage before showing the successful verification of C.
976 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
977 look up a TXT record in a specific list after matching in a combined
980 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
981 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
982 they consult the DNS. I had assumed they would set it the way they
983 wanted; and indeed my experiments on Linux seem to show that in some
984 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
985 To be on the safe side, however, I have now made the interface to
986 host_find_byname() similar to host_find_bydns(), with an argument
987 containing the DNS resolver options. The host_find_byname() function now
988 sets these options at its start, just as host_find_bydns() does. The smtp
989 transport options dns_qualify_single and dns_search_parents are passed to
990 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
991 of host_find_byname() use the default settings of RES_DEFNAMES
992 (qualify_single) but not RES_DNSRCH (search_parents).
994 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
995 spool_read_header() do less string testing, by means of a preliminary
996 switch on the second character of optional "-foo" lines. (This is
997 overdue, caused by the large number of possibilities that now exist.
998 Originally there were few.) While I was there, I also converted the
999 str(n)cmp tests so they don't re-test the leading "-" and the first
1000 character, in the hope this might squeeze out yet more improvement.
1002 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1003 flag allowing group syntax was set by the header_syntax check but not
1004 turned off, possible causing trouble later; (2) The flag was not being
1005 set at all for the header_verify test, causing "group"-style headers to
1006 be rejected. I have now set it in this case, and also caused header_
1007 verify to ignore an empty address taken from a group. While doing this, I
1008 came across some other cases where the code for allowing group syntax
1009 while scanning a header line wasn't quite right (mostly, not resetting
1010 the flag correctly in the right place). These bugs could have caused
1011 trouble for malformed header lines. I hope it is now all correct.
1013 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1014 with the "reply" argument non-NULL. The code, however (which originally
1015 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1016 but it didn't always do it. This confused somebody who was copying the
1017 code for some other use. I have removed all the tests.
1019 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1020 feature that was used to support insecure browsers during the U.S. crypto
1021 embargo. It requires special client support, and Exim is probably the
1022 only MTA that supported it -- and would never use it because real RSA is
1023 always available. This code has been removed, because it had the bad
1024 effect of slowing Exim down by computing (never used) parameters for the
1025 RSA_EXPORT functionality.
1027 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1028 authenticator to fail if there's a tab character in the incoming data
1029 (there should never be unless someone is messing about, as it's supposed
1030 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1031 if the connection is using TLS or if the remote IP is the same as the
1032 local IP, and the "valid-client-cert option" if a client certificate has
1035 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1036 authenticators. This can be used for authorization after authentication
1037 succeeds. (In the case of plaintext, it servers for both authentication
1040 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1041 if any retry times were supplied.
1043 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1044 connection, where there is no client IP address to check. In this
1045 situation, the verify now always succeeds.
1047 PH/25 Applied John Jetmore's -Mset patch.
1049 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1051 PH/27 In a string expansion for a processed (not raw) header when multiple
1052 headers of the same name were present, leading whitespace was being
1053 removed from all of them, but trailing whitespace was being removed only
1054 from the last one. Now trailing whitespace is removed from each header
1055 before concatenation. Completely empty headers in a concatenation (as
1056 before) are ignored.
1058 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1059 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1061 PH/29 [Removed. This was a change that I later backed out, and forgot to
1062 correct the ChangeLog entry (that I had efficiently created) before
1063 committing the later change.]
1065 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1066 address errors (4xx response to RCPT) over the same connection as other
1067 messages routed to the same hosts. Such deliveries are always "forced",
1068 so retry times are not inspected. This resulted in far too many retries
1069 for the affected addresses. The effect occurred only when there were more
1070 hosts than the hosts_max_try setting in the smtp transport when it had
1071 the 4xx errors. Those hosts that it had tried were not added to the list
1072 of hosts for which the message was waiting, so if all were tried, there
1073 was no problem. Two fixes have been applied:
1075 (i) If there are any address or message errors in an SMTP delivery, none
1076 of the hosts (tried or untried) are now added to the list of hosts
1077 for which the message is waiting, so the message should not be a
1078 candidate for sending over the same connection that was used for a
1079 successful delivery of some other message. This seems entirely
1080 reasonable: after all the message is NOT "waiting for some host".
1081 This is so "obvious" that I'm not sure why it wasn't done
1082 previously. Hope I haven't missed anything, but it can't do any
1083 harm, as the worst effect is to miss an optimization.
1085 (ii) If, despite (i), such a delivery is accidentally attempted, the
1086 routing retry time is respected, so at least it doesn't keep
1087 hammering the server.
1089 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1090 in ${readsocket because some servers need this prod.
1092 PH/32 Added some extra debug output when updating a wait-xxx database.
1094 PH/33 The hint "could be header name not terminated by colon", which has been
1095 given for certain expansion errors for a long time, was not being given
1096 for the ${if def:h_colon_omitted{... case.
1098 PH/34 The spec says: "With one important exception, whenever a domain list is
1099 being scanned, $domain contains the subject domain." There was at least
1100 one case where this was not true.
1102 PH/35 The error "getsockname() failed: connection reset by peer" was being
1103 written to the panic log as well as the main log, but it isn't really
1104 panic-worthy as it just means the connection died rather early on. I have
1105 removed the panic log writing for the ECONNRESET error when getsockname()
1108 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1109 runs only) independently of the message's sender address. This meant
1110 that, if the 4xx error was in fact related to the sender, a different
1111 message to the same recipient with a different sender could confuse
1112 things. In particualar, this can happen when sending to a greylisting
1113 server, but other circumstances could also provoke similar problems.
1114 I have changed the default so that the retry time for these errors is now
1115 based a combination of the sender and recipient addresses. This change
1116 can be overridden by setting address_retry_include_sender=false in the
1119 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1120 remote server are returned as part of bounce messages. This was not
1121 happening for LMTP over a pipe (the lmtp transport), but now it is the
1122 same for both kinds of LMTP.
1124 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1125 in header lines that were in fact CNAMEs. This is no longer the case.
1127 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1128 and queue runs started by the daemon processed all messages. This has
1129 been fixed so that -R and -S can now usefully be given with -q<time>.
1131 PH/40 Import PCRE release 6.7 (fixes some bugs).
1133 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1135 PH/42 Give an error if -q is specified more than once.
1137 PH/43 Renamed the variables $interface_address and $interface_port as
1138 $received_ip_address and $received_port, to make it clear that these
1139 values apply to message reception, and not to the outgoing interface when
1140 a message is delivered. (The old names remain recognized, of course.)
1142 PH/44 There was no timeout on the connect() call when using a Unix domain
1143 socket in the ${readsocket expansion. There now is.
1145 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1146 be meaningful with "accept".
1148 SC/01 Eximstats V1.43
1149 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1151 SC/02 Eximstats V1.44
1152 Use a glob alias rather than an array ref in the generated
1153 parser. This improves both readability and performance.
1155 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1156 Collect SpamAssassin and rejection statistics.
1157 Don't display local sender or destination tables unless
1158 there is data to show.
1159 Added average volumes into the top table text output.
1161 SC/04 Eximstats V1.46
1162 Collect data on the number of addresses (recipients)
1163 as well as the number of messages.
1165 SC/05 Eximstats V1.47
1166 Added 'Message too big' to the list of mail rejection
1167 reasons (thanks to Marco Gaiarin).
1169 SC/06 Eximstats V1.48
1170 Mainlog lines which have GMT offsets and are too short to
1171 have a flag are now skipped.
1173 SC/07 Eximstats V1.49 (Alain Williams)
1174 Added the -emptyok flag.
1176 SC/08 Eximstats V1.50
1177 Fixes for obtaining the IP address from reject messages.
1179 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1180 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1181 whitesspace changes from 4.64-PH/27
1183 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1186 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1187 are found, allow negative numbers in numeric criteria)
1189 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1191 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1192 to match changes made in 4.64-PH/43
1194 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1196 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1197 "domain" is actually an IP address, and force a failure. This locks out
1198 those revolvers/nameservers that support "A-for-A" lookups, in
1199 contravention of the specifications.
1201 PH/48 When a host name was looked up from an IP address, and the subsequent
1202 forward lookup of the name timed out, the host name was left in
1203 $sender_host_name, contrary to the specification.
1205 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1206 restricted to single-key lookups, Exim was not diagnosing an error if
1207 * or *@ was used with a query-style lookup.
1209 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1211 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1212 long ago, but noone interested enough thought of it. Let's just say that
1213 the "1.1" means that there are some new functions that weren't there at
1214 some point in the past.
1216 PH/51 Error processing for expansion failure of helo_data from an smtp
1217 transport during callout processing was broken.
1219 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1220 tested/used via the -bh/-bhc/-bs options.
1222 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1223 bug, fixed in subsequent PCRE releases).
1225 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1226 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1228 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1234 SC/01 Use a glob alias rather than an array ref in eximstats generated
1235 parser. This improves both readability and performance.
1237 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1238 Don't display local sender or destination tables in eximstats unless
1239 there is data to show.
1240 Added average volumes into the eximstats top table text output.
1242 SC/03 Collect data on the number of addresses (recipients) as well
1243 as the number of messages in eximstats.
1245 TF/01 Correct an error in the documentation for the redirect router. Exim
1246 does (usually) call initgroups() when daemonizing.
1248 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1249 with consistent privilege compared to when running as a daemon.
1251 TF/03 Note in the spec that $authenticated_id is not set for local
1252 submissions from trusted users.
1254 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1255 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1257 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1258 by adding some example configuration directives to the default
1259 configuration file. A little bit of work is required to uncomment the
1260 directives and define how usernames and passwords are checked, but
1261 there is now a framework to start from.
1263 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1264 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1265 without this. I don't know how relevant this is to other LDAP libraries.
1267 PH/02 Add the verb name to the "unknown ACL verb" error.
1269 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1271 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1273 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1274 directory not expanded when it should be if an expanded home directory
1275 was set for the address (which is overridden by the transport).
1277 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1280 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1281 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1282 because it is too late at that time, and has no effect.
1284 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1285 security issue with \' (bugzilla #107). I could not use the
1286 PQescapeStringConn() function, because it needs a PGconn value as one of
1289 PH/08 When testing addresses using -bt, indicate those final addresses that
1290 are duplicates that would not cause an additional delivery. At least one
1291 person was confused, thinking that -bt output corresponded to deliveries.
1292 (Suppressing duplicates isn't a good idea as you lose the information
1293 about possibly different redirections that led to the duplicates.)
1295 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1296 systems where poll() doesn't work, in particular OS X.
1298 PH/10 Added more information to debugging output for retry time not reached.
1300 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1301 operations in malware.c.
1303 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1306 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1307 syslog_duplication set false, log lines that would normally be written
1308 both the the main log and to the reject log were not written to syslog at
1311 PH/14 In the default configuration, change the use of "message" in ACL warn
1312 statements to "add_header".
1314 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1315 not followed by a command (e.g. "seen endif").
1317 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1318 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1321 PH/17 Added extra conditions to the default value of delay_warning_condition
1325 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1326 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1327 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1330 The Auto-Submitted: and various List- headers are standardised, whereas I
1331 don't think Precedence: ever was.
1333 PH/18 Refactored debugging code in route_finduser() to show more information,
1334 in particular, the error code if getpwnam() issues one.
1336 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1337 This is apparently needed in addition to the PH/07 change above to avoid
1338 any possible encoding problems.
1340 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1341 but not after initializing Perl.
1343 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1344 output them only if debugging. By default they are written stderr,
1345 apparently, which is not desirable.
1347 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1350 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1353 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1355 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1356 authenticated or an ident call has been made. Suppress the default
1357 values for $authenticated_id and $authenticated_sender (but permit -oMai
1358 and -oMas) when testing with -bh.
1360 PH/24 Re-jigged the order of the tests in the default configuration so that the
1361 tests for valid domains and recipients precede the DNS black list and CSA
1362 tests, on the grounds that those ones are more expensive.
1364 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1365 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1366 This bug exists in every version of Exim that I still have, right back to
1369 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1370 However, an attempt to turn on case-sensitivity in a regex key by
1371 including (?-i) didn't work because the subject string was already
1372 lowercased, and the effects were non-intuitive. It turns out that a
1373 one-line patch can be used to allow (?-i) to work as expected.
1379 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1380 other effects) broke the use of negated acl sub-conditions.
1382 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1385 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1386 "Deny" causes Exim to reject the incoming connection with a 554 error.
1387 Unfortunately, if there is a major crisis, such as a disk failure,
1388 tcp-wrappers gives "deny", whereas what one would like would be some
1389 kind of temporary error. A kludge has been added to help with this.
1390 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1391 554 error is used if errno is still zero or contains ENOENT (which occurs
1392 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1395 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1397 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1398 errors. Otherwise a message that provokes a temporary error (when other
1399 messages do not) can cause a whole host to time out.
1401 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1402 addresses were routed directly to files or pipes from a redirect router.
1403 File deliveries just didn't batch; pipe deliveries might have suffered
1406 PH/06 A failure to get a lock for a hints database would erroneously always say
1407 "Failed to get write lock", even when it was really a read lock.
1409 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1410 of 0600. This has been changed to use the value of the lockfile_mode
1411 option (which defaults to 0600).
1413 PH/08 Applied small patch from the Sieve maintainer.
1415 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1416 folder from quota calculations, a direct delivery into this folder messed
1417 up the contents of the maildirsize file. This was because the regex was
1418 used only to exclude .Trash (or whatever) when the size of the mailbox
1419 was calculated. There was no check that a delivery was happening into an
1420 excluded directory. This bug has been fixed by ignoring all quota
1421 processing for deliveries into excluded directories.
1423 PH/10 Added the maildirfolder_create_regex option to appendfile.
1429 PH/01 The code for finding all the local interface addresses on a FreeBSD
1430 system running IPv6 was broken. This may well have applied to all BSD
1431 systems, as well as to others that have similar system calls. The broken
1432 code found IPv4 interfaces correctly, but gave incorrect values for the
1433 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1434 that it would not match correctly against @[] and not recognize the IPv6
1437 PH/02 The ipliteral router was not recognizing addresses of the form user@
1438 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1440 PH/03 Added disable_ipv6.
1442 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1443 decoded form, because it is most often used to construct To: headers
1444 lines in autoreplies, and the decoded form may well be syntactically
1445 invalid. However, $reply_address has leading white space removed, and all
1446 newlines turned into spaces so that the autoreply transport does not
1449 PH/05 If group was specified without a user on a router, and no group or user
1450 was specified on a transport, the group from the router was ignored.
1452 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1453 for visible compile-time settings that can be used to change these
1454 numbers, for those that want even more. Backwards compatibility with old
1455 spool files has been maintained. However, going back to a previous Exim
1456 release will lost any variables that are in spool files.
1458 PH/07 Two small changes when running in the test harness: increase delay when
1459 passing a TCP/IP connection to a new process, in case the original
1460 process has to generate a bounce, and remove special handling of
1461 127.0.0.2 (sic), which is no longer necessary.
1463 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1464 be the same on different OS.
1466 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1469 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1470 whether --show-vars was specified or not
1472 JJ/02 exipick: Added support for new ACL variable spool format introduced
1475 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1476 syntactically invalid From: or Reply-to: line, and a filter used this to
1477 generate an autoreply, and therefore failed to obtain an address for the
1478 autoreply, Exim could try to deliver to a non-existent relative file
1479 name, causing unrelated and misleading errors. What now happens is that
1480 it logs this as a hard delivery error, but does not attempt to create a
1483 PH/11 The exinext utility has a -C option for testing purposes, but although
1484 the given file was scanned by exinext itself; it wasn't being passed on
1485 when Exim was called.
1487 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1488 an end-of-file indication when reading a command response.
1490 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1491 compiled. In many other places in Exim, IPv6 addresses are always
1492 recognized, so I have changed this. It also means that IPv4 domain
1493 literals of the form [IPV4:n.n.n.n] are now always recognized.
1495 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1496 used if the router is not running as root, for example, when verifying at
1497 ACL time, or when using -bh. The debugging output from this situation was
1498 non-existent - all you got was a failure to exec. I have made two
1501 (a) Failures to set uid/gid, the current directory, or a process leader
1502 in a subprocess such as that created by queryprogram now generate
1503 suitable debugging ouput when -d is set.
1505 (b) The queryprogram router detects when it is not running as root,
1506 outputs suitable debugging information if -d is set, and then runs
1507 the subprocess without attempting to change uid/gid.
1509 PH/15 Minor change to Makefile for building test_host (undocumented testing
1512 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1513 additional section of a DNS packet that returns MX or SRV records.
1514 Instead, it always explicitly searches for A/AAAA records. This avoids
1515 major problems that occur when a DNS server includes only records of one
1516 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1517 fixed another bug: if SRV records were looked up and the corresponding
1518 address records were *not* found in the additional section, the port
1519 values from the SRV records were lost.
1521 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1522 using the correct key (the original address) when searching the retry
1523 rules in order to find which one to use for generating the retry hint.
1525 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1526 adding the default one. Similarly, if it contains a Reply-To: header, the
1527 errors_reply_to option, if set, is not used.
1529 PH/19 When calculating a retry time, Exim used to measure the "time since
1530 failure" by looking at the "first failed" field in the retry record. Now
1531 it does not use this if it is later than than the arrival time of the
1532 message. Instead it uses the arrival time. This makes for better
1533 behaviour in cases where some deliveries succeed, thus re-setting the
1534 "first failed" field. An example is a quota failure for a huge message
1535 when small messages continue to be delivered. Without this change, the
1536 "time since failure" will always be short, possible causing more frequent
1537 delivery attempts for the huge message than are intended.
1538 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1540 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1541 $1, $2, $3) because the numerical variables can be reset during some
1542 expansion items (e.g. "match"), thereby losing the authentication data.
1544 PH/21 Make -bV show the size of off_t variables so that the test suite can
1545 decide whether to run tests for quotas > 2G.
1547 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1548 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1549 filecount value is greater than 2G or if a quota value is greater than 2G
1550 on a system where the size of off_t is not greater than 4, a panic error
1553 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1554 never match. The debug and -bh output now contains an explicit error
1555 message indicating a malformed IPv4 address or mask.
1557 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1558 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1559 PH/23 above applies.
1561 PH/25 Do not write to syslog when running in the test harness. The only
1562 occasion when this arises is a failure to open the main or panic logs
1563 (for which there is an explicit test).
1565 PH/26 Added the /no_tell option to "control=freeze".
1567 PH/27 If a host name lookup failed very early in a connection, for example, if
1568 the IP address matched host_lookup and the reverse lookup yielded a name
1569 that did not have a forward lookup, an error message of the form "no IP
1570 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1571 could be logged. Now it outputs the IP address instead of "NULL".
1573 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1574 allows the sender and the authenticated sender to be set when
1575 submitting a message from within Exim. Since child_open_exim() is
1576 documented for local_scan(), the new function should be too.
1578 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1579 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1580 results in an empty string is now treated as unset.
1582 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1584 PH/31 Added sender_verify_fail logging option.
1586 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1587 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1588 tidied the source and removed it altogether.
1590 PH/33 When a queue run was abandoned because the load average was too high, a
1591 log line was always written; now it is written only if the queue_run log
1592 selector is set. In addition, the log line for abandonment now contains
1593 information about the queue run such as the pid. This is always present
1594 in "start" and "stop" lines but was omitted from the "abandon" line.
1596 PH/34 Omit spaces between a header name and the colon in the error message that
1597 is given when verify = headers_syntax fails (if there are lots of them,
1598 the message gets confusing).
1600 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1601 names, as there are now some PTR records that contain slashes. This check
1602 is only to protect against broken name servers that fall over on strange
1603 characters, so the fact that it applies to all lookups doesn't matter.
1605 PH/36 Now that the new test suite is complete, we can remove some of the
1606 special code in Exim that was needed for the old test suite. For example,
1607 sorting DNS records because real resolvers return them in an arbitrary
1608 order. The new test suite's fake resolver always returns records in the
1611 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1612 bounces) except when queue_only is set, to avoid logging races between
1613 the different processes.
1615 PH/38 Panic-die if .include specifies a non-absolute path.
1617 PH/39 A tweak to the "H" retry rule from its user.
1619 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1620 a label. They prevented compilation on older perls.
1622 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1623 a warning to be raised on newish perls.
1625 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1626 on queue. Changes to match documented behaviour of showing count of
1627 messages matching specified criteria.
1629 PH/40 Changed the default ident timeout from 30s to 5s.
1631 PH/41 Added support for the use of login_cap features, on those BSD systems
1632 that have them, for controlling the resources used by pipe deliveries.
1634 PH/42 The content-scanning code uses fopen() to create files in which to put
1635 message data. Previously it was not paying any attention to the mode of
1636 the files. Exim runs with umask(0) because the rest of the code creates
1637 files with open(), and sets the required mode explicitly. Thus, these
1638 files were ending up world-writeable. This was not a big issue, because,
1639 being within the spool directory, they were not world-accessible. I have
1640 created a function called modefopen, which takes an additional mode
1641 argument. It sets umask(777), creates the file, chmods it to the required
1642 mode, then resets the umask. All the relevant calls to fopen() in the
1643 content scanning code have been changed to use this function.
1645 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1646 to 24 hours. This avoids potential overflow problems when processing G
1647 and H retry rules. I suspect nobody ever tinkers with this value.
1649 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1651 PH/45 When the plaintext authenticator is running as a client, the server's
1652 challenges are checked to ensure they are valid base64 strings. By
1653 default, the authentication attempt is cancelled if an invalid string is
1654 received. Setting client_ignore_invalid_base64 true ignores these errors.
1655 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1656 they are received. Thus, the responses can be made to depend on the
1657 challenges. If an invalid string is ignored, an empty string is placed in
1660 PH/46 Messages that are created by the autoreply transport now contains a
1661 References: header, in accordance with RFCs 2822 and 3834.
1663 PH/47 Added authenticated_sender_force to the smtp transport.
1665 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1667 PH/49 Installed latest patch from the Sieve maintainer.
1669 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1670 was also set, the appendfile transport was unnecessarily scanning a
1671 directory of message files (e.g. for maildir delivery) to find the count
1672 of files (along with the size), even though it did not need this
1673 information. It now does the scan only if it needs to find either the
1674 size of the count of files.
1676 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1678 PH/52 Two bugs concerned with error handling when the smtp transport is
1681 (i) Exim was not creating retry information for temporary errors given
1682 for individual recipients after the DATA command when the smtp transport
1683 was used in LMTP mode. This meant that they could be retried too
1684 frequently, and not timed out correctly.
1686 (ii) Exim was setting the flag that allows error details to be returned
1687 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1688 recipients that were returned after the DATA command.
1690 PH/53 This is related to PH/52, but is more general: for any failing address,
1691 when detailed error information was permitted to be returned to the
1692 sender, but the error was temporary, then after the final timeout, only
1693 "retry timeout exceeded" was returned. Now it returns the full error as
1694 well as "retry timeout exceeded".
1696 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1697 do this, and (what is worse) MTAs that accept it.
1699 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1700 will now be deprecated.
1702 PH/56 New os.c-cygwin from the Cygwin maintainer.
1704 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1705 formats (previously only available in exim formats via -bpr, -bpru,
1706 and -bpra. Now also available in native and exiqgrep formats)
1708 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1709 with very large, slow to parse queues
1711 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1713 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1715 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1716 responses to authentication challenges, though it was showing the
1717 challenges; (ii) I've removed the CR characters from the debug output for
1720 PH/58 Allow for the insertion of a newline as well as a space when a string
1721 is turned into more than one encoded-word during RFC 2047 encoding. The
1722 Sieve code now uses this.
1724 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1725 data_4xx, lost_connection, tls_required.
1727 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1728 message was being sent as an SMTP response.
1730 PH/61 Add -l and -k options to exicyclog.
1732 PH/62 When verifying, if an address was redirected to one new address, so that
1733 verification continued, and the new address failed or deferred after
1734 having set something in $address_data, the value of $address_data was not
1735 passed back to the ACL. This was different to the case when no
1736 redirection occurred. The value is now passed back in both cases.
1738 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1739 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1740 use login_cap.h, so on its own it isn't the distinguishing feature. The
1741 new name refers directly to the setclassresources() function.
1743 PH/65 Added configuration files for NetBSD3.
1745 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1747 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1748 is preferred over IPv4.
1750 PH/68 The bounce_return_message and bounce_return_body options were not being
1751 honoured for bounces generated during the reception of non-SMTP messages.
1752 In particular, this applied to messages rejected by the ACL. This bug has
1753 been fixed. However, if bounce_return_message is true and bounce_return_
1754 body is false, the headers that are returned for a non-SMTP message
1755 include only those that have been read before the error was detected.
1756 (In the case of an ACL rejection, they have all been read.)
1758 PH/69 The HTML version of the specification is now built in a directory called
1759 spec_html instead of spec.html, because the latter looks like a path with
1760 a MIME-type, and this confuses some software.
1762 PH/70 Catch two compiler warnings in sieve.c.
1764 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1765 function verify_get_ident() calls ip_connect() to connect a socket, but
1766 if the "connect()" function timed out, ip_connect() used to close the
1767 socket. However, verify_get_ident() also closes the socket later, and in
1768 between Exim writes to the log, which may get opened at this point. When
1769 the socket was closed in ip_connect(), the log could get the same file
1770 descriptor number as the socket. This naturally causes chaos. The fix is
1771 not to close the socket in ip_connect(); the socket should be closed by
1772 the function that creates it. There was only one place in the code where
1773 this was missing, in the iplookup router, which I don't think anybody now
1774 uses, but I've fixed it anyway.
1776 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1777 well as to direct DNS lookups. Otherwise the handling of names in host
1778 lists is inconsistent and therefore confusing.
1784 PH/01 Two changes to the default runtime configuration:
1786 (1) Move the checks for relay_from_hosts and authenticated clients from
1787 after to before the (commented out) DNS black list checks.
1789 (2) Add control=submission to the relay_from_hosts and authenticated
1790 clients checks, on the grounds that messages accepted by these
1791 statements are most likely to be submissions.
1793 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1795 (1) Generate an error if the third argument for the ${prvs expansion is
1798 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1801 (3) Reset the variables that are obtained from the first argument of
1802 ${prvscheck and used in the second argument before leaving the code,
1803 because their memory is reclaimed, so using them afterwards may do
1806 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1807 one (it's much easier than Tom thought :-).
1809 (5) Because of (4), we can now allow for the use of $prvscheck_result
1810 inside the third argument.
1812 PH/03 For some reason, the default setting of PATH when running a command from
1813 a pipe transport was just "/usr/bin". I have changed it to
1816 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1817 anything to be listed in the output from -bV.
1819 PH/05 When a filter generated an autoreply, the entire To: header line was
1820 quoted in the delivery log line, like this:
1822 => >A.N.Other <ano@some.domain> <original@ddress> ...
1824 This has been changed so that it extracts the operative address. There
1825 may be more than one such address. If so, they are comma-separated, like
1828 => >ano@some.domain,ona@other.domain <original@ddress> ...
1830 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1831 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1832 not being looked up in the rDNS to get a host name, Exim was showing the
1833 IP address twice in Received: lines, even though the IP addresses were
1834 identical. For example:
1836 Received: from [1.2.3.4] (helo=[1.2.3.4])
1838 However, if the real host name was known, it was omitting the HELO data
1839 if it matched the actual IP address. This has been tidied up so that it
1840 doesn't show the same IP address twice.
1842 PH/07 When both +timestamp and +memory debugging was on, the value given by
1843 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1844 called by the debug printing, thereby overwriting the timestamp buffer.
1845 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1847 PH/08 When the original message was included in an autoreply transport, it
1848 always said "this is a copy of the message, including all the headers",
1849 even if body_only or headers_only was set. It now gives an appropriate
1852 PH/09 Applied a patch from the Sieve maintainer which:
1854 o fixes some comments
1855 o adds the (disabled) notify extension core
1856 o adds some debug output for the result of if/elsif tests
1857 o points to the current vacation draft in the documentation
1858 and documents the missing references header update
1862 o fixes a bug in processing the envelope test (when testing
1863 multiple envelope elements, the last element determinted the
1866 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1867 Electronic Mail") by including:
1869 Auto-submitted: auto-generated
1871 in the messages that it generates (bounce messages and others, such as
1872 warnings). In the case of bounce messages for non-SMTP mesages, there was
1873 also a typo: it was using "Auto_submitted" (underscore instead of
1874 hyphen). Since every message generated by Exim is necessarily in response
1875 to another message, thes have all been changed to:
1877 Auto-Submitted: auto-replied
1879 in accordance with these statements in the RFC:
1881 The auto-replied keyword:
1883 - SHOULD be used on messages sent in direct response to another
1884 message by an automatic process,
1886 - MUST NOT be used on manually-generated messages,
1888 - MAY be used on Delivery Status Notifications (DSNs) and Message
1889 Disposition Notifications (MDNs),
1891 - MUST NOT be used on messages generated by automatic or periodic
1892 processes, except for messages which are automatic responses to
1895 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1896 to the default Received: header definition.
1898 PH/12 Added log selector acl_warn_skipped (default on).
1900 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1901 variables because (a) they are in the wrong storage pool and (b) even if
1902 they were copied, it wouldn't work properly because of the caching.
1904 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1905 checking when decoding. Apparently there are clients that generate
1906 overlong encoded strings. Why am I not surprised?
1908 PH/15 If the first argument of "${if match_address" was not empty, but did not
1909 contain an "@" character, Exim crashed. Now it writes a panic log message
1910 and treats the condition as false.
1912 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1914 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1915 extension "envelope-auth". The code is finished and in agreement with
1916 other implementations, but there is no documentation so far and in fact,
1917 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1918 not changing the active code.
1920 Print executed "if" and "elsif" statements when debugging is used. This
1921 helps a great deal to understand what a filter does.
1923 Document more things not specified clearly in RFC3028. I had all this
1924 sorted out, when out of a sudden new issues came to my mind. Oops."
1926 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1929 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1930 canonical form (as documented). However, after a host name lookup from
1931 the IP address, check_host() was doing a simple string comparison with
1932 addresses acquired from the DNS when checking that the found name did
1933 have the original IP as one of its addresses. Since any found IPv6
1934 addresses are likely to be in abbreviated form, the comparison could
1935 fail. Luckily, there already exists a function for doing the comparison
1936 by converting both addresses to binary, so now that is used instead of
1937 the text comparison.
1939 PH/20 There was another similar case to PH/19, when a complete host name was
1940 given in a host list; looking up its IP address could give an abbreviated
1941 form, whereas the current host's name might or might not be abbreviated.
1942 The same fix has been applied.
1948 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1949 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1952 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1953 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1955 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1957 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1958 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1959 Also, for "Auto-Submitted", treat anything other than "no" as
1960 non-personal, in accordance with RFC 3834. (Previously it treated
1961 anything starting "auto-" as non-personal.)
1963 TF/01 The control=submission/name=... option had a problem with syntax
1964 errors if the name included a slash character. The /name= option
1965 now slurps the rest of the string, so it can include any characters
1966 but it must come last in the list of options (after /sender_retain
1969 PH/05 Some modifications to the interface to the fake nameserver for the new
1977 TK/01 Added the "success_on_redirect" address verification option. See
1978 NewStuff for rationale and an example.
1980 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1982 PH/02 Patch to exigrep to allow it to work on syslog lines.
1984 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1985 fread() to skip over the body file's header line, because in Cygwin the
1986 header line is locked and is inaccessible.
1988 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1989 co-exist for some time) to make it clear that it is the Exim ID that is
1990 referenced, not the Message-ID: header line.
1992 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1993 string_format() function, because snprintf() does not exist on all
1996 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1997 ${stat: expansion item.
1999 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2000 protocol synchronization error", to keep the pedants happy.
2002 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2003 well as for IRIX systems, when gcc is being used. See the host.c source
2006 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2008 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2011 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2012 transport and to the smtp transport in LMTP mode.
2014 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2016 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2017 run a filter in a subprocess. This could lead to confusion in subsequent
2018 lookups in the parent process. There should also be a search_tidyup() at
2019 the end of the subprocess.
2021 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2022 only if the host matched helo_try_verify_hosts, which caused the
2023 verification to occur when the EHLO/HELO command was issued. The ACL just
2024 tested the remembered result. Now, if a previous verification attempt has
2025 not happened, "verify = helo" does it there and then.
2027 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2029 TK/03 Fix log output including CR from clamd.
2031 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2032 exist provoked a memory error which could cause a segfault.
2034 PH/15 Installed PCRE 6.2
2036 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2038 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2039 of the problem. Specifically, suggested +O2 rather than +O1 for the
2042 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2044 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2045 record, the port was not correctly logged when the outgoing_port log
2046 selector was set (it logged the transort's default port).
2048 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2049 fallback_hosts, and "hosts" in the smtp transport.
2051 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2052 host errors such as "Connection refused".
2054 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2055 authentication with radiusclient 0.4.9:
2057 - Error returned from rc_read_config was caught wrongly
2058 - Username/password not passed on to radius server due to wrong length.
2060 The presumption is that some radiusclient API changes for 4.51/PH/17
2061 were not taken care of correctly. The code is still untested by me (my
2062 Linux distribution still has 0.3.2 of radiusclient), but it was
2063 contributed by a Radius user.
2065 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2066 expanding the "port" option of the smtp transport.
2068 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2069 while reading a MIME header. Thanks to Tom Hughes for a patch.
2071 PH/24 Include config.h inside local_scan.h so that configuration settings are
2074 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2075 that in an ACL for RCPT (for example), you can examine exactly what was
2078 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2079 commands, but it was not correctly comparing the address with the actual
2080 client host address. Thus, it would show the EHLO address in Received:
2081 header lines when this was not necessary.
2083 PH/27 Added the % operator to ${eval:}.
2085 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2086 it should be ignoring failures (because with -C, for example, it has lost
2087 privilege). It wasn't ignoring creation failures other than "already
2090 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2093 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2094 because an input line was too long, either on its own, or by virtue of
2095 too many continuations, the temporary file was not being removed, and the
2096 return code was incorrect.
2098 PH/31 Missing "BOOL" in function definition in filtertest.c.
2100 PH/32 Applied Sieve patches from the maintainer.
2102 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2104 PH/33 Added "verify = not_blind".
2106 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2107 Local/Makefile (with some defaults set). These are used in built scripts
2108 such as exicyclog, but they have never been used in the exim_install
2109 script (though there are many overriding facilities there). I have
2110 arranged that the exim_install script now takes note of these two
2113 PH/35 Installed configuration files for Dragonfly.
2115 PH/36 When a locally submitted message by a trusted user did not contain a
2116 From: header, and the sender address was obtained from -f or from an SMTP
2117 MAIL command, and the trusted user did not use -F to supply a sender
2118 name, $originator_name was incorrectly used when constructing a From:
2119 header. Furthermore, $originator_name was used for submission mode
2120 messages from external hosts without From: headers in a similar way,
2121 which is clearly wrong.
2123 PH/37 Added control=suppress_local_fixups.
2125 PH/38 When log_selector = +received_sender was set, and the addition of the
2126 sender made the log line's construction buffer exactly full, or one byte
2127 less than full, an overflow happened when the terminating "\n" was
2130 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2131 when the result of a list match is failure because a DNS lookup failed.
2133 PH/40 RM_COMMAND is now used in the building process.
2135 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2136 the "build-* directories that it finds.
2138 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2139 address in a domain literal was a prefix of an interface address.
2141 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2142 when verifying a sender address, unless rewrite_headers is false.
2144 PH/44 Wrote a long comment about why errors_to addresses are verified as
2145 recipients, not senders.
2147 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2148 the ratelimit ACL was added.
2150 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2152 PH/46 Added extra information about PostgreSQL errors to the error string.
2154 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2155 avoiding the need to install special zones in a real server. This is
2156 backwards compatible; if it can't find the fake resolver, it drops back.
2157 Thus, both old and new test suites can be run.
2159 TF/02 Added util/ratelimit.pl
2161 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2162 clock is set back in time.
2164 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2165 Candler <B.Candler@pobox.com>.
2167 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2168 OK for addresses that are the result of redirections.
2170 PH/48 A number of further additions for the benefit of the new test suite,
2171 including a fake gethostbyname() that interfaces to the fake DNS resolver
2174 TF/06 The fix for widen_domains has also been applied to qualify_single and
2175 search_parents which are the other dnslookup options that can cause
2178 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2181 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2183 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2184 service" if it encountered an MX record with an empty target hostname.
2185 The message is now "an MX or SRV record indicated no SMTP service".
2187 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2188 if the DNS of the sending site is misconfigured. This is quite a
2189 common situation. This change restores the behaviour of treating a
2190 helo verification defer as a failure.
2192 PH/51 If self=fail was set on a router, the bounce message did not include the
2193 actual error message.
2199 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2201 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2202 command itself ended in error, the underlying message about the transport
2203 filter timeout was being overwritten with the pipe command error. Now the
2204 underlying error message should be appended to the second error message.
2206 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2208 PH/02 Reduce the amount of output that "make" produces by default. Full output
2209 can still be requested.
2211 PH/03 The warning log line about a condition test deferring for a "warn" verb
2212 was being output only once per connection, rather than after each
2213 occurrence (because it was using the same function as for successful
2214 "warn" verbs). This seems wrong, so I have changed it.
2216 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2217 it should not have, which might have caused a crash in the right
2218 circumstances, but probably never did.
2220 PH/04 Installed a modified version of Tony Finch's patch to make submission
2221 mode fix the return path as well as the Sender: header line, and to
2222 add a /name= option so that you can make the user's friendly name appear
2225 TF/03 Added the control = fakedefer ACL modifier.
2227 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2228 Mark Lowes for thorough testing.
2230 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2232 TK/03 Merged latest SRS patch from Miles Wilton.
2234 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2235 with the definition in sysexits.h (which is #included earlier).
2236 Fortunately, Exim does not actually use EX_OK. The code used to try to
2237 preserve the sysexits.h value, by assumimg that macro definitions were
2238 scanned for macro replacements. I have been disabused of this notion,
2239 so now the code just undefines EX_OK before #including unistd.h.
2241 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2242 in the smtp transport. When a block could not be written in a single
2243 write() function, the timeout was being re-applied to each part-write.
2244 This seems wrong - if the receiver was accepting one byte at a time it
2245 would take for ever. The timeout is now adjusted when this happens. It
2246 doesn't have to be particularly precise.
2248 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2249 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2251 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2252 without a domain if the check to <postmaster@domain> fails.
2254 SC/01 Eximstats: added -xls and the ability to specify output files
2255 (patch written by Frank Heydlauf).
2257 SC/02 Eximstats: use FileHandles for outputing results.
2259 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2261 SC/04 Eximstats: fixed display of large numbers with -nvr option
2263 SC/05 Eximstats: fixed merging of reports with empty tables.
2265 SC/06 Eximstats: added the -include_original_destination flag
2267 SC/07 Eximstats: removed tabs and trailing whitespace.
2269 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2271 TK/06 MBOX spool code: Add real "From " MBOX separator line
2272 so the .eml file is really in mbox format (even though
2273 most programs do not really care). Patch from Alex Miller.
2275 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2276 The latter is generated from $received_to and is only set if the
2277 message has one envelope recipient. SA can use these headers,
2278 obviously out-of-the-box. Patch from Alex Miller.
2280 PH/08 The ${def test on a variable was returning false if the variable's
2281 value was "0", contrary to what the specification has always said!
2282 The result should be true unless the variable is empty.
2284 PH/09 The syntax error of a character other than { following "${if
2285 def:variable_name" (after optional whitespace) was not being diagnosed.
2286 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2287 accidental colon was present, for example, could give incorrect results.
2289 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2290 result is used (not including appendfile, where other changes are about
2293 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2294 This involved changing a lot of size variables from int to off_t. It
2295 should work with maildirs and everything.
2297 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2298 spamd dying while we are connected to it.
2300 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2303 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2304 function for BDB 4.3.
2306 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2308 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2309 the sources for PCRE have been re-arranged and more files are now
2312 PH/15 The code I had for printing potentially long long variables in PH/11
2313 above was not the best (it lost precision). The length of off_t variables
2314 is now inspected at build time, and an appropriate printing format (%ld
2315 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2316 to be "long long int" or "long int". This is needed for the internal
2317 formatting function string_vformat().
2319 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2320 the configuration file to be ":syslog", then the script "guesses" where
2321 the logs files are, rather than using the compiled in default. In our
2322 case the guess is not the same as the compiled default, so the script
2323 suddenly stopped working when I started to use syslog. The patch checks
2324 to see if log_file_path is "". If so, it attempts to read it from exim
2325 with no configuration file to get the compiled in version, before it
2326 falls back to the previous guessing code."
2328 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2329 implementing BATV in an Exim configuration. See NewStuff for the gory
2332 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2333 Makefile that are specific to HP-UX.
2335 PH/18 If the "use_postmaster" option was set for a recipient callout together
2336 with the "random" option, the postmaster address was used as the MAIL
2337 FROM address for the random test, but not for the subsequent recipient
2338 test. It is now used for both.
2340 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2341 patch removes a few documentation additions to RFC 3028, because the
2342 latest draft now contains them. It adds the new en;ascii-case comparator
2343 and a new error check for 8bit text in MIME parts. Comparator and
2344 require names are now matched exactly. I enabled the subaddress
2345 extension, but it is not well tested yet (read: it works for me)."
2347 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2348 rework some of the code of TK/09 above to avoid the hardwired use of
2349 "%lld" and "long long". Replaced the call to snprintf() with a call to
2352 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2353 records point to non-existent hosts", "retry timeout exceeded", and
2354 "retry time not reached for any host after a long failure period".
2356 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2357 experimental DomainKeys support:
2359 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2360 (2) On an error such as an illegally used "control", the wrong name for
2361 the control was given.
2363 These problems did NOT occur unless DomainKeys support was compiled.
2365 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2367 PH/24 Added ${if match_ip condition.
2369 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2370 databases so that it will be absolutely obvious if a crash occurs in the
2371 DB library. This is a regular occurrence (often caused by mis-matched
2374 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2375 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2376 picked up on a user's system that detects such things. There doesn't seem
2377 to be a gcc warning option for this - only an attribute that has to be
2378 put on the function's prototype. It seems that in Fedora Core 4 they have
2379 set this on a number of new functions. No doubt there will be more in due
2382 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2383 specify a transport. However, if an address that was verified by such a
2384 router was the subject of a callout, Exim crashed because it tried to
2385 read the rcpt_include_affixes from the non-existent transport. Now it
2386 just assumes that the setting of that option is false. This bug was
2387 introduced by 4.51/PH/31.
2389 PH/28 Changed -d+all to exclude +memory, because that information is very
2390 rarely of interest, but it makes the output a lot bigger. People tend to
2391 do -d+all out of habit.
2393 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2394 code in os-type was giving problems when libc.so lives in lib64, like on
2397 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2398 aren't the modern standard, and it seems that some systems' include files
2399 don't always have them. Exim was already checking for some of the newer
2400 ones like T_AAAA, and defining it itself. I've added checks for all the
2401 record types that Exim uses.
2403 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2404 not automatically generating a new one, as it is supposed to. This
2405 prevented TLS from working. If the file did exist, but contained invalid
2406 data, a new version was generated, as expected. It was only the case of a
2407 non-existent file that was broken.
2409 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2410 with a change in libdomainkeys > 0.64.
2412 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2413 from DNS. If the selector record carries the flag, it now has
2414 precedence over the domain-wide flag.
2416 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2418 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2419 the use of a port name that isn't defined in /etc/services) occurred, the
2420 message was deferred as in a normal delivery, and thus remained on the
2421 spool, instead of being failed because of the mua_wrapper setting. This
2422 is now fixed, and I tidied up some of the mua_wrapper messages at the
2425 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2426 the messages in a hash of arrays rather than using individual hashes.
2427 This is a bit cleaner and results in dramatic memory savings, albeit
2428 at a slight CPU cost.
2430 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2431 as requested by Marc Sherman.
2433 SC/10 Eximstats: added histograms for user specified patterns as requested
2436 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2438 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2439 fopen() in the content-scanning modules that did not already have it.
2445 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2446 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2448 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2450 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2452 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2453 to test Sieve filters that use "vacation".
2455 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2456 that changes the way the GnuTLS parameters are stored in the cache file.
2457 The new format can be generated externally. For backward compatibility,
2458 if the data in the cache doesn't make sense, Exim assumes it has read an
2459 old-format file, and it generates new data and writes a new file. This
2460 means that you can't go back to an older release without removing the
2463 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2464 work if there are any delivery delays because "one_time" forces the
2465 parent to be marked "delivered", so its unseen clone is never tried
2466 again. For this reason, Exim now forbids the simultaneous setting of
2469 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2470 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2471 entry if you want to know the details. The fix, however, neglected to
2472 consider the case when local delivery batching is involved. The test for
2473 "previously delivered" was not happening when checking to see if an
2474 address could be batched with a previous (undelivered) one; under
2475 certain circumstances this could lead to multiple deliveries to the same
2478 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2479 in its include files, and this causes problems building Exim.
2481 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2482 header_syntax) but Exim was just ignoring anything given after a slash.
2483 In particular, this caused confusion with an attempt to use "verify =
2484 reverse_host_lookup/defer_ok". An error is now given when options are
2485 supplied for verify items that do not have them. (Maybe reverse_host_
2486 lookup should have a defer_ok option, but that's a different point.)
2488 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2489 defined by RFC 821) to 2048, because there were problems with some AUTH
2490 commands, and RFC 1869 says the size should be increased for extended
2491 SMTP commands that take arguments.
2493 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2496 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2497 "unknown" error; now it says that the functionality isn't in the binary.
2499 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2500 an address' error message when a string expansion fails (syntax or
2501 whatever). Otherwise the password may appear in the log. Following change
2502 PH/42 below, there is no longer a chance of it appearing in a bounce
2505 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2507 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2508 of its addresses was ever tried. (Bugzilla bug #2.)
2510 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2511 the result incorrectly in the debug output. (It correctly added a newline
2512 to what was transported.)
2514 TF/01 Added $received_time.
2516 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2517 commented out examples of how to interface to a virus scanner and to
2518 SpamAssassin. Also added commented examples of av_scanner and
2519 spamd_address settings.
2521 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2522 and controls are allowed in which ACLs. There were a couple of minor
2523 errors. Some of the entries in the conditions table (which is a table of
2524 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2525 negation of where the condition IS allowed.
2527 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2529 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2530 header file does not have a version number, so I've had to invent a new
2531 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2532 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2533 radiusclient), but it was contributed by a Radius user.
2535 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2536 files or directories, for OpenSSL.
2538 PH/19 When an Exim process that is running as root has to create an Exim log
2539 file, it does so in a subprocess that runs as exim:exim so as to get the
2540 ownership right at creation (otherwise, other Exim processes might see
2541 the file with the wrong ownership). There was no test for failure of this
2542 fork() call, which would lead to the process getting stuck as it waited
2543 for a non-existent subprocess. Forks do occasionally fail when resources
2544 run out. I reviewed all the other calls to fork(); they all seem to check
2547 PH/20 When checking for unexpected SMTP input at connect time (before writing
2548 the banner), Exim was not dealing correctly with a non-positive return
2549 from the read() function. If the client had disconnected by this time,
2550 the result was a log entry for a synchronization error with an empty
2551 string after "input=" when read() returned zero. If read() returned -1
2552 (an event I could not check), uninitialized data bytes were printed.
2553 There were reports of junk text (parts of files, etc) appearing after
2556 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2558 PH/22 Added support for macro redefinition, and (re)definition in between
2559 driver and ACL definitions.
2561 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2562 forgetting to use the resulting value; it was using the unexpanded value.
2564 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2565 hadn't been configured. The fix is from Juergen Kreileder, who
2566 understands it better than I do:
2568 "Here's what I see happening with three configured cyrus_sasl
2569 authenticators configured (plain, login, cram-md5):
2571 On startup auth_cyrus_sasl_init() gets called for each of these.
2572 This means three calls to sasl_listmech() without a specified mech_list.
2573 => SASL tests which mechs of all available mechs actually work
2574 => three warnings about OTP not working
2575 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2577 With the patch, sasl_listmech() also gets called three times. But now
2578 SASL's mech_list option is set to the server_mech specified in the the
2579 authenticator. Or in other words, the answer from sasl_listmech()
2580 gets limited to just the mech you're testing for (which is different
2582 => the return list contains just 'plain' or 'login', 'cram-md5' or
2583 nothing depending on the value of ob->server_mech.
2585 I've just tested the patch: Authentication still works fine,
2586 unavailable mechs specified in the exim configuration are still
2587 caught, and the auth.log warnings about OTP are gone."
2589 PH/25 When debugging is enabled, the contents of the command line are added
2590 to the debugging output, even when log_selector=+arguments is not
2593 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2594 answer is "GNU", and only if the return is "GNU/something" is the answer
2597 PH/27 $acl_verify_message is now set immediately after the failure of a
2598 verification in an ACL, and so is available in subsequent modifiers. In
2599 particular, the message can be preserved by coding like this:
2601 warn !verify = sender
2602 set acl_m0 = $acl_verify_message
2604 Previously, $acl_verify_message was set only while expanding "message"
2605 and "log_message" when a very denied access.
2607 PH/28 Modified OS/os.c-Linux with
2609 -#ifndef OS_LOAD_AVERAGE
2610 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2612 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2613 nomenclature these days.)
2615 PH/29 Installed patch from the Sieve maintainer that adds the options
2616 sieve_useraddress and sieve_subaddress to the redirect router.
2618 PH/30 In these circumstances:
2619 . Two addresses routed to the same list of hosts;
2620 . First host does not offer TLS;
2621 . First host accepts first address;
2622 . First host gives temporary error to second address;
2623 . Second host offers TLS and a TLS session is established;
2624 . Second host accepts second address.
2625 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2626 and peerdn, if requested) that were in fact used only for the second
2629 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2630 attention to any local part prefix or suffix that was matched by the
2631 router that accepted the address. It now behaves in the same way as it
2632 does for delivery: the affixes are removed from the local part unless
2633 rcpt_include_affixes is set on the transport.
2635 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2636 timeout during the DATA phase of an incoming message.
2638 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2639 applied a patch sanctioned by the Sieve maintainer.
2641 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2642 the uid or gid is negative. A case of a negative gid caused this to be
2643 noticed. The fix allows for either to be negative.
2645 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2646 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2649 PH/36 The change PH/12 above was broken. Fixed it.
2651 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2652 the grounds that routing the same address twice would always produce the
2653 same answer. This might have been true once, but it is certainly no
2654 longer true now. Routing a child address may depend on the previous
2655 routing that produced that child. Some complicated redirection strategies
2656 went wrong when messages had multiple recipients, and made Exim's
2657 behaviour dependent on the order in which the addresses were given.
2659 I have moved the duplicate checking until after the routing is complete.
2660 Exim scans the addresses that are assigned to local and remote
2661 transports, and removes any duplicates. This means that more work will be
2662 done, as duplicates will always all be routed, but duplicates are
2663 presumably rare, so I don't expect this is of any significance.
2665 For deliveries to pipes, files, and autoreplies, the duplicate checking
2666 still happens during the routing process, since they are not going to be
2669 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2670 It corrects a timeout issue with spamd. This is Ian's comment: "The
2671 background is that sometimes spamd either never reads data from a
2672 connection it has accepted, or it never writes response data. The exiscan
2673 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2674 blindly assumes that writes won't block so it may never time out."
2676 PH/39 Allow G after quota size as well as K and M.
2678 PH/40 The value set for $authenticated_id in an authenticator may not contain
2679 binary zeroes or newlines because the value is written to log lines and
2680 to spool files. There was no check on this. Now the value is run through
2681 the string_printing() function so that such characters are converted to
2682 printable escape sequences.
2684 PH/41 $message_linecount is a new variable that contains the total number of
2685 lines in the message. Compare $body_linecount, which is the count for the
2688 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2689 bounce and delay warning messages, except in certain special cases, which
2692 (a) An SMTP error message from a remote host;
2693 (b) A message specified in a :fail: redirection;
2694 (c) A message specified in a "fail" command in a system filter;
2695 (d) A message specified in a FAIL return from the queryprogram router;
2696 (e) A message specified by the cannot_route_message router option.
2698 In these cases only, Exim does include the error details in bounce and
2699 warning messages. There are also a few cases where bland messages such
2700 as "unrouteable address" or "local delivery error" are given.
2702 PH/43 $value is now also set for the "else" part of a ${run expansion.
2704 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2705 being worked on, but at least Exim now implements the latest version to
2708 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2709 process to complete was treated as a delivery failure, a timeout while
2710 writing the message to the pipe was logged, but erroneously treated as a
2711 successful delivery. Such timeouts include transport filter timeouts. For
2712 consistency with the overall process timeout, these timeouts are now
2713 treated as errors, giving rise to delivery failures by default. However,
2714 there is now a new Boolean option for the pipe transport called
2715 timeout_defer, which, if set TRUE, converts the failures into defers for
2716 both kinds of timeout. A transport filter timeout is now identified in
2719 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2720 systems where "make" and "gmake" are different, calling "gmake" at top
2721 level broke things. I've arranged for the value of $(MAKE) to be passed
2722 from the Makefile to this script so that it can call the same version of
2726 A note about Exim versions 4.44 and 4.50
2727 ----------------------------------------
2729 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2730 changes of various kinds. As a consequence, a big documentation update was
2731 needed. This delayed the release for rather longer than seemed good, especially
2732 in the light of a couple of (minor) security issues. Therefore, the changes
2733 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2734 release. So 4.44 and 4.50 are in effect two different branches that both start
2737 I have left the 4.50 change log unchanged; it contains all the changes since
2738 4.43. The change log for 4.44 is below; many of its items are identical to
2739 those for 4.50. This seems to be the most sensible way to preserve the
2740 historical information.
2746 1. Minor wording change to the doc/README.SIEVE file.
2748 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2749 computation of the current number of files was incorrect.
2751 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2752 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2755 4. Give more explanation in the error message when the command for a transport
2756 filter fails to execute.
2758 5. There are several places where Exim runs a non-Exim command in a
2759 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2760 was being done only for the command run by the queryprogram router. It is
2761 now done for all such subprocesses. The other cases are: ${run, transport
2762 filters, and the commands run by the lmtp and pipe transports.
2764 6. Added CONFIGURE_GROUP build-time option.
2766 7. Some older OS have a limit of 256 on the maximum number of file
2767 descriptors. Exim was using setrlimit() to set 1000 as a large value
2768 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2769 systems. I've change it so that if it can't get 1000, it tries for 256.
2771 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2772 was an oversight, and furthermore, ever since the addition of extra
2773 controls (e.g. 4.43/32), the checks on when to allow different forms of
2774 "control" were broken. There should now be diagnostics for all cases when a
2775 control that does not make sense is encountered.
2777 9. Added the /retain_sender option to "control=submission".
2779 10. $recipients is now available in the predata ACL (oversight).
2781 11. Tidy the search cache before the fork to do a delivery from a message
2782 received from the command line. Otherwise the child will trigger a lookup
2783 failure and thereby defer the delivery if it tries to use (for example) a
2784 cached ldap connection that the parent has called unbind on.
2786 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2787 of $address_data from the recipient verification was clobbered by the
2788 sender verification.
2790 13. The value of address_data from a sender verification is now available in
2791 $sender_address_data in subsequent conditions in the ACL statement.
2793 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2795 15. Added a new option "connect=<time>" to callout options, to set a different
2798 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2799 was its contents. (It was OK if the option was not defined at all.)
2801 17. A "Completed" log line is now written for messages that are removed from
2802 the spool by the -Mrm option.
2804 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2805 information about exactly what failed.
2807 19. Added -dd to debug only the daemon process.
2809 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2810 handles timeouts, both on the server side and network timeouts. Renamed the
2811 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2813 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2814 It is now set to "smtps".
2816 22. $host_address is now set to the target address during the checking of
2817 ignore_target_hosts.
2819 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2820 being passed; this would have caused $sender_host_name to have been used if
2821 matching the list had actually called for a host name (not very likely,
2822 since this list is usually IP addresses). A host name is now passed as
2825 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2826 code that specifies a non-restarting handler (typically sigaction() in
2827 modern systems) in an attempt to fix a rare and obscure crash bug.
2829 25. Narrowed the window for a race in the daemon that could cause it to ignore
2830 SIGCHLD signals. This is not a major problem, because they are used only to
2831 wake it up if nothing else does.
2833 26. A malformed maildirsize file could cause Exim to calculate negative values
2834 for the mailbox size or file count. Odd effects could occur as a result.
2835 The maildirsize information is now recalculated if the size or filecount
2838 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2839 support for a long time. Removed HAVE_SYS_VFS_H.
2841 28. Installed the latest version of exipick from John Jetmore.
2843 29. In an address list, if the pattern was not a regular expression, an empty
2844 subject address (from a bounce message) matched only if the pattern was an
2845 empty string. Non-empty patterns were not even tested. This was the wrong
2846 because it is perfectly reasonable to use an empty address as part of a
2847 database query. An empty address is now tested by patterns that are
2848 lookups. However, all the other forms of pattern expect the subject to
2849 contain a local part and a domain, and therefore, for them, an empty
2850 address still always fails if the pattern is not itself empty.
2852 30. Exim went into a mad DNS loop when attempting to do a callout where the
2853 host was specified on an smtp transport, and looking it up yielded more
2854 than one IP address.
2856 31. Re-factored the code for checking spool and log partition space into a
2857 function that finds that data and another that does the check. The former
2858 is then used to implement four new variables: $spool_space, $log_space,
2859 $spool_inodes, and $log_inodes.
2861 32. The RFC2047 encoding function was originally intended for short strings
2862 such as real names; it was not keeping to the 75-character limit for
2863 encoded words that the RFC imposes. It now respects the limit, and
2864 generates multiple encoded words if necessary. To be on the safe side, I
2865 have increased the buffer size for the ${rfc2047: expansion operator from
2868 33. It is now permitted to omit both strings after an "if" condition; if the
2869 condition is true, the result is "true". As before, when the second string
2870 is omitted, a false condition yields an empty string. This makes it less
2871 cumbersome to write custom ACL and router conditions.
2873 34. Failure to deliver a bounce message always caused it to be frozen, even if
2874 there was an errors_to setting on the router. The errors_to setting is now
2877 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2878 canonical form (fully expanded) before being placed in
2879 $sender_host_address.
2881 36. The table in the code that translates DNS record types into text (T_A to
2882 "A" for instance) was missing entries for NS and CNAME. It is just possible
2883 that this could have caused confusion if both these types were looked up
2884 for the same domain, because the text type is used as part of Exim's
2885 per-process caching. But the chance of anyone hitting this buglet seems
2888 37. The dnsdb lookup has been extended in a number of ways.
2890 (1) There is a new type, "zns", which walks up the domain tree until it
2891 finds some nameserver records. It should be used with care.
2893 (2) There is a new type, "mxh", which is like "mx" except that it returns
2894 just the host names, not the priorities.
2896 (3) It is now possible to give a list of domains (or IP addresses) to be
2897 looked up. The behaviour when one of the lookups defers can be
2898 controlled by a keyword.
2900 (4) It is now possible to specify the separator character for use when
2901 multiple records are returned.
2903 38. The dnslists ACL condition has been extended: it is now possible to supply
2904 a list of IP addresses and/or domains to be looked up in a particular DNS
2907 39. Added log_selector=+queue_time_overall.
2909 40. When running the queue in the test harness, wait just a tad after forking a
2910 delivery process, to get repeatability of debugging output.
2912 41. Include certificate and key file names in error message when GnuTLS fails
2913 to set them up, because the GnuTLS error message doesn't include the name
2914 of the failing file when there is a problem reading it.
2916 42. Allow both -bf and -bF in the same test run.
2918 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2920 44. The "Exiscan patch" is now merged into the mainline Exim source.
2922 45. Sometimes the final signoff response after QUIT could fail to get
2923 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2924 < 0 before doing a fflush(). This bug looks as though it goes back to the
2925 introduction of TLS in release 3.20, but "sometimes" must have been rare
2926 because the tests only now provoked it.
2928 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2929 (this can affect the format of dates).
2931 47. exim_tidydb, when checking for the continued existence of a message for
2932 which it has found a message-specific retry record, was not finding
2933 messages that were in split spool directories. Consequently, it was
2934 deleting retry records that should have stayed in existence.
2936 48. Steve fixed some bugs in eximstats.
2938 49. The SPA authentication driver was not abandoning authentication and moving
2939 on to the next authenticator when an expansion was forced to fail,
2940 contradicting the general specification for all authenticators. Instead it
2941 was generating a temporary error. It now behaves as specified.
2943 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2944 (the order specifies the preference for clients). The order is now AES256,
2945 AES128, 3DES, ARCFOUR128.
2947 51. Small patch to Sieve code - explicitly set From: when generating an
2950 52. Exim crashed if a remote delivery caused a very long error message to be
2951 recorded - for instance if somebody sent an entire SpamAssassin report back
2952 as a large number of 550 error lines. This bug was coincidentally fixed by
2953 increasing the size of one of Exim's internal buffers (big_buffer) that
2954 happened as part of the Exiscan merge. However, to be on the safe side, I
2955 have made the code more robust (and fixed the comments that describe what
2958 53. Now that there can be additional text after "Completed" in log lines (if
2959 the queue_time_overall log selector is set), a one-byte patch to exigrep
2960 was needed to allow it to recognize "Completed" as not the last thing in
2963 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2964 patch that reportedly fixes this has been added. I am not expert enough to
2965 create a test for it. This is what the patch creator wrote:
2967 "I found a little strange behaviour of ldap code when working with
2968 Windows 2003 AD Domain, where users was placed in more than one
2969 Organization Units. When I tried to give exim partial DN, the exit code
2970 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2971 But simultaneously result of request was absolutely normal ldap result,
2972 so I produce this patch..."
2974 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2975 REFERENCE, so I have modified the code to exclude the patch when that macro
2978 55. Some experimental protocols are using DNS PTR records for new purposes. The
2979 keys for these records are domain names, not reversed IP addresses. The
2980 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2981 leaves it alone. Component reversal etc. now happens only for IP addresses.
2984 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2986 57. Double the size of the debug message buffer (to 2048) so that more of very
2987 long debug lines gets shown.
2989 58. The exicyclog utility now does better if the number of log files to keep
2990 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2992 59. Two changes related to the smtp_active_hostname option:
2994 (1) $smtp_active_hostname is now available as a variable.
2995 (2) The default for smtp_banner uses $smtp_active_hostname instead
2996 of $primary_hostname.
2998 60. The host_aton() function is supposed to be passed a string that is known
2999 to be a valid IP address. However, in the case of IPv6 addresses, it was
3000 not checking this. This is a hostage to fortune. Exim now panics and dies
3001 if the condition is not met. A case was found where this could be provoked
3002 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3003 components; fortuitously, this particular loophole had already been fixed
3004 by change 4.50/55 above.
3006 If there are any other similar loopholes, the new check in host_aton()
3007 itself should stop them being exploited. The report I received stated that
3008 data on the command line could provoke the exploit when Exim was running as
3009 exim, but did not say which command line option was involved. All I could
3010 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3011 running as the user.
3014 61. There was a buffer overflow vulnerability in the SPA authentication code
3015 (which came originally from the Samba project). I have added a test to the
3016 spa_base64_to_bits() function which I hope fixes it.
3019 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3020 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3022 63. The daemon start-up calls getloadavg() while still root for those OS that
3023 need the first call to be done as root, but it missed one case: when
3024 deliver_queue_load_max is set with deliver_drop_privilege. This is
3025 necessary for the benefit of the queue runner, because there is no re-exec
3026 when deliver_drop_privilege is set.
3028 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3029 This has been fixed.
3031 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3032 was in use, was not putting the data itself into the right store pool;
3033 consequently, it could be overwritten for a subsequent message in the same
3034 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3037 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3039 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3040 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3041 the function were treating the return as a boolean value, which happened to
3042 work because 0=false and not-0=true, but is not correct code.
3044 68. The host_aton() function was not handling scoped IPv6 addresses (those
3045 with, for example, "%eth0" on the end) correctly.
3047 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3048 negated items (that is, ~something) in unsigned ints. Some compilers
3049 apparently mutter when there is no cast.
3051 70. If an address verification called from an ACL failed, and did not produce a
3052 user-specific message (i.e. there was only a "system" message), nothing was
3053 put in $acl_verify_message. In this situation, it now puts the system
3056 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3057 session; change 4.31/43 added the unwanted input to the log line - except
3058 that it did not do this in the start of session case. It now does.
3060 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3061 This is wrong and can cause the other end to generate a synchronization
3062 error if it is another Exim or anything else that does the synchronization
3063 check. A QUIT command is no longer sent after a timeout.
3065 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3066 during host lookups.
3068 74. The defer_ok option of callout verification was not working if it was used
3069 when verifying addresses in header lines, that is, for this case:
3071 verify = header_sender/callout=defer_ok
3073 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3074 those file descriptors could be used for SMTP connections. If anything
3075 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3076 could be sent to the SMTP client, causing chaos. The daemon now opens
3077 stdin, stdout, and stderr to /dev/null when it puts itself into the
3080 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3081 log by default. The user can override this with suitable Perl magic.
3083 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3084 the log message when discard triggers, was not working for the DATA ACL or
3085 for the non-SMTP ACL.
3087 78. Error message wording change in sieve.c.
3089 79. If smtp_accept_max_per_host was set, the number of connections could be
3090 restricted to fewer than expected, because the daemon was trying to set up
3091 a new connection before checking whether the processes handling previous
3092 connections had finished. The check for completed processes is now done
3093 earlier. On busy systems, this bug wouldn't be noticed because something
3094 else would have woken the daemon, and it would have reaped the completed
3097 80. If a message was submitted locally by a user whose login name contained one
3098 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3099 It caused a spool format error. I have fixed the spool reading code. A
3100 related problem was that the "from" clause in the Received: line became
3101 illegal because of the space(s). It is now covered by ${quote_local_part.
3103 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3106 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3108 83. Patch from Sieve maintainer to fix unterminated string problem in
3109 "vacation" handling.
3111 84. Some minor changes to the Linux configuration files to help with other
3112 OS variants using glibc.
3114 85. One more patch for Sieve to update vacation handling to latest spec.
3117 ----------------------------------------------------
3118 See the note above about the 4.44 and 4.50 releases.
3119 ----------------------------------------------------
3125 1. Change 4.43/35 introduced a bug that caused file counts to be
3126 incorrectly computed when quota_filecount was set in an appendfile
3129 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3130 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3133 3. Give more explanation in the error message when the command for a transport
3134 filter fails to execute.
3136 4. There are several places where Exim runs a non-Exim command in a
3137 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3138 was being done only for the command run by the queryprogram router. It is
3139 now done for all such subprocesses. The other cases are: ${run, transport
3140 filters, and the commands run by the lmtp and pipe transports.
3142 5. Some older OS have a limit of 256 on the maximum number of file
3143 descriptors. Exim was using setrlimit() to set 1000 as a large value
3144 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3145 systems. I've change it so that if it can't get 1000, it tries for 256.
3147 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3148 was an oversight, and furthermore, ever since the addition of extra
3149 controls (e.g. 4.43/32), the checks on when to allow different forms of
3150 "control" were broken. There should now be diagnostics for all cases when a
3151 control that does not make sense is encountered.
3153 7. $recipients is now available in the predata ACL (oversight).
3155 8. Tidy the search cache before the fork to do a delivery from a message
3156 received from the command line. Otherwise the child will trigger a lookup
3157 failure and thereby defer the delivery if it tries to use (for example) a
3158 cached ldap connection that the parent has called unbind on.
3160 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3161 of $address_data from the recipient verification was clobbered by the
3162 sender verification.
3164 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3165 was its contents. (It was OK if the option was not defined at all.)
3167 11. A "Completed" log line is now written for messages that are removed from
3168 the spool by the -Mrm option.
3170 12. $host_address is now set to the target address during the checking of
3171 ignore_target_hosts.
3173 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3174 being passed; this would have caused $sender_host_name to have been used if
3175 matching the list had actually called for a host name (not very likely,
3176 since this list is usually IP addresses). A host name is now passed as
3179 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3180 code that specifies a non-restarting handler (typically sigaction() in
3181 modern systems) in an attempt to fix a rare and obscure crash bug.
3183 15. Narrowed the window for a race in the daemon that could cause it to ignore
3184 SIGCHLD signals. This is not a major problem, because they are used only to
3185 wake it up if nothing else does.
3187 16. A malformed maildirsize file could cause Exim to calculate negative values
3188 for the mailbox size or file count. Odd effects could occur as a result.
3189 The maildirsize information is now recalculated if the size or filecount
3192 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3193 support for a long time. Removed HAVE_SYS_VFS_H.
3195 18. Updated exipick to current release from John Jetmore.
3197 19. Allow an empty sender to be matched against a lookup in an address list.
3198 Previously the only cases considered were a regular expression, or an
3201 20. Exim went into a mad DNS lookup loop when doing a callout where the
3202 host was specified on the transport, if the DNS lookup yielded more than
3205 21. The RFC2047 encoding function was originally intended for short strings
3206 such as real names; it was not keeping to the 75-character limit for
3207 encoded words that the RFC imposes. It now respects the limit, and
3208 generates multiple encoded words if necessary. To be on the safe side, I
3209 have increased the buffer size for the ${rfc2047: expansion operator from
3212 22. Failure to deliver a bounce message always caused it to be frozen, even if
3213 there was an errors_to setting on the router. The errors_to setting is now
3216 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3217 canonical form (fully expanded) before being placed in
3218 $sender_host_address.
3220 24. Updated eximstats to version 1.33
3222 25. Include certificate and key file names in error message when GnuTLS fails
3223 to set them up, because the GnuTLS error message doesn't include the name
3224 of the failing file when there is a problem reading it.
3226 26. Expand error message when OpenSSL has problems setting up cert/key files.
3229 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3230 (this can affect the format of dates).
3232 28. exim_tidydb, when checking for the continued existence of a message for
3233 which it has found a message-specific retry record, was not finding
3234 messages that were in split spool directories. Consequently, it was
3235 deleting retry records that should have stayed in existence.
3237 29. eximstats updated to version 1.35
3238 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3239 1.35 - bugfix such that pie charts by volume are generated correctly
3241 30. The SPA authentication driver was not abandoning authentication and moving
3242 on to the next authenticator when an expansion was forced to fail,
3243 contradicting the general specification for all authenticators. Instead it
3244 was generating a temporary error. It now behaves as specified.
3246 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3247 (the order specifies the preference for clients). The order is now AES256,
3248 AES128, 3DES, ARCFOUR128.
3250 31. Small patch to Sieve code - explicitly set From: when generating an
3253 32. Exim crashed if a remote delivery caused a very long error message to be
3254 recorded - for instance if somebody sent an entire SpamAssassin report back
3255 as a large number of 550 error lines. This bug was coincidentally fixed by
3256 increasing the size of one of Exim's internal buffers (big_buffer) that
3257 happened as part of the Exiscan merge. However, to be on the safe side, I
3258 have made the code more robust (and fixed the comments that describe what
3261 33. Some experimental protocols are using DNS PTR records for new purposes. The
3262 keys for these records are domain names, not reversed IP addresses. The
3263 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3264 leaves it alone. Component reversal etc. now happens only for IP addresses.
3267 34. The host_aton() function is supposed to be passed a string that is known
3268 to be a valid IP address. However, in the case of IPv6 addresses, it was
3269 not checking this. This is a hostage to fortune. Exim now panics and dies
3270 if the condition is not met. A case was found where this could be provoked
3271 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3272 components; fortuitously, this particular loophole had already been fixed
3273 by change 4.50/55 or 4.44/33 above.
3275 If there are any other similar loopholes, the new check in host_aton()
3276 itself should stop them being exploited. The report I received stated that
3277 data on the command line could provoke the exploit when Exim was running as
3278 exim, but did not say which command line option was involved. All I could
3279 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3280 running as the user.
3283 35. There was a buffer overflow vulnerability in the SPA authentication code
3284 (which came originally from the Samba project). I have added a test to the
3285 spa_base64_to_bits() function which I hope fixes it.
3288 36. The daemon start-up calls getloadavg() while still root for those OS that
3289 need the first call to be done as root, but it missed one case: when
3290 deliver_queue_load_max is set with deliver_drop_privilege. This is
3291 necessary for the benefit of the queue runner, because there is no re-exec
3292 when deliver_drop_privilege is set.
3294 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3295 was in use, was not putting the data itself into the right store pool;
3296 consequently, it could be overwritten for a subsequent message in the same
3297 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3300 38. Sometimes the final signoff response after QUIT could fail to get
3301 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3302 < 0 before doing a fflush(). This bug looks as though it goes back to the
3303 introduction of TLS in release 3.20, but "sometimes" must have been rare
3304 because the tests only now provoked it.
3310 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3311 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3312 Now it returns an integer. A number of calls were still expecting a T/F
3313 return. Fortuitously, in all cases, the tests worked in OK situations,
3314 which is the norm. However, things would have gone wrong on any write
3315 failures on the smtp file descriptor. This function is used when sending
3316 messages over SMTP and also when doing verify callouts.
3318 2. When Exim is called to do synchronous delivery of a locally submitted
3319 message (the -odf or -odi options), it no longer closes stderr before doing
3322 3. Implemented the mua_wrapper option.
3324 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3326 5. Implemented the functions header_remove(), header_testname(),
3327 header_add_at_position(), and receive_remove_recipient(), and exported them
3330 6. If an ACL "warn" statement specified the addition of headers, Exim already
3331 inserted X-ACL-Warn: at the start if there was no header name. However, it
3332 was not making this test for the second and subsequent header lines if
3333 there were newlines in the string. This meant that an invalid header could
3334 be inserted if Exim was badly configured.
3336 7. Allow an ACL "warn" statement to add header lines at the start or after all
3337 the Received: headers, as well as at the end.
3339 8. Added the rcpt_4xx retry error code.
3341 9. Added postmaster_mailfrom=xxx to callout verification option.
3343 10. Added mailfrom=xxxx to the callout verification option, for verify=
3346 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3347 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3349 12. Inserted some casts to stop certain compilers warning when using pointer
3350 differences as field lengths or precisions in printf-type calls (mostly
3351 affecting debugging statements).
3353 13. Added optional readline() support for -be (dynamically loaded).
3355 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3356 same clock tick as a message's arrival, so that its received time was the
3357 same as the "first fail" time on the retry record, and that message
3358 remained on the queue past the ultimate address timeout, every queue runner
3359 would try a delivery (because it was past the ultimate address timeout) but
3360 after another failure, the ultimate address timeout, which should have then
3361 bounced the address, did not kick in. This was a "< instead of <=" error;
3362 in most cases the first failure would have been in the next clock tick
3363 after the received time, and all would be well.
3365 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3366 being recognized when the domain list was tested by the match_domain
3367 condition in an expansion string.
3369 16. Added the ${str2b64: operator.
3371 17. Exim was always calling setrlimit() to set a large limit for the number of
3372 processes, without checking whether the existing limit was already
3373 adequate. (It did check for the limit on file descriptors.) Furthermore,
3374 errors from getrlimit() and setrlimit() were being ignored. Now they are
3375 logged to the main and panic logs, but Exim does carry on, to try to do its
3376 job under whatever limits there are.
3378 18. Imported PCRE 5.0.
3380 19. Trivial typo in log message " temporarily refused connection" (the leading
3383 20. If the log selector return_path_on_delivery was set and an address was
3384 redirected to /dev/null, the delivery process crashed because it assumed
3385 that a return path would always be set for a "successful" delivery. In this
3386 case, the whole delivery is bypassed as an optimization, and therefore no
3389 21. Internal re-arrangement: the function for sending a challenge and reading
3390 a response while authentication was assuming a zero-terminated challenge
3391 string. It's now changed to take a pointer and a length, to allow for
3392 binary data in such strings.
3394 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3396 23. Exim was not respecting finduser_retries when seeking the login of the
3397 uid under which it was called; it was always trying 10 times. (The default
3398 setting of finduser_retries is zero.) Also, it was sleeping after the final
3399 failure, which is pointless.
3401 24. Implemented tls_on_connect_ports.
3403 25. Implemented acl_smtp_predata.
3405 26. If the domain in control=submission is set empty, Exim assumes that the
3406 authenticated id is a complete email address when it generates From: or
3407 Sender: header lines.
3409 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3410 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3411 chown and chgrp in /bin and hostname in /usr/bin.
3413 28. Exim was keeping the "process log" file open after each use, just as it
3414 does for the main log. This opens the possibility of it remaining open for
3415 long periods when the USR1 signal hits a daemon. Occasional processlog
3416 errors were reported, that could have been caused by this. Anyway, it seems
3417 much more sensible not to leave this file open at all, so that is what now
3420 29. The long-running daemon process does not normally write to the log once it
3421 has entered its main loop, and it closes the log before doing so. This is
3422 so that log files can straightforwardly be renamed and moved. However,
3423 there are a couple of unusual error situations where the daemon does write
3424 log entries, and I had neglected to close the log afterwards.
3426 30. The text of an SMTP error response that was received during a remote
3427 delivery was being truncated at 512 bytes. This is too short for some of
3428 the long messages that one sometimes sees. I've increased the limit to
3431 31. It is now possible to make retry rules that apply only when a message has a
3432 specific sender, in particular, an empty sender.
3434 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3435 it possible to be selective about when SMTP synchronization is enforced.
3437 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3439 32. Implemented hosts_connection_nolog.
3441 33. Added an ACL for QUIT.
3443 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3446 35. Added mailbox_size and mailbox_filecount to appendfile.
3448 36. Added control = no_multiline_responses to ACLs.
3450 37. There was a bug in the logic of the code that waits for the clock to tick
3451 in the case where the clock went backwards by a substantial amount such
3452 that the microsecond fraction of "now" was more than the microsecond
3453 fraction of "then" (but the whole seconds number was less).
3455 38. Added support for the libradius Radius client library this is found on
3456 FreeBSD (previously only the radiusclient library was supported).
3462 1. When certain lookups returned multiple values in the form name=value, the
3463 quoting of the values was not always being done properly. Specifically:
3464 (a) If the value started with a double quote, but contained no whitespace,
3466 (b) If the value contained whitespace other than a space character (i.e.
3467 tabs or newlines or carriage returns) it was not quoted.
3468 This fix has been applied to the mysql and pgsql lookups by writing a
3469 separate quoting function and calling it from the lookup code. The fix
3470 should probably also be applied to nisplus, ibase and oracle lookups, but
3471 since I cannot test any of those, I have not disturbed their existing code.
3473 2. A hit in the callout cache for a specific address caused a log line with no
3474 reason for rejecting RCPT. Now it says "Previous (cached) callout
3475 verification failure".
3477 3. There was an off-by-one bug in the queryprogram router. An over-long
3478 return line was truncated at 256 instead of 255 characters, thereby
3479 overflowing its buffer with the terminating zero. As well as fixing this, I
3480 have increased the buffer size to 1024 (and made a note to document this).
3482 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3483 when Exim is waiting for an SMTP response from a remote server, Exim
3484 restarts its select() call on the socket, thereby resetting its timeout.
3485 This is not a problem when such interrupts are rare. Somebody set up a cron
3486 job to run exiwhat every 2 minutes, which is less than the normal select()
3487 timeout (5 or 10 minutes). This meant that the select() timeout never
3488 kicked in because it was always reset. I have fixed this by comparing the
3489 time when an interrupt arrives with the time at the start of the first call
3490 to select(). If more time than the timeout has elapsed, the interrupt is
3491 treated as a timeout.
3493 5. Some internal re-factoring in preparation for the addition of Sieve
3494 extensions (by MH). In particular, the "personal" test is moved to a
3495 separate function, and given an option for scanning Cc: and Bcc: (which is
3496 not set for Exim filters).
3498 6. When Exim created an email address using the login of the caller as the
3499 local part (e.g. when creating a From: or Sender: header line), it was not
3500 quoting the local part when it contained special characters such as @.
3502 7. Installed new OpenBSD configuration files.
3504 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3505 try to make them clearer.
3507 9. Callout options, other than the timeout value, were being ignored when
3508 verifying sender addresses in header lines. For example, when using
3510 verify = header_sender/callout=no_cache
3512 the cache was (incorrectly) being used.
3514 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3515 only the Cygwin environment.
3517 11. When return_path_on_delivery was set as a log selector, if different remote
3518 addresses in the same message used different return paths and parallel
3519 remote delivery occurred, the wrong values would sometimes be logged.
3520 (Whenever a remote delivery process finished, the return path value from
3521 the most recently started remote delivery process was logged.)
3523 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3524 lines when AUTH and/or TLS are in use. This is the "received protocol"
3525 field. Exim used to use "asmtp" for authenticated SMTP, without any
3526 indication (in the protocol name) for TLS use. Now it follows the RFC and
3527 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3528 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3529 names appear in log lines as well as in Received: header lines.
3531 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3532 extensions, and comparison tests, and to fix some bugs.
3534 14. Changes to the "personal" filter test:
3536 (1) The test was buggy in that it was just doing the equivalent of
3537 "contains" tests on header lines. For example, if a user's address was
3538 anne@some.where, the "personal" test would incorrectly be true for
3540 To: susanne@some.where
3542 This test is now done by extracting each address from the header in turn,
3543 and checking the entire address. Other tests that are part of "personal"
3544 are now done using regular expressions (for example, to check local parts
3545 of addresses in From: header lines).
3547 (2) The list of non-personal local parts in From: addresses has been
3548 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3549 taken from the Sieve specification recommendations.
3551 (3) If the message contains any header line starting with "List-" it is
3552 treated as non-personal.
3554 (4) The test for "circular" in the Subject: header line has been removed
3555 because it now seems ill-conceived.
3557 15. Minor typos in src/EDITME comments corrected.
3559 16. Installed latest exipick from John Jetmore.
3561 17. If headers_add on a router specified a text string that was too long for
3562 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3563 of string_sprintf() is now avoided.
3565 18. $message_body_size was not set (it was always zero) when running the DATA
3566 ACL and the local_scan() function.
3568 19. For the "mail" command in an Exim filter, no default was being set for
3569 the once_repeat time, causing a random time value to be used if "once" was
3570 specified. (If the value happened to be <= 0, no repeat happened.) The
3571 default is now 0s, meaning "never repeat". The "vacation" command was OK
3572 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3573 (I found it when inspecting the code).
3575 20. There is now an overall timeout for performing a callout verification. It
3576 defaults to 4 times the callout timeout, which applies to individual SMTP
3577 commands during the callout. The overall timeout applies when there is more
3578 than one host that can be tried. The timeout is checked before trying the
3579 next host. This prevents very long delays if there are a large number of
3580 hosts and all are timing out (e.g. when the network connections are timing
3581 out). The value of the overall timeout can be changed by specifying an
3582 additional sub-option for "callout", called "maxwait". For example:
3584 verify = sender/callout=5s,maxwait=20s
3586 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3587 to the end before writing, but this should make it even safer).
3589 22. Exim was forgetting that it had advertised PIPELINING for the second and
3590 subsequent messages on an SMTP connection. It was also not resetting its
3591 memory on STARTTLS and an internal HELO.
3593 23. When Exim logs an SMTP synchronization error within a session, it now
3594 records whether PIPELINING has been advertised or not.
3596 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3597 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3598 rather than long int.
3600 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3602 26. Added the never_mail option to autoreply.
3608 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3609 crash if the getsockname() call failed; this can happen if a connection is
3610 closed very soon after it is established. The problem was simply in the
3611 order in which certain operations were done, causing Exim to try to write
3612 to the SMTP stream before it had set up the file descriptor. The bug has
3613 been fixed by making things happen in the correct order.
3619 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3620 before the connection was closed, thus losing the rejection response.
3622 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3623 some early Solaris releases, but causes trouble in current releases where
3624 socklen_t is defined.
3626 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3629 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3632 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3633 that releasing the top of it at the end releases what was used for sub-
3634 expansions (unless the block got too big). However, discard this block if
3635 the first thing is a variable or header, so that we can use its block when
3636 it is dynamic (useful for very large $message_headers, for example).
3638 6. Lookups now cache *every* query, not just the most recent. A new, separate
3639 store pool is used for this. It can be recovered when all lookup caches are
3640 flushed. Lookups now release memory at the end of their result strings.
3641 This has involved some general refactoring of the lookup sources.
3643 7. Some code has been added to the store_xxx() functions to reduce the amount
3644 of flapping under certain conditions.
3646 8. log_incoming_interface used to affect only the <= reception log lines. Now
3647 it causes the local interface and port to be added to several more SMTP log
3648 lines, for example "SMTP connection from", and rejection lines.
3650 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3652 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3654 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3655 could be overwritten at the end of the current message (or the start of a
3656 new message if it was set in a HELO ACL). The value is now preserved for
3657 the duration of the SMTP connection.
3659 12. If a transport had a headers_rewrite setting, and a matching header line
3660 contained an unqualified address, that address was qualified, even if it
3661 did not match any rewriting rules. The underlying bug was that the values
3662 of the flags that permit the existence of unqualified sender and recipient
3663 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3664 non-local messages, and by -bnq for local messages) were not being
3665 preserved with the message after it was received.
3667 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3668 "next input=" as part of the text comprising the host identity instead of
3669 the correct text. The code was using the same buffer for two different
3670 strings. However, depending on which order the printing function evaluated
3671 its arguments, the bug did not always show up. Under Linux, for example, my
3672 test suite worked just fine.
3674 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3675 doesn't work with some older versions of Perl. It has been changed to "my",
3676 which in any case is probably the better facility to use.
3678 15. A really picky compiler found some instances of statements for creating
3679 error messages that either had too many or two few arguments for the format
3682 16. The size of the buffer for calls to the DNS resolver has been increased
3683 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3684 for addresses that have a lot of PTR records. This alleviates a problem; it
3685 does not fully solve it.
3687 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3688 buffer now truncates the list and logs the incident, which is the same
3689 action as happens when Exim is looking up a host name and its aliases.
3690 Previously in this situation something unpredictable would happen;
3691 sometimes it was "internal error: store_reset failed".
3693 18. If a server dropped the connection unexpectedly when an Exim client was
3694 using GnuTLS and trying to read a response, the client delivery process
3695 crashed while trying to generate an error log message.
3697 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3698 string, for example:
3700 warn message = H1: something\nH2: something
3702 the text was added as a single header line from Exim's point of view
3703 though it ended up OK in the delivered message. However, searching for the
3704 second and subsequent header lines using $h_h2: did not work. This has been
3705 fixed. Similarly, if a system filter added multiple headers in this way,
3706 the routers could not see them.
3708 20. Expanded the error message when iplsearch is called with an invalid key to
3709 suggest using net-iplsearch in a host list.
3711 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3712 ACLs are no longer actually imposed (and a message to that effect is
3715 22. If a "gecos" field in a passwd entry contained escaped characters, in
3716 particular, if it contained a \" sequence, Exim got it wrong when building
3717 a From: or a Sender: header from that name. A second bug also caused
3718 incorrect handling when an unquoted " was present following a character
3719 that needed quoting.
3721 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3722 was not being matched caselessly.
3724 24. Arranged for all hyphens in the exim.8 source to be escaped with
3727 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3728 properly. Recipient callout cache records were still being keyed to include
3729 the sender, even when use_sender was set false. This led to far more
3730 callouts that were necessary. The sender is no longer included in the key
3731 when use_sender is false.
3733 26. Added "control = submission" modifier to ACLs.
3735 27. Added the ${base62d: operator to decode base 62 numbers.
3737 28. dnsdb lookups can now access SRV records.
3739 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3740 the configuration file.
3742 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3743 (-v) mode. This makes the output for a verbose queue run more intelligible.
3745 31. Added a use_postmaster feature to recipient callouts.
3747 32. Added the $body_zerocount variable, containing the number of binary zero
3748 bytes in the message body.
3750 33. The time of last modification of the "new" subdirectory is now used as the
3751 "mailbox time last read" when there is a quota error for a maildir
3754 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3756 35. Added +ignore_unknown as a special item in host lists.
3758 36. Code for decoding IPv6 addresses in host lists is now included, even if
3759 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3760 address was recognized as an IP address, but was then not correctly decoded
3761 into binary, causing unexpected and incorrect effects when compared with
3768 1. Very minor rewording of debugging text in manualroute to say "list of
3769 hosts" instead of "hostlist".
3771 2. If verify=header_syntax was set, and a header line with an unqualified
3772 address (no domain) and a large number of spaces between the end of the
3773 name and the colon was received, the reception process suffered a buffer
3774 overflow, and (when I tested it) crashed. This was caused by some obsolete
3775 code that should have been removed. The fix is to remove it!
3777 3. When running in the test harness, delay a bit after writing a bounce
3778 message to get a bit more predictability in the log output.
3780 4. Added a call to search_tidyup() just before forking a reception process. In
3781 theory, someone could use a lookup in the expansion of smtp_accept_max_
3782 per_host which, without the tidyup, could leave open a database connection.
3784 5. Added the variables $recipient_data and $sender_data which get set from a
3785 lookup success in an ACL "recipients" or "senders" condition, or a router
3786 "senders" option, similar to $domain_data and $local_part_data.
3788 6. Moved the writing of debug_print from before to after the "senders" test
3791 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3792 problems for message scanning, either using a data ACL, or using
3793 local_scan() because the Received: header was not generated till after they
3794 were called (in order to set the time as the time of reception completion).
3795 I have revised the way this works. The header is now generated after the
3796 body is received, but before the ACL or local_scan() are called. After they
3797 are run, the timestamp in the header is updated.
3803 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3804 before starting a queue runner without re-exec. This happened only when
3805 deliver_drop_privilege was set or when the Exim user was set to root. The
3806 effect of the bug was that timeouts during subsequent deliveries caused
3807 crashes instead of being properly handled. The handler is now left at its
3808 default (and expected) setting.
3810 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3811 message, again when deliver_drop_privilege is set or Exim is run as root.
3812 The bug described in (1) was not present in this case, but the tidying up
3813 of the other signals was missing. I have made the two cases consistent.
3815 3. The ignore_target_hosts setting on a manualroute router was being ignored
3816 for hosts that were looked up using the /MX notation.
3818 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3821 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3822 operated on the sender address. After changing the $sender_address to <>
3823 for the sender address verify, Exim was re-instated it as the original
3824 (before rewriting) address, but remembering that it had rewritten it, so it
3825 wasn't rewriting it again. This bug also had the effect of breaking the
3826 sender address verification caching when the sender address was rewritten.
3828 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3829 This has been changed so that if the ip literal address matches
3830 ignore_target_hosts, the router declines.
3832 7. Added expansion conditions match_domain, match_address, and match_local_
3833 part (NOT match_host).
3835 8. The placeholder for the Received: header didn't have a length field set.
3837 9. Added code to Exim itself and to exim_lock to test for a specific race
3838 condition that could lead to file corruption when using MBX delivery. The
3839 issue is with the lockfile that is created in /tmp. If this file is removed
3840 after a process has opened it but before that process has acquired a lock,
3841 there is the potential for a second process to recreate the file and also
3842 acquire a lock. This could lead to two Exim processes writing to the file
3843 at the same time. The added code performs the same test as UW imapd; it
3844 checks after acquiring the lock that its file descriptor still refers to
3845 the same named file.
3847 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3848 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3851 11. Added the smtp_active_hostname option. If used, this will typically be made
3852 to depend on the incoming interface address. Because $interface_address is
3853 not set up until the daemon has forked a reception process, error responses
3854 that can happen earlier (such as "too many connections") no longer contain
3857 12. If an expansion in a condition on a "warn" statement fails because a lookup
3858 defers, the "warn" statement is abandoned, and the next ACL statement is
3859 processed. Previously this caused the whole ACL to be aborted.
3861 13. Added the iplsearch lookup type.
3863 14. Added ident_timeout as a log selector.
3865 15. Added tls_certificate_verified as a log selector.
3867 16. Added a global option tls_require_ciphers (compare the smtp transport
3868 option of the same name). This controls incoming TLS connections.
3870 17. I finally figured out how to make tls_require_ciphers do a similar thing
3871 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3872 before starting the TLS session.
3874 18. Tabs are now shown as \t in -bP output.
3876 19. If the log selector return_path_on_delivery was set, Exim crashed when
3877 bouncing a message because it had too many Received: header lines.
3879 20. If two routers both had headers_remove settings, and the first one included
3880 a superfluous trailing colon, the final name in the first list and the
3881 first name in the second list were incorrectly joined into one item (with a
3882 colon in the middle).
3888 1. Added -C and -D options to the exinext utility, mainly to make it easier
3889 to include in the automated testing, but these could be helpful when
3890 multiple configurations are in use.
3892 2. The exinext utility was not formatting the output nicely when there was
3893 an alternate port involved in the retry record key, nor when there was a
3894 message id as well (for retries that were specific to a specific message
3895 and a specific host). It was also confused by IPv6 addresses, because of
3896 the additional colons they contain. I have fixed the IPv4 problem, and
3897 patched it up to do a reasonable job for IPv6.
3899 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3900 delivery, the log line now contains "pipelined" if PIPELINING was used.
3902 4. An SMTP transport process used to panic and die if the bind() call to set
3903 an explicit outgoing interface failed. This has been changed; it is now
3904 treated in the same way as a connect() failure.
3906 5. A reference to $sender_host_name in the part of a conditional expansion
3907 that was being skipped was still causing a DNS lookup. This no longer
3910 6. The def: expansion condition was not recognizing references to header lines
3911 that used bh_ and bheader_.
3913 7. Added the _cache feature to named lists.
3915 8. The code for checking quota_filecount in the appendfile transport was
3916 allowing one more file than it should have been.
3918 9. For compatibility with Sendmail, the command line option
3926 and sets the incoming protocol and host name (for trusted callers). The
3927 host name and its colon can be omitted when only the protocol is to be set.
3928 Note the Exim already has two private options, -pd and -ps, that refer to
3929 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3930 "s", but I don't think that's a major issue.
3932 10. A number of refactoring changes to the code, none of which should affect
3935 (a) The number of logging options was getting close to filling up the
3936 32-bit word that was used as a bit map. I have split them into two classes:
3937 those that are passed in the argument to log_write(), and those that are
3938 only ever tested independently outside of that function. These are now in
3939 separate 32-bit words, so there is plenty of room for expansion again.
3940 There is no change in the user interface or the logging behaviour.
3942 (b) When building, for example, log lines, the code previously used a
3943 macro that called string_cat() twice, in order to add two strings. This is
3944 not really sufficiently general. Furthermore, there was one instance where
3945 it was actually wrong because one of the argument was used twice, and in
3946 one call a function was used. (As it happened, calling the function twice
3947 did not affect the overall behaviour.) The macro has been replaced by a
3948 function that can join an arbitrary number of extra strings onto a growing
3951 (c) The code for expansion conditions now uses a table and a binary chop
3952 instead of a serial search (which was left over from when there were very
3953 few conditions). Also, it now recognizes conditions like "pam" even when
3954 the relevant support is not compiled in: a suitably worded error message is
3955 given if an attempt is made to use such a condition.
3957 11. Added ${time_interval:xxxxx}.
3959 12. A bug was causing one of the ddress fields not to be passed back correctly
3960 from remote delivery subprocesses. The field in question was not being
3961 subsequently used, so this caused to problems in practice.
3963 13. Added new log selectors queue_time and deliver_time.
3965 14. Might have fixed a bug in maildirsizefile handling that threw up
3966 "unexpected character" debug warnings, and recalculated the data
3967 unnecessarily. In any case, I expanded the warning message to give more
3970 15. Added the message "Restricted characters in address" to the statements in
3971 the default ACL that block characters like @ and % in local parts.
3973 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3974 Three changes have been made:
3976 (a) There was a serious bug; a negative response to MAIL caused the whole
3977 recipient domain to be cached as invalid, thereby blocking all messages
3978 to all local parts at the same domain, from all senders. This bug has
3979 been fixed. The domain is no longer cached after a negative response to
3980 MAIL if the sender used is not empty.
3982 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3985 (c) A new callout option, "use_sender" has been added for people who want
3986 the modified behaviour.
3992 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3995 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3996 indeed breaks things for older releases.
3998 3. Added additional logging to the case where there is a problem reading data
3999 from a filter that is running in a subprocess using a pipe, in order to
4000 try to track down a specific problem.
4002 4. Testing facility fudge: when running in the test harness and attempting
4003 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4004 getting "No route to host". Convert this to a timeout.
4006 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4009 6. Some OS don't have socklen_t but use size_t instead. This affects the
4010 fifth argument of getsockopt() amongst other things. This is now
4011 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4012 can be set for individual OS. I have set it for SunOS5, OSF1, and
4013 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4014 some earlier ones do not.
4016 7. Change 4.30/15 was not doing the test caselessly.
4018 8. The standard form for an IPv6 address literal was being rejected by address
4019 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4020 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4021 this, as well as the form without the "IPv6" on the front (but only when
4022 address literals are enabled, of course).
4024 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4026 10. Exim crashed if a message with an empty sender address specified by -f
4027 encountered a router with an errors_to setting. This could be provoked only
4028 by a command such as
4032 where an empty string was supplied; "<>" did not hit this bug.
4034 11. Installed PCRE release 4.5.
4036 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4037 remained set. It is now erased.
4039 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4040 times from message ids (which are base 36 rather than the normal 62).
4042 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4043 were being counted as actual protocol errors, and logged if the log
4044 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4045 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4046 and DATA following a set of rejected RCPTs do not count as protocol errors.
4047 In other words, Exim assumes they were pipelined, though this may not
4048 actually be the case. Of course, in all cases the client gets an
4049 appropriate error code.
4051 15. If a lookup fails in an ACL condition, a message about the failure may
4052 be available; it is used if testing the ACL cannot continue, because most
4053 such messages specify what the cause of the deferral is. However, some
4054 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4055 that caused an old message to be retained and used if a later statement
4056 caused a defer, replacing the real cause of the deferral.
4058 16. If an IP address had so many PTR records that the DNS lookup buffer
4059 was not large enough to hold them, Exim could crash while trying to process
4060 the truncated data. It now detects and logs this case.
4062 17. Further to 4.21/58, another change has been made: if (and only if) the
4063 first line of a message (the first header line) ends with CRLF, a bare LF
4064 in a subsequent header line has a space inserted after it, so as not to
4065 terminate the header.
4067 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4068 unnecessarily, used atoi() instead of strtol(), and didn't check the
4069 termination when getting file sizes from file names by regex.
4071 19. Completely re-implemented the support for maildirsize files, in the light
4072 of a number of problems with the previous contributed implementation
4073 (4.30/29). In particular:
4075 . If the quota is zero, the maildirsize file is maintained, but no quota is
4078 . If the maildir directory does not exist, it is created before any attempt
4079 to write a maildirsize file.
4081 . The quota value in the file is just a cache; if the quota is changed in
4082 the transport, the new value overrides.
4084 . A regular expression is available for excluding directories from the
4087 20. The autoreply transport checks the characters in options that define the
4088 message's headers; it allows continued headers, but it was checking with
4089 isspace() after an embedded newline instead of explicitly looking for a
4092 21. If all the "regular" hosts to which an address was routed had passed their
4093 expiry times, and had not reached their retry times, the address was
4094 bounced, even if fallback hosts were defined. Now Exim should go on to try
4097 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4098 equivalent code in the SMTP transport. Some hosts send humungous responses
4099 to HELO/EHLO, more than 1024 it seems.
4101 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4102 gives compiler warnings in some environments. I've now done it "properly",
4105 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4106 (because of problems with the built-in one) was declared to return uschar *
4107 instead of char *, causing compiler failure.
4109 25. Fixed a file descriptor leak when processing alias/forward files.
4111 26. Fixed a minor format string issue in dbfn.c.
4113 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4115 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4116 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4119 29. When a sender address is verified, it is cached, to save repeating the test
4120 when there is more than one recipient in a message. However, when the
4121 verification involves a callout, it is possible for different callout
4122 options to be set for different recipients. It is too complicated to keep
4123 track of this in the cache, so now Exim always runs a verification when a
4124 callout is required, relying on the callout cache for the optimization.
4125 The overhead is duplication of the address routing, but this should not be
4128 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4129 to be verified with callout=postmaster, and the main callout worked but the
4130 postmaster check failed, the verification correctly failed. However, if a
4131 subsequent RCPT command asked for sender verification *without* the
4132 postmaster check, incorrect caching caused this verification also to fail,
4135 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4136 it was not caching the DNS options (qualify_single, search_parents) that
4137 were used when the lookup failed. A subsequent lookup with different
4138 options therefore always gave the same answer, though there were cases
4139 where it should not have. (Example: a "domains = !$mx_any" option on a
4140 dnslookup router: the "domains" option is always processed without any
4141 widening, but the router might have qualify_single set.) Now Exim uses the
4142 cached value only when the same options are set.
4144 32. Added John Jetmore's "exipick" utility to the distribution.
4146 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4147 than a timeout (e.g. a certificate is required, and is not provided), an
4148 Exim server now closes the connection immediately. Previously it waited for
4149 the client to close - but if the client is SSL, it seems that they each
4150 wait for each other, leading to a delay before one of them times out.
4152 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4153 maintained 0.8.x compatibility because I don't think many are using it, and
4154 it is clearly obsolete.
4156 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4159 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4160 client certificate was expired. A simple patch fixes this, though I don't
4161 understand the full logic of why the verify callback is called multiple
4164 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4165 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4166 which causes problems with some clients (such as the Certicom SSL Plus
4167 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4168 disables the coutermeasure allowing Eudora to connect."
4170 38. Exim was not checking that a write() to a log file succeeded. This could
4171 lead to Bad Things if a log got too big, in particular if it hit a file
4172 size limit. Exim now panics and dies if it cannot write to a log file, just
4173 as it does if it cannot open a log file.
4175 39. Modified OS/Makefile-Linux so that it now contains
4177 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4179 The two -D definitions ensure that Exim is compiled with large file
4180 support, which makes it possible to handle log files that are bigger than
4183 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4184 instance) a domain was checked against a named list that involved a lookup,
4185 causing $domain_data to be set, then another domain was checked against the
4186 same list, then the first domain was re-checked, the value of $domain_data
4187 after the final check could be wrong. In particular, if the second check
4188 failed, it could be set empty. This bug probably also applied to
4191 41. The strip_trailing_dot option was not being applied to the address given
4192 with the -f command-line option.
4194 42. The code for reading a message's header from the spool was incrementing
4195 $received_count, but never initializing it. This meant that the value was
4196 incorrect (doubled) while delivering a message in the same process in which
4197 it was received. In the most common configuration of Exim, this never
4198 happens - a fresh exec is done - but it can happen when
4199 deliver_drop_privilege is set.
4201 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4202 it now includes up to 150 characters of the unexpected data in the log
4205 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4206 and building data strings. The size of both of these buffers was 10 000
4207 bytes - far larger than anybody would *ever* want, thought I. Needless to
4208 say, somebody hit the limit. I have increased the maximum line length to
4209 20 000 and the maximum data length of concatenated lines to 100 000. I have
4210 also fixed two bugs, because there was no checking on these buffers. Tsk,
4211 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4212 buffer is too small.
4214 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4215 lsearch lookups. Now it does.
4217 46. When parsing a route_list item in a manualroute router, a fixed-length
4218 buffer was used for the list of hosts. I made this 1024 bytes long,
4219 thinking that nobody would ever have a list of hosts that long. Wrong.
4220 Somebody had a whole pile of complicated expansion conditions, and the
4221 string was silently truncated, leading to an expansion error. It turns out
4222 that it is easier to change to an unlimited length (owing to other changes
4223 that have happened since this code was originally written) than to build
4224 structure for giving a limitation error. The length of the item that
4225 expands into the list of hosts is now unlimited.
4227 47. The lsearch lookup could not handle data where the length of text line was
4228 more than 4095 characters. Such lines were truncated, leading to shortened
4229 data being returned. It should now handle lines of any length.
4231 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4232 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4235 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4236 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4237 no longer happens in comment lines. A list of replacements is now placed
4238 at the head of all of the source files, except those whose only change is
4239 to replace PERL_COMMAND in the very first #! line.
4241 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4242 messages on the queue, with a bottom-up merge sort, using code contributed
4243 by Michael Haardt. This should make operations like -bp somewhat faster on
4244 large queues. It won't affect queue runners, except when queue_run_in_order
4247 51. Installed eximstats 1.31 in the distribution.
4249 52. Added support for SRV lookups to the dnslookup router.
4251 53. If an ACL referred to $message_body or $message_body_end, the value was not
4252 reset for any messages that followed in the same SMTP session.
4254 54. The store-handling optimization for building very long strings was not
4255 differentiating between the different store pools. I don't think this
4256 actually made any difference in practice, but I've tidied it.
4258 55. While running the routers to verify a sender address, $sender_address
4259 was still set to the sender address. This is wrong, because when routing to
4260 send a bounce to the sender, it would be empty. Therefore, I have changed
4261 it so that, while verifying a sender address, $sender_address is set to <>.
4262 (There is no change to what happens when verifying a recipient address.)
4264 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4265 target A or AAAA records (if not already returned) without resetting the
4266 qualify_single or search_parents options of the DNS resolver. These are
4267 inappropriate in this case because the targets of MX and SRV records must
4268 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4269 target that, when qualified, matched something in the local domain. These
4270 two options are now turned off when doing these lookups.
4272 57. It seems that at least some releases of Reiserfs (which does not have the
4273 concept of a fixed number of inodes) returns zero and not -1 for the
4274 number of available inodes. This interacted badly with check_spool_inodes,
4275 which assumed that -1 was the "no such thing" setting. What I have done is
4276 to check that the total number of inodes is greater than zero before doing
4277 the test of how many are available.
4279 58. When a "warn" ACL statement has a log_message modifier, the message is
4280 remembered, and not repeated. This is to avoid a lot of repetition when a
4281 message has many recipients that cause the same warning to be written.
4282 Howewer, Exim was preserving the list of already written lines for an
4283 entire SMTP session, which doesn't seem right. The memory is now reset if a
4284 new message is started.
4286 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4287 debugging output unless log_rewrite was also set.
4289 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4290 of (int)(handle) when we know that handle contains (void *)(-1).
4292 61. The Exim daemon panic-logs an error return when it closes the incoming
4293 connection. However "connection reset by peer" seems to be common, and
4294 isn't really an error worthy of noting specially, so that particular error
4297 62. When Exim is trying to find all the local interfaces, it used to panic and
4298 die if the ioctl to get the interface flags failed. However, it seems that
4299 on at least one OS (Solaris 9) it is possible to have an interface that is
4300 included in the list of interfaces, but for which you get a failure error
4301 for this call. This happens when the interface is not "plumbed" into a
4302 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4303 failure of the "get flags" call assumes that the interface is down.
4305 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4306 makes life easier for people who are doing arithmetic on fields extracted
4307 from dates, where you often get leading zeros that should not be
4308 interpreted as octal.
4310 64. Added qualify_domain to the redirect router, to override the global
4313 65. If a pathologically long header line contained very many addresses (the
4314 report of this problem mentioned 10 000) and each of them was rewritten,
4315 Exim could use up a very large amount of memory. (It kept on making new
4316 copies of the header line as it rewrote, and never released the old ones.)
4317 At the expense of a bit more processing, the header rewriting function has
4318 been changed so that it no longer eats memory in this way.
4320 66. The generation of the Received: header has been moved from the time that a
4321 message starts to be received, to the time that it finishes. The timestamp
4322 in the Received: header should now be very close to that of the <= log
4323 line. There are two side-effects of this change:
4325 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4326 logged header lines no longer include the local Received: line, because
4327 it has not yet been created. The same applies to a copy of the message
4328 that is returned to a non-SMTP sender when a message is rejected.
4330 (b) When a filter file is tested using -bf, no additional Received: header
4331 is added to the test message. After some thought, I decided that this
4334 This change does not affect the value of $received_for. It is still set
4335 after address rewriting, but before local_scan() is called.
4337 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4339 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4340 gave an unhelpful panic error message, and a defer error. I have managed to
4341 change this behaviour so that it now rejects any supplied certificate,
4342 which seems right, as the list of acceptable certificates is empty.
4344 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4345 gave an unhelpful defer error. I have not managed to make this reject any
4346 supplied certificates, but the error message it gives is "no certificate
4347 supplied", which is not helpful.
4349 70. exigrep's output now also includes lines that are not associated with any
4350 message, but which match the given pattern. Implemented by a patch from
4351 Martin Sluka, which also tidied up the Perl a bit.
4353 71. Recipient callout verification, like sender verification, was using <> in
4354 the MAIL FROM command. This isn't really the right thing, since the actual
4355 sender may affect whether the remote host accepts the recipient or not. I
4356 have changed it to use the actual sender in the callout; this means that
4357 the cache record is now keyed on a recipient/sender pair, not just the
4358 recipient address. There doesn't seem to be a real danger of callout loops,
4359 since a callout by the remote host to check the sender would use <>.
4360 [SEE ABOVE: changed after hitting problems.]
4362 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4363 temporary errors. However, in the case of such a code being given after
4364 the end of a data transmission (i.e. after ".") Exim was failing to write
4365 a retry record for the message. (Yes, there was some broken host that was
4366 actually sending 8xx at this point.)
4368 73. An unknown lookup type in a host list could cause Exim to panic-die when
4369 the list was checked. (An example that provoked this was putting <; in the
4370 middle of a list instead of at the start.) If this happened during a DATA
4371 ACL check, a -D file could be left lying around. This kind of configuration
4372 error no longer causes Exim to die; instead it causes a defer errror. The
4373 incident is still logged to the main and panic logs.
4375 74. Buglet left over from Exim 3 conversion. The message "too many messages
4376 in one connection" was written to the rejectlog but not the mainlog, except
4377 when address rewriting (yes!) was being logged.
4379 75. Added write_rejectlog option.
4381 76. When a system filter was run not as root (that is, when system_filter_user
4382 was set), the values of the $n variables were not being returned to the
4383 main process; thus, they were not subsequently available in the $sn
4386 77. Added +return_path_on_delivery log selector.
4388 78. A connection timeout was being treated differently from recipients deferred
4389 when testing hosts_max_try with a message that was older than the host's
4390 retry timeout. (The host should not be counted, thus allowing all hosts to
4391 be tried at least once before bouncing.) This may have been the cause of an
4392 occasionally reported bug whereby a message would remain on the queue
4393 longer than the retry timeout, but would be bounced if a delivery was
4394 forced. I say "may" because I never totally pinned down the problem;
4395 setting up timeout/retry tests is difficult. See also the next item.
4397 79. The ultimate address timeout was not being applied to errors that involved
4398 a combination of host plus message (for example, a timeout on a MAIL
4399 command). When an address resolved to a number of possible hosts, and they
4400 were not all tried for each delivery (e.g. because of hosts_max_try), a
4401 message could remain on the queue longer than the retry timeout.
4403 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4406 81. Fixed an obscure SMTP outgoing bug which required at least the following
4407 conditions: (a) there was another message waiting for the same server;
4408 (b) the server returned 5xx to all RCPT commands in the first message so
4409 that the message was not completed; (c) the server dropped the connection
4410 or gave a negative response to the RSET that Exim sends to abort the
4411 transaction. The observed case was a dropped connection after DATA that had
4412 been sent in pipelining mode. That is, the server had advertised PIPELINING
4413 but was not implementing it correctly. The effect of the bug was incorrect
4414 behaviour, such as trying another host, and this could lead to a crash.
4420 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4421 and daemon.c were passed as pointers to ints; they should have been
4422 pointers to socklen_t variables (which are typically unsigned ints).
4424 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4427 3. Fixed a really odd bug that affected only the testing scheme; patching a
4428 certain fixed string in the binary changed the value of another string that
4429 happened to be identical to the end of the original first string.
4431 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4432 name", it returns that address as the IP address. On some operating
4433 systems (e.g. Solaris), it also passes back the IP address string as the
4434 "host name". However, on others (e.g. Linux), it passes back an empty
4435 string. Exim wasn't checking for this, and was changing the host name to an
4436 empty string, assuming it had been canonicized.
4438 5. Although rare, it is permitted to have more than one PTR record for a given
4439 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4440 all the names associated with an address, because they do in Solaris.
4441 However, it seems that they do not in Linux for data that comes from the
4442 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4443 I found this out when I moved to a new Linux workstation and tried to run
4444 the Exim test suite.
4446 To get round this problem I have changed the code so that it now does its
4447 own call to the DNS to look up PTR records when searching for a host name.
4448 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4449 addresses that are only in /etc/hosts are still found.
4451 This behaviour is, however, controlled by an option called host_lookup_
4452 order, which defaults to "bydns:byaddr". If people want to use the other
4453 order, or indeed, just use one or the other means of lookup, they can
4454 specify it in this variable.
4456 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4457 some operating systems, this comes back from gethostbyaddr() as an empty
4458 string, and this is what Exim used to test for. However, it seems that in
4459 other systems, "." is yielded. Exim now tests for this case too.
4461 7. The values of check_spool_space and check_log_space are now held internally
4462 as a number of kilobytes instead of an absolute number of bytes. If a
4463 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4464 kilobyte. This means that much larger values can be stored.
4466 8. Exim monitor: an attempt to get the action menu when not actually pointing
4467 at a message produces an empty menu entitled "No message selected". This
4468 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4469 no entries in it ("Shell widget menu has zero width and/or height"). So I
4470 have added a single, blank menu entry in this case.
4472 9. Added ${quote_local_part.
4474 10. MIME decoding is now applied to the contents of Subject: header lines when
4477 11. Now that a reference to $sender_host_address automatically causes a reverse
4478 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4479 host lookup before query-style lookups in lists that might use this
4480 variable. This has therefore been abolished, and the "net-" prefix is no
4481 longer necessary for query-style lookups.
4483 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4484 have been a typo for LFLAGS, so it has been changed.
4486 13. The install script calls Exim with "-C /dev/null" in order to find the
4487 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4488 to be output. Howeve, since Exim outputs its version number before the
4489 error, it didn't break the script. It just looked ugly. I fixed this by
4490 always allowing "-C /dev/null" if the caller is root.
4492 14. Ignore overlarge ACL variable number when reading spool file - insurance
4493 against a later release with more variables having written the file.
4495 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4496 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4497 this, as well as the form without the "IPv6" on the front.
4499 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4500 OS/Makefile-Darwin file.
4502 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4503 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4505 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4506 variable. However, some OS use ints for inodes. I've added cast to long int
4507 to get rid of the compiler warning.
4509 19. I had forgotten to lock out "/../" in configuration file names when
4510 ALT_CONFIG_PREFIX was set.
4512 20. Routers used for verification do not need to specify transports. However,
4513 if such a router generated a host list, and callout was configured, Exim
4514 crashed, because it could not find a port number from the (non-existent)
4515 transport. It now assumes port 25 in this circumstance.
4517 21. Added the -t option to exigrep.
4519 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4520 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4521 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4522 them is set, LOOKUP_LSEARCH is forced.
4524 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4526 24. Added sender and host information to the "rejected by local_scan()" log
4527 line; previously there was no indication of these.
4529 25. Added .include_if_exists.
4531 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4532 Linux. It turns out that not all file systems support this. Apparently some
4533 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4534 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4535 not supported on the file descriptor, is now ignored when Exim is trying to
4536 sync a directory. This applies only to Linux.
4538 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4540 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4541 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4544 29. Added support for maildirsize files from supplied patch (modified a bit).
4546 30. The use of :fail: followed by an empty string could lead Exim to respond to
4547 sender verification failures with (e.g.):
4549 550 Verification failed for <xxx>
4550 550 Sender verify failed
4552 where the first response line was missing the '-' that indicates it is not
4553 the final line of the response.
4555 31. The loop for finding the name of the user that called Exim had a hardwired
4556 limit of 10; it now uses the value of finduser_retries, which is used for
4557 all other user lookups.
4559 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4562 33. Exim was neglecting to zero errno before one call of strtol() when
4563 expanding a string and expecting an integer value. On some systems this
4564 resulted in spurious "integer overflow" errors. Also, it was casting the
4565 result into an int without checking.
4567 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4568 did not work. The code looks as if it has *never* worked, though it appears
4569 to have been documented since at least releast 1.62. I have made it work.
4571 35. The "timeout_DNS" error in retry rules, also documented since at least
4572 1.62, also never worked. As it isn't clear exactly what this means, and
4573 clearly it isn't a major issue, I have abolished the feature by treating it
4574 as "timeout", and writing a warning to the main and panic logs.
4576 36. The display of retry rules for -brt wasn't always showing the error code
4579 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4580 timeout_connect_A, timeout_connect_MX.
4582 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4583 to the empty sender.
4585 39. The daemon was not analysing the content of -oX till after it had closed
4586 stderr and disconnected from the controlling terminal. This meant that any
4587 syntax errors were only noted on the panic log, and the return code from
4588 the command was 0. By re-arranging the code a little, I've made the
4589 decoding happen first, so such errors now appear on stderr, and the return
4590 code is 1. However, the actual setting up of the sockets still happens in
4591 the disconnected process, so errors there are still only recorded on the
4594 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4595 connections (as happens on some IP stacks) was logged at start up time as
4596 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4597 it from "IPv6 and IPv4", which means that two separate sockets are being
4600 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4601 says whether AF_INET or AF_INET6 was passed as an argument.
4603 42. Exiwhat output was messed up when time zones were included in log
4606 43. Exiwhat now gives more information about the daemon's listening ports,
4607 and whether -tls-on-connect was used.
4609 44. The "port" option of the smtp transport is now expanded.
4611 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4612 silently ignored. Now an error message is written to the main and panic
4615 46. There's a new ACL modifier called "logwrite" which writes to a log file
4616 as soon as it is encountered.
4618 47. Added $local_user_uid and $local_user_gid at routing time.
4620 48. Exim crashed when trying to verify a sender address that was being
4623 49. Exim was recognizing only a space character after ".include". It now also
4624 recognizes a tab character.
4626 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4627 extracting the relevant information from the specification. The man page no
4628 longer contains scrambled data for the -d option, and I've added a section
4629 at the front about calling Exim under different names.
4631 51. Added "extra_headers" argument to the "mail" command in filter files.
4633 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4636 53. Installed eximstats 1.29.
4638 54. Added transport_filter_timeout as a generic transport option.
4640 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4641 Cc: header lines. This was required by RFC 822, but it not required by RFC
4644 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4645 incoming messages that did not have them. Now it does so only if the
4646 message originates locally, that is, if there is no associated remote host
4647 address. When Resent- header lines are present, this applies to the Resent-
4648 lines rather than the non-Resent- lines.
4650 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4651 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4653 58. Messages for configuration errors now include the name of the main
4654 configuration file - useful now that there may be more than one file in a
4655 list (.included file names were always shown).
4657 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4658 for those rare installations that do not start the daemon as root or run it
4659 setuid root. I've cut out the call to initgroups() if the daemon is not
4662 60. The Exim user and group can now be bound into the binary as text strings
4663 that are looked up at the start of Exim's processing.
4665 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4667 62. Added $mailstore_basename variable.
4669 63. Installed patch to sieve.c from Michael Haardt.
4671 64. When Exim failed to open the panic log after failing to open the main log,
4672 the original message it was trying to log was written to stderr and debug
4673 output, but if they were not available (the usual case in production), it
4674 was lost. Now it is written to syslog before the two lines that record the
4675 failures to open the logs.
4677 65. Users' Exim filters run in subprocesses under the user's uid. It is
4678 possible for a "deliver" command or an alias in a "personal" command to
4679 provoke an address rewrite. If logging of address rewriting is configured,
4680 this fails because the process is not running as root or exim. There may be
4681 a better way of dealing with this, but for the moment (because 4.30 needs
4682 to be released), I have disabled address rewrite logging when running a
4683 filter in a non-root, non-exim process.
4689 1. The buildconfig auxiliary program wasn't quoting the value set for
4690 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4691 not defined. This bug was masked in 4.22 by the effect that was fixed in
4694 2. Some messages that were rejected after a message id was allocated were
4695 shown as "incomplete" by exigrep. It no longer does this for messages that
4696 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4698 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4699 have allow_domain_literals set, the ID did not get logged in the <= line.
4700 Domain literals are now always recognized in Message-ID: header lines.
4702 4. The first argument for a ${extract expansion item is the key name or field
4703 number. Leading and trailing spaces in this item were not being ignored,
4704 causing some misleading effects.
4706 5. When deliver_drop_privilege was set, single queue runner processes started
4707 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4708 same command in the process it spins off) were not dropping privilege.
4710 6. When the daemon running as "exim" started a queue runner, it always
4711 re-executed Exim in the spun-off process. This is a waste of effort when
4712 deliver_drop_privilege is set. The new process now just calls the
4713 queue-runner function directly.
4719 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4722 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4723 ignored. Though the use of -oP was forcing the writing of a pid file, it
4724 was always written to the default place.
4726 3. If the message "no IP address found for host xxxx" is generated during
4727 incoming verification, it is now followed by identification of the incoming
4728 connection (so you can more easily find what provoked it).
4730 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4732 5. Added some features to "harden" Exim a bit more against certain attacks:
4734 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4735 be put in Local/Makefile. This is like the never_users runtime option,
4736 but it cannot be overridden. The default setting is "root".
4738 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4739 prefix string with which any file named in a -C command line option
4742 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4743 is retained for -C and -D only if the caller of Exim is root. Without
4744 it, the exim user may also use -C and -D and retain privilege.
4746 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4747 command line option is disabled.
4749 6. Macro names set by the -D option must start with an upper case letter, just
4750 like macro names defined in the configuration file.
4752 7. Added "dereference=" facility to LDAP.
4754 8. Two instances of the typo "uknown" in the source files are fixed.
4756 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4757 the Configure-Makefile script screwed up while processing it.
4759 10. Incorporated PCRE 4.4.
4761 11. The SMTP synchronization check was not operating right at the start of an
4762 SMTP session. For example, it could not catch a HELO sent before the client
4763 waited for the greeting. There is now a check for outstanding input at the
4764 point when the greeting is written. Because of the duplex, asynchronous
4765 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4766 way, but not yet received, when the check is performed.
4768 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4769 on TCP/IP sockets, because this apparently causes some broken clients to
4772 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4773 unchanged) from the Cygwin maintainer.
4775 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4776 was supported instead of testing for mbx. Effectively a typo.
4778 15. The spa authenticator server code was not checking that the input it
4779 received was valid base64.
4781 16. The debug output line for the "set" modifier in ACLs was not showing the
4782 name of the variable that was being set.
4784 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4786 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4787 lookup on its own. Something else was needed to trigger the lookup. For
4788 example, a match in host_lookup or the need for a host name in a host list.
4789 Now, if $sender_host_name is referenced and the host name has not yet been
4790 looked up, a lookup is performed. If the lookup fails, the variable remains
4791 empty, and $host_lookup_failed is set to "1".
4793 19. Added "eqi" as a case-independent comparison operator.
4795 20. The saslauthd authentication condition could segfault if neither service
4796 nor realm was specified.
4798 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4799 error message that was logged was misleading, and incoming SMTP
4800 connections were dropped. The message is now more accurate, and temporary
4801 errors are given to SMTP connections.
4803 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4804 (except RSET) with a 421 error, until QUIT is received. However, it was
4805 failing to send a response to QUIT.
4807 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4808 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4809 ACL is now run after the helo_try_verify_hosts code.
4811 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4812 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4813 case-independent, so other case variants are also recognized). Apparently
4814 some systems use these upper case variants.
4816 25. If more than two messages were waiting for the same host, and a transport
4817 filter was specified for the transport, Exim sent two messages over the
4818 same TCP/IP connection, and then failed with "socket operation on non-
4819 socket" when it tried to send the third.
4821 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4823 27. The extern definition of crypt16() in expand.c was not being excluded when
4824 the OS had its own crypt16() function.
4826 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4827 as a preferred synonym for return_size_limit, both as an option and as an
4830 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4832 30. Changed the default configuration ACL to relax the local part checking rule
4833 for addresses that are not in any local domains. For these addresses,
4834 slashes and pipe symbols are allowed within local parts, but the sequence
4835 /../ is explicitly forbidden.
4837 31. SPA server authentication was not clearing the challenge buffer before
4840 32. log_message in a "warn" ACL statement was writing to the reject log as
4841 well as to the main log, which contradicts the documentation and doesn't
4842 seem right (because no rejection is happening). So I have stopped it.
4844 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4845 However, I am unable to do any testing of this.
4847 34. Fixed an infelicity in the appendfile transport. When checking directories
4848 for a mailbox, to see if any needed to be created, it was accidentally
4849 using path names with one or more superfluous leading slashes; tracing
4850 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4852 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4853 contents are added to the log line that is written for every discarded
4854 recipient. (Previously a log_message setting was ignored.)
4856 36. The ${quote: operator now quotes the string if it is empty.
4858 37. The install script runs exim in order to find its version number. If for
4859 some reason other than non-existence or emptiness, which it checks, it
4860 could not run './exim', it was installing it with an empty version number,
4861 i.e. as "exim-". This error state is now caught, and the installation is
4864 38. An argument was missing from the function that creates an error message
4865 when Exim fails to connect to the socket for saslauthd authentication.
4866 This could cause Exim to crash, or give a corrupted message.
4868 39. Added isip, isip4, and isip6 to ${if conditions.
4870 40. The ACL variables $acl_xx are now saved with the message, and can be
4871 accessed later in routers, transports, and filters.
4873 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4874 strings in the file are not string-expanded.
4876 42. If a MAIL command specified a SIZE value that was too large to fit into an
4877 int variable, the check against message_size_limit failed. Such values are
4878 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4879 day this will have to be increased, but I don't think I want to be around
4880 when emails are that large.
4887 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4888 iconv() is not standard in FreeBSD.
4890 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4891 IPv6 enabled. The observed symptom was a segmentation fault on return from
4892 the function os_common_find_running_interfaces() in src/os.c.
4894 3. In the check_special_case() function in daemon.c I had used "errno" as an
4895 argument name, which causes warnings on some systems. This was basically a
4896 typo, since it was named "eno" in the comments!
4898 4. The code that waits for the clock to tick (at a resolution of some fraction
4899 of a second) so as to ensure message-id uniqueness was always waiting for
4900 at least one whole tick, when it could have waited for less. [This is
4901 almost certainly not relevant at current processor speeds, where it is
4902 unlikely to ever wait at all. But we try to future-proof.]
4904 5. The function that sleeps for a time interval that includes fractions of a
4905 second contained a race. It did not block SIGALRM between setting the
4906 timer, and suspending (a couple of lines later). If the interval was short
4907 and the sigsuspend() was delayed until after it had expired, the suspension
4908 never ended. On busy systems this could lead to processes getting stuck for
4911 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4912 process, before it forks any delivery processes. The open lookup caching
4913 mechanism meant that the open file or database connection was passed into
4914 the delivery process. The problem was that delivery processes always tidy
4915 up cached lookup data. This could cause a problem for the next delivery
4916 process started by the queue runner, because the external queue runner
4917 process does not know about the closure. So the next delivery process
4918 still has data in the lookup cache. In the case of a file lookup, there was
4919 no problem because closing a file descriptor in a subprocess doesn't affect
4920 the parent. However, if the lookup was caching a connection to a database,
4921 the connection was closed, and the second delivery process was likely to
4922 see errors such as "PGSQL: query failed: server closed the connection
4923 unexpectedly". The problem has been fixed by closing all cached lookups
4924 in a queue runner before running a delivery process.
4926 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4927 seem to have the "const" qualifier which it has on other OS. I've
4930 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4931 *of the same type* (client or server) with the same public name that an
4932 error should be diagnosed.
4934 9. When Exim looked up a host name for an IP address, but failed to find the
4935 original IP address when looking up the host name (a safety check), it
4936 output the message "<ip address> does not match any IP for NULL", which was
4937 confusing, to say the least. The bug was that the host name should have
4938 appeared instead of "NULL".
4940 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4941 user that is built into the binary, and the -C or -D options is used, root
4942 privilege is dropped before the configuration file is read. In addition,
4943 logging is switched to stderr instead of the normal log files. If the
4944 configuration then re-defines the Exim user, the unprivileged environment
4945 is probably not what is expected, so Exim logs a panic warning message (but
4948 However, if deliver_drop_privilege is set, the unprivileged state may well
4949 be exactly what is intended, so the warning has been cut out in that case,
4950 and Exim is allowed to try to write to its normal log files.
4956 1. smtp_return_error_details was not giving details for temporary sender
4957 or receiver verification errors.
4959 2. Diagnose a configuration error if two authenticators have the same public
4962 3. Exim used not to create the message log file for a message until the first
4963 delivery attempt. This could be confusing when incoming messages were held
4964 for policy or load reasons. The message log file is now created at the time
4965 the message is received, and an initial "Received" line is written to it.
4967 4. The automatically generated man page for command line options had a minor
4968 bug that caused no ill effects; however, a more serious problem was that
4969 the procedure for building the man page automatically didn't always
4970 operate. Consequently, release 4.20 contains an out-of-date version. This
4971 shouldn't happen again.
4973 5. When building Exim with embedded Perl support, the script that builds the
4974 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4975 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4977 6. The freeze_tell option was not being used for messages that were frozen on
4978 arrival, either by an ACL or by local_scan().
4980 7. Added the smtp_incomplete_transaction log selector.
4982 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4983 was accepting AUTH without a new EHLO.
4985 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4986 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4989 10. Logging of TCP/IP connections (when configured) now happens in the main
4990 daemon process instead of the child process, so that the TCP/IP connection
4991 count is more accurate (but it can never be perfect).
4993 11. The use of "drop" in a nested ACL was not being handled correctly in the
4994 outer ACL. Now, if condition failure induced by the nested "drop" causes
4995 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4996 or "require"), the connection is dropped.
4998 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4999 that yield "discard" can now be used with an "accept" or a "discard" verb,
5000 but an error is generated for any others (because I can't see a useful way
5001 to define what should happen).
5003 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5004 are now processed in the same way as lines in the Exim configuration file.
5005 In particular, continuation lines are supported.
5007 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5009 15. Added -ti meaning -t -i.
5011 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5012 domains, and warn by logging if others are found.
5014 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5015 returned from GIFCONF when Exim is trying to find the list of interfaces on
5016 a host. The code in os.c has been modified to copy each ifreq to an aligned
5017 structure in all cases.
5019 Also, in some cases, the returned ifreq's were being copied to a 'struct
5020 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5021 means the last couple of bytes of an IPv6 address could be chopped if the
5022 ifreq contained only a normal sockaddr (14 bytes storage).
5024 18. Named domain lists were not supported in the hosts_treat_as_local option.
5025 An entry such as +xxxx was not recognized, and was treated as a literal
5028 19. Ensure that header lines added by a DATA ACL are included in the reject log
5029 if the ACL subsequently rejects the message.
5031 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5032 MD5 (which is deprecated).
5034 21. When testing a filter file using -bf, Exim was writing a message when it
5035 took the sender from a "From " line in the message, but it was not doing so
5036 when it took $return_path from a Return-Path: header line. It now does.
5038 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5039 with a valid header line field name (a series of printing characters
5040 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5042 23. Changed "disc" in the source to "disk" to conform to the documentation and
5043 the book and for uniformity.
5045 24. Ignore Sendmail's -Ooption=value command line item.
5047 25. When execve() failed while trying to run a command in a pipe transport,
5048 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5049 could be confused with a return value of 69 from the command itself. This
5050 has been changed to 127, the value the shell returns if it is asked to run
5051 a non-existent command. The wording for the related log line suggests a
5052 non-existent command as the problem.
5054 26. If received_header_text expands to an empty string, do not add a Received:
5055 header line to the message. (Well, it adds a token one on the spool, but
5056 marks it "old" so that it doesn't get used or transmitted.)
5058 27. Installed eximstats 1.28 (addition of -nt option).
5060 28. There was no check for failure on the call to getsockname() in the daemon
5061 code. This can fail if there is a shortage of resources on the system, with
5062 ENOMEM, for example. A temporary error is now given on failure.
5064 29. Contrary to the C standard, it seems that in some environments, the
5065 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5066 program. Exim now does this explicitly; it affects the formatting of
5067 timestamps using strftime().
5069 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5070 complaints. I've now initialized all the variables, to avoid this.
5072 32. Header lines added by a system filter were not being "seen" during
5073 transport-time rewrites.
5075 33. The info_callback() function passed to OpenSSL is set up with type void
5076 (*)(SSL *, int, int), as described somewhere. However, when calling the
5077 function (actually a macro) that sets it up, the type void(*)() is
5078 expected. I've put in a cast to prevent warnings from picky compilers.
5080 34. If a DNS black list lookup found a CNAME record, but there were no A
5081 records associated with the domain it pointed at, Exim crashed.
5083 35. If a DNS black list lookup returned more than one A record, Exim ignored
5084 all but the first. It now scans all returned addresses if a particular IP
5085 value is being sought. In this situation, the contents of the
5086 $dnslist_value variable are a list of all the addresses, separated by a
5089 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5090 to accept a host name and all its aliases if the forward lookup for any of
5091 them yielded the IP address of the incoming connection. Now it accepts only
5092 those names whose forward lookup yields the correct IP address. Any other
5093 names are discarded. This closes a loophole whereby a rogue DNS
5094 administrator could create reverse DNS records to break through a
5095 wildcarded host restriction in an ACL.
5097 37. If a user filter or a system filter that ran in a subprocess used any of
5098 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5099 the wrong values were passed to the pipe command ($thisaddress had the
5100 value of $0, $0 had the value of $1, etc). This bug was introduced by
5101 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5103 38. Improved the line breaking for long SMTP error messages from ACLs.
5104 Previously, if there was no break point between 40 and 75 characters, Exim
5105 left the rest of the message alone. Two changes have been made: (a) I've
5106 reduced the minimum length to 35 characters; (b) if it can't find a break
5107 point between 35 and 75 characters, it looks ahead and uses the first one
5108 that it finds. This may give the occasional overlong line, but at least the
5109 remaining text gets split now.
5111 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5112 file descriptors might be low, and that setting 1000 would always raise it.
5113 It turns out that in some environments, the limit is already over 1000 and
5114 that lowering it causes trouble. So now Exim takes care not to decrease it.
5116 40. When delivering a message, the value of $return_path is set to $sender_
5117 address at the start of routing (routers may change the value). By an
5118 oversight, this default was not being set up when an address was tested by
5119 -bt or -bv, which affected the outcome if any router or filter referred to
5122 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5123 log only when "message" or "log_message" are set. However, if one of the
5124 conditions was an address verification, or a call to a nested ACL, the
5125 messages generated by the underlying test were being passed through. This
5126 no longer happens. The underlying message is available in $acl_verify_
5127 message for both "message" and "log_message" expansions, so it can be
5128 passed through if needed.
5130 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5131 new expansion $bh_ to give the encoded byte string without charset
5132 translation. Translation happens only if iconv() is available; HAVE_ICONV
5133 indicates this at build time. HEADERS_CHARSET gives the charset to
5134 translate to; headers_charset can change it in the configuration, and
5135 "headers charset" can change it in an individual filter file.
5137 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5138 that creates RFC 2047 encoded "words" labels them as that charset instead
5139 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5140 expansion operator; (ii) when Exim creates a From: line for a local
5141 message; (iii) when a header line is rewritten to include a "phrase" part.
5143 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5144 buggy, causing it to skip the first lines of messages whose message ID
5145 ended in 'D'. This would not have bitten before Exim release 4.14, because
5146 message IDs were unlikely to end in 'D' before then. The effect was to have
5147 incorrect size information for certain domains.
5149 45. #include "config.h" was missing at the start of the crypt16.c module. This
5150 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5153 46. If there was a timeout during a "random" callout check, Exim treated it as
5154 a failure of the random address, and carried on sending RSET and the real
5155 address. If the delay was just some slowness somewhere, the response to the
5156 original RCPT would be taken as a response to RSET and so on, causing
5157 mayhem of various kinds.
5159 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5160 when I implemented it. It didn't allow for the fact that some option values
5161 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5162 the right test for positive values.
5164 48. Domain names in DNS records are case-independent. Exim always looks them up
5165 in lower case. Some resolvers return domain names in exactly the case they
5166 appear in the zone file, that is, they may contain uppercase letters. Not
5167 all resolvers do this - some return always lower case. Exim was treating a
5168 change of case by a resolver as a change of domain, similar to a widening
5169 of a domain abbreviation. This triggered its re-routing code and so it was
5170 trying to route what was effectively the same domain again. This normally
5171 caused routing to fail (because the router wouldn't handle the domain
5172 twice). Now Exim checks for this case specially, and just changes the
5173 casing of the domain that it ultimately uses when it transmits the message
5176 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5179 50. If a filter generated a file delivery with a non-absolute name (possible if
5180 no home directory exists for the router), the forbid_file option was not
5183 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5184 the existing equality matching.
5186 52. Exim was using ints instead of ino_t variables in some places where it was
5187 dealing with inode numbers.
5189 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5190 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5191 called TMPDIR, and if it finds it is different, it changes its value.
5193 54. The smtp_printf() function is now made available to local_scan() so
5194 additional output lines can be written before returning. There is also an
5195 smtp_fflush() function to enable the detection of a dropped connection.
5196 The variables smtp_input and smtp_batched_input are exported to
5199 55. Changed the default runtime configuration: the message "Unknown user"
5200 has been removed from the ACL, and instead placed on the localuser router,
5201 using the cannot_route_message feature. This means that any verification
5202 failures that generate their own messages won't get overridden. Similarly,
5203 the "Unrouteable address" message that was in the ACL for unverifiable
5204 relay addresses has also been removed.
5206 56. Added hosts_avoid_esmtp to the smtp transport.
5208 57. The exicyclog script was not checking for the esoteric option
5209 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5210 will work only if exicyclog is run under the appropriate euid.
5212 58. Following a discussion on the list, the rules by which Exim recognises line
5213 endings on incoming messages have been changed. The -dropcr and drop_cr
5214 options are now no-ops, retained only for backwards compatibility. The
5215 following line terminators are recognized: LF CRLF CR. However, special
5216 processing applies to CR:
5218 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5219 nor a local message in the state where . is a terminator.
5221 (ii) If a bare CR is encountered in a header line, an extra space is added
5222 after the line terminator so as not to end the header. The reasoning
5223 behind this is that bare CRs in header lines are most likely either
5224 to be mistakes, or people trying to play silly games.
5226 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5227 was being incorrectly given as 18 bytes larger than it should have been.
5228 This is a VOB (very old bug).
5230 60. This may never have affected anything current, but just in case it has:
5231 When the local host is found other than at the start of a list of hosts,
5232 the local host, those with the same MX, and any that follow, are discarded.
5233 When the list in question was part of a longer list of hosts, the following
5234 hosts (not currently being processed) were also being discarded. This no
5235 longer happens. I'm not sure if this situation could ever has previously
5238 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5241 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5242 header. This is something that is recommended in a new Internet Draft, and
5243 is something that is documented as being done by Sendmail. There are two
5244 possible values. For messages generated by the autoreply transport, Exim
5247 Auto-Submitted: auto-replied
5249 whereas for all other generated messages (e.g. bounces) it adds
5251 Auto-Submitted: auto-generated
5253 63. The "personal" condition in filters now includes a test for the
5254 Auto-Submitted: header. If it contains the string "auto-" the message it
5255 not considered personal.
5257 64. Added rcpt_include_affixes as a generic transport option.
5259 65. Added queue_only_override (default true).
5261 66. Added the syslog_duplication option.
5263 67. If what should have been the first header line of a message consisted of
5264 a space followed by a colon, Exim was mis-interpreting it as a header line.
5265 It isn't of course - it is syntactically invalid and should therefore be
5266 treated as the start of the message body. The misbehaviour could have
5267 caused a number of strange effects, including loss of data in subsequent
5268 header lines, and spool format errors.
5270 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5271 client host had authenticated. This control can now be exercised by an ACL
5272 for more flexibility.
5274 69. By default, callouts do not happen when testing with -bh. There is now a
5275 variant, -bhc, which does actually run the callout code, including
5276 consulting and updating the callout cache.
5278 70. Added support for saslauthd authentication, courtesy of Alexander
5281 71. If statvfs() failed on the spool or log directories while checking their
5282 size for availability, Exim confusingly gave the error "space shortage".
5283 Furthermore, in debugging mode it crashed with a floating point exception.
5284 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5285 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5286 serious problem, Exim now writes to the main and panic logs when this
5287 happens, with details of the failure. It then refuses to accept the
5288 incoming message, giving the message "spool directory problem" or "log
5289 directory problem" with a 421 code for SMTP messages.
5291 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5292 0, 1, and 2 exist, because some OS complain for execs without them (see
5293 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5294 descriptors. However, the code omitted to check that the open succeeded,
5295 causing mysterious errors if for some reason the permissions on /dev/null
5296 got screwed. Now Exim writes a message to the main and panic logs, and
5297 bombs out if it can't open /dev/null.
5299 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5300 interact so that it is all more flexible. It is supposed to remain
5301 backwards compatible. Also added extra_local_interfaces.
5303 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5304 to bomb out with an assertion failure - to the client this appears as a
5305 connection drop. This problem occurs in the part of the code that was taken
5306 from the Samba project. Fortunately, the assertion is in a very simple
5307 function, so I have fixed this by reproducing the function inline in the
5308 one place where it is called, and arranging for authentication to fail
5309 instead of killing the process with assert().
5311 75. The SPA client code was not working when the server requested OEM rather
5312 than Unicode encoding.
5314 76. Added code to make require_files with a specific uid setting more usable in
5315 the case where statting the file as root fails - usually a non-root-mounted
5316 NFS file system. When this happens and the failure is EACCES, Exim now
5317 forks a subprocess and does the per-uid checking as the relevant uid.
5319 77. Added process_log_path.
5321 78. If log_file_path was not explicitly set, a setting of check_log_space or
5322 check_log_inodes was ignored.
5324 79. If a space check for the spool or log partitions fails, the incident is now
5325 logged. Of course, in the latter case the data may get lost...
5327 80. Added the %p formatting code to string_format() so that it can be used to
5328 print addresses in debug_print(). Adjusted all the address printing in the
5329 debugging in store.c to use %p rather than %d.
5331 81. There was a concern that a line of code in smtp_in.c could overflow a
5332 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5333 initially expressed, the concern was not well-founded, because trailing
5334 spaces are removed early. However, if the trailing spaces were followed by
5335 a NULL, they did not get removed, so the overflow was possible. Two fixes
5338 (a) I re-wrote the offending code in a cleaner fashion.
5339 (b) If an incoming SMTP command contains a NULL character, it is rejected
5342 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5343 runs initgroups(), so that if the Exim user is in any additional groups,
5344 they will be used during message reception.
5350 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob