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 Solaris build fix for Oracle's LDAP libraries.
10 Bugzilla 1109, patch from Stephen Usher.
12 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
14 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
20 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
22 PP/02 Harmonised TLS library version reporting; only show if debugging.
23 Layout now matches that introduced for other libraries in 4.74 PP/03.
25 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
27 PP/04 New "dns_use_edns0" global option.
29 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
32 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
33 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
35 TK/01 Updated PolarSSL code to 0.14.2.
36 Bugzilla 1097. Patch from Andreas Metzler.
38 PP/07 Catch divide-by-zero in ${eval:...}.
41 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
44 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
45 format-string attack -- SECURITY: remote arbitrary code execution.
47 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
48 time unintentionally subject to list matching rules, letting the header
49 cause arbitrary Exim lookups (of items which can occur in lists, *not*
50 arbitrary string expansion). This allowed for information disclosure.
52 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
53 INT_MIN/-1 -- value coerced to INT_MAX.
59 NM/01 Workround for PCRE version dependency in version reporting
62 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
63 This fixes portability to compilers other than gcc, notably
64 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
66 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
67 makefiles for portability to HP-UX and POSIX correctness.
69 PP/01 Permit LOOKUP_foo enabling on the make command-line.
70 Also via indented variable definition in the Makefile.
71 (Debugging by Oliver Heesakkers).
73 PP/02 Restore caching of spamd results with expanded spamd_address.
74 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
76 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
77 Improves build reliability. Fix from: Frank Elsner
79 NM/02 Fix wide character breakage in the rfc2047 coding
80 Fixes bug 1064. Patch from Andrey N. Oktyabrski
82 NM/03 Allow underscore in dnslist lookups
83 Fixes bug 1026. Patch from Graeme Fowler
85 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
86 Code patches from Adam Ciarcinski of NetBSD.
88 NM/04 Fixed exiqgrep to cope with mailq missing size issue
91 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
92 is logged, to avoid truncation. Patch from John Horne.
94 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
95 Patch from Jakob Hirsch.
97 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
98 SQL string expansion failure details.
99 Patch from Andrey Oktyabrski.
101 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
102 Patch from Simon Arlott.
104 PP/09 New lookups functionality failed to compile on old gcc which rejects
105 extern declarations in function scope.
106 Patch from Oliver Fleischmann
108 PP/10 Use sig_atomic_t for flags set from signal handlers.
109 Check getgroups() return and improve debugging.
110 Fixed developed for diagnosis in bug 927 (which turned out to be
113 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
114 Patch from Mark Zealey.
116 PP/12 Bugzilla 1056: Improved spamd server selection.
117 Patch from Mark Zealey.
119 PP/13 Bugzilla 1086: Deal with maildir quota file races.
120 Based on patch from Heiko Schlittermann.
122 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
123 Patch from Uwe Doering, sign-off by Michael Haardt.
125 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
126 variable declaration deep within a block. Bug and patch from
129 PP/15 lookups-Makefile IRIX compatibilty coercion.
131 PP/16 Make DISABLE_DKIM build knob functional.
133 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
134 Patch by Simon Arlott
136 TF/03 Fix valgrind.h portability to C89 compilers that do not support
137 variable argument macros. Our copy now differs from upstream.
143 TF/01 Failure to get a lock on a hints database can have serious
144 consequences so log it to the panic log.
146 TF/02 Log LMTP confirmation messages in the same way as SMTP,
147 controlled using the smtp_confirmation log selector.
149 TF/03 Include the error message when we fail to unlink a spool file.
151 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
152 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
153 for maintaining out-of-tree patches for some time.
155 PP/01 Bugzilla 139: Documentation and portability issues.
156 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
157 Handle per-OS dynamic-module compilation flags.
159 PP/02 Let /dev/null have normal permissions.
160 The 4.73 fixes were a little too stringent and complained about the
161 permissions on /dev/null. Exempt it from some checks.
162 Reported by Andreas M. Kirchwitz.
164 PP/03 Report version information for many libraries, including
165 Exim version information for dynamically loaded libraries. Created
166 version.h, now support a version extension string for distributors
167 who patch heavily. Dynamic module ABI change.
169 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
170 privilege escalation vulnerability whereby the Exim run-time user
171 can cause root to append content of the attacker's choosing to
174 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
177 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
178 If dropping privileges for untrusted macros, we disabled normal logging
179 on the basis that it would fail; for the Exim run-time user, this is not
180 the case, and it resulted in successful deliveries going unlogged.
181 Fixed. Reported by Andreas Metzler.
187 PP/01 Date: & Message-Id: revert to normally being appended to a message,
188 only prepend for the Resent-* case. Fixes regression introduced in
189 Exim 4.70 by NM/22 for Bugzilla 607.
191 PP/02 Include check_rfc2047_length in configure.default because we're seeing
192 increasing numbers of administrators be bitten by this.
194 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
196 PP/03 Bugzilla 994: added openssl_options main configuration option.
198 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
200 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
202 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
204 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
205 without a peer certificate, leading to a segfault because of an
206 assumption that peers always have certificates. Be a little more
207 paranoid. Problem reported by Martin Tscholak.
209 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
210 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
211 NB: ClamAV planning to remove STREAM in "middle of 2010".
212 CL also introduces -bmalware, various -d+acl logging additions and
213 more caution in buffer sizes.
215 PP/09 Implemented reverse_ip expansion operator.
217 PP/10 Bugzilla 937: provide a "debug" ACL control.
219 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
221 PP/12 Bugzilla 973: Implement --version.
223 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
225 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
227 PP/15 Bugzilla 816: support multiple condition rules on Routers.
229 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
230 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
231 ignore trailing whitespace.
233 JJ/02 prevent non-panic DKIM error from being sent to paniclog
235 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
238 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
239 Notification from Dr Andrew Aitchison.
241 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
242 ExtendedDetectionInfo response format.
243 Notification from John Horne.
245 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
248 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
249 XSL and documented dependency on system catalogs, with examples of how
252 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
255 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
256 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
257 configuration file which is writeable by the Exim user or group.
259 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
260 of configuration files to cover files specified with the -C option if
261 they are going to be used with root privileges, not just the default
264 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
265 option (effectively making it always true).
267 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
268 files to be used while preserving root privileges.
270 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
271 that rogue child processes cannot use them.
273 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
274 run-time user, instead of root.
276 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
277 Exim run-time user without dropping privileges.
279 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
280 result string, instead of calling string_vformat() twice with the same
283 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
284 for other users. Others should always drop root privileges if they use
285 -C on the command line, even for a whitelisted configure file.
287 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
289 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
295 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
296 $data_path, and $header_path variables; fixed documentation bugs and
299 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
300 exipick to access non-standard spools, including the "frozen" queue
303 NM/01 Bugzilla 965: Support mysql stored procedures.
304 Patch from Alain Williams
306 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
308 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
309 Patch from Andreas Metzler
311 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
312 Patch from Kirill Miazine
314 NM/05 Bugzilla 671: Added umask to procmail example.
316 JJ/03 installed exipick 20100323.0, fixing doc bug
318 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
319 directory. Notification and patch from Dan Rosenberg.
321 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
323 TK/02 Improve log output when DKIM signing operation fails.
325 MH/01 Treat the transport option dkim_domain as a colon separated
326 list, not as a single string, and sign the message with each element,
327 omitting multiple occurences of the same signer.
329 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
330 Bugzilla 985, 986. Patch by Simon Arlott
332 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
333 Patch by Simon Arlott
335 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
336 MBX locking. Notification from Dan Rosenberg.
342 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
344 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
346 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
348 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
350 NM/04 Bugzilla 915: Flag broken perl installation during build.
356 TK/01 Added patch by Johannes Berg that expands the main option
357 "spamd_address" if it starts with a dollar sign.
359 TK/02 Write list of recipients to X-Envelope-Sender header when building
360 the mbox-format spool file for content scanning (suggested by Jakob
363 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
364 (http://www.dcc-servers.net/) support via dccifd. Activated by
365 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
367 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
368 by Mark Daniel Reidel <mr@df.eu>.
370 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
371 When building exim an external PCRE library is now needed -
372 PCRE is a system library on the majority of modern systems.
373 See entry on PCRE_LIBS in EDITME file.
375 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
376 conversation. Added nologin parameter to request.
377 Patch contributed by Kirill Miazine.
379 TF/01 Do not log submission mode rewrites if they do not change the address.
381 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
383 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
384 log files in place. Contributed by Roberto Lima.
386 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
388 TF/03 Bugzilla 615: When checking the local_parts router precondition
389 after a local_part_suffix or local_part_prefix option, Exim now
390 does not use the address's named list lookup cache, since this
391 contains cached lookups for the whole local part.
393 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
394 Robert Millan. Documentation is in experimental-spec.txt.
396 TF/04 Bugzilla 668: Fix parallel build (make -j).
398 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
400 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
401 Patch provided by Jan Srzednicki.
403 TF/05 Leading white space used to be stripped from $spam_report which
404 wrecked the formatting. Now it is preserved.
406 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
407 that they are available at delivery time.
409 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
411 TF/08 TLS error reporting now respects the incoming_interface and
412 incoming_port log selectors.
414 TF/09 Produce a more useful error message if an SMTP transport's hosts
415 setting expands to an empty string.
417 NM/06 Bugzilla 744: EXPN did not work under TLS.
418 Patch provided by Phil Pennock.
420 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
421 Patch provided by Richard Godbee.
423 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
424 acl_smtp_notquit, added index entry.
426 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
427 Patch provided by Eugene Bujak.
429 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
430 accept(). Patch provided by Maxim Dounin.
432 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
433 Patch provided by Phil Pennock.
435 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
437 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
438 Patch provided by Brad "anomie" Jorsch.
440 NM/14 Bugzilla 622: Added timeout setting to transport filter.
441 Patch provided by Dean Brooks.
443 TK/05 Add native DKIM support (does not depend on external libraries).
445 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
446 Patch provided by Graeme Fowler.
448 NM/16 Bugzilla 851: Documentation example syntax fix.
450 NM/17 Changed NOTICE file to remove references to embedded PCRE.
452 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
455 NM/19 Bugzilla 745: TLS version reporting.
456 Patch provided by Phil Pennock.
458 NM/20 Bugzilla 167: bool: condition support.
459 Patch provided by Phil Pennock.
461 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
462 clients. Patch provided by Phil Pennock.
464 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
465 Patch provided by Brad "anomie" Jorsch.
467 NM/23 Bugzilla 687: Fix misparses in eximstats.
468 Patch provided by Heiko Schlittermann.
470 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
471 Patch provided by Heiko Schlittermann.
473 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
474 plus update to original patch.
476 NM/26 Bugzilla 799: Documentation correction for ratelimit.
478 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
479 Patch provided by David Brownlee.
481 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
483 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
485 NM/30 Bugzilla 888: TLS documentation bugfixes.
487 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
489 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
490 Unlike the original bugzilla I have changed all shell scripts in src tree.
492 NM/33 Bugzilla 898: Transport filter timeout fix.
493 Patch by Todd Rinaldo.
495 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
496 Patch by Serge Demonchaux.
498 NM/35 Bugzilla 39: Base64 decode bug fixes.
499 Patch by Jakob Hirsch.
501 NM/36 Bugzilla 909: Correct connect() call in dcc code.
503 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
505 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
507 NM/39 Bugzilla 911: Fixed MakeLinks build script.
513 TK/01 Add preliminary DKIM support. Currently requires a forked version of
514 ALT-N's libdkim that I have put here:
515 http://duncanthrax.net/exim-experimental/
517 Note to Michael Haardt: I had to rename some vars in sieve.c. They
518 were called 'true' and it seems that C99 defines that as a reserved
519 keyword to be used with 'bool' variable types. That means you could
520 not include C99-style headers which use bools without triggering
521 build errors in sieve.c.
523 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
524 as mailq or other aliases. Changed the --help handling significantly
525 to do whats expected. exim_usage() emits usage/help information.
527 SC/01 Added the -bylocaldomain option to eximstats.
529 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
531 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
533 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
539 PH/01 Another patch from the Sieve maintainer.
541 PH/02 When an IPv6 address is converted to a string for single-key lookup
542 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
543 dots are used instead of colons so that keys in lsearch files need not
544 contain colons. This was done some time before quoting was made available
545 in lsearch files. However, iplsearch files do require colons in IPv6 keys
546 (notated using the quote facility) so as to distinguish them from IPv4
547 keys. This meant that lookups for IP addresses in host lists did not work
548 for iplsearch lookups.
550 This has been fixed by arranging for IPv6 addresses to be expressed with
551 colons if the lookup type is iplsearch. This is not incompatible, because
552 previously such lookups could never work.
554 The situation is now rather anomolous, since one *can* have colons in
555 ordinary lsearch keys. However, making the change in all cases is
556 incompatible and would probably break a number of configurations.
558 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
561 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
562 conversion specification without a maximum field width, thereby enabling
563 a rogue spamd server to cause a buffer overflow. While nobody in their
564 right mind would setup Exim to query an untrusted spamd server, an
565 attacker that gains access to a server running spamd could potentially
566 exploit this vulnerability to run arbitrary code as the Exim user.
568 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
569 $primary_hostname instead of what libspf2 thinks the hosts name is.
571 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
572 a directory entry by the name of the lookup key. Previously, if a
573 symlink pointed to a non-existing file or a file in a directory that
574 Exim lacked permissions to read, a lookup for a key matching that
575 symlink would fail. Now it is enough that a matching directory entry
576 exists, symlink or not. (Bugzilla 503.)
578 PH/03 The body_linecount and body_zerocount variables are now exported in the
581 PH/04 Added the $dnslist_matched variable.
583 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
584 This means they are set thereafter only if the connection becomes
587 PH/06 Added the client_condition to authenticators so that some can be skipped
588 by clients under certain conditions.
590 PH/07 The error message for a badly-placed control=no_multiline_responses left
591 "_responses" off the end of the name.
593 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
595 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
596 (without spaces) instead of just copying the configuration text.
598 PH/10 Added the /noupdate option to the ratelimit ACL condition.
600 PH/11 Added $max_received_linelength.
602 PH/12 Added +ignore_defer and +include_defer to host lists.
604 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
605 way in which PCRE > 7.0 is built.
607 PH/14 Implemented queue_only_load_latch.
609 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
610 MAIL command. The effect was to mangle the value on 64-bit systems.
612 PH/16 Another patch from the Sieve maintainer.
614 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
616 PH/18 If a system quota error occurred while trying to create the file for
617 a maildir delivery, the message "Mailbox is full" was not appended to the
618 bounce if the delivery eventually timed out. Change 4.67/27 below applied
619 only to a quota excession during the actual writing of the file.
621 PH/19 It seems that peer DN values may contain newlines (and other non-printing
622 characters?) which causes problems in log lines. The DN values are now
623 passed through string_printing() before being added to log lines.
625 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
626 and InterBase are left for another time.)
628 PH/21 Added message_body_newlines option.
630 PH/22 Guard against possible overflow in moan_check_errorcopy().
632 PH/23 POSIX allows open() to be a macro; guard against that.
634 PH/24 If the recipient of an error message contained an @ in the local part
635 (suitably quoted, of course), incorrect values were put in $domain and
636 $local_part during the evaluation of errors_copy.
642 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
643 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
646 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
647 issue a MAIL command.
649 PH/02 In an ACL statement such as
651 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
653 if a client was not listed at all, or was listed with a value other than
654 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
655 the condition was not true (as it should be), so access was not denied.
656 The bug was that the ! inversion was incorrectly passed on to the second
657 item. This has been fixed.
659 PH/03 Added additional dnslists conditions == and =& which are different from
660 = and & when the dns lookup returns more than one IP address.
662 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
663 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
665 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
666 FSYNC, which compiles an option called disable_fsync that allows for
667 bypassing fsync(). The documentation is heavily laced with warnings.
669 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
671 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
672 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
673 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
674 including adding "make clean"; (3) Added -fPIC when compiling the test
675 dynamically loaded module, to get rid of a warning.
677 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
678 message fails, move_frozen_messages = true and ignore_bounce_errors_after
679 = 0s. The bug is otherwise harmless.
681 PH/07 There was a bug in the dovecot authenticator such that the value of
682 $auth1 could be overwritten, and so not correctly preserved, after a
683 successful authentication. This usually meant that the value preserved by
684 the server_setid option was incorrect.
686 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
688 PH/09 Installed PCRE release 7.0.
690 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
691 run for batched SMTP input. It is now run at the start of every message
692 in the batch. While fixing this I discovered that the process information
693 (output by running exiwhat) was not always getting set for -bs and -bS
694 input. This is fixed, and it now also says "batched" for BSMTP.
696 PH/11 Added control=no_pipelining.
698 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
699 patch, slightly modified), and move the expansion of helo_data till after
700 the connection is made in the smtp transport (so it can use these
703 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
705 PH/14 Added log_selector = +pid.
707 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
709 PH/16 Add ${if forany and ${if forall.
711 PH/17 Added dsn_from option to vary the From: line in DSNs.
713 PH/18 Flush SMTP output before performing a callout, unless control =
714 no_callout_flush is set.
716 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
717 was true (the default) a successful delivery failed to delete the retry
718 item, thus causing premature timeout of the address. The bug is now
721 PH/20 Added hosts_avoid_pipelining to the smtp transport.
723 PH/21 Long custom messages for fakedefer and fakereject are now split up
724 into multiline reponses in the same way that messages for "deny" and
725 other ACL rejections are.
727 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
728 with slight modification.
730 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
731 draft, changing the syntax and factoring some duplicate code".
733 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
734 for deliveries of the second and subsequent messages over the same SMTP
737 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
738 ${reduce, with only minor "tidies".
740 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
742 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
743 expansion side effects.
745 PH/27 When a message times out after an over-quota error from an Exim-imposed
746 quota, the bounce message says "mailbox is full". This message was not
747 being given when it was a system quota that was exceeded. It now should
750 MH/03 Made $recipients available in local_scan(). local_scan() already has
751 better access to the recipient list through recipients_list[], but
752 $recipients can be useful in postmaster-provided expansion strings.
754 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
755 in the case of a MAIL command with additional options following the
756 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
757 were accidentally chopped off.
759 PH/29 SMTP synchronization checks are implemented when a command is read -
760 there is a check that no more input is waiting when there shouldn't be
761 any. However, for some commands, a delay in an ACL can mean that it is
762 some time before the response is written. In this time, more input might
763 arrive, invalidly. So now there are extra checks after an ACL has run for
764 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
765 pipelining has not been advertised.
767 PH/30 MH's patch to allow iscntrl() characters to be list separators.
769 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
770 returned in the SMTP response when smtp_return_error_details was false.
773 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
774 instead of the C I/O that was originally supplied, because problems were
777 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
778 Exim which did not show up earlier: it was assuming that a call to
779 SSL_CTX_set_info_callback() might give an error value. In fact, there is
780 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
781 was a macro that became an assignment, so it seemed to work. This has
782 changed to a proper function call with a void return, hence the compile
783 error. Exim's code has been fixed.
785 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
788 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
790 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
792 PH/37 If a message is not accepted after it has had an id assigned (e.g.
793 because it turns out to be too big or there is a timeout) there is no
794 "Completed" line in the log. When some messages of this type were
795 selected by exigrep, they were listed as "not completed". Others were
796 picked up by some special patterns. I have improved the selection
797 criteria to be more general.
799 PH/38 The host_find_failed option in the manualroute router can now be set
800 to "ignore", to completely ignore a host whose IP address cannot be
801 found. If all hosts are ignored, the behaviour is controlled by the new
802 host_all_ignored option.
804 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
805 homing or because of multiple MX records with /mx) generated more than
806 one IP address, and the following item turned out to be the local host,
807 all the secondary addresses of the first item were incorrectly removed
808 from the list, along with the local host and any following hosts (which
809 is what is supposed to happen).
811 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
812 whoever called Exim into the -H file. In the case of the daemon it was
813 behaving confusingly. When first started, it used values for whoever
814 started the daemon, but after a SIGHUP it used the Exim user (because it
815 calls itself on a restart). I have changed the code so that it now always
818 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
819 message are rejected with the same error (e.g. no authentication or bad
820 sender address), and a DATA command is nevertheless sent (as can happen
821 with PIPELINING or a stupid MUA), the error message that was given to the
822 RCPT commands is included in the rejection of the DATA command. This is
823 intended to be helpful for MUAs that show only the final error to their
826 PH/42 Another patch from the Sieve maintainer.
828 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
829 Eximstats - Fixed some broken HTML links and added missing column headers
831 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
832 columns for Rejects, Temp Rejects, Ham, and Spam rows.
834 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
836 PH/43 Yet another patch from the Sieve maintainer.
838 PH/44 I found a way to check for a TCP/IP connection going away before sending
839 the response to the final '.' that terminates a message, but only in the
840 case where the client has not sent further data following the '.'
841 (unfortunately, this is allowed). However, in many cases there won't be
842 any further data because there won't be any more messages to send. A call
843 to select() can be used: if it shows that the input is "ready", there is
844 either input waiting, or the socket has been closed. An attempt to read
845 the next input character can distinguish the two cases. Previously, Exim
846 would have sent an OK response which the client would never have see.
847 This could lead to message repetition. This fix should cure that, at
848 least in a lot of common cases.
850 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
851 advertised in response to EHLO.
857 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
858 fixed by 4.65/MH/01 (is this a record?) are fixed:
860 (i) An empty string was always treated as zero by the numeric comparison
861 operators. This behaviour has been restored.
863 (ii) It is documented that the numeric comparison operators always treat
864 their arguments as decimal numbers. This was broken in that numbers
865 starting with 0 were being interpreted as octal.
867 While fixing these problems I realized that there was another issue that
868 hadn't been noticed. Values of message_size_limit (both the global option
869 and the transport option) were treated as octal if they started with 0.
870 The documentation was vague. These values are now always treated as
871 decimal, and I will make that clear in the documentation.
877 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
878 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
881 MH/01 Don't check that the operands of numeric comparison operators are
882 integers when their expansion is in "skipping" mode (fixes bug
883 introduced by 4.64-PH/07).
885 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
886 child addresses, Exim now panics and dies. Previously, because the count
887 is held in a short int, deliveries were likely to be lost. As such a
888 large number of recipients for a single message is ridiculous
889 (performance will be very, very poor), I have chosen to impose a limit
890 rather than extend the field.
896 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
897 leftover -K file (the existence of which was triggered by #402).
898 While we were at it, introduced process PID as part of the -K
899 filename. This should rule out race conditions when creating
902 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
903 processing considerably. Previous code took too long for large mails,
904 triggering a timeout which in turn triggers #401.
906 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
907 in the DK code in transports.c. sendfile() is not really portable,
908 hence the _LINUX specificness.
910 TF/01 In the add_headers option to the mail command in an Exim filter,
911 there was a bug that Exim would claim a syntax error in any
912 header after the first one which had an odd number of characters
915 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
916 callout verification, Exim cached a "reject" for the entire domain. This
917 is correct for most verifications, but it is not correct for a recipient
918 verification with use_sender or use_postmaster set, because in that case
919 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
920 case of MAIL FROM:<> rejection from other early rejections (e.g.
921 rejection of HELO). When verifying a recipient using a non-null MAIL
922 address, the cache is ignored if it shows MAIL FROM:<> rejection.
923 Whatever the result of the callout, the value of the domain cache is
924 left unchanged (for any other kind of callout, getting as far as trying
925 RCPT means that the domain itself is ok).
927 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
930 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
931 manifest itself as EPIPE rather than ECONNECT. When tidying away a
932 session, the daemon ignores ECONNECT errors and logs others; it now
933 ignores EPIPE as well.
935 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
936 (quoted-printable decoding).
938 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
939 later the small subsequent patch to fix an introduced bug.
941 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
943 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
945 PH/08 An error is now given if message_size_limit is specified negative.
947 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
948 to be given (somewhat) arbitrary names.
950 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
953 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
954 miscellaneous code fixes
956 PH/10 Added the log_reject_target ACL modifier to specify where to log
959 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
960 hostname. This is wrong, because it relates to the incoming message (and
961 probably the interface on which it is arriving) and not to the outgoing
962 callout (which could be using a different interface). This has been
963 changed to use the value of the helo_data option from the smtp transport
964 instead - this is what is used when a message is actually being sent. If
965 there is no remote transport (possible with a router that sets up host
966 addresses), $smtp_active_hostname is used.
968 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
969 tweaks were necessary in order to get it to work (see also 21 below):
970 (a) The code assumed that strncpy() returns a negative number on buffer
971 overflow, which isn't the case. Replaced with Exim's string_format()
973 (b) There were several signed/unsigned issues. I just did the minimum
974 hacking in of casts. There is scope for a larger refactoring.
975 (c) The code used strcasecmp() which is not a standard C function.
976 Replaced with Exim's strcmpic() function.
977 (d) The code set only $1; it now sets $auth1 as well.
978 (e) A simple test gave the error "authentication client didn't specify
979 service in request". It would seem that Dovecot has changed its
980 interface. Fortunately there's a specification; I followed it and
981 changed what the client sends and it appears to be working now.
983 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
986 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
987 address A is aliased to B and C, where B exists and C does not. Without
988 -v the output is "A verified" because verification stops after a
989 successful redirection if more than one address is generated. However,
990 with -v the child addresses are also verified. Exim was outputting "A
991 failed to verify" and then showing the successful verification for C,
992 with its parentage. It now outputs "B failed to verify", showing B's
993 parentage before showing the successful verification of C.
995 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
996 look up a TXT record in a specific list after matching in a combined
999 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1000 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1001 they consult the DNS. I had assumed they would set it the way they
1002 wanted; and indeed my experiments on Linux seem to show that in some
1003 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1004 To be on the safe side, however, I have now made the interface to
1005 host_find_byname() similar to host_find_bydns(), with an argument
1006 containing the DNS resolver options. The host_find_byname() function now
1007 sets these options at its start, just as host_find_bydns() does. The smtp
1008 transport options dns_qualify_single and dns_search_parents are passed to
1009 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1010 of host_find_byname() use the default settings of RES_DEFNAMES
1011 (qualify_single) but not RES_DNSRCH (search_parents).
1013 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1014 spool_read_header() do less string testing, by means of a preliminary
1015 switch on the second character of optional "-foo" lines. (This is
1016 overdue, caused by the large number of possibilities that now exist.
1017 Originally there were few.) While I was there, I also converted the
1018 str(n)cmp tests so they don't re-test the leading "-" and the first
1019 character, in the hope this might squeeze out yet more improvement.
1021 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1022 flag allowing group syntax was set by the header_syntax check but not
1023 turned off, possible causing trouble later; (2) The flag was not being
1024 set at all for the header_verify test, causing "group"-style headers to
1025 be rejected. I have now set it in this case, and also caused header_
1026 verify to ignore an empty address taken from a group. While doing this, I
1027 came across some other cases where the code for allowing group syntax
1028 while scanning a header line wasn't quite right (mostly, not resetting
1029 the flag correctly in the right place). These bugs could have caused
1030 trouble for malformed header lines. I hope it is now all correct.
1032 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1033 with the "reply" argument non-NULL. The code, however (which originally
1034 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1035 but it didn't always do it. This confused somebody who was copying the
1036 code for some other use. I have removed all the tests.
1038 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1039 feature that was used to support insecure browsers during the U.S. crypto
1040 embargo. It requires special client support, and Exim is probably the
1041 only MTA that supported it -- and would never use it because real RSA is
1042 always available. This code has been removed, because it had the bad
1043 effect of slowing Exim down by computing (never used) parameters for the
1044 RSA_EXPORT functionality.
1046 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1047 authenticator to fail if there's a tab character in the incoming data
1048 (there should never be unless someone is messing about, as it's supposed
1049 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1050 if the connection is using TLS or if the remote IP is the same as the
1051 local IP, and the "valid-client-cert option" if a client certificate has
1054 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1055 authenticators. This can be used for authorization after authentication
1056 succeeds. (In the case of plaintext, it servers for both authentication
1059 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1060 if any retry times were supplied.
1062 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1063 connection, where there is no client IP address to check. In this
1064 situation, the verify now always succeeds.
1066 PH/25 Applied John Jetmore's -Mset patch.
1068 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1070 PH/27 In a string expansion for a processed (not raw) header when multiple
1071 headers of the same name were present, leading whitespace was being
1072 removed from all of them, but trailing whitespace was being removed only
1073 from the last one. Now trailing whitespace is removed from each header
1074 before concatenation. Completely empty headers in a concatenation (as
1075 before) are ignored.
1077 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1078 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1080 PH/29 [Removed. This was a change that I later backed out, and forgot to
1081 correct the ChangeLog entry (that I had efficiently created) before
1082 committing the later change.]
1084 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1085 address errors (4xx response to RCPT) over the same connection as other
1086 messages routed to the same hosts. Such deliveries are always "forced",
1087 so retry times are not inspected. This resulted in far too many retries
1088 for the affected addresses. The effect occurred only when there were more
1089 hosts than the hosts_max_try setting in the smtp transport when it had
1090 the 4xx errors. Those hosts that it had tried were not added to the list
1091 of hosts for which the message was waiting, so if all were tried, there
1092 was no problem. Two fixes have been applied:
1094 (i) If there are any address or message errors in an SMTP delivery, none
1095 of the hosts (tried or untried) are now added to the list of hosts
1096 for which the message is waiting, so the message should not be a
1097 candidate for sending over the same connection that was used for a
1098 successful delivery of some other message. This seems entirely
1099 reasonable: after all the message is NOT "waiting for some host".
1100 This is so "obvious" that I'm not sure why it wasn't done
1101 previously. Hope I haven't missed anything, but it can't do any
1102 harm, as the worst effect is to miss an optimization.
1104 (ii) If, despite (i), such a delivery is accidentally attempted, the
1105 routing retry time is respected, so at least it doesn't keep
1106 hammering the server.
1108 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1109 in ${readsocket because some servers need this prod.
1111 PH/32 Added some extra debug output when updating a wait-xxx database.
1113 PH/33 The hint "could be header name not terminated by colon", which has been
1114 given for certain expansion errors for a long time, was not being given
1115 for the ${if def:h_colon_omitted{... case.
1117 PH/34 The spec says: "With one important exception, whenever a domain list is
1118 being scanned, $domain contains the subject domain." There was at least
1119 one case where this was not true.
1121 PH/35 The error "getsockname() failed: connection reset by peer" was being
1122 written to the panic log as well as the main log, but it isn't really
1123 panic-worthy as it just means the connection died rather early on. I have
1124 removed the panic log writing for the ECONNRESET error when getsockname()
1127 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1128 runs only) independently of the message's sender address. This meant
1129 that, if the 4xx error was in fact related to the sender, a different
1130 message to the same recipient with a different sender could confuse
1131 things. In particualar, this can happen when sending to a greylisting
1132 server, but other circumstances could also provoke similar problems.
1133 I have changed the default so that the retry time for these errors is now
1134 based a combination of the sender and recipient addresses. This change
1135 can be overridden by setting address_retry_include_sender=false in the
1138 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1139 remote server are returned as part of bounce messages. This was not
1140 happening for LMTP over a pipe (the lmtp transport), but now it is the
1141 same for both kinds of LMTP.
1143 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1144 in header lines that were in fact CNAMEs. This is no longer the case.
1146 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1147 and queue runs started by the daemon processed all messages. This has
1148 been fixed so that -R and -S can now usefully be given with -q<time>.
1150 PH/40 Import PCRE release 6.7 (fixes some bugs).
1152 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1154 PH/42 Give an error if -q is specified more than once.
1156 PH/43 Renamed the variables $interface_address and $interface_port as
1157 $received_ip_address and $received_port, to make it clear that these
1158 values apply to message reception, and not to the outgoing interface when
1159 a message is delivered. (The old names remain recognized, of course.)
1161 PH/44 There was no timeout on the connect() call when using a Unix domain
1162 socket in the ${readsocket expansion. There now is.
1164 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1165 be meaningful with "accept".
1167 SC/01 Eximstats V1.43
1168 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1170 SC/02 Eximstats V1.44
1171 Use a glob alias rather than an array ref in the generated
1172 parser. This improves both readability and performance.
1174 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1175 Collect SpamAssassin and rejection statistics.
1176 Don't display local sender or destination tables unless
1177 there is data to show.
1178 Added average volumes into the top table text output.
1180 SC/04 Eximstats V1.46
1181 Collect data on the number of addresses (recipients)
1182 as well as the number of messages.
1184 SC/05 Eximstats V1.47
1185 Added 'Message too big' to the list of mail rejection
1186 reasons (thanks to Marco Gaiarin).
1188 SC/06 Eximstats V1.48
1189 Mainlog lines which have GMT offsets and are too short to
1190 have a flag are now skipped.
1192 SC/07 Eximstats V1.49 (Alain Williams)
1193 Added the -emptyok flag.
1195 SC/08 Eximstats V1.50
1196 Fixes for obtaining the IP address from reject messages.
1198 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1199 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1200 whitesspace changes from 4.64-PH/27
1202 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1205 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1206 are found, allow negative numbers in numeric criteria)
1208 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1210 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1211 to match changes made in 4.64-PH/43
1213 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1215 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1216 "domain" is actually an IP address, and force a failure. This locks out
1217 those revolvers/nameservers that support "A-for-A" lookups, in
1218 contravention of the specifications.
1220 PH/48 When a host name was looked up from an IP address, and the subsequent
1221 forward lookup of the name timed out, the host name was left in
1222 $sender_host_name, contrary to the specification.
1224 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1225 restricted to single-key lookups, Exim was not diagnosing an error if
1226 * or *@ was used with a query-style lookup.
1228 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1230 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1231 long ago, but noone interested enough thought of it. Let's just say that
1232 the "1.1" means that there are some new functions that weren't there at
1233 some point in the past.
1235 PH/51 Error processing for expansion failure of helo_data from an smtp
1236 transport during callout processing was broken.
1238 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1239 tested/used via the -bh/-bhc/-bs options.
1241 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1242 bug, fixed in subsequent PCRE releases).
1244 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1245 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1247 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1253 SC/01 Use a glob alias rather than an array ref in eximstats generated
1254 parser. This improves both readability and performance.
1256 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1257 Don't display local sender or destination tables in eximstats unless
1258 there is data to show.
1259 Added average volumes into the eximstats top table text output.
1261 SC/03 Collect data on the number of addresses (recipients) as well
1262 as the number of messages in eximstats.
1264 TF/01 Correct an error in the documentation for the redirect router. Exim
1265 does (usually) call initgroups() when daemonizing.
1267 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1268 with consistent privilege compared to when running as a daemon.
1270 TF/03 Note in the spec that $authenticated_id is not set for local
1271 submissions from trusted users.
1273 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1274 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1276 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1277 by adding some example configuration directives to the default
1278 configuration file. A little bit of work is required to uncomment the
1279 directives and define how usernames and passwords are checked, but
1280 there is now a framework to start from.
1282 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1283 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1284 without this. I don't know how relevant this is to other LDAP libraries.
1286 PH/02 Add the verb name to the "unknown ACL verb" error.
1288 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1290 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1292 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1293 directory not expanded when it should be if an expanded home directory
1294 was set for the address (which is overridden by the transport).
1296 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1299 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1300 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1301 because it is too late at that time, and has no effect.
1303 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1304 security issue with \' (bugzilla #107). I could not use the
1305 PQescapeStringConn() function, because it needs a PGconn value as one of
1308 PH/08 When testing addresses using -bt, indicate those final addresses that
1309 are duplicates that would not cause an additional delivery. At least one
1310 person was confused, thinking that -bt output corresponded to deliveries.
1311 (Suppressing duplicates isn't a good idea as you lose the information
1312 about possibly different redirections that led to the duplicates.)
1314 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1315 systems where poll() doesn't work, in particular OS X.
1317 PH/10 Added more information to debugging output for retry time not reached.
1319 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1320 operations in malware.c.
1322 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1325 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1326 syslog_duplication set false, log lines that would normally be written
1327 both the the main log and to the reject log were not written to syslog at
1330 PH/14 In the default configuration, change the use of "message" in ACL warn
1331 statements to "add_header".
1333 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1334 not followed by a command (e.g. "seen endif").
1336 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1337 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1340 PH/17 Added extra conditions to the default value of delay_warning_condition
1344 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1345 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1346 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1349 The Auto-Submitted: and various List- headers are standardised, whereas I
1350 don't think Precedence: ever was.
1352 PH/18 Refactored debugging code in route_finduser() to show more information,
1353 in particular, the error code if getpwnam() issues one.
1355 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1356 This is apparently needed in addition to the PH/07 change above to avoid
1357 any possible encoding problems.
1359 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1360 but not after initializing Perl.
1362 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1363 output them only if debugging. By default they are written stderr,
1364 apparently, which is not desirable.
1366 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1369 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1372 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1374 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1375 authenticated or an ident call has been made. Suppress the default
1376 values for $authenticated_id and $authenticated_sender (but permit -oMai
1377 and -oMas) when testing with -bh.
1379 PH/24 Re-jigged the order of the tests in the default configuration so that the
1380 tests for valid domains and recipients precede the DNS black list and CSA
1381 tests, on the grounds that those ones are more expensive.
1383 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1384 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1385 This bug exists in every version of Exim that I still have, right back to
1388 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1389 However, an attempt to turn on case-sensitivity in a regex key by
1390 including (?-i) didn't work because the subject string was already
1391 lowercased, and the effects were non-intuitive. It turns out that a
1392 one-line patch can be used to allow (?-i) to work as expected.
1398 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1399 other effects) broke the use of negated acl sub-conditions.
1401 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1404 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1405 "Deny" causes Exim to reject the incoming connection with a 554 error.
1406 Unfortunately, if there is a major crisis, such as a disk failure,
1407 tcp-wrappers gives "deny", whereas what one would like would be some
1408 kind of temporary error. A kludge has been added to help with this.
1409 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1410 554 error is used if errno is still zero or contains ENOENT (which occurs
1411 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1414 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1416 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1417 errors. Otherwise a message that provokes a temporary error (when other
1418 messages do not) can cause a whole host to time out.
1420 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1421 addresses were routed directly to files or pipes from a redirect router.
1422 File deliveries just didn't batch; pipe deliveries might have suffered
1425 PH/06 A failure to get a lock for a hints database would erroneously always say
1426 "Failed to get write lock", even when it was really a read lock.
1428 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1429 of 0600. This has been changed to use the value of the lockfile_mode
1430 option (which defaults to 0600).
1432 PH/08 Applied small patch from the Sieve maintainer.
1434 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1435 folder from quota calculations, a direct delivery into this folder messed
1436 up the contents of the maildirsize file. This was because the regex was
1437 used only to exclude .Trash (or whatever) when the size of the mailbox
1438 was calculated. There was no check that a delivery was happening into an
1439 excluded directory. This bug has been fixed by ignoring all quota
1440 processing for deliveries into excluded directories.
1442 PH/10 Added the maildirfolder_create_regex option to appendfile.
1448 PH/01 The code for finding all the local interface addresses on a FreeBSD
1449 system running IPv6 was broken. This may well have applied to all BSD
1450 systems, as well as to others that have similar system calls. The broken
1451 code found IPv4 interfaces correctly, but gave incorrect values for the
1452 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1453 that it would not match correctly against @[] and not recognize the IPv6
1456 PH/02 The ipliteral router was not recognizing addresses of the form user@
1457 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1459 PH/03 Added disable_ipv6.
1461 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1462 decoded form, because it is most often used to construct To: headers
1463 lines in autoreplies, and the decoded form may well be syntactically
1464 invalid. However, $reply_address has leading white space removed, and all
1465 newlines turned into spaces so that the autoreply transport does not
1468 PH/05 If group was specified without a user on a router, and no group or user
1469 was specified on a transport, the group from the router was ignored.
1471 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1472 for visible compile-time settings that can be used to change these
1473 numbers, for those that want even more. Backwards compatibility with old
1474 spool files has been maintained. However, going back to a previous Exim
1475 release will lost any variables that are in spool files.
1477 PH/07 Two small changes when running in the test harness: increase delay when
1478 passing a TCP/IP connection to a new process, in case the original
1479 process has to generate a bounce, and remove special handling of
1480 127.0.0.2 (sic), which is no longer necessary.
1482 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1483 be the same on different OS.
1485 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1488 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1489 whether --show-vars was specified or not
1491 JJ/02 exipick: Added support for new ACL variable spool format introduced
1494 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1495 syntactically invalid From: or Reply-to: line, and a filter used this to
1496 generate an autoreply, and therefore failed to obtain an address for the
1497 autoreply, Exim could try to deliver to a non-existent relative file
1498 name, causing unrelated and misleading errors. What now happens is that
1499 it logs this as a hard delivery error, but does not attempt to create a
1502 PH/11 The exinext utility has a -C option for testing purposes, but although
1503 the given file was scanned by exinext itself; it wasn't being passed on
1504 when Exim was called.
1506 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1507 an end-of-file indication when reading a command response.
1509 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1510 compiled. In many other places in Exim, IPv6 addresses are always
1511 recognized, so I have changed this. It also means that IPv4 domain
1512 literals of the form [IPV4:n.n.n.n] are now always recognized.
1514 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1515 used if the router is not running as root, for example, when verifying at
1516 ACL time, or when using -bh. The debugging output from this situation was
1517 non-existent - all you got was a failure to exec. I have made two
1520 (a) Failures to set uid/gid, the current directory, or a process leader
1521 in a subprocess such as that created by queryprogram now generate
1522 suitable debugging ouput when -d is set.
1524 (b) The queryprogram router detects when it is not running as root,
1525 outputs suitable debugging information if -d is set, and then runs
1526 the subprocess without attempting to change uid/gid.
1528 PH/15 Minor change to Makefile for building test_host (undocumented testing
1531 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1532 additional section of a DNS packet that returns MX or SRV records.
1533 Instead, it always explicitly searches for A/AAAA records. This avoids
1534 major problems that occur when a DNS server includes only records of one
1535 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1536 fixed another bug: if SRV records were looked up and the corresponding
1537 address records were *not* found in the additional section, the port
1538 values from the SRV records were lost.
1540 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1541 using the correct key (the original address) when searching the retry
1542 rules in order to find which one to use for generating the retry hint.
1544 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1545 adding the default one. Similarly, if it contains a Reply-To: header, the
1546 errors_reply_to option, if set, is not used.
1548 PH/19 When calculating a retry time, Exim used to measure the "time since
1549 failure" by looking at the "first failed" field in the retry record. Now
1550 it does not use this if it is later than than the arrival time of the
1551 message. Instead it uses the arrival time. This makes for better
1552 behaviour in cases where some deliveries succeed, thus re-setting the
1553 "first failed" field. An example is a quota failure for a huge message
1554 when small messages continue to be delivered. Without this change, the
1555 "time since failure" will always be short, possible causing more frequent
1556 delivery attempts for the huge message than are intended.
1557 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1559 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1560 $1, $2, $3) because the numerical variables can be reset during some
1561 expansion items (e.g. "match"), thereby losing the authentication data.
1563 PH/21 Make -bV show the size of off_t variables so that the test suite can
1564 decide whether to run tests for quotas > 2G.
1566 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1567 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1568 filecount value is greater than 2G or if a quota value is greater than 2G
1569 on a system where the size of off_t is not greater than 4, a panic error
1572 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1573 never match. The debug and -bh output now contains an explicit error
1574 message indicating a malformed IPv4 address or mask.
1576 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1577 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1578 PH/23 above applies.
1580 PH/25 Do not write to syslog when running in the test harness. The only
1581 occasion when this arises is a failure to open the main or panic logs
1582 (for which there is an explicit test).
1584 PH/26 Added the /no_tell option to "control=freeze".
1586 PH/27 If a host name lookup failed very early in a connection, for example, if
1587 the IP address matched host_lookup and the reverse lookup yielded a name
1588 that did not have a forward lookup, an error message of the form "no IP
1589 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1590 could be logged. Now it outputs the IP address instead of "NULL".
1592 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1593 allows the sender and the authenticated sender to be set when
1594 submitting a message from within Exim. Since child_open_exim() is
1595 documented for local_scan(), the new function should be too.
1597 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1598 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1599 results in an empty string is now treated as unset.
1601 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1603 PH/31 Added sender_verify_fail logging option.
1605 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1606 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1607 tidied the source and removed it altogether.
1609 PH/33 When a queue run was abandoned because the load average was too high, a
1610 log line was always written; now it is written only if the queue_run log
1611 selector is set. In addition, the log line for abandonment now contains
1612 information about the queue run such as the pid. This is always present
1613 in "start" and "stop" lines but was omitted from the "abandon" line.
1615 PH/34 Omit spaces between a header name and the colon in the error message that
1616 is given when verify = headers_syntax fails (if there are lots of them,
1617 the message gets confusing).
1619 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1620 names, as there are now some PTR records that contain slashes. This check
1621 is only to protect against broken name servers that fall over on strange
1622 characters, so the fact that it applies to all lookups doesn't matter.
1624 PH/36 Now that the new test suite is complete, we can remove some of the
1625 special code in Exim that was needed for the old test suite. For example,
1626 sorting DNS records because real resolvers return them in an arbitrary
1627 order. The new test suite's fake resolver always returns records in the
1630 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1631 bounces) except when queue_only is set, to avoid logging races between
1632 the different processes.
1634 PH/38 Panic-die if .include specifies a non-absolute path.
1636 PH/39 A tweak to the "H" retry rule from its user.
1638 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1639 a label. They prevented compilation on older perls.
1641 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1642 a warning to be raised on newish perls.
1644 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1645 on queue. Changes to match documented behaviour of showing count of
1646 messages matching specified criteria.
1648 PH/40 Changed the default ident timeout from 30s to 5s.
1650 PH/41 Added support for the use of login_cap features, on those BSD systems
1651 that have them, for controlling the resources used by pipe deliveries.
1653 PH/42 The content-scanning code uses fopen() to create files in which to put
1654 message data. Previously it was not paying any attention to the mode of
1655 the files. Exim runs with umask(0) because the rest of the code creates
1656 files with open(), and sets the required mode explicitly. Thus, these
1657 files were ending up world-writeable. This was not a big issue, because,
1658 being within the spool directory, they were not world-accessible. I have
1659 created a function called modefopen, which takes an additional mode
1660 argument. It sets umask(777), creates the file, chmods it to the required
1661 mode, then resets the umask. All the relevant calls to fopen() in the
1662 content scanning code have been changed to use this function.
1664 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1665 to 24 hours. This avoids potential overflow problems when processing G
1666 and H retry rules. I suspect nobody ever tinkers with this value.
1668 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1670 PH/45 When the plaintext authenticator is running as a client, the server's
1671 challenges are checked to ensure they are valid base64 strings. By
1672 default, the authentication attempt is cancelled if an invalid string is
1673 received. Setting client_ignore_invalid_base64 true ignores these errors.
1674 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1675 they are received. Thus, the responses can be made to depend on the
1676 challenges. If an invalid string is ignored, an empty string is placed in
1679 PH/46 Messages that are created by the autoreply transport now contains a
1680 References: header, in accordance with RFCs 2822 and 3834.
1682 PH/47 Added authenticated_sender_force to the smtp transport.
1684 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1686 PH/49 Installed latest patch from the Sieve maintainer.
1688 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1689 was also set, the appendfile transport was unnecessarily scanning a
1690 directory of message files (e.g. for maildir delivery) to find the count
1691 of files (along with the size), even though it did not need this
1692 information. It now does the scan only if it needs to find either the
1693 size of the count of files.
1695 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1697 PH/52 Two bugs concerned with error handling when the smtp transport is
1700 (i) Exim was not creating retry information for temporary errors given
1701 for individual recipients after the DATA command when the smtp transport
1702 was used in LMTP mode. This meant that they could be retried too
1703 frequently, and not timed out correctly.
1705 (ii) Exim was setting the flag that allows error details to be returned
1706 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1707 recipients that were returned after the DATA command.
1709 PH/53 This is related to PH/52, but is more general: for any failing address,
1710 when detailed error information was permitted to be returned to the
1711 sender, but the error was temporary, then after the final timeout, only
1712 "retry timeout exceeded" was returned. Now it returns the full error as
1713 well as "retry timeout exceeded".
1715 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1716 do this, and (what is worse) MTAs that accept it.
1718 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1719 will now be deprecated.
1721 PH/56 New os.c-cygwin from the Cygwin maintainer.
1723 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1724 formats (previously only available in exim formats via -bpr, -bpru,
1725 and -bpra. Now also available in native and exiqgrep formats)
1727 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1728 with very large, slow to parse queues
1730 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1732 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1734 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1735 responses to authentication challenges, though it was showing the
1736 challenges; (ii) I've removed the CR characters from the debug output for
1739 PH/58 Allow for the insertion of a newline as well as a space when a string
1740 is turned into more than one encoded-word during RFC 2047 encoding. The
1741 Sieve code now uses this.
1743 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1744 data_4xx, lost_connection, tls_required.
1746 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1747 message was being sent as an SMTP response.
1749 PH/61 Add -l and -k options to exicyclog.
1751 PH/62 When verifying, if an address was redirected to one new address, so that
1752 verification continued, and the new address failed or deferred after
1753 having set something in $address_data, the value of $address_data was not
1754 passed back to the ACL. This was different to the case when no
1755 redirection occurred. The value is now passed back in both cases.
1757 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1758 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1759 use login_cap.h, so on its own it isn't the distinguishing feature. The
1760 new name refers directly to the setclassresources() function.
1762 PH/65 Added configuration files for NetBSD3.
1764 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1766 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1767 is preferred over IPv4.
1769 PH/68 The bounce_return_message and bounce_return_body options were not being
1770 honoured for bounces generated during the reception of non-SMTP messages.
1771 In particular, this applied to messages rejected by the ACL. This bug has
1772 been fixed. However, if bounce_return_message is true and bounce_return_
1773 body is false, the headers that are returned for a non-SMTP message
1774 include only those that have been read before the error was detected.
1775 (In the case of an ACL rejection, they have all been read.)
1777 PH/69 The HTML version of the specification is now built in a directory called
1778 spec_html instead of spec.html, because the latter looks like a path with
1779 a MIME-type, and this confuses some software.
1781 PH/70 Catch two compiler warnings in sieve.c.
1783 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1784 function verify_get_ident() calls ip_connect() to connect a socket, but
1785 if the "connect()" function timed out, ip_connect() used to close the
1786 socket. However, verify_get_ident() also closes the socket later, and in
1787 between Exim writes to the log, which may get opened at this point. When
1788 the socket was closed in ip_connect(), the log could get the same file
1789 descriptor number as the socket. This naturally causes chaos. The fix is
1790 not to close the socket in ip_connect(); the socket should be closed by
1791 the function that creates it. There was only one place in the code where
1792 this was missing, in the iplookup router, which I don't think anybody now
1793 uses, but I've fixed it anyway.
1795 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1796 well as to direct DNS lookups. Otherwise the handling of names in host
1797 lists is inconsistent and therefore confusing.
1803 PH/01 Two changes to the default runtime configuration:
1805 (1) Move the checks for relay_from_hosts and authenticated clients from
1806 after to before the (commented out) DNS black list checks.
1808 (2) Add control=submission to the relay_from_hosts and authenticated
1809 clients checks, on the grounds that messages accepted by these
1810 statements are most likely to be submissions.
1812 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1814 (1) Generate an error if the third argument for the ${prvs expansion is
1817 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1820 (3) Reset the variables that are obtained from the first argument of
1821 ${prvscheck and used in the second argument before leaving the code,
1822 because their memory is reclaimed, so using them afterwards may do
1825 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1826 one (it's much easier than Tom thought :-).
1828 (5) Because of (4), we can now allow for the use of $prvscheck_result
1829 inside the third argument.
1831 PH/03 For some reason, the default setting of PATH when running a command from
1832 a pipe transport was just "/usr/bin". I have changed it to
1835 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1836 anything to be listed in the output from -bV.
1838 PH/05 When a filter generated an autoreply, the entire To: header line was
1839 quoted in the delivery log line, like this:
1841 => >A.N.Other <ano@some.domain> <original@ddress> ...
1843 This has been changed so that it extracts the operative address. There
1844 may be more than one such address. If so, they are comma-separated, like
1847 => >ano@some.domain,ona@other.domain <original@ddress> ...
1849 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1850 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1851 not being looked up in the rDNS to get a host name, Exim was showing the
1852 IP address twice in Received: lines, even though the IP addresses were
1853 identical. For example:
1855 Received: from [1.2.3.4] (helo=[1.2.3.4])
1857 However, if the real host name was known, it was omitting the HELO data
1858 if it matched the actual IP address. This has been tidied up so that it
1859 doesn't show the same IP address twice.
1861 PH/07 When both +timestamp and +memory debugging was on, the value given by
1862 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1863 called by the debug printing, thereby overwriting the timestamp buffer.
1864 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1866 PH/08 When the original message was included in an autoreply transport, it
1867 always said "this is a copy of the message, including all the headers",
1868 even if body_only or headers_only was set. It now gives an appropriate
1871 PH/09 Applied a patch from the Sieve maintainer which:
1873 o fixes some comments
1874 o adds the (disabled) notify extension core
1875 o adds some debug output for the result of if/elsif tests
1876 o points to the current vacation draft in the documentation
1877 and documents the missing references header update
1881 o fixes a bug in processing the envelope test (when testing
1882 multiple envelope elements, the last element determinted the
1885 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1886 Electronic Mail") by including:
1888 Auto-submitted: auto-generated
1890 in the messages that it generates (bounce messages and others, such as
1891 warnings). In the case of bounce messages for non-SMTP mesages, there was
1892 also a typo: it was using "Auto_submitted" (underscore instead of
1893 hyphen). Since every message generated by Exim is necessarily in response
1894 to another message, thes have all been changed to:
1896 Auto-Submitted: auto-replied
1898 in accordance with these statements in the RFC:
1900 The auto-replied keyword:
1902 - SHOULD be used on messages sent in direct response to another
1903 message by an automatic process,
1905 - MUST NOT be used on manually-generated messages,
1907 - MAY be used on Delivery Status Notifications (DSNs) and Message
1908 Disposition Notifications (MDNs),
1910 - MUST NOT be used on messages generated by automatic or periodic
1911 processes, except for messages which are automatic responses to
1914 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1915 to the default Received: header definition.
1917 PH/12 Added log selector acl_warn_skipped (default on).
1919 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1920 variables because (a) they are in the wrong storage pool and (b) even if
1921 they were copied, it wouldn't work properly because of the caching.
1923 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1924 checking when decoding. Apparently there are clients that generate
1925 overlong encoded strings. Why am I not surprised?
1927 PH/15 If the first argument of "${if match_address" was not empty, but did not
1928 contain an "@" character, Exim crashed. Now it writes a panic log message
1929 and treats the condition as false.
1931 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1933 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1934 extension "envelope-auth". The code is finished and in agreement with
1935 other implementations, but there is no documentation so far and in fact,
1936 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1937 not changing the active code.
1939 Print executed "if" and "elsif" statements when debugging is used. This
1940 helps a great deal to understand what a filter does.
1942 Document more things not specified clearly in RFC3028. I had all this
1943 sorted out, when out of a sudden new issues came to my mind. Oops."
1945 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1948 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1949 canonical form (as documented). However, after a host name lookup from
1950 the IP address, check_host() was doing a simple string comparison with
1951 addresses acquired from the DNS when checking that the found name did
1952 have the original IP as one of its addresses. Since any found IPv6
1953 addresses are likely to be in abbreviated form, the comparison could
1954 fail. Luckily, there already exists a function for doing the comparison
1955 by converting both addresses to binary, so now that is used instead of
1956 the text comparison.
1958 PH/20 There was another similar case to PH/19, when a complete host name was
1959 given in a host list; looking up its IP address could give an abbreviated
1960 form, whereas the current host's name might or might not be abbreviated.
1961 The same fix has been applied.
1967 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1968 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1971 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1972 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1974 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1976 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1977 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1978 Also, for "Auto-Submitted", treat anything other than "no" as
1979 non-personal, in accordance with RFC 3834. (Previously it treated
1980 anything starting "auto-" as non-personal.)
1982 TF/01 The control=submission/name=... option had a problem with syntax
1983 errors if the name included a slash character. The /name= option
1984 now slurps the rest of the string, so it can include any characters
1985 but it must come last in the list of options (after /sender_retain
1988 PH/05 Some modifications to the interface to the fake nameserver for the new
1996 TK/01 Added the "success_on_redirect" address verification option. See
1997 NewStuff for rationale and an example.
1999 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2001 PH/02 Patch to exigrep to allow it to work on syslog lines.
2003 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2004 fread() to skip over the body file's header line, because in Cygwin the
2005 header line is locked and is inaccessible.
2007 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2008 co-exist for some time) to make it clear that it is the Exim ID that is
2009 referenced, not the Message-ID: header line.
2011 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2012 string_format() function, because snprintf() does not exist on all
2015 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2016 ${stat: expansion item.
2018 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2019 protocol synchronization error", to keep the pedants happy.
2021 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2022 well as for IRIX systems, when gcc is being used. See the host.c source
2025 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2027 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2030 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2031 transport and to the smtp transport in LMTP mode.
2033 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2035 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2036 run a filter in a subprocess. This could lead to confusion in subsequent
2037 lookups in the parent process. There should also be a search_tidyup() at
2038 the end of the subprocess.
2040 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2041 only if the host matched helo_try_verify_hosts, which caused the
2042 verification to occur when the EHLO/HELO command was issued. The ACL just
2043 tested the remembered result. Now, if a previous verification attempt has
2044 not happened, "verify = helo" does it there and then.
2046 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2048 TK/03 Fix log output including CR from clamd.
2050 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2051 exist provoked a memory error which could cause a segfault.
2053 PH/15 Installed PCRE 6.2
2055 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2057 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2058 of the problem. Specifically, suggested +O2 rather than +O1 for the
2061 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2063 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2064 record, the port was not correctly logged when the outgoing_port log
2065 selector was set (it logged the transort's default port).
2067 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2068 fallback_hosts, and "hosts" in the smtp transport.
2070 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2071 host errors such as "Connection refused".
2073 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2074 authentication with radiusclient 0.4.9:
2076 - Error returned from rc_read_config was caught wrongly
2077 - Username/password not passed on to radius server due to wrong length.
2079 The presumption is that some radiusclient API changes for 4.51/PH/17
2080 were not taken care of correctly. The code is still untested by me (my
2081 Linux distribution still has 0.3.2 of radiusclient), but it was
2082 contributed by a Radius user.
2084 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2085 expanding the "port" option of the smtp transport.
2087 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2088 while reading a MIME header. Thanks to Tom Hughes for a patch.
2090 PH/24 Include config.h inside local_scan.h so that configuration settings are
2093 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2094 that in an ACL for RCPT (for example), you can examine exactly what was
2097 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2098 commands, but it was not correctly comparing the address with the actual
2099 client host address. Thus, it would show the EHLO address in Received:
2100 header lines when this was not necessary.
2102 PH/27 Added the % operator to ${eval:}.
2104 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2105 it should be ignoring failures (because with -C, for example, it has lost
2106 privilege). It wasn't ignoring creation failures other than "already
2109 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2112 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2113 because an input line was too long, either on its own, or by virtue of
2114 too many continuations, the temporary file was not being removed, and the
2115 return code was incorrect.
2117 PH/31 Missing "BOOL" in function definition in filtertest.c.
2119 PH/32 Applied Sieve patches from the maintainer.
2121 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2123 PH/33 Added "verify = not_blind".
2125 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2126 Local/Makefile (with some defaults set). These are used in built scripts
2127 such as exicyclog, but they have never been used in the exim_install
2128 script (though there are many overriding facilities there). I have
2129 arranged that the exim_install script now takes note of these two
2132 PH/35 Installed configuration files for Dragonfly.
2134 PH/36 When a locally submitted message by a trusted user did not contain a
2135 From: header, and the sender address was obtained from -f or from an SMTP
2136 MAIL command, and the trusted user did not use -F to supply a sender
2137 name, $originator_name was incorrectly used when constructing a From:
2138 header. Furthermore, $originator_name was used for submission mode
2139 messages from external hosts without From: headers in a similar way,
2140 which is clearly wrong.
2142 PH/37 Added control=suppress_local_fixups.
2144 PH/38 When log_selector = +received_sender was set, and the addition of the
2145 sender made the log line's construction buffer exactly full, or one byte
2146 less than full, an overflow happened when the terminating "\n" was
2149 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2150 when the result of a list match is failure because a DNS lookup failed.
2152 PH/40 RM_COMMAND is now used in the building process.
2154 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2155 the "build-* directories that it finds.
2157 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2158 address in a domain literal was a prefix of an interface address.
2160 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2161 when verifying a sender address, unless rewrite_headers is false.
2163 PH/44 Wrote a long comment about why errors_to addresses are verified as
2164 recipients, not senders.
2166 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2167 the ratelimit ACL was added.
2169 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2171 PH/46 Added extra information about PostgreSQL errors to the error string.
2173 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2174 avoiding the need to install special zones in a real server. This is
2175 backwards compatible; if it can't find the fake resolver, it drops back.
2176 Thus, both old and new test suites can be run.
2178 TF/02 Added util/ratelimit.pl
2180 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2181 clock is set back in time.
2183 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2184 Candler <B.Candler@pobox.com>.
2186 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2187 OK for addresses that are the result of redirections.
2189 PH/48 A number of further additions for the benefit of the new test suite,
2190 including a fake gethostbyname() that interfaces to the fake DNS resolver
2193 TF/06 The fix for widen_domains has also been applied to qualify_single and
2194 search_parents which are the other dnslookup options that can cause
2197 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2200 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2202 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2203 service" if it encountered an MX record with an empty target hostname.
2204 The message is now "an MX or SRV record indicated no SMTP service".
2206 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2207 if the DNS of the sending site is misconfigured. This is quite a
2208 common situation. This change restores the behaviour of treating a
2209 helo verification defer as a failure.
2211 PH/51 If self=fail was set on a router, the bounce message did not include the
2212 actual error message.
2218 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2220 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2221 command itself ended in error, the underlying message about the transport
2222 filter timeout was being overwritten with the pipe command error. Now the
2223 underlying error message should be appended to the second error message.
2225 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2227 PH/02 Reduce the amount of output that "make" produces by default. Full output
2228 can still be requested.
2230 PH/03 The warning log line about a condition test deferring for a "warn" verb
2231 was being output only once per connection, rather than after each
2232 occurrence (because it was using the same function as for successful
2233 "warn" verbs). This seems wrong, so I have changed it.
2235 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2236 it should not have, which might have caused a crash in the right
2237 circumstances, but probably never did.
2239 PH/04 Installed a modified version of Tony Finch's patch to make submission
2240 mode fix the return path as well as the Sender: header line, and to
2241 add a /name= option so that you can make the user's friendly name appear
2244 TF/03 Added the control = fakedefer ACL modifier.
2246 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2247 Mark Lowes for thorough testing.
2249 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2251 TK/03 Merged latest SRS patch from Miles Wilton.
2253 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2254 with the definition in sysexits.h (which is #included earlier).
2255 Fortunately, Exim does not actually use EX_OK. The code used to try to
2256 preserve the sysexits.h value, by assumimg that macro definitions were
2257 scanned for macro replacements. I have been disabused of this notion,
2258 so now the code just undefines EX_OK before #including unistd.h.
2260 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2261 in the smtp transport. When a block could not be written in a single
2262 write() function, the timeout was being re-applied to each part-write.
2263 This seems wrong - if the receiver was accepting one byte at a time it
2264 would take for ever. The timeout is now adjusted when this happens. It
2265 doesn't have to be particularly precise.
2267 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2268 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2270 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2271 without a domain if the check to <postmaster@domain> fails.
2273 SC/01 Eximstats: added -xls and the ability to specify output files
2274 (patch written by Frank Heydlauf).
2276 SC/02 Eximstats: use FileHandles for outputing results.
2278 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2280 SC/04 Eximstats: fixed display of large numbers with -nvr option
2282 SC/05 Eximstats: fixed merging of reports with empty tables.
2284 SC/06 Eximstats: added the -include_original_destination flag
2286 SC/07 Eximstats: removed tabs and trailing whitespace.
2288 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2290 TK/06 MBOX spool code: Add real "From " MBOX separator line
2291 so the .eml file is really in mbox format (even though
2292 most programs do not really care). Patch from Alex Miller.
2294 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2295 The latter is generated from $received_to and is only set if the
2296 message has one envelope recipient. SA can use these headers,
2297 obviously out-of-the-box. Patch from Alex Miller.
2299 PH/08 The ${def test on a variable was returning false if the variable's
2300 value was "0", contrary to what the specification has always said!
2301 The result should be true unless the variable is empty.
2303 PH/09 The syntax error of a character other than { following "${if
2304 def:variable_name" (after optional whitespace) was not being diagnosed.
2305 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2306 accidental colon was present, for example, could give incorrect results.
2308 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2309 result is used (not including appendfile, where other changes are about
2312 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2313 This involved changing a lot of size variables from int to off_t. It
2314 should work with maildirs and everything.
2316 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2317 spamd dying while we are connected to it.
2319 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2322 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2323 function for BDB 4.3.
2325 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2327 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2328 the sources for PCRE have been re-arranged and more files are now
2331 PH/15 The code I had for printing potentially long long variables in PH/11
2332 above was not the best (it lost precision). The length of off_t variables
2333 is now inspected at build time, and an appropriate printing format (%ld
2334 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2335 to be "long long int" or "long int". This is needed for the internal
2336 formatting function string_vformat().
2338 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2339 the configuration file to be ":syslog", then the script "guesses" where
2340 the logs files are, rather than using the compiled in default. In our
2341 case the guess is not the same as the compiled default, so the script
2342 suddenly stopped working when I started to use syslog. The patch checks
2343 to see if log_file_path is "". If so, it attempts to read it from exim
2344 with no configuration file to get the compiled in version, before it
2345 falls back to the previous guessing code."
2347 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2348 implementing BATV in an Exim configuration. See NewStuff for the gory
2351 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2352 Makefile that are specific to HP-UX.
2354 PH/18 If the "use_postmaster" option was set for a recipient callout together
2355 with the "random" option, the postmaster address was used as the MAIL
2356 FROM address for the random test, but not for the subsequent recipient
2357 test. It is now used for both.
2359 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2360 patch removes a few documentation additions to RFC 3028, because the
2361 latest draft now contains them. It adds the new en;ascii-case comparator
2362 and a new error check for 8bit text in MIME parts. Comparator and
2363 require names are now matched exactly. I enabled the subaddress
2364 extension, but it is not well tested yet (read: it works for me)."
2366 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2367 rework some of the code of TK/09 above to avoid the hardwired use of
2368 "%lld" and "long long". Replaced the call to snprintf() with a call to
2371 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2372 records point to non-existent hosts", "retry timeout exceeded", and
2373 "retry time not reached for any host after a long failure period".
2375 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2376 experimental DomainKeys support:
2378 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2379 (2) On an error such as an illegally used "control", the wrong name for
2380 the control was given.
2382 These problems did NOT occur unless DomainKeys support was compiled.
2384 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2386 PH/24 Added ${if match_ip condition.
2388 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2389 databases so that it will be absolutely obvious if a crash occurs in the
2390 DB library. This is a regular occurrence (often caused by mis-matched
2393 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2394 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2395 picked up on a user's system that detects such things. There doesn't seem
2396 to be a gcc warning option for this - only an attribute that has to be
2397 put on the function's prototype. It seems that in Fedora Core 4 they have
2398 set this on a number of new functions. No doubt there will be more in due
2401 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2402 specify a transport. However, if an address that was verified by such a
2403 router was the subject of a callout, Exim crashed because it tried to
2404 read the rcpt_include_affixes from the non-existent transport. Now it
2405 just assumes that the setting of that option is false. This bug was
2406 introduced by 4.51/PH/31.
2408 PH/28 Changed -d+all to exclude +memory, because that information is very
2409 rarely of interest, but it makes the output a lot bigger. People tend to
2410 do -d+all out of habit.
2412 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2413 code in os-type was giving problems when libc.so lives in lib64, like on
2416 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2417 aren't the modern standard, and it seems that some systems' include files
2418 don't always have them. Exim was already checking for some of the newer
2419 ones like T_AAAA, and defining it itself. I've added checks for all the
2420 record types that Exim uses.
2422 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2423 not automatically generating a new one, as it is supposed to. This
2424 prevented TLS from working. If the file did exist, but contained invalid
2425 data, a new version was generated, as expected. It was only the case of a
2426 non-existent file that was broken.
2428 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2429 with a change in libdomainkeys > 0.64.
2431 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2432 from DNS. If the selector record carries the flag, it now has
2433 precedence over the domain-wide flag.
2435 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2437 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2438 the use of a port name that isn't defined in /etc/services) occurred, the
2439 message was deferred as in a normal delivery, and thus remained on the
2440 spool, instead of being failed because of the mua_wrapper setting. This
2441 is now fixed, and I tidied up some of the mua_wrapper messages at the
2444 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2445 the messages in a hash of arrays rather than using individual hashes.
2446 This is a bit cleaner and results in dramatic memory savings, albeit
2447 at a slight CPU cost.
2449 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2450 as requested by Marc Sherman.
2452 SC/10 Eximstats: added histograms for user specified patterns as requested
2455 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2457 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2458 fopen() in the content-scanning modules that did not already have it.
2464 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2465 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2467 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2469 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2471 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2472 to test Sieve filters that use "vacation".
2474 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2475 that changes the way the GnuTLS parameters are stored in the cache file.
2476 The new format can be generated externally. For backward compatibility,
2477 if the data in the cache doesn't make sense, Exim assumes it has read an
2478 old-format file, and it generates new data and writes a new file. This
2479 means that you can't go back to an older release without removing the
2482 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2483 work if there are any delivery delays because "one_time" forces the
2484 parent to be marked "delivered", so its unseen clone is never tried
2485 again. For this reason, Exim now forbids the simultaneous setting of
2488 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2489 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2490 entry if you want to know the details. The fix, however, neglected to
2491 consider the case when local delivery batching is involved. The test for
2492 "previously delivered" was not happening when checking to see if an
2493 address could be batched with a previous (undelivered) one; under
2494 certain circumstances this could lead to multiple deliveries to the same
2497 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2498 in its include files, and this causes problems building Exim.
2500 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2501 header_syntax) but Exim was just ignoring anything given after a slash.
2502 In particular, this caused confusion with an attempt to use "verify =
2503 reverse_host_lookup/defer_ok". An error is now given when options are
2504 supplied for verify items that do not have them. (Maybe reverse_host_
2505 lookup should have a defer_ok option, but that's a different point.)
2507 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2508 defined by RFC 821) to 2048, because there were problems with some AUTH
2509 commands, and RFC 1869 says the size should be increased for extended
2510 SMTP commands that take arguments.
2512 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2515 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2516 "unknown" error; now it says that the functionality isn't in the binary.
2518 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2519 an address' error message when a string expansion fails (syntax or
2520 whatever). Otherwise the password may appear in the log. Following change
2521 PH/42 below, there is no longer a chance of it appearing in a bounce
2524 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2526 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2527 of its addresses was ever tried. (Bugzilla bug #2.)
2529 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2530 the result incorrectly in the debug output. (It correctly added a newline
2531 to what was transported.)
2533 TF/01 Added $received_time.
2535 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2536 commented out examples of how to interface to a virus scanner and to
2537 SpamAssassin. Also added commented examples of av_scanner and
2538 spamd_address settings.
2540 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2541 and controls are allowed in which ACLs. There were a couple of minor
2542 errors. Some of the entries in the conditions table (which is a table of
2543 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2544 negation of where the condition IS allowed.
2546 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2548 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2549 header file does not have a version number, so I've had to invent a new
2550 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2551 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2552 radiusclient), but it was contributed by a Radius user.
2554 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2555 files or directories, for OpenSSL.
2557 PH/19 When an Exim process that is running as root has to create an Exim log
2558 file, it does so in a subprocess that runs as exim:exim so as to get the
2559 ownership right at creation (otherwise, other Exim processes might see
2560 the file with the wrong ownership). There was no test for failure of this
2561 fork() call, which would lead to the process getting stuck as it waited
2562 for a non-existent subprocess. Forks do occasionally fail when resources
2563 run out. I reviewed all the other calls to fork(); they all seem to check
2566 PH/20 When checking for unexpected SMTP input at connect time (before writing
2567 the banner), Exim was not dealing correctly with a non-positive return
2568 from the read() function. If the client had disconnected by this time,
2569 the result was a log entry for a synchronization error with an empty
2570 string after "input=" when read() returned zero. If read() returned -1
2571 (an event I could not check), uninitialized data bytes were printed.
2572 There were reports of junk text (parts of files, etc) appearing after
2575 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2577 PH/22 Added support for macro redefinition, and (re)definition in between
2578 driver and ACL definitions.
2580 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2581 forgetting to use the resulting value; it was using the unexpanded value.
2583 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2584 hadn't been configured. The fix is from Juergen Kreileder, who
2585 understands it better than I do:
2587 "Here's what I see happening with three configured cyrus_sasl
2588 authenticators configured (plain, login, cram-md5):
2590 On startup auth_cyrus_sasl_init() gets called for each of these.
2591 This means three calls to sasl_listmech() without a specified mech_list.
2592 => SASL tests which mechs of all available mechs actually work
2593 => three warnings about OTP not working
2594 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2596 With the patch, sasl_listmech() also gets called three times. But now
2597 SASL's mech_list option is set to the server_mech specified in the the
2598 authenticator. Or in other words, the answer from sasl_listmech()
2599 gets limited to just the mech you're testing for (which is different
2601 => the return list contains just 'plain' or 'login', 'cram-md5' or
2602 nothing depending on the value of ob->server_mech.
2604 I've just tested the patch: Authentication still works fine,
2605 unavailable mechs specified in the exim configuration are still
2606 caught, and the auth.log warnings about OTP are gone."
2608 PH/25 When debugging is enabled, the contents of the command line are added
2609 to the debugging output, even when log_selector=+arguments is not
2612 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2613 answer is "GNU", and only if the return is "GNU/something" is the answer
2616 PH/27 $acl_verify_message is now set immediately after the failure of a
2617 verification in an ACL, and so is available in subsequent modifiers. In
2618 particular, the message can be preserved by coding like this:
2620 warn !verify = sender
2621 set acl_m0 = $acl_verify_message
2623 Previously, $acl_verify_message was set only while expanding "message"
2624 and "log_message" when a very denied access.
2626 PH/28 Modified OS/os.c-Linux with
2628 -#ifndef OS_LOAD_AVERAGE
2629 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2631 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2632 nomenclature these days.)
2634 PH/29 Installed patch from the Sieve maintainer that adds the options
2635 sieve_useraddress and sieve_subaddress to the redirect router.
2637 PH/30 In these circumstances:
2638 . Two addresses routed to the same list of hosts;
2639 . First host does not offer TLS;
2640 . First host accepts first address;
2641 . First host gives temporary error to second address;
2642 . Second host offers TLS and a TLS session is established;
2643 . Second host accepts second address.
2644 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2645 and peerdn, if requested) that were in fact used only for the second
2648 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2649 attention to any local part prefix or suffix that was matched by the
2650 router that accepted the address. It now behaves in the same way as it
2651 does for delivery: the affixes are removed from the local part unless
2652 rcpt_include_affixes is set on the transport.
2654 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2655 timeout during the DATA phase of an incoming message.
2657 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2658 applied a patch sanctioned by the Sieve maintainer.
2660 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2661 the uid or gid is negative. A case of a negative gid caused this to be
2662 noticed. The fix allows for either to be negative.
2664 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2665 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2668 PH/36 The change PH/12 above was broken. Fixed it.
2670 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2671 the grounds that routing the same address twice would always produce the
2672 same answer. This might have been true once, but it is certainly no
2673 longer true now. Routing a child address may depend on the previous
2674 routing that produced that child. Some complicated redirection strategies
2675 went wrong when messages had multiple recipients, and made Exim's
2676 behaviour dependent on the order in which the addresses were given.
2678 I have moved the duplicate checking until after the routing is complete.
2679 Exim scans the addresses that are assigned to local and remote
2680 transports, and removes any duplicates. This means that more work will be
2681 done, as duplicates will always all be routed, but duplicates are
2682 presumably rare, so I don't expect this is of any significance.
2684 For deliveries to pipes, files, and autoreplies, the duplicate checking
2685 still happens during the routing process, since they are not going to be
2688 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2689 It corrects a timeout issue with spamd. This is Ian's comment: "The
2690 background is that sometimes spamd either never reads data from a
2691 connection it has accepted, or it never writes response data. The exiscan
2692 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2693 blindly assumes that writes won't block so it may never time out."
2695 PH/39 Allow G after quota size as well as K and M.
2697 PH/40 The value set for $authenticated_id in an authenticator may not contain
2698 binary zeroes or newlines because the value is written to log lines and
2699 to spool files. There was no check on this. Now the value is run through
2700 the string_printing() function so that such characters are converted to
2701 printable escape sequences.
2703 PH/41 $message_linecount is a new variable that contains the total number of
2704 lines in the message. Compare $body_linecount, which is the count for the
2707 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2708 bounce and delay warning messages, except in certain special cases, which
2711 (a) An SMTP error message from a remote host;
2712 (b) A message specified in a :fail: redirection;
2713 (c) A message specified in a "fail" command in a system filter;
2714 (d) A message specified in a FAIL return from the queryprogram router;
2715 (e) A message specified by the cannot_route_message router option.
2717 In these cases only, Exim does include the error details in bounce and
2718 warning messages. There are also a few cases where bland messages such
2719 as "unrouteable address" or "local delivery error" are given.
2721 PH/43 $value is now also set for the "else" part of a ${run expansion.
2723 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2724 being worked on, but at least Exim now implements the latest version to
2727 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2728 process to complete was treated as a delivery failure, a timeout while
2729 writing the message to the pipe was logged, but erroneously treated as a
2730 successful delivery. Such timeouts include transport filter timeouts. For
2731 consistency with the overall process timeout, these timeouts are now
2732 treated as errors, giving rise to delivery failures by default. However,
2733 there is now a new Boolean option for the pipe transport called
2734 timeout_defer, which, if set TRUE, converts the failures into defers for
2735 both kinds of timeout. A transport filter timeout is now identified in
2738 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2739 systems where "make" and "gmake" are different, calling "gmake" at top
2740 level broke things. I've arranged for the value of $(MAKE) to be passed
2741 from the Makefile to this script so that it can call the same version of
2745 A note about Exim versions 4.44 and 4.50
2746 ----------------------------------------
2748 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2749 changes of various kinds. As a consequence, a big documentation update was
2750 needed. This delayed the release for rather longer than seemed good, especially
2751 in the light of a couple of (minor) security issues. Therefore, the changes
2752 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2753 release. So 4.44 and 4.50 are in effect two different branches that both start
2756 I have left the 4.50 change log unchanged; it contains all the changes since
2757 4.43. The change log for 4.44 is below; many of its items are identical to
2758 those for 4.50. This seems to be the most sensible way to preserve the
2759 historical information.
2765 1. Minor wording change to the doc/README.SIEVE file.
2767 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2768 computation of the current number of files was incorrect.
2770 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2771 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2774 4. Give more explanation in the error message when the command for a transport
2775 filter fails to execute.
2777 5. There are several places where Exim runs a non-Exim command in a
2778 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2779 was being done only for the command run by the queryprogram router. It is
2780 now done for all such subprocesses. The other cases are: ${run, transport
2781 filters, and the commands run by the lmtp and pipe transports.
2783 6. Added CONFIGURE_GROUP build-time option.
2785 7. Some older OS have a limit of 256 on the maximum number of file
2786 descriptors. Exim was using setrlimit() to set 1000 as a large value
2787 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2788 systems. I've change it so that if it can't get 1000, it tries for 256.
2790 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2791 was an oversight, and furthermore, ever since the addition of extra
2792 controls (e.g. 4.43/32), the checks on when to allow different forms of
2793 "control" were broken. There should now be diagnostics for all cases when a
2794 control that does not make sense is encountered.
2796 9. Added the /retain_sender option to "control=submission".
2798 10. $recipients is now available in the predata ACL (oversight).
2800 11. Tidy the search cache before the fork to do a delivery from a message
2801 received from the command line. Otherwise the child will trigger a lookup
2802 failure and thereby defer the delivery if it tries to use (for example) a
2803 cached ldap connection that the parent has called unbind on.
2805 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2806 of $address_data from the recipient verification was clobbered by the
2807 sender verification.
2809 13. The value of address_data from a sender verification is now available in
2810 $sender_address_data in subsequent conditions in the ACL statement.
2812 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2814 15. Added a new option "connect=<time>" to callout options, to set a different
2817 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2818 was its contents. (It was OK if the option was not defined at all.)
2820 17. A "Completed" log line is now written for messages that are removed from
2821 the spool by the -Mrm option.
2823 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2824 information about exactly what failed.
2826 19. Added -dd to debug only the daemon process.
2828 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2829 handles timeouts, both on the server side and network timeouts. Renamed the
2830 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2832 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2833 It is now set to "smtps".
2835 22. $host_address is now set to the target address during the checking of
2836 ignore_target_hosts.
2838 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2839 being passed; this would have caused $sender_host_name to have been used if
2840 matching the list had actually called for a host name (not very likely,
2841 since this list is usually IP addresses). A host name is now passed as
2844 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2845 code that specifies a non-restarting handler (typically sigaction() in
2846 modern systems) in an attempt to fix a rare and obscure crash bug.
2848 25. Narrowed the window for a race in the daemon that could cause it to ignore
2849 SIGCHLD signals. This is not a major problem, because they are used only to
2850 wake it up if nothing else does.
2852 26. A malformed maildirsize file could cause Exim to calculate negative values
2853 for the mailbox size or file count. Odd effects could occur as a result.
2854 The maildirsize information is now recalculated if the size or filecount
2857 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2858 support for a long time. Removed HAVE_SYS_VFS_H.
2860 28. Installed the latest version of exipick from John Jetmore.
2862 29. In an address list, if the pattern was not a regular expression, an empty
2863 subject address (from a bounce message) matched only if the pattern was an
2864 empty string. Non-empty patterns were not even tested. This was the wrong
2865 because it is perfectly reasonable to use an empty address as part of a
2866 database query. An empty address is now tested by patterns that are
2867 lookups. However, all the other forms of pattern expect the subject to
2868 contain a local part and a domain, and therefore, for them, an empty
2869 address still always fails if the pattern is not itself empty.
2871 30. Exim went into a mad DNS loop when attempting to do a callout where the
2872 host was specified on an smtp transport, and looking it up yielded more
2873 than one IP address.
2875 31. Re-factored the code for checking spool and log partition space into a
2876 function that finds that data and another that does the check. The former
2877 is then used to implement four new variables: $spool_space, $log_space,
2878 $spool_inodes, and $log_inodes.
2880 32. The RFC2047 encoding function was originally intended for short strings
2881 such as real names; it was not keeping to the 75-character limit for
2882 encoded words that the RFC imposes. It now respects the limit, and
2883 generates multiple encoded words if necessary. To be on the safe side, I
2884 have increased the buffer size for the ${rfc2047: expansion operator from
2887 33. It is now permitted to omit both strings after an "if" condition; if the
2888 condition is true, the result is "true". As before, when the second string
2889 is omitted, a false condition yields an empty string. This makes it less
2890 cumbersome to write custom ACL and router conditions.
2892 34. Failure to deliver a bounce message always caused it to be frozen, even if
2893 there was an errors_to setting on the router. The errors_to setting is now
2896 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2897 canonical form (fully expanded) before being placed in
2898 $sender_host_address.
2900 36. The table in the code that translates DNS record types into text (T_A to
2901 "A" for instance) was missing entries for NS and CNAME. It is just possible
2902 that this could have caused confusion if both these types were looked up
2903 for the same domain, because the text type is used as part of Exim's
2904 per-process caching. But the chance of anyone hitting this buglet seems
2907 37. The dnsdb lookup has been extended in a number of ways.
2909 (1) There is a new type, "zns", which walks up the domain tree until it
2910 finds some nameserver records. It should be used with care.
2912 (2) There is a new type, "mxh", which is like "mx" except that it returns
2913 just the host names, not the priorities.
2915 (3) It is now possible to give a list of domains (or IP addresses) to be
2916 looked up. The behaviour when one of the lookups defers can be
2917 controlled by a keyword.
2919 (4) It is now possible to specify the separator character for use when
2920 multiple records are returned.
2922 38. The dnslists ACL condition has been extended: it is now possible to supply
2923 a list of IP addresses and/or domains to be looked up in a particular DNS
2926 39. Added log_selector=+queue_time_overall.
2928 40. When running the queue in the test harness, wait just a tad after forking a
2929 delivery process, to get repeatability of debugging output.
2931 41. Include certificate and key file names in error message when GnuTLS fails
2932 to set them up, because the GnuTLS error message doesn't include the name
2933 of the failing file when there is a problem reading it.
2935 42. Allow both -bf and -bF in the same test run.
2937 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2939 44. The "Exiscan patch" is now merged into the mainline Exim source.
2941 45. Sometimes the final signoff response after QUIT could fail to get
2942 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2943 < 0 before doing a fflush(). This bug looks as though it goes back to the
2944 introduction of TLS in release 3.20, but "sometimes" must have been rare
2945 because the tests only now provoked it.
2947 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2948 (this can affect the format of dates).
2950 47. exim_tidydb, when checking for the continued existence of a message for
2951 which it has found a message-specific retry record, was not finding
2952 messages that were in split spool directories. Consequently, it was
2953 deleting retry records that should have stayed in existence.
2955 48. Steve fixed some bugs in eximstats.
2957 49. The SPA authentication driver was not abandoning authentication and moving
2958 on to the next authenticator when an expansion was forced to fail,
2959 contradicting the general specification for all authenticators. Instead it
2960 was generating a temporary error. It now behaves as specified.
2962 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2963 (the order specifies the preference for clients). The order is now AES256,
2964 AES128, 3DES, ARCFOUR128.
2966 51. Small patch to Sieve code - explicitly set From: when generating an
2969 52. Exim crashed if a remote delivery caused a very long error message to be
2970 recorded - for instance if somebody sent an entire SpamAssassin report back
2971 as a large number of 550 error lines. This bug was coincidentally fixed by
2972 increasing the size of one of Exim's internal buffers (big_buffer) that
2973 happened as part of the Exiscan merge. However, to be on the safe side, I
2974 have made the code more robust (and fixed the comments that describe what
2977 53. Now that there can be additional text after "Completed" in log lines (if
2978 the queue_time_overall log selector is set), a one-byte patch to exigrep
2979 was needed to allow it to recognize "Completed" as not the last thing in
2982 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2983 patch that reportedly fixes this has been added. I am not expert enough to
2984 create a test for it. This is what the patch creator wrote:
2986 "I found a little strange behaviour of ldap code when working with
2987 Windows 2003 AD Domain, where users was placed in more than one
2988 Organization Units. When I tried to give exim partial DN, the exit code
2989 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2990 But simultaneously result of request was absolutely normal ldap result,
2991 so I produce this patch..."
2993 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2994 REFERENCE, so I have modified the code to exclude the patch when that macro
2997 55. Some experimental protocols are using DNS PTR records for new purposes. The
2998 keys for these records are domain names, not reversed IP addresses. The
2999 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3000 leaves it alone. Component reversal etc. now happens only for IP addresses.
3003 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3005 57. Double the size of the debug message buffer (to 2048) so that more of very
3006 long debug lines gets shown.
3008 58. The exicyclog utility now does better if the number of log files to keep
3009 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3011 59. Two changes related to the smtp_active_hostname option:
3013 (1) $smtp_active_hostname is now available as a variable.
3014 (2) The default for smtp_banner uses $smtp_active_hostname instead
3015 of $primary_hostname.
3017 60. The host_aton() function is supposed to be passed a string that is known
3018 to be a valid IP address. However, in the case of IPv6 addresses, it was
3019 not checking this. This is a hostage to fortune. Exim now panics and dies
3020 if the condition is not met. A case was found where this could be provoked
3021 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3022 components; fortuitously, this particular loophole had already been fixed
3023 by change 4.50/55 above.
3025 If there are any other similar loopholes, the new check in host_aton()
3026 itself should stop them being exploited. The report I received stated that
3027 data on the command line could provoke the exploit when Exim was running as
3028 exim, but did not say which command line option was involved. All I could
3029 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3030 running as the user.
3033 61. There was a buffer overflow vulnerability in the SPA authentication code
3034 (which came originally from the Samba project). I have added a test to the
3035 spa_base64_to_bits() function which I hope fixes it.
3038 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3039 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3041 63. The daemon start-up calls getloadavg() while still root for those OS that
3042 need the first call to be done as root, but it missed one case: when
3043 deliver_queue_load_max is set with deliver_drop_privilege. This is
3044 necessary for the benefit of the queue runner, because there is no re-exec
3045 when deliver_drop_privilege is set.
3047 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3048 This has been fixed.
3050 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3051 was in use, was not putting the data itself into the right store pool;
3052 consequently, it could be overwritten for a subsequent message in the same
3053 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3056 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3058 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3059 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3060 the function were treating the return as a boolean value, which happened to
3061 work because 0=false and not-0=true, but is not correct code.
3063 68. The host_aton() function was not handling scoped IPv6 addresses (those
3064 with, for example, "%eth0" on the end) correctly.
3066 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3067 negated items (that is, ~something) in unsigned ints. Some compilers
3068 apparently mutter when there is no cast.
3070 70. If an address verification called from an ACL failed, and did not produce a
3071 user-specific message (i.e. there was only a "system" message), nothing was
3072 put in $acl_verify_message. In this situation, it now puts the system
3075 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3076 session; change 4.31/43 added the unwanted input to the log line - except
3077 that it did not do this in the start of session case. It now does.
3079 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3080 This is wrong and can cause the other end to generate a synchronization
3081 error if it is another Exim or anything else that does the synchronization
3082 check. A QUIT command is no longer sent after a timeout.
3084 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3085 during host lookups.
3087 74. The defer_ok option of callout verification was not working if it was used
3088 when verifying addresses in header lines, that is, for this case:
3090 verify = header_sender/callout=defer_ok
3092 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3093 those file descriptors could be used for SMTP connections. If anything
3094 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3095 could be sent to the SMTP client, causing chaos. The daemon now opens
3096 stdin, stdout, and stderr to /dev/null when it puts itself into the
3099 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3100 log by default. The user can override this with suitable Perl magic.
3102 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3103 the log message when discard triggers, was not working for the DATA ACL or
3104 for the non-SMTP ACL.
3106 78. Error message wording change in sieve.c.
3108 79. If smtp_accept_max_per_host was set, the number of connections could be
3109 restricted to fewer than expected, because the daemon was trying to set up
3110 a new connection before checking whether the processes handling previous
3111 connections had finished. The check for completed processes is now done
3112 earlier. On busy systems, this bug wouldn't be noticed because something
3113 else would have woken the daemon, and it would have reaped the completed
3116 80. If a message was submitted locally by a user whose login name contained one
3117 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3118 It caused a spool format error. I have fixed the spool reading code. A
3119 related problem was that the "from" clause in the Received: line became
3120 illegal because of the space(s). It is now covered by ${quote_local_part.
3122 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3125 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3127 83. Patch from Sieve maintainer to fix unterminated string problem in
3128 "vacation" handling.
3130 84. Some minor changes to the Linux configuration files to help with other
3131 OS variants using glibc.
3133 85. One more patch for Sieve to update vacation handling to latest spec.
3136 ----------------------------------------------------
3137 See the note above about the 4.44 and 4.50 releases.
3138 ----------------------------------------------------
3144 1. Change 4.43/35 introduced a bug that caused file counts to be
3145 incorrectly computed when quota_filecount was set in an appendfile
3148 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3149 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3152 3. Give more explanation in the error message when the command for a transport
3153 filter fails to execute.
3155 4. There are several places where Exim runs a non-Exim command in a
3156 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3157 was being done only for the command run by the queryprogram router. It is
3158 now done for all such subprocesses. The other cases are: ${run, transport
3159 filters, and the commands run by the lmtp and pipe transports.
3161 5. Some older OS have a limit of 256 on the maximum number of file
3162 descriptors. Exim was using setrlimit() to set 1000 as a large value
3163 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3164 systems. I've change it so that if it can't get 1000, it tries for 256.
3166 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3167 was an oversight, and furthermore, ever since the addition of extra
3168 controls (e.g. 4.43/32), the checks on when to allow different forms of
3169 "control" were broken. There should now be diagnostics for all cases when a
3170 control that does not make sense is encountered.
3172 7. $recipients is now available in the predata ACL (oversight).
3174 8. Tidy the search cache before the fork to do a delivery from a message
3175 received from the command line. Otherwise the child will trigger a lookup
3176 failure and thereby defer the delivery if it tries to use (for example) a
3177 cached ldap connection that the parent has called unbind on.
3179 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3180 of $address_data from the recipient verification was clobbered by the
3181 sender verification.
3183 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3184 was its contents. (It was OK if the option was not defined at all.)
3186 11. A "Completed" log line is now written for messages that are removed from
3187 the spool by the -Mrm option.
3189 12. $host_address is now set to the target address during the checking of
3190 ignore_target_hosts.
3192 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3193 being passed; this would have caused $sender_host_name to have been used if
3194 matching the list had actually called for a host name (not very likely,
3195 since this list is usually IP addresses). A host name is now passed as
3198 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3199 code that specifies a non-restarting handler (typically sigaction() in
3200 modern systems) in an attempt to fix a rare and obscure crash bug.
3202 15. Narrowed the window for a race in the daemon that could cause it to ignore
3203 SIGCHLD signals. This is not a major problem, because they are used only to
3204 wake it up if nothing else does.
3206 16. A malformed maildirsize file could cause Exim to calculate negative values
3207 for the mailbox size or file count. Odd effects could occur as a result.
3208 The maildirsize information is now recalculated if the size or filecount
3211 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3212 support for a long time. Removed HAVE_SYS_VFS_H.
3214 18. Updated exipick to current release from John Jetmore.
3216 19. Allow an empty sender to be matched against a lookup in an address list.
3217 Previously the only cases considered were a regular expression, or an
3220 20. Exim went into a mad DNS lookup loop when doing a callout where the
3221 host was specified on the transport, if the DNS lookup yielded more than
3224 21. The RFC2047 encoding function was originally intended for short strings
3225 such as real names; it was not keeping to the 75-character limit for
3226 encoded words that the RFC imposes. It now respects the limit, and
3227 generates multiple encoded words if necessary. To be on the safe side, I
3228 have increased the buffer size for the ${rfc2047: expansion operator from
3231 22. Failure to deliver a bounce message always caused it to be frozen, even if
3232 there was an errors_to setting on the router. The errors_to setting is now
3235 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3236 canonical form (fully expanded) before being placed in
3237 $sender_host_address.
3239 24. Updated eximstats to version 1.33
3241 25. Include certificate and key file names in error message when GnuTLS fails
3242 to set them up, because the GnuTLS error message doesn't include the name
3243 of the failing file when there is a problem reading it.
3245 26. Expand error message when OpenSSL has problems setting up cert/key files.
3248 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3249 (this can affect the format of dates).
3251 28. exim_tidydb, when checking for the continued existence of a message for
3252 which it has found a message-specific retry record, was not finding
3253 messages that were in split spool directories. Consequently, it was
3254 deleting retry records that should have stayed in existence.
3256 29. eximstats updated to version 1.35
3257 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3258 1.35 - bugfix such that pie charts by volume are generated correctly
3260 30. The SPA authentication driver was not abandoning authentication and moving
3261 on to the next authenticator when an expansion was forced to fail,
3262 contradicting the general specification for all authenticators. Instead it
3263 was generating a temporary error. It now behaves as specified.
3265 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3266 (the order specifies the preference for clients). The order is now AES256,
3267 AES128, 3DES, ARCFOUR128.
3269 31. Small patch to Sieve code - explicitly set From: when generating an
3272 32. Exim crashed if a remote delivery caused a very long error message to be
3273 recorded - for instance if somebody sent an entire SpamAssassin report back
3274 as a large number of 550 error lines. This bug was coincidentally fixed by
3275 increasing the size of one of Exim's internal buffers (big_buffer) that
3276 happened as part of the Exiscan merge. However, to be on the safe side, I
3277 have made the code more robust (and fixed the comments that describe what
3280 33. Some experimental protocols are using DNS PTR records for new purposes. The
3281 keys for these records are domain names, not reversed IP addresses. The
3282 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3283 leaves it alone. Component reversal etc. now happens only for IP addresses.
3286 34. The host_aton() function is supposed to be passed a string that is known
3287 to be a valid IP address. However, in the case of IPv6 addresses, it was
3288 not checking this. This is a hostage to fortune. Exim now panics and dies
3289 if the condition is not met. A case was found where this could be provoked
3290 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3291 components; fortuitously, this particular loophole had already been fixed
3292 by change 4.50/55 or 4.44/33 above.
3294 If there are any other similar loopholes, the new check in host_aton()
3295 itself should stop them being exploited. The report I received stated that
3296 data on the command line could provoke the exploit when Exim was running as
3297 exim, but did not say which command line option was involved. All I could
3298 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3299 running as the user.
3302 35. There was a buffer overflow vulnerability in the SPA authentication code
3303 (which came originally from the Samba project). I have added a test to the
3304 spa_base64_to_bits() function which I hope fixes it.
3307 36. The daemon start-up calls getloadavg() while still root for those OS that
3308 need the first call to be done as root, but it missed one case: when
3309 deliver_queue_load_max is set with deliver_drop_privilege. This is
3310 necessary for the benefit of the queue runner, because there is no re-exec
3311 when deliver_drop_privilege is set.
3313 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3314 was in use, was not putting the data itself into the right store pool;
3315 consequently, it could be overwritten for a subsequent message in the same
3316 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3319 38. Sometimes the final signoff response after QUIT could fail to get
3320 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3321 < 0 before doing a fflush(). This bug looks as though it goes back to the
3322 introduction of TLS in release 3.20, but "sometimes" must have been rare
3323 because the tests only now provoked it.
3329 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3330 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3331 Now it returns an integer. A number of calls were still expecting a T/F
3332 return. Fortuitously, in all cases, the tests worked in OK situations,
3333 which is the norm. However, things would have gone wrong on any write
3334 failures on the smtp file descriptor. This function is used when sending
3335 messages over SMTP and also when doing verify callouts.
3337 2. When Exim is called to do synchronous delivery of a locally submitted
3338 message (the -odf or -odi options), it no longer closes stderr before doing
3341 3. Implemented the mua_wrapper option.
3343 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3345 5. Implemented the functions header_remove(), header_testname(),
3346 header_add_at_position(), and receive_remove_recipient(), and exported them
3349 6. If an ACL "warn" statement specified the addition of headers, Exim already
3350 inserted X-ACL-Warn: at the start if there was no header name. However, it
3351 was not making this test for the second and subsequent header lines if
3352 there were newlines in the string. This meant that an invalid header could
3353 be inserted if Exim was badly configured.
3355 7. Allow an ACL "warn" statement to add header lines at the start or after all
3356 the Received: headers, as well as at the end.
3358 8. Added the rcpt_4xx retry error code.
3360 9. Added postmaster_mailfrom=xxx to callout verification option.
3362 10. Added mailfrom=xxxx to the callout verification option, for verify=
3365 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3366 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3368 12. Inserted some casts to stop certain compilers warning when using pointer
3369 differences as field lengths or precisions in printf-type calls (mostly
3370 affecting debugging statements).
3372 13. Added optional readline() support for -be (dynamically loaded).
3374 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3375 same clock tick as a message's arrival, so that its received time was the
3376 same as the "first fail" time on the retry record, and that message
3377 remained on the queue past the ultimate address timeout, every queue runner
3378 would try a delivery (because it was past the ultimate address timeout) but
3379 after another failure, the ultimate address timeout, which should have then
3380 bounced the address, did not kick in. This was a "< instead of <=" error;
3381 in most cases the first failure would have been in the next clock tick
3382 after the received time, and all would be well.
3384 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3385 being recognized when the domain list was tested by the match_domain
3386 condition in an expansion string.
3388 16. Added the ${str2b64: operator.
3390 17. Exim was always calling setrlimit() to set a large limit for the number of
3391 processes, without checking whether the existing limit was already
3392 adequate. (It did check for the limit on file descriptors.) Furthermore,
3393 errors from getrlimit() and setrlimit() were being ignored. Now they are
3394 logged to the main and panic logs, but Exim does carry on, to try to do its
3395 job under whatever limits there are.
3397 18. Imported PCRE 5.0.
3399 19. Trivial typo in log message " temporarily refused connection" (the leading
3402 20. If the log selector return_path_on_delivery was set and an address was
3403 redirected to /dev/null, the delivery process crashed because it assumed
3404 that a return path would always be set for a "successful" delivery. In this
3405 case, the whole delivery is bypassed as an optimization, and therefore no
3408 21. Internal re-arrangement: the function for sending a challenge and reading
3409 a response while authentication was assuming a zero-terminated challenge
3410 string. It's now changed to take a pointer and a length, to allow for
3411 binary data in such strings.
3413 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3415 23. Exim was not respecting finduser_retries when seeking the login of the
3416 uid under which it was called; it was always trying 10 times. (The default
3417 setting of finduser_retries is zero.) Also, it was sleeping after the final
3418 failure, which is pointless.
3420 24. Implemented tls_on_connect_ports.
3422 25. Implemented acl_smtp_predata.
3424 26. If the domain in control=submission is set empty, Exim assumes that the
3425 authenticated id is a complete email address when it generates From: or
3426 Sender: header lines.
3428 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3429 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3430 chown and chgrp in /bin and hostname in /usr/bin.
3432 28. Exim was keeping the "process log" file open after each use, just as it
3433 does for the main log. This opens the possibility of it remaining open for
3434 long periods when the USR1 signal hits a daemon. Occasional processlog
3435 errors were reported, that could have been caused by this. Anyway, it seems
3436 much more sensible not to leave this file open at all, so that is what now
3439 29. The long-running daemon process does not normally write to the log once it
3440 has entered its main loop, and it closes the log before doing so. This is
3441 so that log files can straightforwardly be renamed and moved. However,
3442 there are a couple of unusual error situations where the daemon does write
3443 log entries, and I had neglected to close the log afterwards.
3445 30. The text of an SMTP error response that was received during a remote
3446 delivery was being truncated at 512 bytes. This is too short for some of
3447 the long messages that one sometimes sees. I've increased the limit to
3450 31. It is now possible to make retry rules that apply only when a message has a
3451 specific sender, in particular, an empty sender.
3453 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3454 it possible to be selective about when SMTP synchronization is enforced.
3456 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3458 32. Implemented hosts_connection_nolog.
3460 33. Added an ACL for QUIT.
3462 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3465 35. Added mailbox_size and mailbox_filecount to appendfile.
3467 36. Added control = no_multiline_responses to ACLs.
3469 37. There was a bug in the logic of the code that waits for the clock to tick
3470 in the case where the clock went backwards by a substantial amount such
3471 that the microsecond fraction of "now" was more than the microsecond
3472 fraction of "then" (but the whole seconds number was less).
3474 38. Added support for the libradius Radius client library this is found on
3475 FreeBSD (previously only the radiusclient library was supported).
3481 1. When certain lookups returned multiple values in the form name=value, the
3482 quoting of the values was not always being done properly. Specifically:
3483 (a) If the value started with a double quote, but contained no whitespace,
3485 (b) If the value contained whitespace other than a space character (i.e.
3486 tabs or newlines or carriage returns) it was not quoted.
3487 This fix has been applied to the mysql and pgsql lookups by writing a
3488 separate quoting function and calling it from the lookup code. The fix
3489 should probably also be applied to nisplus, ibase and oracle lookups, but
3490 since I cannot test any of those, I have not disturbed their existing code.
3492 2. A hit in the callout cache for a specific address caused a log line with no
3493 reason for rejecting RCPT. Now it says "Previous (cached) callout
3494 verification failure".
3496 3. There was an off-by-one bug in the queryprogram router. An over-long
3497 return line was truncated at 256 instead of 255 characters, thereby
3498 overflowing its buffer with the terminating zero. As well as fixing this, I
3499 have increased the buffer size to 1024 (and made a note to document this).
3501 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3502 when Exim is waiting for an SMTP response from a remote server, Exim
3503 restarts its select() call on the socket, thereby resetting its timeout.
3504 This is not a problem when such interrupts are rare. Somebody set up a cron
3505 job to run exiwhat every 2 minutes, which is less than the normal select()
3506 timeout (5 or 10 minutes). This meant that the select() timeout never
3507 kicked in because it was always reset. I have fixed this by comparing the
3508 time when an interrupt arrives with the time at the start of the first call
3509 to select(). If more time than the timeout has elapsed, the interrupt is
3510 treated as a timeout.
3512 5. Some internal re-factoring in preparation for the addition of Sieve
3513 extensions (by MH). In particular, the "personal" test is moved to a
3514 separate function, and given an option for scanning Cc: and Bcc: (which is
3515 not set for Exim filters).
3517 6. When Exim created an email address using the login of the caller as the
3518 local part (e.g. when creating a From: or Sender: header line), it was not
3519 quoting the local part when it contained special characters such as @.
3521 7. Installed new OpenBSD configuration files.
3523 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3524 try to make them clearer.
3526 9. Callout options, other than the timeout value, were being ignored when
3527 verifying sender addresses in header lines. For example, when using
3529 verify = header_sender/callout=no_cache
3531 the cache was (incorrectly) being used.
3533 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3534 only the Cygwin environment.
3536 11. When return_path_on_delivery was set as a log selector, if different remote
3537 addresses in the same message used different return paths and parallel
3538 remote delivery occurred, the wrong values would sometimes be logged.
3539 (Whenever a remote delivery process finished, the return path value from
3540 the most recently started remote delivery process was logged.)
3542 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3543 lines when AUTH and/or TLS are in use. This is the "received protocol"
3544 field. Exim used to use "asmtp" for authenticated SMTP, without any
3545 indication (in the protocol name) for TLS use. Now it follows the RFC and
3546 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3547 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3548 names appear in log lines as well as in Received: header lines.
3550 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3551 extensions, and comparison tests, and to fix some bugs.
3553 14. Changes to the "personal" filter test:
3555 (1) The test was buggy in that it was just doing the equivalent of
3556 "contains" tests on header lines. For example, if a user's address was
3557 anne@some.where, the "personal" test would incorrectly be true for
3559 To: susanne@some.where
3561 This test is now done by extracting each address from the header in turn,
3562 and checking the entire address. Other tests that are part of "personal"
3563 are now done using regular expressions (for example, to check local parts
3564 of addresses in From: header lines).
3566 (2) The list of non-personal local parts in From: addresses has been
3567 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3568 taken from the Sieve specification recommendations.
3570 (3) If the message contains any header line starting with "List-" it is
3571 treated as non-personal.
3573 (4) The test for "circular" in the Subject: header line has been removed
3574 because it now seems ill-conceived.
3576 15. Minor typos in src/EDITME comments corrected.
3578 16. Installed latest exipick from John Jetmore.
3580 17. If headers_add on a router specified a text string that was too long for
3581 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3582 of string_sprintf() is now avoided.
3584 18. $message_body_size was not set (it was always zero) when running the DATA
3585 ACL and the local_scan() function.
3587 19. For the "mail" command in an Exim filter, no default was being set for
3588 the once_repeat time, causing a random time value to be used if "once" was
3589 specified. (If the value happened to be <= 0, no repeat happened.) The
3590 default is now 0s, meaning "never repeat". The "vacation" command was OK
3591 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3592 (I found it when inspecting the code).
3594 20. There is now an overall timeout for performing a callout verification. It
3595 defaults to 4 times the callout timeout, which applies to individual SMTP
3596 commands during the callout. The overall timeout applies when there is more
3597 than one host that can be tried. The timeout is checked before trying the
3598 next host. This prevents very long delays if there are a large number of
3599 hosts and all are timing out (e.g. when the network connections are timing
3600 out). The value of the overall timeout can be changed by specifying an
3601 additional sub-option for "callout", called "maxwait". For example:
3603 verify = sender/callout=5s,maxwait=20s
3605 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3606 to the end before writing, but this should make it even safer).
3608 22. Exim was forgetting that it had advertised PIPELINING for the second and
3609 subsequent messages on an SMTP connection. It was also not resetting its
3610 memory on STARTTLS and an internal HELO.
3612 23. When Exim logs an SMTP synchronization error within a session, it now
3613 records whether PIPELINING has been advertised or not.
3615 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3616 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3617 rather than long int.
3619 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3621 26. Added the never_mail option to autoreply.
3627 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3628 crash if the getsockname() call failed; this can happen if a connection is
3629 closed very soon after it is established. The problem was simply in the
3630 order in which certain operations were done, causing Exim to try to write
3631 to the SMTP stream before it had set up the file descriptor. The bug has
3632 been fixed by making things happen in the correct order.
3638 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3639 before the connection was closed, thus losing the rejection response.
3641 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3642 some early Solaris releases, but causes trouble in current releases where
3643 socklen_t is defined.
3645 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3648 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3651 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3652 that releasing the top of it at the end releases what was used for sub-
3653 expansions (unless the block got too big). However, discard this block if
3654 the first thing is a variable or header, so that we can use its block when
3655 it is dynamic (useful for very large $message_headers, for example).
3657 6. Lookups now cache *every* query, not just the most recent. A new, separate
3658 store pool is used for this. It can be recovered when all lookup caches are
3659 flushed. Lookups now release memory at the end of their result strings.
3660 This has involved some general refactoring of the lookup sources.
3662 7. Some code has been added to the store_xxx() functions to reduce the amount
3663 of flapping under certain conditions.
3665 8. log_incoming_interface used to affect only the <= reception log lines. Now
3666 it causes the local interface and port to be added to several more SMTP log
3667 lines, for example "SMTP connection from", and rejection lines.
3669 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3671 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3673 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3674 could be overwritten at the end of the current message (or the start of a
3675 new message if it was set in a HELO ACL). The value is now preserved for
3676 the duration of the SMTP connection.
3678 12. If a transport had a headers_rewrite setting, and a matching header line
3679 contained an unqualified address, that address was qualified, even if it
3680 did not match any rewriting rules. The underlying bug was that the values
3681 of the flags that permit the existence of unqualified sender and recipient
3682 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3683 non-local messages, and by -bnq for local messages) were not being
3684 preserved with the message after it was received.
3686 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3687 "next input=" as part of the text comprising the host identity instead of
3688 the correct text. The code was using the same buffer for two different
3689 strings. However, depending on which order the printing function evaluated
3690 its arguments, the bug did not always show up. Under Linux, for example, my
3691 test suite worked just fine.
3693 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3694 doesn't work with some older versions of Perl. It has been changed to "my",
3695 which in any case is probably the better facility to use.
3697 15. A really picky compiler found some instances of statements for creating
3698 error messages that either had too many or two few arguments for the format
3701 16. The size of the buffer for calls to the DNS resolver has been increased
3702 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3703 for addresses that have a lot of PTR records. This alleviates a problem; it
3704 does not fully solve it.
3706 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3707 buffer now truncates the list and logs the incident, which is the same
3708 action as happens when Exim is looking up a host name and its aliases.
3709 Previously in this situation something unpredictable would happen;
3710 sometimes it was "internal error: store_reset failed".
3712 18. If a server dropped the connection unexpectedly when an Exim client was
3713 using GnuTLS and trying to read a response, the client delivery process
3714 crashed while trying to generate an error log message.
3716 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3717 string, for example:
3719 warn message = H1: something\nH2: something
3721 the text was added as a single header line from Exim's point of view
3722 though it ended up OK in the delivered message. However, searching for the
3723 second and subsequent header lines using $h_h2: did not work. This has been
3724 fixed. Similarly, if a system filter added multiple headers in this way,
3725 the routers could not see them.
3727 20. Expanded the error message when iplsearch is called with an invalid key to
3728 suggest using net-iplsearch in a host list.
3730 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3731 ACLs are no longer actually imposed (and a message to that effect is
3734 22. If a "gecos" field in a passwd entry contained escaped characters, in
3735 particular, if it contained a \" sequence, Exim got it wrong when building
3736 a From: or a Sender: header from that name. A second bug also caused
3737 incorrect handling when an unquoted " was present following a character
3738 that needed quoting.
3740 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3741 was not being matched caselessly.
3743 24. Arranged for all hyphens in the exim.8 source to be escaped with
3746 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3747 properly. Recipient callout cache records were still being keyed to include
3748 the sender, even when use_sender was set false. This led to far more
3749 callouts that were necessary. The sender is no longer included in the key
3750 when use_sender is false.
3752 26. Added "control = submission" modifier to ACLs.
3754 27. Added the ${base62d: operator to decode base 62 numbers.
3756 28. dnsdb lookups can now access SRV records.
3758 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3759 the configuration file.
3761 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3762 (-v) mode. This makes the output for a verbose queue run more intelligible.
3764 31. Added a use_postmaster feature to recipient callouts.
3766 32. Added the $body_zerocount variable, containing the number of binary zero
3767 bytes in the message body.
3769 33. The time of last modification of the "new" subdirectory is now used as the
3770 "mailbox time last read" when there is a quota error for a maildir
3773 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3775 35. Added +ignore_unknown as a special item in host lists.
3777 36. Code for decoding IPv6 addresses in host lists is now included, even if
3778 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3779 address was recognized as an IP address, but was then not correctly decoded
3780 into binary, causing unexpected and incorrect effects when compared with
3787 1. Very minor rewording of debugging text in manualroute to say "list of
3788 hosts" instead of "hostlist".
3790 2. If verify=header_syntax was set, and a header line with an unqualified
3791 address (no domain) and a large number of spaces between the end of the
3792 name and the colon was received, the reception process suffered a buffer
3793 overflow, and (when I tested it) crashed. This was caused by some obsolete
3794 code that should have been removed. The fix is to remove it!
3796 3. When running in the test harness, delay a bit after writing a bounce
3797 message to get a bit more predictability in the log output.
3799 4. Added a call to search_tidyup() just before forking a reception process. In
3800 theory, someone could use a lookup in the expansion of smtp_accept_max_
3801 per_host which, without the tidyup, could leave open a database connection.
3803 5. Added the variables $recipient_data and $sender_data which get set from a
3804 lookup success in an ACL "recipients" or "senders" condition, or a router
3805 "senders" option, similar to $domain_data and $local_part_data.
3807 6. Moved the writing of debug_print from before to after the "senders" test
3810 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3811 problems for message scanning, either using a data ACL, or using
3812 local_scan() because the Received: header was not generated till after they
3813 were called (in order to set the time as the time of reception completion).
3814 I have revised the way this works. The header is now generated after the
3815 body is received, but before the ACL or local_scan() are called. After they
3816 are run, the timestamp in the header is updated.
3822 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3823 before starting a queue runner without re-exec. This happened only when
3824 deliver_drop_privilege was set or when the Exim user was set to root. The
3825 effect of the bug was that timeouts during subsequent deliveries caused
3826 crashes instead of being properly handled. The handler is now left at its
3827 default (and expected) setting.
3829 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3830 message, again when deliver_drop_privilege is set or Exim is run as root.
3831 The bug described in (1) was not present in this case, but the tidying up
3832 of the other signals was missing. I have made the two cases consistent.
3834 3. The ignore_target_hosts setting on a manualroute router was being ignored
3835 for hosts that were looked up using the /MX notation.
3837 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3840 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3841 operated on the sender address. After changing the $sender_address to <>
3842 for the sender address verify, Exim was re-instated it as the original
3843 (before rewriting) address, but remembering that it had rewritten it, so it
3844 wasn't rewriting it again. This bug also had the effect of breaking the
3845 sender address verification caching when the sender address was rewritten.
3847 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3848 This has been changed so that if the ip literal address matches
3849 ignore_target_hosts, the router declines.
3851 7. Added expansion conditions match_domain, match_address, and match_local_
3852 part (NOT match_host).
3854 8. The placeholder for the Received: header didn't have a length field set.
3856 9. Added code to Exim itself and to exim_lock to test for a specific race
3857 condition that could lead to file corruption when using MBX delivery. The
3858 issue is with the lockfile that is created in /tmp. If this file is removed
3859 after a process has opened it but before that process has acquired a lock,
3860 there is the potential for a second process to recreate the file and also
3861 acquire a lock. This could lead to two Exim processes writing to the file
3862 at the same time. The added code performs the same test as UW imapd; it
3863 checks after acquiring the lock that its file descriptor still refers to
3864 the same named file.
3866 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3867 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3870 11. Added the smtp_active_hostname option. If used, this will typically be made
3871 to depend on the incoming interface address. Because $interface_address is
3872 not set up until the daemon has forked a reception process, error responses
3873 that can happen earlier (such as "too many connections") no longer contain
3876 12. If an expansion in a condition on a "warn" statement fails because a lookup
3877 defers, the "warn" statement is abandoned, and the next ACL statement is
3878 processed. Previously this caused the whole ACL to be aborted.
3880 13. Added the iplsearch lookup type.
3882 14. Added ident_timeout as a log selector.
3884 15. Added tls_certificate_verified as a log selector.
3886 16. Added a global option tls_require_ciphers (compare the smtp transport
3887 option of the same name). This controls incoming TLS connections.
3889 17. I finally figured out how to make tls_require_ciphers do a similar thing
3890 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3891 before starting the TLS session.
3893 18. Tabs are now shown as \t in -bP output.
3895 19. If the log selector return_path_on_delivery was set, Exim crashed when
3896 bouncing a message because it had too many Received: header lines.
3898 20. If two routers both had headers_remove settings, and the first one included
3899 a superfluous trailing colon, the final name in the first list and the
3900 first name in the second list were incorrectly joined into one item (with a
3901 colon in the middle).
3907 1. Added -C and -D options to the exinext utility, mainly to make it easier
3908 to include in the automated testing, but these could be helpful when
3909 multiple configurations are in use.
3911 2. The exinext utility was not formatting the output nicely when there was
3912 an alternate port involved in the retry record key, nor when there was a
3913 message id as well (for retries that were specific to a specific message
3914 and a specific host). It was also confused by IPv6 addresses, because of
3915 the additional colons they contain. I have fixed the IPv4 problem, and
3916 patched it up to do a reasonable job for IPv6.
3918 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3919 delivery, the log line now contains "pipelined" if PIPELINING was used.
3921 4. An SMTP transport process used to panic and die if the bind() call to set
3922 an explicit outgoing interface failed. This has been changed; it is now
3923 treated in the same way as a connect() failure.
3925 5. A reference to $sender_host_name in the part of a conditional expansion
3926 that was being skipped was still causing a DNS lookup. This no longer
3929 6. The def: expansion condition was not recognizing references to header lines
3930 that used bh_ and bheader_.
3932 7. Added the _cache feature to named lists.
3934 8. The code for checking quota_filecount in the appendfile transport was
3935 allowing one more file than it should have been.
3937 9. For compatibility with Sendmail, the command line option
3945 and sets the incoming protocol and host name (for trusted callers). The
3946 host name and its colon can be omitted when only the protocol is to be set.
3947 Note the Exim already has two private options, -pd and -ps, that refer to
3948 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3949 "s", but I don't think that's a major issue.
3951 10. A number of refactoring changes to the code, none of which should affect
3954 (a) The number of logging options was getting close to filling up the
3955 32-bit word that was used as a bit map. I have split them into two classes:
3956 those that are passed in the argument to log_write(), and those that are
3957 only ever tested independently outside of that function. These are now in
3958 separate 32-bit words, so there is plenty of room for expansion again.
3959 There is no change in the user interface or the logging behaviour.
3961 (b) When building, for example, log lines, the code previously used a
3962 macro that called string_cat() twice, in order to add two strings. This is
3963 not really sufficiently general. Furthermore, there was one instance where
3964 it was actually wrong because one of the argument was used twice, and in
3965 one call a function was used. (As it happened, calling the function twice
3966 did not affect the overall behaviour.) The macro has been replaced by a
3967 function that can join an arbitrary number of extra strings onto a growing
3970 (c) The code for expansion conditions now uses a table and a binary chop
3971 instead of a serial search (which was left over from when there were very
3972 few conditions). Also, it now recognizes conditions like "pam" even when
3973 the relevant support is not compiled in: a suitably worded error message is
3974 given if an attempt is made to use such a condition.
3976 11. Added ${time_interval:xxxxx}.
3978 12. A bug was causing one of the ddress fields not to be passed back correctly
3979 from remote delivery subprocesses. The field in question was not being
3980 subsequently used, so this caused to problems in practice.
3982 13. Added new log selectors queue_time and deliver_time.
3984 14. Might have fixed a bug in maildirsizefile handling that threw up
3985 "unexpected character" debug warnings, and recalculated the data
3986 unnecessarily. In any case, I expanded the warning message to give more
3989 15. Added the message "Restricted characters in address" to the statements in
3990 the default ACL that block characters like @ and % in local parts.
3992 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3993 Three changes have been made:
3995 (a) There was a serious bug; a negative response to MAIL caused the whole
3996 recipient domain to be cached as invalid, thereby blocking all messages
3997 to all local parts at the same domain, from all senders. This bug has
3998 been fixed. The domain is no longer cached after a negative response to
3999 MAIL if the sender used is not empty.
4001 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4004 (c) A new callout option, "use_sender" has been added for people who want
4005 the modified behaviour.
4011 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4014 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4015 indeed breaks things for older releases.
4017 3. Added additional logging to the case where there is a problem reading data
4018 from a filter that is running in a subprocess using a pipe, in order to
4019 try to track down a specific problem.
4021 4. Testing facility fudge: when running in the test harness and attempting
4022 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4023 getting "No route to host". Convert this to a timeout.
4025 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4028 6. Some OS don't have socklen_t but use size_t instead. This affects the
4029 fifth argument of getsockopt() amongst other things. This is now
4030 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4031 can be set for individual OS. I have set it for SunOS5, OSF1, and
4032 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4033 some earlier ones do not.
4035 7. Change 4.30/15 was not doing the test caselessly.
4037 8. The standard form for an IPv6 address literal was being rejected by address
4038 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4039 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4040 this, as well as the form without the "IPv6" on the front (but only when
4041 address literals are enabled, of course).
4043 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4045 10. Exim crashed if a message with an empty sender address specified by -f
4046 encountered a router with an errors_to setting. This could be provoked only
4047 by a command such as
4051 where an empty string was supplied; "<>" did not hit this bug.
4053 11. Installed PCRE release 4.5.
4055 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4056 remained set. It is now erased.
4058 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4059 times from message ids (which are base 36 rather than the normal 62).
4061 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4062 were being counted as actual protocol errors, and logged if the log
4063 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4064 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4065 and DATA following a set of rejected RCPTs do not count as protocol errors.
4066 In other words, Exim assumes they were pipelined, though this may not
4067 actually be the case. Of course, in all cases the client gets an
4068 appropriate error code.
4070 15. If a lookup fails in an ACL condition, a message about the failure may
4071 be available; it is used if testing the ACL cannot continue, because most
4072 such messages specify what the cause of the deferral is. However, some
4073 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4074 that caused an old message to be retained and used if a later statement
4075 caused a defer, replacing the real cause of the deferral.
4077 16. If an IP address had so many PTR records that the DNS lookup buffer
4078 was not large enough to hold them, Exim could crash while trying to process
4079 the truncated data. It now detects and logs this case.
4081 17. Further to 4.21/58, another change has been made: if (and only if) the
4082 first line of a message (the first header line) ends with CRLF, a bare LF
4083 in a subsequent header line has a space inserted after it, so as not to
4084 terminate the header.
4086 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4087 unnecessarily, used atoi() instead of strtol(), and didn't check the
4088 termination when getting file sizes from file names by regex.
4090 19. Completely re-implemented the support for maildirsize files, in the light
4091 of a number of problems with the previous contributed implementation
4092 (4.30/29). In particular:
4094 . If the quota is zero, the maildirsize file is maintained, but no quota is
4097 . If the maildir directory does not exist, it is created before any attempt
4098 to write a maildirsize file.
4100 . The quota value in the file is just a cache; if the quota is changed in
4101 the transport, the new value overrides.
4103 . A regular expression is available for excluding directories from the
4106 20. The autoreply transport checks the characters in options that define the
4107 message's headers; it allows continued headers, but it was checking with
4108 isspace() after an embedded newline instead of explicitly looking for a
4111 21. If all the "regular" hosts to which an address was routed had passed their
4112 expiry times, and had not reached their retry times, the address was
4113 bounced, even if fallback hosts were defined. Now Exim should go on to try
4116 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4117 equivalent code in the SMTP transport. Some hosts send humungous responses
4118 to HELO/EHLO, more than 1024 it seems.
4120 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4121 gives compiler warnings in some environments. I've now done it "properly",
4124 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4125 (because of problems with the built-in one) was declared to return uschar *
4126 instead of char *, causing compiler failure.
4128 25. Fixed a file descriptor leak when processing alias/forward files.
4130 26. Fixed a minor format string issue in dbfn.c.
4132 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4134 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4135 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4138 29. When a sender address is verified, it is cached, to save repeating the test
4139 when there is more than one recipient in a message. However, when the
4140 verification involves a callout, it is possible for different callout
4141 options to be set for different recipients. It is too complicated to keep
4142 track of this in the cache, so now Exim always runs a verification when a
4143 callout is required, relying on the callout cache for the optimization.
4144 The overhead is duplication of the address routing, but this should not be
4147 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4148 to be verified with callout=postmaster, and the main callout worked but the
4149 postmaster check failed, the verification correctly failed. However, if a
4150 subsequent RCPT command asked for sender verification *without* the
4151 postmaster check, incorrect caching caused this verification also to fail,
4154 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4155 it was not caching the DNS options (qualify_single, search_parents) that
4156 were used when the lookup failed. A subsequent lookup with different
4157 options therefore always gave the same answer, though there were cases
4158 where it should not have. (Example: a "domains = !$mx_any" option on a
4159 dnslookup router: the "domains" option is always processed without any
4160 widening, but the router might have qualify_single set.) Now Exim uses the
4161 cached value only when the same options are set.
4163 32. Added John Jetmore's "exipick" utility to the distribution.
4165 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4166 than a timeout (e.g. a certificate is required, and is not provided), an
4167 Exim server now closes the connection immediately. Previously it waited for
4168 the client to close - but if the client is SSL, it seems that they each
4169 wait for each other, leading to a delay before one of them times out.
4171 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4172 maintained 0.8.x compatibility because I don't think many are using it, and
4173 it is clearly obsolete.
4175 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4178 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4179 client certificate was expired. A simple patch fixes this, though I don't
4180 understand the full logic of why the verify callback is called multiple
4183 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4184 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4185 which causes problems with some clients (such as the Certicom SSL Plus
4186 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4187 disables the coutermeasure allowing Eudora to connect."
4189 38. Exim was not checking that a write() to a log file succeeded. This could
4190 lead to Bad Things if a log got too big, in particular if it hit a file
4191 size limit. Exim now panics and dies if it cannot write to a log file, just
4192 as it does if it cannot open a log file.
4194 39. Modified OS/Makefile-Linux so that it now contains
4196 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4198 The two -D definitions ensure that Exim is compiled with large file
4199 support, which makes it possible to handle log files that are bigger than
4202 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4203 instance) a domain was checked against a named list that involved a lookup,
4204 causing $domain_data to be set, then another domain was checked against the
4205 same list, then the first domain was re-checked, the value of $domain_data
4206 after the final check could be wrong. In particular, if the second check
4207 failed, it could be set empty. This bug probably also applied to
4210 41. The strip_trailing_dot option was not being applied to the address given
4211 with the -f command-line option.
4213 42. The code for reading a message's header from the spool was incrementing
4214 $received_count, but never initializing it. This meant that the value was
4215 incorrect (doubled) while delivering a message in the same process in which
4216 it was received. In the most common configuration of Exim, this never
4217 happens - a fresh exec is done - but it can happen when
4218 deliver_drop_privilege is set.
4220 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4221 it now includes up to 150 characters of the unexpected data in the log
4224 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4225 and building data strings. The size of both of these buffers was 10 000
4226 bytes - far larger than anybody would *ever* want, thought I. Needless to
4227 say, somebody hit the limit. I have increased the maximum line length to
4228 20 000 and the maximum data length of concatenated lines to 100 000. I have
4229 also fixed two bugs, because there was no checking on these buffers. Tsk,
4230 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4231 buffer is too small.
4233 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4234 lsearch lookups. Now it does.
4236 46. When parsing a route_list item in a manualroute router, a fixed-length
4237 buffer was used for the list of hosts. I made this 1024 bytes long,
4238 thinking that nobody would ever have a list of hosts that long. Wrong.
4239 Somebody had a whole pile of complicated expansion conditions, and the
4240 string was silently truncated, leading to an expansion error. It turns out
4241 that it is easier to change to an unlimited length (owing to other changes
4242 that have happened since this code was originally written) than to build
4243 structure for giving a limitation error. The length of the item that
4244 expands into the list of hosts is now unlimited.
4246 47. The lsearch lookup could not handle data where the length of text line was
4247 more than 4095 characters. Such lines were truncated, leading to shortened
4248 data being returned. It should now handle lines of any length.
4250 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4251 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4254 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4255 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4256 no longer happens in comment lines. A list of replacements is now placed
4257 at the head of all of the source files, except those whose only change is
4258 to replace PERL_COMMAND in the very first #! line.
4260 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4261 messages on the queue, with a bottom-up merge sort, using code contributed
4262 by Michael Haardt. This should make operations like -bp somewhat faster on
4263 large queues. It won't affect queue runners, except when queue_run_in_order
4266 51. Installed eximstats 1.31 in the distribution.
4268 52. Added support for SRV lookups to the dnslookup router.
4270 53. If an ACL referred to $message_body or $message_body_end, the value was not
4271 reset for any messages that followed in the same SMTP session.
4273 54. The store-handling optimization for building very long strings was not
4274 differentiating between the different store pools. I don't think this
4275 actually made any difference in practice, but I've tidied it.
4277 55. While running the routers to verify a sender address, $sender_address
4278 was still set to the sender address. This is wrong, because when routing to
4279 send a bounce to the sender, it would be empty. Therefore, I have changed
4280 it so that, while verifying a sender address, $sender_address is set to <>.
4281 (There is no change to what happens when verifying a recipient address.)
4283 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4284 target A or AAAA records (if not already returned) without resetting the
4285 qualify_single or search_parents options of the DNS resolver. These are
4286 inappropriate in this case because the targets of MX and SRV records must
4287 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4288 target that, when qualified, matched something in the local domain. These
4289 two options are now turned off when doing these lookups.
4291 57. It seems that at least some releases of Reiserfs (which does not have the
4292 concept of a fixed number of inodes) returns zero and not -1 for the
4293 number of available inodes. This interacted badly with check_spool_inodes,
4294 which assumed that -1 was the "no such thing" setting. What I have done is
4295 to check that the total number of inodes is greater than zero before doing
4296 the test of how many are available.
4298 58. When a "warn" ACL statement has a log_message modifier, the message is
4299 remembered, and not repeated. This is to avoid a lot of repetition when a
4300 message has many recipients that cause the same warning to be written.
4301 Howewer, Exim was preserving the list of already written lines for an
4302 entire SMTP session, which doesn't seem right. The memory is now reset if a
4303 new message is started.
4305 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4306 debugging output unless log_rewrite was also set.
4308 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4309 of (int)(handle) when we know that handle contains (void *)(-1).
4311 61. The Exim daemon panic-logs an error return when it closes the incoming
4312 connection. However "connection reset by peer" seems to be common, and
4313 isn't really an error worthy of noting specially, so that particular error
4316 62. When Exim is trying to find all the local interfaces, it used to panic and
4317 die if the ioctl to get the interface flags failed. However, it seems that
4318 on at least one OS (Solaris 9) it is possible to have an interface that is
4319 included in the list of interfaces, but for which you get a failure error
4320 for this call. This happens when the interface is not "plumbed" into a
4321 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4322 failure of the "get flags" call assumes that the interface is down.
4324 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4325 makes life easier for people who are doing arithmetic on fields extracted
4326 from dates, where you often get leading zeros that should not be
4327 interpreted as octal.
4329 64. Added qualify_domain to the redirect router, to override the global
4332 65. If a pathologically long header line contained very many addresses (the
4333 report of this problem mentioned 10 000) and each of them was rewritten,
4334 Exim could use up a very large amount of memory. (It kept on making new
4335 copies of the header line as it rewrote, and never released the old ones.)
4336 At the expense of a bit more processing, the header rewriting function has
4337 been changed so that it no longer eats memory in this way.
4339 66. The generation of the Received: header has been moved from the time that a
4340 message starts to be received, to the time that it finishes. The timestamp
4341 in the Received: header should now be very close to that of the <= log
4342 line. There are two side-effects of this change:
4344 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4345 logged header lines no longer include the local Received: line, because
4346 it has not yet been created. The same applies to a copy of the message
4347 that is returned to a non-SMTP sender when a message is rejected.
4349 (b) When a filter file is tested using -bf, no additional Received: header
4350 is added to the test message. After some thought, I decided that this
4353 This change does not affect the value of $received_for. It is still set
4354 after address rewriting, but before local_scan() is called.
4356 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4358 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4359 gave an unhelpful panic error message, and a defer error. I have managed to
4360 change this behaviour so that it now rejects any supplied certificate,
4361 which seems right, as the list of acceptable certificates is empty.
4363 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4364 gave an unhelpful defer error. I have not managed to make this reject any
4365 supplied certificates, but the error message it gives is "no certificate
4366 supplied", which is not helpful.
4368 70. exigrep's output now also includes lines that are not associated with any
4369 message, but which match the given pattern. Implemented by a patch from
4370 Martin Sluka, which also tidied up the Perl a bit.
4372 71. Recipient callout verification, like sender verification, was using <> in
4373 the MAIL FROM command. This isn't really the right thing, since the actual
4374 sender may affect whether the remote host accepts the recipient or not. I
4375 have changed it to use the actual sender in the callout; this means that
4376 the cache record is now keyed on a recipient/sender pair, not just the
4377 recipient address. There doesn't seem to be a real danger of callout loops,
4378 since a callout by the remote host to check the sender would use <>.
4379 [SEE ABOVE: changed after hitting problems.]
4381 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4382 temporary errors. However, in the case of such a code being given after
4383 the end of a data transmission (i.e. after ".") Exim was failing to write
4384 a retry record for the message. (Yes, there was some broken host that was
4385 actually sending 8xx at this point.)
4387 73. An unknown lookup type in a host list could cause Exim to panic-die when
4388 the list was checked. (An example that provoked this was putting <; in the
4389 middle of a list instead of at the start.) If this happened during a DATA
4390 ACL check, a -D file could be left lying around. This kind of configuration
4391 error no longer causes Exim to die; instead it causes a defer errror. The
4392 incident is still logged to the main and panic logs.
4394 74. Buglet left over from Exim 3 conversion. The message "too many messages
4395 in one connection" was written to the rejectlog but not the mainlog, except
4396 when address rewriting (yes!) was being logged.
4398 75. Added write_rejectlog option.
4400 76. When a system filter was run not as root (that is, when system_filter_user
4401 was set), the values of the $n variables were not being returned to the
4402 main process; thus, they were not subsequently available in the $sn
4405 77. Added +return_path_on_delivery log selector.
4407 78. A connection timeout was being treated differently from recipients deferred
4408 when testing hosts_max_try with a message that was older than the host's
4409 retry timeout. (The host should not be counted, thus allowing all hosts to
4410 be tried at least once before bouncing.) This may have been the cause of an
4411 occasionally reported bug whereby a message would remain on the queue
4412 longer than the retry timeout, but would be bounced if a delivery was
4413 forced. I say "may" because I never totally pinned down the problem;
4414 setting up timeout/retry tests is difficult. See also the next item.
4416 79. The ultimate address timeout was not being applied to errors that involved
4417 a combination of host plus message (for example, a timeout on a MAIL
4418 command). When an address resolved to a number of possible hosts, and they
4419 were not all tried for each delivery (e.g. because of hosts_max_try), a
4420 message could remain on the queue longer than the retry timeout.
4422 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4425 81. Fixed an obscure SMTP outgoing bug which required at least the following
4426 conditions: (a) there was another message waiting for the same server;
4427 (b) the server returned 5xx to all RCPT commands in the first message so
4428 that the message was not completed; (c) the server dropped the connection
4429 or gave a negative response to the RSET that Exim sends to abort the
4430 transaction. The observed case was a dropped connection after DATA that had
4431 been sent in pipelining mode. That is, the server had advertised PIPELINING
4432 but was not implementing it correctly. The effect of the bug was incorrect
4433 behaviour, such as trying another host, and this could lead to a crash.
4439 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4440 and daemon.c were passed as pointers to ints; they should have been
4441 pointers to socklen_t variables (which are typically unsigned ints).
4443 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4446 3. Fixed a really odd bug that affected only the testing scheme; patching a
4447 certain fixed string in the binary changed the value of another string that
4448 happened to be identical to the end of the original first string.
4450 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4451 name", it returns that address as the IP address. On some operating
4452 systems (e.g. Solaris), it also passes back the IP address string as the
4453 "host name". However, on others (e.g. Linux), it passes back an empty
4454 string. Exim wasn't checking for this, and was changing the host name to an
4455 empty string, assuming it had been canonicized.
4457 5. Although rare, it is permitted to have more than one PTR record for a given
4458 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4459 all the names associated with an address, because they do in Solaris.
4460 However, it seems that they do not in Linux for data that comes from the
4461 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4462 I found this out when I moved to a new Linux workstation and tried to run
4463 the Exim test suite.
4465 To get round this problem I have changed the code so that it now does its
4466 own call to the DNS to look up PTR records when searching for a host name.
4467 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4468 addresses that are only in /etc/hosts are still found.
4470 This behaviour is, however, controlled by an option called host_lookup_
4471 order, which defaults to "bydns:byaddr". If people want to use the other
4472 order, or indeed, just use one or the other means of lookup, they can
4473 specify it in this variable.
4475 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4476 some operating systems, this comes back from gethostbyaddr() as an empty
4477 string, and this is what Exim used to test for. However, it seems that in
4478 other systems, "." is yielded. Exim now tests for this case too.
4480 7. The values of check_spool_space and check_log_space are now held internally
4481 as a number of kilobytes instead of an absolute number of bytes. If a
4482 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4483 kilobyte. This means that much larger values can be stored.
4485 8. Exim monitor: an attempt to get the action menu when not actually pointing
4486 at a message produces an empty menu entitled "No message selected". This
4487 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4488 no entries in it ("Shell widget menu has zero width and/or height"). So I
4489 have added a single, blank menu entry in this case.
4491 9. Added ${quote_local_part.
4493 10. MIME decoding is now applied to the contents of Subject: header lines when
4496 11. Now that a reference to $sender_host_address automatically causes a reverse
4497 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4498 host lookup before query-style lookups in lists that might use this
4499 variable. This has therefore been abolished, and the "net-" prefix is no
4500 longer necessary for query-style lookups.
4502 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4503 have been a typo for LFLAGS, so it has been changed.
4505 13. The install script calls Exim with "-C /dev/null" in order to find the
4506 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4507 to be output. Howeve, since Exim outputs its version number before the
4508 error, it didn't break the script. It just looked ugly. I fixed this by
4509 always allowing "-C /dev/null" if the caller is root.
4511 14. Ignore overlarge ACL variable number when reading spool file - insurance
4512 against a later release with more variables having written the file.
4514 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4515 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4516 this, as well as the form without the "IPv6" on the front.
4518 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4519 OS/Makefile-Darwin file.
4521 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4522 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4524 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4525 variable. However, some OS use ints for inodes. I've added cast to long int
4526 to get rid of the compiler warning.
4528 19. I had forgotten to lock out "/../" in configuration file names when
4529 ALT_CONFIG_PREFIX was set.
4531 20. Routers used for verification do not need to specify transports. However,
4532 if such a router generated a host list, and callout was configured, Exim
4533 crashed, because it could not find a port number from the (non-existent)
4534 transport. It now assumes port 25 in this circumstance.
4536 21. Added the -t option to exigrep.
4538 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4539 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4540 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4541 them is set, LOOKUP_LSEARCH is forced.
4543 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4545 24. Added sender and host information to the "rejected by local_scan()" log
4546 line; previously there was no indication of these.
4548 25. Added .include_if_exists.
4550 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4551 Linux. It turns out that not all file systems support this. Apparently some
4552 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4553 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4554 not supported on the file descriptor, is now ignored when Exim is trying to
4555 sync a directory. This applies only to Linux.
4557 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4559 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4560 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4563 29. Added support for maildirsize files from supplied patch (modified a bit).
4565 30. The use of :fail: followed by an empty string could lead Exim to respond to
4566 sender verification failures with (e.g.):
4568 550 Verification failed for <xxx>
4569 550 Sender verify failed
4571 where the first response line was missing the '-' that indicates it is not
4572 the final line of the response.
4574 31. The loop for finding the name of the user that called Exim had a hardwired
4575 limit of 10; it now uses the value of finduser_retries, which is used for
4576 all other user lookups.
4578 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4581 33. Exim was neglecting to zero errno before one call of strtol() when
4582 expanding a string and expecting an integer value. On some systems this
4583 resulted in spurious "integer overflow" errors. Also, it was casting the
4584 result into an int without checking.
4586 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4587 did not work. The code looks as if it has *never* worked, though it appears
4588 to have been documented since at least releast 1.62. I have made it work.
4590 35. The "timeout_DNS" error in retry rules, also documented since at least
4591 1.62, also never worked. As it isn't clear exactly what this means, and
4592 clearly it isn't a major issue, I have abolished the feature by treating it
4593 as "timeout", and writing a warning to the main and panic logs.
4595 36. The display of retry rules for -brt wasn't always showing the error code
4598 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4599 timeout_connect_A, timeout_connect_MX.
4601 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4602 to the empty sender.
4604 39. The daemon was not analysing the content of -oX till after it had closed
4605 stderr and disconnected from the controlling terminal. This meant that any
4606 syntax errors were only noted on the panic log, and the return code from
4607 the command was 0. By re-arranging the code a little, I've made the
4608 decoding happen first, so such errors now appear on stderr, and the return
4609 code is 1. However, the actual setting up of the sockets still happens in
4610 the disconnected process, so errors there are still only recorded on the
4613 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4614 connections (as happens on some IP stacks) was logged at start up time as
4615 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4616 it from "IPv6 and IPv4", which means that two separate sockets are being
4619 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4620 says whether AF_INET or AF_INET6 was passed as an argument.
4622 42. Exiwhat output was messed up when time zones were included in log
4625 43. Exiwhat now gives more information about the daemon's listening ports,
4626 and whether -tls-on-connect was used.
4628 44. The "port" option of the smtp transport is now expanded.
4630 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4631 silently ignored. Now an error message is written to the main and panic
4634 46. There's a new ACL modifier called "logwrite" which writes to a log file
4635 as soon as it is encountered.
4637 47. Added $local_user_uid and $local_user_gid at routing time.
4639 48. Exim crashed when trying to verify a sender address that was being
4642 49. Exim was recognizing only a space character after ".include". It now also
4643 recognizes a tab character.
4645 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4646 extracting the relevant information from the specification. The man page no
4647 longer contains scrambled data for the -d option, and I've added a section
4648 at the front about calling Exim under different names.
4650 51. Added "extra_headers" argument to the "mail" command in filter files.
4652 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4655 53. Installed eximstats 1.29.
4657 54. Added transport_filter_timeout as a generic transport option.
4659 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4660 Cc: header lines. This was required by RFC 822, but it not required by RFC
4663 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4664 incoming messages that did not have them. Now it does so only if the
4665 message originates locally, that is, if there is no associated remote host
4666 address. When Resent- header lines are present, this applies to the Resent-
4667 lines rather than the non-Resent- lines.
4669 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4670 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4672 58. Messages for configuration errors now include the name of the main
4673 configuration file - useful now that there may be more than one file in a
4674 list (.included file names were always shown).
4676 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4677 for those rare installations that do not start the daemon as root or run it
4678 setuid root. I've cut out the call to initgroups() if the daemon is not
4681 60. The Exim user and group can now be bound into the binary as text strings
4682 that are looked up at the start of Exim's processing.
4684 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4686 62. Added $mailstore_basename variable.
4688 63. Installed patch to sieve.c from Michael Haardt.
4690 64. When Exim failed to open the panic log after failing to open the main log,
4691 the original message it was trying to log was written to stderr and debug
4692 output, but if they were not available (the usual case in production), it
4693 was lost. Now it is written to syslog before the two lines that record the
4694 failures to open the logs.
4696 65. Users' Exim filters run in subprocesses under the user's uid. It is
4697 possible for a "deliver" command or an alias in a "personal" command to
4698 provoke an address rewrite. If logging of address rewriting is configured,
4699 this fails because the process is not running as root or exim. There may be
4700 a better way of dealing with this, but for the moment (because 4.30 needs
4701 to be released), I have disabled address rewrite logging when running a
4702 filter in a non-root, non-exim process.
4708 1. The buildconfig auxiliary program wasn't quoting the value set for
4709 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4710 not defined. This bug was masked in 4.22 by the effect that was fixed in
4713 2. Some messages that were rejected after a message id was allocated were
4714 shown as "incomplete" by exigrep. It no longer does this for messages that
4715 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4717 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4718 have allow_domain_literals set, the ID did not get logged in the <= line.
4719 Domain literals are now always recognized in Message-ID: header lines.
4721 4. The first argument for a ${extract expansion item is the key name or field
4722 number. Leading and trailing spaces in this item were not being ignored,
4723 causing some misleading effects.
4725 5. When deliver_drop_privilege was set, single queue runner processes started
4726 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4727 same command in the process it spins off) were not dropping privilege.
4729 6. When the daemon running as "exim" started a queue runner, it always
4730 re-executed Exim in the spun-off process. This is a waste of effort when
4731 deliver_drop_privilege is set. The new process now just calls the
4732 queue-runner function directly.
4738 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4741 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4742 ignored. Though the use of -oP was forcing the writing of a pid file, it
4743 was always written to the default place.
4745 3. If the message "no IP address found for host xxxx" is generated during
4746 incoming verification, it is now followed by identification of the incoming
4747 connection (so you can more easily find what provoked it).
4749 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4751 5. Added some features to "harden" Exim a bit more against certain attacks:
4753 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4754 be put in Local/Makefile. This is like the never_users runtime option,
4755 but it cannot be overridden. The default setting is "root".
4757 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4758 prefix string with which any file named in a -C command line option
4761 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4762 is retained for -C and -D only if the caller of Exim is root. Without
4763 it, the exim user may also use -C and -D and retain privilege.
4765 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4766 command line option is disabled.
4768 6. Macro names set by the -D option must start with an upper case letter, just
4769 like macro names defined in the configuration file.
4771 7. Added "dereference=" facility to LDAP.
4773 8. Two instances of the typo "uknown" in the source files are fixed.
4775 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4776 the Configure-Makefile script screwed up while processing it.
4778 10. Incorporated PCRE 4.4.
4780 11. The SMTP synchronization check was not operating right at the start of an
4781 SMTP session. For example, it could not catch a HELO sent before the client
4782 waited for the greeting. There is now a check for outstanding input at the
4783 point when the greeting is written. Because of the duplex, asynchronous
4784 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4785 way, but not yet received, when the check is performed.
4787 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4788 on TCP/IP sockets, because this apparently causes some broken clients to
4791 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4792 unchanged) from the Cygwin maintainer.
4794 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4795 was supported instead of testing for mbx. Effectively a typo.
4797 15. The spa authenticator server code was not checking that the input it
4798 received was valid base64.
4800 16. The debug output line for the "set" modifier in ACLs was not showing the
4801 name of the variable that was being set.
4803 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4805 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4806 lookup on its own. Something else was needed to trigger the lookup. For
4807 example, a match in host_lookup or the need for a host name in a host list.
4808 Now, if $sender_host_name is referenced and the host name has not yet been
4809 looked up, a lookup is performed. If the lookup fails, the variable remains
4810 empty, and $host_lookup_failed is set to "1".
4812 19. Added "eqi" as a case-independent comparison operator.
4814 20. The saslauthd authentication condition could segfault if neither service
4815 nor realm was specified.
4817 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4818 error message that was logged was misleading, and incoming SMTP
4819 connections were dropped. The message is now more accurate, and temporary
4820 errors are given to SMTP connections.
4822 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4823 (except RSET) with a 421 error, until QUIT is received. However, it was
4824 failing to send a response to QUIT.
4826 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4827 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4828 ACL is now run after the helo_try_verify_hosts code.
4830 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4831 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4832 case-independent, so other case variants are also recognized). Apparently
4833 some systems use these upper case variants.
4835 25. If more than two messages were waiting for the same host, and a transport
4836 filter was specified for the transport, Exim sent two messages over the
4837 same TCP/IP connection, and then failed with "socket operation on non-
4838 socket" when it tried to send the third.
4840 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4842 27. The extern definition of crypt16() in expand.c was not being excluded when
4843 the OS had its own crypt16() function.
4845 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4846 as a preferred synonym for return_size_limit, both as an option and as an
4849 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4851 30. Changed the default configuration ACL to relax the local part checking rule
4852 for addresses that are not in any local domains. For these addresses,
4853 slashes and pipe symbols are allowed within local parts, but the sequence
4854 /../ is explicitly forbidden.
4856 31. SPA server authentication was not clearing the challenge buffer before
4859 32. log_message in a "warn" ACL statement was writing to the reject log as
4860 well as to the main log, which contradicts the documentation and doesn't
4861 seem right (because no rejection is happening). So I have stopped it.
4863 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4864 However, I am unable to do any testing of this.
4866 34. Fixed an infelicity in the appendfile transport. When checking directories
4867 for a mailbox, to see if any needed to be created, it was accidentally
4868 using path names with one or more superfluous leading slashes; tracing
4869 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4871 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4872 contents are added to the log line that is written for every discarded
4873 recipient. (Previously a log_message setting was ignored.)
4875 36. The ${quote: operator now quotes the string if it is empty.
4877 37. The install script runs exim in order to find its version number. If for
4878 some reason other than non-existence or emptiness, which it checks, it
4879 could not run './exim', it was installing it with an empty version number,
4880 i.e. as "exim-". This error state is now caught, and the installation is
4883 38. An argument was missing from the function that creates an error message
4884 when Exim fails to connect to the socket for saslauthd authentication.
4885 This could cause Exim to crash, or give a corrupted message.
4887 39. Added isip, isip4, and isip6 to ${if conditions.
4889 40. The ACL variables $acl_xx are now saved with the message, and can be
4890 accessed later in routers, transports, and filters.
4892 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4893 strings in the file are not string-expanded.
4895 42. If a MAIL command specified a SIZE value that was too large to fit into an
4896 int variable, the check against message_size_limit failed. Such values are
4897 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4898 day this will have to be increased, but I don't think I want to be around
4899 when emails are that large.
4906 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4907 iconv() is not standard in FreeBSD.
4909 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4910 IPv6 enabled. The observed symptom was a segmentation fault on return from
4911 the function os_common_find_running_interfaces() in src/os.c.
4913 3. In the check_special_case() function in daemon.c I had used "errno" as an
4914 argument name, which causes warnings on some systems. This was basically a
4915 typo, since it was named "eno" in the comments!
4917 4. The code that waits for the clock to tick (at a resolution of some fraction
4918 of a second) so as to ensure message-id uniqueness was always waiting for
4919 at least one whole tick, when it could have waited for less. [This is
4920 almost certainly not relevant at current processor speeds, where it is
4921 unlikely to ever wait at all. But we try to future-proof.]
4923 5. The function that sleeps for a time interval that includes fractions of a
4924 second contained a race. It did not block SIGALRM between setting the
4925 timer, and suspending (a couple of lines later). If the interval was short
4926 and the sigsuspend() was delayed until after it had expired, the suspension
4927 never ended. On busy systems this could lead to processes getting stuck for
4930 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4931 process, before it forks any delivery processes. The open lookup caching
4932 mechanism meant that the open file or database connection was passed into
4933 the delivery process. The problem was that delivery processes always tidy
4934 up cached lookup data. This could cause a problem for the next delivery
4935 process started by the queue runner, because the external queue runner
4936 process does not know about the closure. So the next delivery process
4937 still has data in the lookup cache. In the case of a file lookup, there was
4938 no problem because closing a file descriptor in a subprocess doesn't affect
4939 the parent. However, if the lookup was caching a connection to a database,
4940 the connection was closed, and the second delivery process was likely to
4941 see errors such as "PGSQL: query failed: server closed the connection
4942 unexpectedly". The problem has been fixed by closing all cached lookups
4943 in a queue runner before running a delivery process.
4945 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4946 seem to have the "const" qualifier which it has on other OS. I've
4949 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4950 *of the same type* (client or server) with the same public name that an
4951 error should be diagnosed.
4953 9. When Exim looked up a host name for an IP address, but failed to find the
4954 original IP address when looking up the host name (a safety check), it
4955 output the message "<ip address> does not match any IP for NULL", which was
4956 confusing, to say the least. The bug was that the host name should have
4957 appeared instead of "NULL".
4959 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4960 user that is built into the binary, and the -C or -D options is used, root
4961 privilege is dropped before the configuration file is read. In addition,
4962 logging is switched to stderr instead of the normal log files. If the
4963 configuration then re-defines the Exim user, the unprivileged environment
4964 is probably not what is expected, so Exim logs a panic warning message (but
4967 However, if deliver_drop_privilege is set, the unprivileged state may well
4968 be exactly what is intended, so the warning has been cut out in that case,
4969 and Exim is allowed to try to write to its normal log files.
4975 1. smtp_return_error_details was not giving details for temporary sender
4976 or receiver verification errors.
4978 2. Diagnose a configuration error if two authenticators have the same public
4981 3. Exim used not to create the message log file for a message until the first
4982 delivery attempt. This could be confusing when incoming messages were held
4983 for policy or load reasons. The message log file is now created at the time
4984 the message is received, and an initial "Received" line is written to it.
4986 4. The automatically generated man page for command line options had a minor
4987 bug that caused no ill effects; however, a more serious problem was that
4988 the procedure for building the man page automatically didn't always
4989 operate. Consequently, release 4.20 contains an out-of-date version. This
4990 shouldn't happen again.
4992 5. When building Exim with embedded Perl support, the script that builds the
4993 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4994 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4996 6. The freeze_tell option was not being used for messages that were frozen on
4997 arrival, either by an ACL or by local_scan().
4999 7. Added the smtp_incomplete_transaction log selector.
5001 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5002 was accepting AUTH without a new EHLO.
5004 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5005 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5008 10. Logging of TCP/IP connections (when configured) now happens in the main
5009 daemon process instead of the child process, so that the TCP/IP connection
5010 count is more accurate (but it can never be perfect).
5012 11. The use of "drop" in a nested ACL was not being handled correctly in the
5013 outer ACL. Now, if condition failure induced by the nested "drop" causes
5014 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5015 or "require"), the connection is dropped.
5017 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5018 that yield "discard" can now be used with an "accept" or a "discard" verb,
5019 but an error is generated for any others (because I can't see a useful way
5020 to define what should happen).
5022 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5023 are now processed in the same way as lines in the Exim configuration file.
5024 In particular, continuation lines are supported.
5026 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5028 15. Added -ti meaning -t -i.
5030 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5031 domains, and warn by logging if others are found.
5033 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5034 returned from GIFCONF when Exim is trying to find the list of interfaces on
5035 a host. The code in os.c has been modified to copy each ifreq to an aligned
5036 structure in all cases.
5038 Also, in some cases, the returned ifreq's were being copied to a 'struct
5039 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5040 means the last couple of bytes of an IPv6 address could be chopped if the
5041 ifreq contained only a normal sockaddr (14 bytes storage).
5043 18. Named domain lists were not supported in the hosts_treat_as_local option.
5044 An entry such as +xxxx was not recognized, and was treated as a literal
5047 19. Ensure that header lines added by a DATA ACL are included in the reject log
5048 if the ACL subsequently rejects the message.
5050 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5051 MD5 (which is deprecated).
5053 21. When testing a filter file using -bf, Exim was writing a message when it
5054 took the sender from a "From " line in the message, but it was not doing so
5055 when it took $return_path from a Return-Path: header line. It now does.
5057 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5058 with a valid header line field name (a series of printing characters
5059 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5061 23. Changed "disc" in the source to "disk" to conform to the documentation and
5062 the book and for uniformity.
5064 24. Ignore Sendmail's -Ooption=value command line item.
5066 25. When execve() failed while trying to run a command in a pipe transport,
5067 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5068 could be confused with a return value of 69 from the command itself. This
5069 has been changed to 127, the value the shell returns if it is asked to run
5070 a non-existent command. The wording for the related log line suggests a
5071 non-existent command as the problem.
5073 26. If received_header_text expands to an empty string, do not add a Received:
5074 header line to the message. (Well, it adds a token one on the spool, but
5075 marks it "old" so that it doesn't get used or transmitted.)
5077 27. Installed eximstats 1.28 (addition of -nt option).
5079 28. There was no check for failure on the call to getsockname() in the daemon
5080 code. This can fail if there is a shortage of resources on the system, with
5081 ENOMEM, for example. A temporary error is now given on failure.
5083 29. Contrary to the C standard, it seems that in some environments, the
5084 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5085 program. Exim now does this explicitly; it affects the formatting of
5086 timestamps using strftime().
5088 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5089 complaints. I've now initialized all the variables, to avoid this.
5091 32. Header lines added by a system filter were not being "seen" during
5092 transport-time rewrites.
5094 33. The info_callback() function passed to OpenSSL is set up with type void
5095 (*)(SSL *, int, int), as described somewhere. However, when calling the
5096 function (actually a macro) that sets it up, the type void(*)() is
5097 expected. I've put in a cast to prevent warnings from picky compilers.
5099 34. If a DNS black list lookup found a CNAME record, but there were no A
5100 records associated with the domain it pointed at, Exim crashed.
5102 35. If a DNS black list lookup returned more than one A record, Exim ignored
5103 all but the first. It now scans all returned addresses if a particular IP
5104 value is being sought. In this situation, the contents of the
5105 $dnslist_value variable are a list of all the addresses, separated by a
5108 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5109 to accept a host name and all its aliases if the forward lookup for any of
5110 them yielded the IP address of the incoming connection. Now it accepts only
5111 those names whose forward lookup yields the correct IP address. Any other
5112 names are discarded. This closes a loophole whereby a rogue DNS
5113 administrator could create reverse DNS records to break through a
5114 wildcarded host restriction in an ACL.
5116 37. If a user filter or a system filter that ran in a subprocess used any of
5117 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5118 the wrong values were passed to the pipe command ($thisaddress had the
5119 value of $0, $0 had the value of $1, etc). This bug was introduced by
5120 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5122 38. Improved the line breaking for long SMTP error messages from ACLs.
5123 Previously, if there was no break point between 40 and 75 characters, Exim
5124 left the rest of the message alone. Two changes have been made: (a) I've
5125 reduced the minimum length to 35 characters; (b) if it can't find a break
5126 point between 35 and 75 characters, it looks ahead and uses the first one
5127 that it finds. This may give the occasional overlong line, but at least the
5128 remaining text gets split now.
5130 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5131 file descriptors might be low, and that setting 1000 would always raise it.
5132 It turns out that in some environments, the limit is already over 1000 and
5133 that lowering it causes trouble. So now Exim takes care not to decrease it.
5135 40. When delivering a message, the value of $return_path is set to $sender_
5136 address at the start of routing (routers may change the value). By an
5137 oversight, this default was not being set up when an address was tested by
5138 -bt or -bv, which affected the outcome if any router or filter referred to
5141 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5142 log only when "message" or "log_message" are set. However, if one of the
5143 conditions was an address verification, or a call to a nested ACL, the
5144 messages generated by the underlying test were being passed through. This
5145 no longer happens. The underlying message is available in $acl_verify_
5146 message for both "message" and "log_message" expansions, so it can be
5147 passed through if needed.
5149 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5150 new expansion $bh_ to give the encoded byte string without charset
5151 translation. Translation happens only if iconv() is available; HAVE_ICONV
5152 indicates this at build time. HEADERS_CHARSET gives the charset to
5153 translate to; headers_charset can change it in the configuration, and
5154 "headers charset" can change it in an individual filter file.
5156 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5157 that creates RFC 2047 encoded "words" labels them as that charset instead
5158 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5159 expansion operator; (ii) when Exim creates a From: line for a local
5160 message; (iii) when a header line is rewritten to include a "phrase" part.
5162 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5163 buggy, causing it to skip the first lines of messages whose message ID
5164 ended in 'D'. This would not have bitten before Exim release 4.14, because
5165 message IDs were unlikely to end in 'D' before then. The effect was to have
5166 incorrect size information for certain domains.
5168 45. #include "config.h" was missing at the start of the crypt16.c module. This
5169 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5172 46. If there was a timeout during a "random" callout check, Exim treated it as
5173 a failure of the random address, and carried on sending RSET and the real
5174 address. If the delay was just some slowness somewhere, the response to the
5175 original RCPT would be taken as a response to RSET and so on, causing
5176 mayhem of various kinds.
5178 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5179 when I implemented it. It didn't allow for the fact that some option values
5180 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5181 the right test for positive values.
5183 48. Domain names in DNS records are case-independent. Exim always looks them up
5184 in lower case. Some resolvers return domain names in exactly the case they
5185 appear in the zone file, that is, they may contain uppercase letters. Not
5186 all resolvers do this - some return always lower case. Exim was treating a
5187 change of case by a resolver as a change of domain, similar to a widening
5188 of a domain abbreviation. This triggered its re-routing code and so it was
5189 trying to route what was effectively the same domain again. This normally
5190 caused routing to fail (because the router wouldn't handle the domain
5191 twice). Now Exim checks for this case specially, and just changes the
5192 casing of the domain that it ultimately uses when it transmits the message
5195 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5198 50. If a filter generated a file delivery with a non-absolute name (possible if
5199 no home directory exists for the router), the forbid_file option was not
5202 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5203 the existing equality matching.
5205 52. Exim was using ints instead of ino_t variables in some places where it was
5206 dealing with inode numbers.
5208 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5209 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5210 called TMPDIR, and if it finds it is different, it changes its value.
5212 54. The smtp_printf() function is now made available to local_scan() so
5213 additional output lines can be written before returning. There is also an
5214 smtp_fflush() function to enable the detection of a dropped connection.
5215 The variables smtp_input and smtp_batched_input are exported to
5218 55. Changed the default runtime configuration: the message "Unknown user"
5219 has been removed from the ACL, and instead placed on the localuser router,
5220 using the cannot_route_message feature. This means that any verification
5221 failures that generate their own messages won't get overridden. Similarly,
5222 the "Unrouteable address" message that was in the ACL for unverifiable
5223 relay addresses has also been removed.
5225 56. Added hosts_avoid_esmtp to the smtp transport.
5227 57. The exicyclog script was not checking for the esoteric option
5228 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5229 will work only if exicyclog is run under the appropriate euid.
5231 58. Following a discussion on the list, the rules by which Exim recognises line
5232 endings on incoming messages have been changed. The -dropcr and drop_cr
5233 options are now no-ops, retained only for backwards compatibility. The
5234 following line terminators are recognized: LF CRLF CR. However, special
5235 processing applies to CR:
5237 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5238 nor a local message in the state where . is a terminator.
5240 (ii) If a bare CR is encountered in a header line, an extra space is added
5241 after the line terminator so as not to end the header. The reasoning
5242 behind this is that bare CRs in header lines are most likely either
5243 to be mistakes, or people trying to play silly games.
5245 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5246 was being incorrectly given as 18 bytes larger than it should have been.
5247 This is a VOB (very old bug).
5249 60. This may never have affected anything current, but just in case it has:
5250 When the local host is found other than at the start of a list of hosts,
5251 the local host, those with the same MX, and any that follow, are discarded.
5252 When the list in question was part of a longer list of hosts, the following
5253 hosts (not currently being processed) were also being discarded. This no
5254 longer happens. I'm not sure if this situation could ever has previously
5257 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5260 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5261 header. This is something that is recommended in a new Internet Draft, and
5262 is something that is documented as being done by Sendmail. There are two
5263 possible values. For messages generated by the autoreply transport, Exim
5266 Auto-Submitted: auto-replied
5268 whereas for all other generated messages (e.g. bounces) it adds
5270 Auto-Submitted: auto-generated
5272 63. The "personal" condition in filters now includes a test for the
5273 Auto-Submitted: header. If it contains the string "auto-" the message it
5274 not considered personal.
5276 64. Added rcpt_include_affixes as a generic transport option.
5278 65. Added queue_only_override (default true).
5280 66. Added the syslog_duplication option.
5282 67. If what should have been the first header line of a message consisted of
5283 a space followed by a colon, Exim was mis-interpreting it as a header line.
5284 It isn't of course - it is syntactically invalid and should therefore be
5285 treated as the start of the message body. The misbehaviour could have
5286 caused a number of strange effects, including loss of data in subsequent
5287 header lines, and spool format errors.
5289 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5290 client host had authenticated. This control can now be exercised by an ACL
5291 for more flexibility.
5293 69. By default, callouts do not happen when testing with -bh. There is now a
5294 variant, -bhc, which does actually run the callout code, including
5295 consulting and updating the callout cache.
5297 70. Added support for saslauthd authentication, courtesy of Alexander
5300 71. If statvfs() failed on the spool or log directories while checking their
5301 size for availability, Exim confusingly gave the error "space shortage".
5302 Furthermore, in debugging mode it crashed with a floating point exception.
5303 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5304 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5305 serious problem, Exim now writes to the main and panic logs when this
5306 happens, with details of the failure. It then refuses to accept the
5307 incoming message, giving the message "spool directory problem" or "log
5308 directory problem" with a 421 code for SMTP messages.
5310 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5311 0, 1, and 2 exist, because some OS complain for execs without them (see
5312 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5313 descriptors. However, the code omitted to check that the open succeeded,
5314 causing mysterious errors if for some reason the permissions on /dev/null
5315 got screwed. Now Exim writes a message to the main and panic logs, and
5316 bombs out if it can't open /dev/null.
5318 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5319 interact so that it is all more flexible. It is supposed to remain
5320 backwards compatible. Also added extra_local_interfaces.
5322 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5323 to bomb out with an assertion failure - to the client this appears as a
5324 connection drop. This problem occurs in the part of the code that was taken
5325 from the Samba project. Fortunately, the assertion is in a very simple
5326 function, so I have fixed this by reproducing the function inline in the
5327 one place where it is called, and arranging for authentication to fail
5328 instead of killing the process with assert().
5330 75. The SPA client code was not working when the server requested OEM rather
5331 than Unicode encoding.
5333 76. Added code to make require_files with a specific uid setting more usable in
5334 the case where statting the file as root fails - usually a non-root-mounted
5335 NFS file system. When this happens and the failure is EACCES, Exim now
5336 forks a subprocess and does the per-uid checking as the relevant uid.
5338 77. Added process_log_path.
5340 78. If log_file_path was not explicitly set, a setting of check_log_space or
5341 check_log_inodes was ignored.
5343 79. If a space check for the spool or log partitions fails, the incident is now
5344 logged. Of course, in the latter case the data may get lost...
5346 80. Added the %p formatting code to string_format() so that it can be used to
5347 print addresses in debug_print(). Adjusted all the address printing in the
5348 debugging in store.c to use %p rather than %d.
5350 81. There was a concern that a line of code in smtp_in.c could overflow a
5351 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5352 initially expressed, the concern was not well-founded, because trailing
5353 spaces are removed early. However, if the trailing spaces were followed by
5354 a NULL, they did not get removed, so the overflow was possible. Two fixes
5357 (a) I re-wrote the offending code in a cleaner fashion.
5358 (b) If an incoming SMTP command contains a NULL character, it is rejected
5361 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5362 runs initgroups(), so that if the Exim user is in any additional groups,
5363 they will be used during message reception.
5369 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob