1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
15 NM/01 Workround for PCRE version dependency in version reporting
18 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
19 This fixes portability to compilers other than gcc, notably
20 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
22 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
23 makefiles for portability to HP-UX and POSIX correctness.
25 PP/01 Permit LOOKUP_foo enabling on the make command-line.
26 Also via indented variable definition in the Makefile.
27 (Debugging by Oliver Heesakkers).
29 PP/02 Restore caching of spamd results with expanded spamd_address.
30 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
32 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
33 Improves build reliability. Fix from: Frank Elsner
35 NM/02 Fix wide character breakage in the rfc2047 coding
36 Fixes bug 1064. Patch from Andrey N. Oktyabrski
38 NM/03 Allow underscore in dnslist lookups
39 Fixes bug 1026. Patch from Graeme Fowler
41 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
42 Code patches from Adam Ciarcinski of NetBSD.
44 NM/04 Fixed exiqgrep to cope with mailq missing size issue
47 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
48 is logged, to avoid truncation. Patch from John Horne.
50 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
51 Patch from Jakob Hirsch.
53 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
54 SQL string expansion failure details.
55 Patch from Andrey Oktyabrski.
57 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
58 Patch from Simon Arlott.
60 PP/09 New lookups functionality failed to compile on old gcc which rejects
61 extern declarations in function scope.
62 Patch from Oliver Fleischmann
64 PP/10 Use sig_atomic_t for flags set from signal handlers.
65 Check getgroups() return and improve debugging.
66 Fixed developed for diagnosis in bug 927 (which turned out to be
69 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
70 Patch from Mark Zealey.
72 PP/12 Bugzilla 1056: Improved spamd server selection.
73 Patch from Mark Zealey.
75 PP/13 Bugzilla 1086: Deal with maildir quota file races.
76 Based on patch from Heiko Schlittermann.
78 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
79 Patch from Uwe Doering, sign-off by Michael Haardt.
81 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
82 variable declaration deep within a block. Bug and patch from
85 PP/15 lookups-Makefile IRIX compatibilty coercion.
87 PP/16 Make DISABLE_DKIM build knob functional.
89 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
92 TF/03 Fix valgrind.h portability to C89 compilers that do not support
93 variable argument macros. Our copy now differs from upstream.
99 TF/01 Failure to get a lock on a hints database can have serious
100 consequences so log it to the panic log.
102 TF/02 Log LMTP confirmation messages in the same way as SMTP,
103 controlled using the smtp_confirmation log selector.
105 TF/03 Include the error message when we fail to unlink a spool file.
107 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
108 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
109 for maintaining out-of-tree patches for some time.
111 PP/01 Bugzilla 139: Documentation and portability issues.
112 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
113 Handle per-OS dynamic-module compilation flags.
115 PP/02 Let /dev/null have normal permissions.
116 The 4.73 fixes were a little too stringent and complained about the
117 permissions on /dev/null. Exempt it from some checks.
118 Reported by Andreas M. Kirchwitz.
120 PP/03 Report version information for many libraries, including
121 Exim version information for dynamically loaded libraries. Created
122 version.h, now support a version extension string for distributors
123 who patch heavily. Dynamic module ABI change.
125 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
126 privilege escalation vulnerability whereby the Exim run-time user
127 can cause root to append content of the attacker's choosing to
130 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
133 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
134 If dropping privileges for untrusted macros, we disabled normal logging
135 on the basis that it would fail; for the Exim run-time user, this is not
136 the case, and it resulted in successful deliveries going unlogged.
137 Fixed. Reported by Andreas Metzler.
143 PP/01 Date: & Message-Id: revert to normally being appended to a message,
144 only prepend for the Resent-* case. Fixes regression introduced in
145 Exim 4.70 by NM/22 for Bugzilla 607.
147 PP/02 Include check_rfc2047_length in configure.default because we're seeing
148 increasing numbers of administrators be bitten by this.
150 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
152 PP/03 Bugzilla 994: added openssl_options main configuration option.
154 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
156 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
158 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
160 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
161 without a peer certificate, leading to a segfault because of an
162 assumption that peers always have certificates. Be a little more
163 paranoid. Problem reported by Martin Tscholak.
165 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
166 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
167 NB: ClamAV planning to remove STREAM in "middle of 2010".
168 CL also introduces -bmalware, various -d+acl logging additions and
169 more caution in buffer sizes.
171 PP/09 Implemented reverse_ip expansion operator.
173 PP/10 Bugzilla 937: provide a "debug" ACL control.
175 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
177 PP/12 Bugzilla 973: Implement --version.
179 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
181 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
183 PP/15 Bugzilla 816: support multiple condition rules on Routers.
185 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
186 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
187 ignore trailing whitespace.
189 JJ/02 prevent non-panic DKIM error from being sent to paniclog
191 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
194 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
195 Notification from Dr Andrew Aitchison.
197 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
198 ExtendedDetectionInfo response format.
199 Notification from John Horne.
201 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
204 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
205 XSL and documented dependency on system catalogs, with examples of how
208 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
211 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
212 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
213 configuration file which is writeable by the Exim user or group.
215 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
216 of configuration files to cover files specified with the -C option if
217 they are going to be used with root privileges, not just the default
220 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
221 option (effectively making it always true).
223 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
224 files to be used while preserving root privileges.
226 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
227 that rogue child processes cannot use them.
229 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
230 run-time user, instead of root.
232 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
233 Exim run-time user without dropping privileges.
235 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
236 result string, instead of calling string_vformat() twice with the same
239 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
240 for other users. Others should always drop root privileges if they use
241 -C on the command line, even for a whitelisted configure file.
243 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
245 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
251 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
252 $data_path, and $header_path variables; fixed documentation bugs and
255 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
256 exipick to access non-standard spools, including the "frozen" queue
259 NM/01 Bugzilla 965: Support mysql stored procedures.
260 Patch from Alain Williams
262 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
264 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
265 Patch from Andreas Metzler
267 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
268 Patch from Kirill Miazine
270 NM/05 Bugzilla 671: Added umask to procmail example.
272 JJ/03 installed exipick 20100323.0, fixing doc bug
274 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
275 directory. Notification and patch from Dan Rosenberg.
277 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
279 TK/02 Improve log output when DKIM signing operation fails.
281 MH/01 Treat the transport option dkim_domain as a colon separated
282 list, not as a single string, and sign the message with each element,
283 omitting multiple occurences of the same signer.
285 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
286 Bugzilla 985, 986. Patch by Simon Arlott
288 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
289 Patch by Simon Arlott
291 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
292 MBX locking. Notification from Dan Rosenberg.
298 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
300 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
302 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
304 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
306 NM/04 Bugzilla 915: Flag broken perl installation during build.
312 TK/01 Added patch by Johannes Berg that expands the main option
313 "spamd_address" if it starts with a dollar sign.
315 TK/02 Write list of recipients to X-Envelope-Sender header when building
316 the mbox-format spool file for content scanning (suggested by Jakob
319 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
320 (http://www.dcc-servers.net/) support via dccifd. Activated by
321 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
323 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
324 by Mark Daniel Reidel <mr@df.eu>.
326 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
327 When building exim an external PCRE library is now needed -
328 PCRE is a system library on the majority of modern systems.
329 See entry on PCRE_LIBS in EDITME file.
331 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
332 conversation. Added nologin parameter to request.
333 Patch contributed by Kirill Miazine.
335 TF/01 Do not log submission mode rewrites if they do not change the address.
337 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
339 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
340 log files in place. Contributed by Roberto Lima.
342 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
344 TF/03 Bugzilla 615: When checking the local_parts router precondition
345 after a local_part_suffix or local_part_prefix option, Exim now
346 does not use the address's named list lookup cache, since this
347 contains cached lookups for the whole local part.
349 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
350 Robert Millan. Documentation is in experimental-spec.txt.
352 TF/04 Bugzilla 668: Fix parallel build (make -j).
354 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
356 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
357 Patch provided by Jan Srzednicki.
359 TF/05 Leading white space used to be stripped from $spam_report which
360 wrecked the formatting. Now it is preserved.
362 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
363 that they are available at delivery time.
365 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
367 TF/08 TLS error reporting now respects the incoming_interface and
368 incoming_port log selectors.
370 TF/09 Produce a more useful error message if an SMTP transport's hosts
371 setting expands to an empty string.
373 NM/06 Bugzilla 744: EXPN did not work under TLS.
374 Patch provided by Phil Pennock.
376 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
377 Patch provided by Richard Godbee.
379 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
380 acl_smtp_notquit, added index entry.
382 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
383 Patch provided by Eugene Bujak.
385 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
386 accept(). Patch provided by Maxim Dounin.
388 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
389 Patch provided by Phil Pennock.
391 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
393 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
394 Patch provided by Brad "anomie" Jorsch.
396 NM/14 Bugzilla 622: Added timeout setting to transport filter.
397 Patch provided by Dean Brooks.
399 TK/05 Add native DKIM support (does not depend on external libraries).
401 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
402 Patch provided by Graeme Fowler.
404 NM/16 Bugzilla 851: Documentation example syntax fix.
406 NM/17 Changed NOTICE file to remove references to embedded PCRE.
408 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
411 NM/19 Bugzilla 745: TLS version reporting.
412 Patch provided by Phil Pennock.
414 NM/20 Bugzilla 167: bool: condition support.
415 Patch provided by Phil Pennock.
417 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
418 clients. Patch provided by Phil Pennock.
420 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
421 Patch provided by Brad "anomie" Jorsch.
423 NM/23 Bugzilla 687: Fix misparses in eximstats.
424 Patch provided by Heiko Schlittermann.
426 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
427 Patch provided by Heiko Schlittermann.
429 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
430 plus update to original patch.
432 NM/26 Bugzilla 799: Documentation correction for ratelimit.
434 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
435 Patch provided by David Brownlee.
437 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
439 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
441 NM/30 Bugzilla 888: TLS documentation bugfixes.
443 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
445 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
446 Unlike the original bugzilla I have changed all shell scripts in src tree.
448 NM/33 Bugzilla 898: Transport filter timeout fix.
449 Patch by Todd Rinaldo.
451 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
452 Patch by Serge Demonchaux.
454 NM/35 Bugzilla 39: Base64 decode bug fixes.
455 Patch by Jakob Hirsch.
457 NM/36 Bugzilla 909: Correct connect() call in dcc code.
459 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
461 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
463 NM/39 Bugzilla 911: Fixed MakeLinks build script.
469 TK/01 Add preliminary DKIM support. Currently requires a forked version of
470 ALT-N's libdkim that I have put here:
471 http://duncanthrax.net/exim-experimental/
473 Note to Michael Haardt: I had to rename some vars in sieve.c. They
474 were called 'true' and it seems that C99 defines that as a reserved
475 keyword to be used with 'bool' variable types. That means you could
476 not include C99-style headers which use bools without triggering
477 build errors in sieve.c.
479 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
480 as mailq or other aliases. Changed the --help handling significantly
481 to do whats expected. exim_usage() emits usage/help information.
483 SC/01 Added the -bylocaldomain option to eximstats.
485 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
487 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
489 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
495 PH/01 Another patch from the Sieve maintainer.
497 PH/02 When an IPv6 address is converted to a string for single-key lookup
498 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
499 dots are used instead of colons so that keys in lsearch files need not
500 contain colons. This was done some time before quoting was made available
501 in lsearch files. However, iplsearch files do require colons in IPv6 keys
502 (notated using the quote facility) so as to distinguish them from IPv4
503 keys. This meant that lookups for IP addresses in host lists did not work
504 for iplsearch lookups.
506 This has been fixed by arranging for IPv6 addresses to be expressed with
507 colons if the lookup type is iplsearch. This is not incompatible, because
508 previously such lookups could never work.
510 The situation is now rather anomolous, since one *can* have colons in
511 ordinary lsearch keys. However, making the change in all cases is
512 incompatible and would probably break a number of configurations.
514 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
517 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
518 conversion specification without a maximum field width, thereby enabling
519 a rogue spamd server to cause a buffer overflow. While nobody in their
520 right mind would setup Exim to query an untrusted spamd server, an
521 attacker that gains access to a server running spamd could potentially
522 exploit this vulnerability to run arbitrary code as the Exim user.
524 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
525 $primary_hostname instead of what libspf2 thinks the hosts name is.
527 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
528 a directory entry by the name of the lookup key. Previously, if a
529 symlink pointed to a non-existing file or a file in a directory that
530 Exim lacked permissions to read, a lookup for a key matching that
531 symlink would fail. Now it is enough that a matching directory entry
532 exists, symlink or not. (Bugzilla 503.)
534 PH/03 The body_linecount and body_zerocount variables are now exported in the
537 PH/04 Added the $dnslist_matched variable.
539 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
540 This means they are set thereafter only if the connection becomes
543 PH/06 Added the client_condition to authenticators so that some can be skipped
544 by clients under certain conditions.
546 PH/07 The error message for a badly-placed control=no_multiline_responses left
547 "_responses" off the end of the name.
549 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
551 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
552 (without spaces) instead of just copying the configuration text.
554 PH/10 Added the /noupdate option to the ratelimit ACL condition.
556 PH/11 Added $max_received_linelength.
558 PH/12 Added +ignore_defer and +include_defer to host lists.
560 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
561 way in which PCRE > 7.0 is built.
563 PH/14 Implemented queue_only_load_latch.
565 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
566 MAIL command. The effect was to mangle the value on 64-bit systems.
568 PH/16 Another patch from the Sieve maintainer.
570 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
572 PH/18 If a system quota error occurred while trying to create the file for
573 a maildir delivery, the message "Mailbox is full" was not appended to the
574 bounce if the delivery eventually timed out. Change 4.67/27 below applied
575 only to a quota excession during the actual writing of the file.
577 PH/19 It seems that peer DN values may contain newlines (and other non-printing
578 characters?) which causes problems in log lines. The DN values are now
579 passed through string_printing() before being added to log lines.
581 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
582 and InterBase are left for another time.)
584 PH/21 Added message_body_newlines option.
586 PH/22 Guard against possible overflow in moan_check_errorcopy().
588 PH/23 POSIX allows open() to be a macro; guard against that.
590 PH/24 If the recipient of an error message contained an @ in the local part
591 (suitably quoted, of course), incorrect values were put in $domain and
592 $local_part during the evaluation of errors_copy.
598 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
599 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
602 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
603 issue a MAIL command.
605 PH/02 In an ACL statement such as
607 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
609 if a client was not listed at all, or was listed with a value other than
610 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
611 the condition was not true (as it should be), so access was not denied.
612 The bug was that the ! inversion was incorrectly passed on to the second
613 item. This has been fixed.
615 PH/03 Added additional dnslists conditions == and =& which are different from
616 = and & when the dns lookup returns more than one IP address.
618 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
619 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
621 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
622 FSYNC, which compiles an option called disable_fsync that allows for
623 bypassing fsync(). The documentation is heavily laced with warnings.
625 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
627 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
628 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
629 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
630 including adding "make clean"; (3) Added -fPIC when compiling the test
631 dynamically loaded module, to get rid of a warning.
633 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
634 message fails, move_frozen_messages = true and ignore_bounce_errors_after
635 = 0s. The bug is otherwise harmless.
637 PH/07 There was a bug in the dovecot authenticator such that the value of
638 $auth1 could be overwritten, and so not correctly preserved, after a
639 successful authentication. This usually meant that the value preserved by
640 the server_setid option was incorrect.
642 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
644 PH/09 Installed PCRE release 7.0.
646 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
647 run for batched SMTP input. It is now run at the start of every message
648 in the batch. While fixing this I discovered that the process information
649 (output by running exiwhat) was not always getting set for -bs and -bS
650 input. This is fixed, and it now also says "batched" for BSMTP.
652 PH/11 Added control=no_pipelining.
654 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
655 patch, slightly modified), and move the expansion of helo_data till after
656 the connection is made in the smtp transport (so it can use these
659 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
661 PH/14 Added log_selector = +pid.
663 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
665 PH/16 Add ${if forany and ${if forall.
667 PH/17 Added dsn_from option to vary the From: line in DSNs.
669 PH/18 Flush SMTP output before performing a callout, unless control =
670 no_callout_flush is set.
672 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
673 was true (the default) a successful delivery failed to delete the retry
674 item, thus causing premature timeout of the address. The bug is now
677 PH/20 Added hosts_avoid_pipelining to the smtp transport.
679 PH/21 Long custom messages for fakedefer and fakereject are now split up
680 into multiline reponses in the same way that messages for "deny" and
681 other ACL rejections are.
683 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
684 with slight modification.
686 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
687 draft, changing the syntax and factoring some duplicate code".
689 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
690 for deliveries of the second and subsequent messages over the same SMTP
693 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
694 ${reduce, with only minor "tidies".
696 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
698 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
699 expansion side effects.
701 PH/27 When a message times out after an over-quota error from an Exim-imposed
702 quota, the bounce message says "mailbox is full". This message was not
703 being given when it was a system quota that was exceeded. It now should
706 MH/03 Made $recipients available in local_scan(). local_scan() already has
707 better access to the recipient list through recipients_list[], but
708 $recipients can be useful in postmaster-provided expansion strings.
710 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
711 in the case of a MAIL command with additional options following the
712 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
713 were accidentally chopped off.
715 PH/29 SMTP synchronization checks are implemented when a command is read -
716 there is a check that no more input is waiting when there shouldn't be
717 any. However, for some commands, a delay in an ACL can mean that it is
718 some time before the response is written. In this time, more input might
719 arrive, invalidly. So now there are extra checks after an ACL has run for
720 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
721 pipelining has not been advertised.
723 PH/30 MH's patch to allow iscntrl() characters to be list separators.
725 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
726 returned in the SMTP response when smtp_return_error_details was false.
729 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
730 instead of the C I/O that was originally supplied, because problems were
733 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
734 Exim which did not show up earlier: it was assuming that a call to
735 SSL_CTX_set_info_callback() might give an error value. In fact, there is
736 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
737 was a macro that became an assignment, so it seemed to work. This has
738 changed to a proper function call with a void return, hence the compile
739 error. Exim's code has been fixed.
741 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
744 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
746 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
748 PH/37 If a message is not accepted after it has had an id assigned (e.g.
749 because it turns out to be too big or there is a timeout) there is no
750 "Completed" line in the log. When some messages of this type were
751 selected by exigrep, they were listed as "not completed". Others were
752 picked up by some special patterns. I have improved the selection
753 criteria to be more general.
755 PH/38 The host_find_failed option in the manualroute router can now be set
756 to "ignore", to completely ignore a host whose IP address cannot be
757 found. If all hosts are ignored, the behaviour is controlled by the new
758 host_all_ignored option.
760 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
761 homing or because of multiple MX records with /mx) generated more than
762 one IP address, and the following item turned out to be the local host,
763 all the secondary addresses of the first item were incorrectly removed
764 from the list, along with the local host and any following hosts (which
765 is what is supposed to happen).
767 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
768 whoever called Exim into the -H file. In the case of the daemon it was
769 behaving confusingly. When first started, it used values for whoever
770 started the daemon, but after a SIGHUP it used the Exim user (because it
771 calls itself on a restart). I have changed the code so that it now always
774 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
775 message are rejected with the same error (e.g. no authentication or bad
776 sender address), and a DATA command is nevertheless sent (as can happen
777 with PIPELINING or a stupid MUA), the error message that was given to the
778 RCPT commands is included in the rejection of the DATA command. This is
779 intended to be helpful for MUAs that show only the final error to their
782 PH/42 Another patch from the Sieve maintainer.
784 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
785 Eximstats - Fixed some broken HTML links and added missing column headers
787 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
788 columns for Rejects, Temp Rejects, Ham, and Spam rows.
790 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
792 PH/43 Yet another patch from the Sieve maintainer.
794 PH/44 I found a way to check for a TCP/IP connection going away before sending
795 the response to the final '.' that terminates a message, but only in the
796 case where the client has not sent further data following the '.'
797 (unfortunately, this is allowed). However, in many cases there won't be
798 any further data because there won't be any more messages to send. A call
799 to select() can be used: if it shows that the input is "ready", there is
800 either input waiting, or the socket has been closed. An attempt to read
801 the next input character can distinguish the two cases. Previously, Exim
802 would have sent an OK response which the client would never have see.
803 This could lead to message repetition. This fix should cure that, at
804 least in a lot of common cases.
806 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
807 advertised in response to EHLO.
813 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
814 fixed by 4.65/MH/01 (is this a record?) are fixed:
816 (i) An empty string was always treated as zero by the numeric comparison
817 operators. This behaviour has been restored.
819 (ii) It is documented that the numeric comparison operators always treat
820 their arguments as decimal numbers. This was broken in that numbers
821 starting with 0 were being interpreted as octal.
823 While fixing these problems I realized that there was another issue that
824 hadn't been noticed. Values of message_size_limit (both the global option
825 and the transport option) were treated as octal if they started with 0.
826 The documentation was vague. These values are now always treated as
827 decimal, and I will make that clear in the documentation.
833 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
834 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
837 MH/01 Don't check that the operands of numeric comparison operators are
838 integers when their expansion is in "skipping" mode (fixes bug
839 introduced by 4.64-PH/07).
841 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
842 child addresses, Exim now panics and dies. Previously, because the count
843 is held in a short int, deliveries were likely to be lost. As such a
844 large number of recipients for a single message is ridiculous
845 (performance will be very, very poor), I have chosen to impose a limit
846 rather than extend the field.
852 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
853 leftover -K file (the existence of which was triggered by #402).
854 While we were at it, introduced process PID as part of the -K
855 filename. This should rule out race conditions when creating
858 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
859 processing considerably. Previous code took too long for large mails,
860 triggering a timeout which in turn triggers #401.
862 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
863 in the DK code in transports.c. sendfile() is not really portable,
864 hence the _LINUX specificness.
866 TF/01 In the add_headers option to the mail command in an Exim filter,
867 there was a bug that Exim would claim a syntax error in any
868 header after the first one which had an odd number of characters
871 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
872 callout verification, Exim cached a "reject" for the entire domain. This
873 is correct for most verifications, but it is not correct for a recipient
874 verification with use_sender or use_postmaster set, because in that case
875 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
876 case of MAIL FROM:<> rejection from other early rejections (e.g.
877 rejection of HELO). When verifying a recipient using a non-null MAIL
878 address, the cache is ignored if it shows MAIL FROM:<> rejection.
879 Whatever the result of the callout, the value of the domain cache is
880 left unchanged (for any other kind of callout, getting as far as trying
881 RCPT means that the domain itself is ok).
883 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
886 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
887 manifest itself as EPIPE rather than ECONNECT. When tidying away a
888 session, the daemon ignores ECONNECT errors and logs others; it now
889 ignores EPIPE as well.
891 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
892 (quoted-printable decoding).
894 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
895 later the small subsequent patch to fix an introduced bug.
897 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
899 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
901 PH/08 An error is now given if message_size_limit is specified negative.
903 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
904 to be given (somewhat) arbitrary names.
906 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
909 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
910 miscellaneous code fixes
912 PH/10 Added the log_reject_target ACL modifier to specify where to log
915 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
916 hostname. This is wrong, because it relates to the incoming message (and
917 probably the interface on which it is arriving) and not to the outgoing
918 callout (which could be using a different interface). This has been
919 changed to use the value of the helo_data option from the smtp transport
920 instead - this is what is used when a message is actually being sent. If
921 there is no remote transport (possible with a router that sets up host
922 addresses), $smtp_active_hostname is used.
924 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
925 tweaks were necessary in order to get it to work (see also 21 below):
926 (a) The code assumed that strncpy() returns a negative number on buffer
927 overflow, which isn't the case. Replaced with Exim's string_format()
929 (b) There were several signed/unsigned issues. I just did the minimum
930 hacking in of casts. There is scope for a larger refactoring.
931 (c) The code used strcasecmp() which is not a standard C function.
932 Replaced with Exim's strcmpic() function.
933 (d) The code set only $1; it now sets $auth1 as well.
934 (e) A simple test gave the error "authentication client didn't specify
935 service in request". It would seem that Dovecot has changed its
936 interface. Fortunately there's a specification; I followed it and
937 changed what the client sends and it appears to be working now.
939 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
942 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
943 address A is aliased to B and C, where B exists and C does not. Without
944 -v the output is "A verified" because verification stops after a
945 successful redirection if more than one address is generated. However,
946 with -v the child addresses are also verified. Exim was outputting "A
947 failed to verify" and then showing the successful verification for C,
948 with its parentage. It now outputs "B failed to verify", showing B's
949 parentage before showing the successful verification of C.
951 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
952 look up a TXT record in a specific list after matching in a combined
955 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
956 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
957 they consult the DNS. I had assumed they would set it the way they
958 wanted; and indeed my experiments on Linux seem to show that in some
959 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
960 To be on the safe side, however, I have now made the interface to
961 host_find_byname() similar to host_find_bydns(), with an argument
962 containing the DNS resolver options. The host_find_byname() function now
963 sets these options at its start, just as host_find_bydns() does. The smtp
964 transport options dns_qualify_single and dns_search_parents are passed to
965 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
966 of host_find_byname() use the default settings of RES_DEFNAMES
967 (qualify_single) but not RES_DNSRCH (search_parents).
969 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
970 spool_read_header() do less string testing, by means of a preliminary
971 switch on the second character of optional "-foo" lines. (This is
972 overdue, caused by the large number of possibilities that now exist.
973 Originally there were few.) While I was there, I also converted the
974 str(n)cmp tests so they don't re-test the leading "-" and the first
975 character, in the hope this might squeeze out yet more improvement.
977 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
978 flag allowing group syntax was set by the header_syntax check but not
979 turned off, possible causing trouble later; (2) The flag was not being
980 set at all for the header_verify test, causing "group"-style headers to
981 be rejected. I have now set it in this case, and also caused header_
982 verify to ignore an empty address taken from a group. While doing this, I
983 came across some other cases where the code for allowing group syntax
984 while scanning a header line wasn't quite right (mostly, not resetting
985 the flag correctly in the right place). These bugs could have caused
986 trouble for malformed header lines. I hope it is now all correct.
988 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
989 with the "reply" argument non-NULL. The code, however (which originally
990 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
991 but it didn't always do it. This confused somebody who was copying the
992 code for some other use. I have removed all the tests.
994 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
995 feature that was used to support insecure browsers during the U.S. crypto
996 embargo. It requires special client support, and Exim is probably the
997 only MTA that supported it -- and would never use it because real RSA is
998 always available. This code has been removed, because it had the bad
999 effect of slowing Exim down by computing (never used) parameters for the
1000 RSA_EXPORT functionality.
1002 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1003 authenticator to fail if there's a tab character in the incoming data
1004 (there should never be unless someone is messing about, as it's supposed
1005 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1006 if the connection is using TLS or if the remote IP is the same as the
1007 local IP, and the "valid-client-cert option" if a client certificate has
1010 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1011 authenticators. This can be used for authorization after authentication
1012 succeeds. (In the case of plaintext, it servers for both authentication
1015 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1016 if any retry times were supplied.
1018 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1019 connection, where there is no client IP address to check. In this
1020 situation, the verify now always succeeds.
1022 PH/25 Applied John Jetmore's -Mset patch.
1024 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1026 PH/27 In a string expansion for a processed (not raw) header when multiple
1027 headers of the same name were present, leading whitespace was being
1028 removed from all of them, but trailing whitespace was being removed only
1029 from the last one. Now trailing whitespace is removed from each header
1030 before concatenation. Completely empty headers in a concatenation (as
1031 before) are ignored.
1033 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1034 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1036 PH/29 [Removed. This was a change that I later backed out, and forgot to
1037 correct the ChangeLog entry (that I had efficiently created) before
1038 committing the later change.]
1040 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1041 address errors (4xx response to RCPT) over the same connection as other
1042 messages routed to the same hosts. Such deliveries are always "forced",
1043 so retry times are not inspected. This resulted in far too many retries
1044 for the affected addresses. The effect occurred only when there were more
1045 hosts than the hosts_max_try setting in the smtp transport when it had
1046 the 4xx errors. Those hosts that it had tried were not added to the list
1047 of hosts for which the message was waiting, so if all were tried, there
1048 was no problem. Two fixes have been applied:
1050 (i) If there are any address or message errors in an SMTP delivery, none
1051 of the hosts (tried or untried) are now added to the list of hosts
1052 for which the message is waiting, so the message should not be a
1053 candidate for sending over the same connection that was used for a
1054 successful delivery of some other message. This seems entirely
1055 reasonable: after all the message is NOT "waiting for some host".
1056 This is so "obvious" that I'm not sure why it wasn't done
1057 previously. Hope I haven't missed anything, but it can't do any
1058 harm, as the worst effect is to miss an optimization.
1060 (ii) If, despite (i), such a delivery is accidentally attempted, the
1061 routing retry time is respected, so at least it doesn't keep
1062 hammering the server.
1064 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1065 in ${readsocket because some servers need this prod.
1067 PH/32 Added some extra debug output when updating a wait-xxx database.
1069 PH/33 The hint "could be header name not terminated by colon", which has been
1070 given for certain expansion errors for a long time, was not being given
1071 for the ${if def:h_colon_omitted{... case.
1073 PH/34 The spec says: "With one important exception, whenever a domain list is
1074 being scanned, $domain contains the subject domain." There was at least
1075 one case where this was not true.
1077 PH/35 The error "getsockname() failed: connection reset by peer" was being
1078 written to the panic log as well as the main log, but it isn't really
1079 panic-worthy as it just means the connection died rather early on. I have
1080 removed the panic log writing for the ECONNRESET error when getsockname()
1083 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1084 runs only) independently of the message's sender address. This meant
1085 that, if the 4xx error was in fact related to the sender, a different
1086 message to the same recipient with a different sender could confuse
1087 things. In particualar, this can happen when sending to a greylisting
1088 server, but other circumstances could also provoke similar problems.
1089 I have changed the default so that the retry time for these errors is now
1090 based a combination of the sender and recipient addresses. This change
1091 can be overridden by setting address_retry_include_sender=false in the
1094 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1095 remote server are returned as part of bounce messages. This was not
1096 happening for LMTP over a pipe (the lmtp transport), but now it is the
1097 same for both kinds of LMTP.
1099 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1100 in header lines that were in fact CNAMEs. This is no longer the case.
1102 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1103 and queue runs started by the daemon processed all messages. This has
1104 been fixed so that -R and -S can now usefully be given with -q<time>.
1106 PH/40 Import PCRE release 6.7 (fixes some bugs).
1108 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1110 PH/42 Give an error if -q is specified more than once.
1112 PH/43 Renamed the variables $interface_address and $interface_port as
1113 $received_ip_address and $received_port, to make it clear that these
1114 values apply to message reception, and not to the outgoing interface when
1115 a message is delivered. (The old names remain recognized, of course.)
1117 PH/44 There was no timeout on the connect() call when using a Unix domain
1118 socket in the ${readsocket expansion. There now is.
1120 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1121 be meaningful with "accept".
1123 SC/01 Eximstats V1.43
1124 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1126 SC/02 Eximstats V1.44
1127 Use a glob alias rather than an array ref in the generated
1128 parser. This improves both readability and performance.
1130 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1131 Collect SpamAssassin and rejection statistics.
1132 Don't display local sender or destination tables unless
1133 there is data to show.
1134 Added average volumes into the top table text output.
1136 SC/04 Eximstats V1.46
1137 Collect data on the number of addresses (recipients)
1138 as well as the number of messages.
1140 SC/05 Eximstats V1.47
1141 Added 'Message too big' to the list of mail rejection
1142 reasons (thanks to Marco Gaiarin).
1144 SC/06 Eximstats V1.48
1145 Mainlog lines which have GMT offsets and are too short to
1146 have a flag are now skipped.
1148 SC/07 Eximstats V1.49 (Alain Williams)
1149 Added the -emptyok flag.
1151 SC/08 Eximstats V1.50
1152 Fixes for obtaining the IP address from reject messages.
1154 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1155 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1156 whitesspace changes from 4.64-PH/27
1158 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1161 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1162 are found, allow negative numbers in numeric criteria)
1164 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1166 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1167 to match changes made in 4.64-PH/43
1169 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1171 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1172 "domain" is actually an IP address, and force a failure. This locks out
1173 those revolvers/nameservers that support "A-for-A" lookups, in
1174 contravention of the specifications.
1176 PH/48 When a host name was looked up from an IP address, and the subsequent
1177 forward lookup of the name timed out, the host name was left in
1178 $sender_host_name, contrary to the specification.
1180 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1181 restricted to single-key lookups, Exim was not diagnosing an error if
1182 * or *@ was used with a query-style lookup.
1184 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1186 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1187 long ago, but noone interested enough thought of it. Let's just say that
1188 the "1.1" means that there are some new functions that weren't there at
1189 some point in the past.
1191 PH/51 Error processing for expansion failure of helo_data from an smtp
1192 transport during callout processing was broken.
1194 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1195 tested/used via the -bh/-bhc/-bs options.
1197 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1198 bug, fixed in subsequent PCRE releases).
1200 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1201 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1203 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1209 SC/01 Use a glob alias rather than an array ref in eximstats generated
1210 parser. This improves both readability and performance.
1212 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1213 Don't display local sender or destination tables in eximstats unless
1214 there is data to show.
1215 Added average volumes into the eximstats top table text output.
1217 SC/03 Collect data on the number of addresses (recipients) as well
1218 as the number of messages in eximstats.
1220 TF/01 Correct an error in the documentation for the redirect router. Exim
1221 does (usually) call initgroups() when daemonizing.
1223 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1224 with consistent privilege compared to when running as a daemon.
1226 TF/03 Note in the spec that $authenticated_id is not set for local
1227 submissions from trusted users.
1229 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1230 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1232 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1233 by adding some example configuration directives to the default
1234 configuration file. A little bit of work is required to uncomment the
1235 directives and define how usernames and passwords are checked, but
1236 there is now a framework to start from.
1238 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1239 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1240 without this. I don't know how relevant this is to other LDAP libraries.
1242 PH/02 Add the verb name to the "unknown ACL verb" error.
1244 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1246 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1248 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1249 directory not expanded when it should be if an expanded home directory
1250 was set for the address (which is overridden by the transport).
1252 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1255 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1256 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1257 because it is too late at that time, and has no effect.
1259 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1260 security issue with \' (bugzilla #107). I could not use the
1261 PQescapeStringConn() function, because it needs a PGconn value as one of
1264 PH/08 When testing addresses using -bt, indicate those final addresses that
1265 are duplicates that would not cause an additional delivery. At least one
1266 person was confused, thinking that -bt output corresponded to deliveries.
1267 (Suppressing duplicates isn't a good idea as you lose the information
1268 about possibly different redirections that led to the duplicates.)
1270 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1271 systems where poll() doesn't work, in particular OS X.
1273 PH/10 Added more information to debugging output for retry time not reached.
1275 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1276 operations in malware.c.
1278 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1281 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1282 syslog_duplication set false, log lines that would normally be written
1283 both the the main log and to the reject log were not written to syslog at
1286 PH/14 In the default configuration, change the use of "message" in ACL warn
1287 statements to "add_header".
1289 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1290 not followed by a command (e.g. "seen endif").
1292 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1293 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1296 PH/17 Added extra conditions to the default value of delay_warning_condition
1300 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1301 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1302 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1305 The Auto-Submitted: and various List- headers are standardised, whereas I
1306 don't think Precedence: ever was.
1308 PH/18 Refactored debugging code in route_finduser() to show more information,
1309 in particular, the error code if getpwnam() issues one.
1311 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1312 This is apparently needed in addition to the PH/07 change above to avoid
1313 any possible encoding problems.
1315 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1316 but not after initializing Perl.
1318 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1319 output them only if debugging. By default they are written stderr,
1320 apparently, which is not desirable.
1322 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1325 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1328 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1330 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1331 authenticated or an ident call has been made. Suppress the default
1332 values for $authenticated_id and $authenticated_sender (but permit -oMai
1333 and -oMas) when testing with -bh.
1335 PH/24 Re-jigged the order of the tests in the default configuration so that the
1336 tests for valid domains and recipients precede the DNS black list and CSA
1337 tests, on the grounds that those ones are more expensive.
1339 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1340 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1341 This bug exists in every version of Exim that I still have, right back to
1344 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1345 However, an attempt to turn on case-sensitivity in a regex key by
1346 including (?-i) didn't work because the subject string was already
1347 lowercased, and the effects were non-intuitive. It turns out that a
1348 one-line patch can be used to allow (?-i) to work as expected.
1354 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1355 other effects) broke the use of negated acl sub-conditions.
1357 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1360 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1361 "Deny" causes Exim to reject the incoming connection with a 554 error.
1362 Unfortunately, if there is a major crisis, such as a disk failure,
1363 tcp-wrappers gives "deny", whereas what one would like would be some
1364 kind of temporary error. A kludge has been added to help with this.
1365 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1366 554 error is used if errno is still zero or contains ENOENT (which occurs
1367 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1370 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1372 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1373 errors. Otherwise a message that provokes a temporary error (when other
1374 messages do not) can cause a whole host to time out.
1376 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1377 addresses were routed directly to files or pipes from a redirect router.
1378 File deliveries just didn't batch; pipe deliveries might have suffered
1381 PH/06 A failure to get a lock for a hints database would erroneously always say
1382 "Failed to get write lock", even when it was really a read lock.
1384 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1385 of 0600. This has been changed to use the value of the lockfile_mode
1386 option (which defaults to 0600).
1388 PH/08 Applied small patch from the Sieve maintainer.
1390 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1391 folder from quota calculations, a direct delivery into this folder messed
1392 up the contents of the maildirsize file. This was because the regex was
1393 used only to exclude .Trash (or whatever) when the size of the mailbox
1394 was calculated. There was no check that a delivery was happening into an
1395 excluded directory. This bug has been fixed by ignoring all quota
1396 processing for deliveries into excluded directories.
1398 PH/10 Added the maildirfolder_create_regex option to appendfile.
1404 PH/01 The code for finding all the local interface addresses on a FreeBSD
1405 system running IPv6 was broken. This may well have applied to all BSD
1406 systems, as well as to others that have similar system calls. The broken
1407 code found IPv4 interfaces correctly, but gave incorrect values for the
1408 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1409 that it would not match correctly against @[] and not recognize the IPv6
1412 PH/02 The ipliteral router was not recognizing addresses of the form user@
1413 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1415 PH/03 Added disable_ipv6.
1417 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1418 decoded form, because it is most often used to construct To: headers
1419 lines in autoreplies, and the decoded form may well be syntactically
1420 invalid. However, $reply_address has leading white space removed, and all
1421 newlines turned into spaces so that the autoreply transport does not
1424 PH/05 If group was specified without a user on a router, and no group or user
1425 was specified on a transport, the group from the router was ignored.
1427 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1428 for visible compile-time settings that can be used to change these
1429 numbers, for those that want even more. Backwards compatibility with old
1430 spool files has been maintained. However, going back to a previous Exim
1431 release will lost any variables that are in spool files.
1433 PH/07 Two small changes when running in the test harness: increase delay when
1434 passing a TCP/IP connection to a new process, in case the original
1435 process has to generate a bounce, and remove special handling of
1436 127.0.0.2 (sic), which is no longer necessary.
1438 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1439 be the same on different OS.
1441 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1444 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1445 whether --show-vars was specified or not
1447 JJ/02 exipick: Added support for new ACL variable spool format introduced
1450 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1451 syntactically invalid From: or Reply-to: line, and a filter used this to
1452 generate an autoreply, and therefore failed to obtain an address for the
1453 autoreply, Exim could try to deliver to a non-existent relative file
1454 name, causing unrelated and misleading errors. What now happens is that
1455 it logs this as a hard delivery error, but does not attempt to create a
1458 PH/11 The exinext utility has a -C option for testing purposes, but although
1459 the given file was scanned by exinext itself; it wasn't being passed on
1460 when Exim was called.
1462 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1463 an end-of-file indication when reading a command response.
1465 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1466 compiled. In many other places in Exim, IPv6 addresses are always
1467 recognized, so I have changed this. It also means that IPv4 domain
1468 literals of the form [IPV4:n.n.n.n] are now always recognized.
1470 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1471 used if the router is not running as root, for example, when verifying at
1472 ACL time, or when using -bh. The debugging output from this situation was
1473 non-existent - all you got was a failure to exec. I have made two
1476 (a) Failures to set uid/gid, the current directory, or a process leader
1477 in a subprocess such as that created by queryprogram now generate
1478 suitable debugging ouput when -d is set.
1480 (b) The queryprogram router detects when it is not running as root,
1481 outputs suitable debugging information if -d is set, and then runs
1482 the subprocess without attempting to change uid/gid.
1484 PH/15 Minor change to Makefile for building test_host (undocumented testing
1487 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1488 additional section of a DNS packet that returns MX or SRV records.
1489 Instead, it always explicitly searches for A/AAAA records. This avoids
1490 major problems that occur when a DNS server includes only records of one
1491 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1492 fixed another bug: if SRV records were looked up and the corresponding
1493 address records were *not* found in the additional section, the port
1494 values from the SRV records were lost.
1496 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1497 using the correct key (the original address) when searching the retry
1498 rules in order to find which one to use for generating the retry hint.
1500 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1501 adding the default one. Similarly, if it contains a Reply-To: header, the
1502 errors_reply_to option, if set, is not used.
1504 PH/19 When calculating a retry time, Exim used to measure the "time since
1505 failure" by looking at the "first failed" field in the retry record. Now
1506 it does not use this if it is later than than the arrival time of the
1507 message. Instead it uses the arrival time. This makes for better
1508 behaviour in cases where some deliveries succeed, thus re-setting the
1509 "first failed" field. An example is a quota failure for a huge message
1510 when small messages continue to be delivered. Without this change, the
1511 "time since failure" will always be short, possible causing more frequent
1512 delivery attempts for the huge message than are intended.
1513 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1515 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1516 $1, $2, $3) because the numerical variables can be reset during some
1517 expansion items (e.g. "match"), thereby losing the authentication data.
1519 PH/21 Make -bV show the size of off_t variables so that the test suite can
1520 decide whether to run tests for quotas > 2G.
1522 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1523 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1524 filecount value is greater than 2G or if a quota value is greater than 2G
1525 on a system where the size of off_t is not greater than 4, a panic error
1528 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1529 never match. The debug and -bh output now contains an explicit error
1530 message indicating a malformed IPv4 address or mask.
1532 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1533 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1534 PH/23 above applies.
1536 PH/25 Do not write to syslog when running in the test harness. The only
1537 occasion when this arises is a failure to open the main or panic logs
1538 (for which there is an explicit test).
1540 PH/26 Added the /no_tell option to "control=freeze".
1542 PH/27 If a host name lookup failed very early in a connection, for example, if
1543 the IP address matched host_lookup and the reverse lookup yielded a name
1544 that did not have a forward lookup, an error message of the form "no IP
1545 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1546 could be logged. Now it outputs the IP address instead of "NULL".
1548 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1549 allows the sender and the authenticated sender to be set when
1550 submitting a message from within Exim. Since child_open_exim() is
1551 documented for local_scan(), the new function should be too.
1553 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1554 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1555 results in an empty string is now treated as unset.
1557 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1559 PH/31 Added sender_verify_fail logging option.
1561 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1562 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1563 tidied the source and removed it altogether.
1565 PH/33 When a queue run was abandoned because the load average was too high, a
1566 log line was always written; now it is written only if the queue_run log
1567 selector is set. In addition, the log line for abandonment now contains
1568 information about the queue run such as the pid. This is always present
1569 in "start" and "stop" lines but was omitted from the "abandon" line.
1571 PH/34 Omit spaces between a header name and the colon in the error message that
1572 is given when verify = headers_syntax fails (if there are lots of them,
1573 the message gets confusing).
1575 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1576 names, as there are now some PTR records that contain slashes. This check
1577 is only to protect against broken name servers that fall over on strange
1578 characters, so the fact that it applies to all lookups doesn't matter.
1580 PH/36 Now that the new test suite is complete, we can remove some of the
1581 special code in Exim that was needed for the old test suite. For example,
1582 sorting DNS records because real resolvers return them in an arbitrary
1583 order. The new test suite's fake resolver always returns records in the
1586 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1587 bounces) except when queue_only is set, to avoid logging races between
1588 the different processes.
1590 PH/38 Panic-die if .include specifies a non-absolute path.
1592 PH/39 A tweak to the "H" retry rule from its user.
1594 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1595 a label. They prevented compilation on older perls.
1597 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1598 a warning to be raised on newish perls.
1600 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1601 on queue. Changes to match documented behaviour of showing count of
1602 messages matching specified criteria.
1604 PH/40 Changed the default ident timeout from 30s to 5s.
1606 PH/41 Added support for the use of login_cap features, on those BSD systems
1607 that have them, for controlling the resources used by pipe deliveries.
1609 PH/42 The content-scanning code uses fopen() to create files in which to put
1610 message data. Previously it was not paying any attention to the mode of
1611 the files. Exim runs with umask(0) because the rest of the code creates
1612 files with open(), and sets the required mode explicitly. Thus, these
1613 files were ending up world-writeable. This was not a big issue, because,
1614 being within the spool directory, they were not world-accessible. I have
1615 created a function called modefopen, which takes an additional mode
1616 argument. It sets umask(777), creates the file, chmods it to the required
1617 mode, then resets the umask. All the relevant calls to fopen() in the
1618 content scanning code have been changed to use this function.
1620 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1621 to 24 hours. This avoids potential overflow problems when processing G
1622 and H retry rules. I suspect nobody ever tinkers with this value.
1624 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1626 PH/45 When the plaintext authenticator is running as a client, the server's
1627 challenges are checked to ensure they are valid base64 strings. By
1628 default, the authentication attempt is cancelled if an invalid string is
1629 received. Setting client_ignore_invalid_base64 true ignores these errors.
1630 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1631 they are received. Thus, the responses can be made to depend on the
1632 challenges. If an invalid string is ignored, an empty string is placed in
1635 PH/46 Messages that are created by the autoreply transport now contains a
1636 References: header, in accordance with RFCs 2822 and 3834.
1638 PH/47 Added authenticated_sender_force to the smtp transport.
1640 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1642 PH/49 Installed latest patch from the Sieve maintainer.
1644 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1645 was also set, the appendfile transport was unnecessarily scanning a
1646 directory of message files (e.g. for maildir delivery) to find the count
1647 of files (along with the size), even though it did not need this
1648 information. It now does the scan only if it needs to find either the
1649 size of the count of files.
1651 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1653 PH/52 Two bugs concerned with error handling when the smtp transport is
1656 (i) Exim was not creating retry information for temporary errors given
1657 for individual recipients after the DATA command when the smtp transport
1658 was used in LMTP mode. This meant that they could be retried too
1659 frequently, and not timed out correctly.
1661 (ii) Exim was setting the flag that allows error details to be returned
1662 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1663 recipients that were returned after the DATA command.
1665 PH/53 This is related to PH/52, but is more general: for any failing address,
1666 when detailed error information was permitted to be returned to the
1667 sender, but the error was temporary, then after the final timeout, only
1668 "retry timeout exceeded" was returned. Now it returns the full error as
1669 well as "retry timeout exceeded".
1671 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1672 do this, and (what is worse) MTAs that accept it.
1674 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1675 will now be deprecated.
1677 PH/56 New os.c-cygwin from the Cygwin maintainer.
1679 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1680 formats (previously only available in exim formats via -bpr, -bpru,
1681 and -bpra. Now also available in native and exiqgrep formats)
1683 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1684 with very large, slow to parse queues
1686 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1688 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1690 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1691 responses to authentication challenges, though it was showing the
1692 challenges; (ii) I've removed the CR characters from the debug output for
1695 PH/58 Allow for the insertion of a newline as well as a space when a string
1696 is turned into more than one encoded-word during RFC 2047 encoding. The
1697 Sieve code now uses this.
1699 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1700 data_4xx, lost_connection, tls_required.
1702 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1703 message was being sent as an SMTP response.
1705 PH/61 Add -l and -k options to exicyclog.
1707 PH/62 When verifying, if an address was redirected to one new address, so that
1708 verification continued, and the new address failed or deferred after
1709 having set something in $address_data, the value of $address_data was not
1710 passed back to the ACL. This was different to the case when no
1711 redirection occurred. The value is now passed back in both cases.
1713 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1714 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1715 use login_cap.h, so on its own it isn't the distinguishing feature. The
1716 new name refers directly to the setclassresources() function.
1718 PH/65 Added configuration files for NetBSD3.
1720 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1722 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1723 is preferred over IPv4.
1725 PH/68 The bounce_return_message and bounce_return_body options were not being
1726 honoured for bounces generated during the reception of non-SMTP messages.
1727 In particular, this applied to messages rejected by the ACL. This bug has
1728 been fixed. However, if bounce_return_message is true and bounce_return_
1729 body is false, the headers that are returned for a non-SMTP message
1730 include only those that have been read before the error was detected.
1731 (In the case of an ACL rejection, they have all been read.)
1733 PH/69 The HTML version of the specification is now built in a directory called
1734 spec_html instead of spec.html, because the latter looks like a path with
1735 a MIME-type, and this confuses some software.
1737 PH/70 Catch two compiler warnings in sieve.c.
1739 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1740 function verify_get_ident() calls ip_connect() to connect a socket, but
1741 if the "connect()" function timed out, ip_connect() used to close the
1742 socket. However, verify_get_ident() also closes the socket later, and in
1743 between Exim writes to the log, which may get opened at this point. When
1744 the socket was closed in ip_connect(), the log could get the same file
1745 descriptor number as the socket. This naturally causes chaos. The fix is
1746 not to close the socket in ip_connect(); the socket should be closed by
1747 the function that creates it. There was only one place in the code where
1748 this was missing, in the iplookup router, which I don't think anybody now
1749 uses, but I've fixed it anyway.
1751 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1752 well as to direct DNS lookups. Otherwise the handling of names in host
1753 lists is inconsistent and therefore confusing.
1759 PH/01 Two changes to the default runtime configuration:
1761 (1) Move the checks for relay_from_hosts and authenticated clients from
1762 after to before the (commented out) DNS black list checks.
1764 (2) Add control=submission to the relay_from_hosts and authenticated
1765 clients checks, on the grounds that messages accepted by these
1766 statements are most likely to be submissions.
1768 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1770 (1) Generate an error if the third argument for the ${prvs expansion is
1773 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1776 (3) Reset the variables that are obtained from the first argument of
1777 ${prvscheck and used in the second argument before leaving the code,
1778 because their memory is reclaimed, so using them afterwards may do
1781 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1782 one (it's much easier than Tom thought :-).
1784 (5) Because of (4), we can now allow for the use of $prvscheck_result
1785 inside the third argument.
1787 PH/03 For some reason, the default setting of PATH when running a command from
1788 a pipe transport was just "/usr/bin". I have changed it to
1791 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1792 anything to be listed in the output from -bV.
1794 PH/05 When a filter generated an autoreply, the entire To: header line was
1795 quoted in the delivery log line, like this:
1797 => >A.N.Other <ano@some.domain> <original@ddress> ...
1799 This has been changed so that it extracts the operative address. There
1800 may be more than one such address. If so, they are comma-separated, like
1803 => >ano@some.domain,ona@other.domain <original@ddress> ...
1805 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1806 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1807 not being looked up in the rDNS to get a host name, Exim was showing the
1808 IP address twice in Received: lines, even though the IP addresses were
1809 identical. For example:
1811 Received: from [1.2.3.4] (helo=[1.2.3.4])
1813 However, if the real host name was known, it was omitting the HELO data
1814 if it matched the actual IP address. This has been tidied up so that it
1815 doesn't show the same IP address twice.
1817 PH/07 When both +timestamp and +memory debugging was on, the value given by
1818 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1819 called by the debug printing, thereby overwriting the timestamp buffer.
1820 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1822 PH/08 When the original message was included in an autoreply transport, it
1823 always said "this is a copy of the message, including all the headers",
1824 even if body_only or headers_only was set. It now gives an appropriate
1827 PH/09 Applied a patch from the Sieve maintainer which:
1829 o fixes some comments
1830 o adds the (disabled) notify extension core
1831 o adds some debug output for the result of if/elsif tests
1832 o points to the current vacation draft in the documentation
1833 and documents the missing references header update
1837 o fixes a bug in processing the envelope test (when testing
1838 multiple envelope elements, the last element determinted the
1841 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1842 Electronic Mail") by including:
1844 Auto-submitted: auto-generated
1846 in the messages that it generates (bounce messages and others, such as
1847 warnings). In the case of bounce messages for non-SMTP mesages, there was
1848 also a typo: it was using "Auto_submitted" (underscore instead of
1849 hyphen). Since every message generated by Exim is necessarily in response
1850 to another message, thes have all been changed to:
1852 Auto-Submitted: auto-replied
1854 in accordance with these statements in the RFC:
1856 The auto-replied keyword:
1858 - SHOULD be used on messages sent in direct response to another
1859 message by an automatic process,
1861 - MUST NOT be used on manually-generated messages,
1863 - MAY be used on Delivery Status Notifications (DSNs) and Message
1864 Disposition Notifications (MDNs),
1866 - MUST NOT be used on messages generated by automatic or periodic
1867 processes, except for messages which are automatic responses to
1870 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1871 to the default Received: header definition.
1873 PH/12 Added log selector acl_warn_skipped (default on).
1875 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1876 variables because (a) they are in the wrong storage pool and (b) even if
1877 they were copied, it wouldn't work properly because of the caching.
1879 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1880 checking when decoding. Apparently there are clients that generate
1881 overlong encoded strings. Why am I not surprised?
1883 PH/15 If the first argument of "${if match_address" was not empty, but did not
1884 contain an "@" character, Exim crashed. Now it writes a panic log message
1885 and treats the condition as false.
1887 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1889 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1890 extension "envelope-auth". The code is finished and in agreement with
1891 other implementations, but there is no documentation so far and in fact,
1892 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1893 not changing the active code.
1895 Print executed "if" and "elsif" statements when debugging is used. This
1896 helps a great deal to understand what a filter does.
1898 Document more things not specified clearly in RFC3028. I had all this
1899 sorted out, when out of a sudden new issues came to my mind. Oops."
1901 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1904 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1905 canonical form (as documented). However, after a host name lookup from
1906 the IP address, check_host() was doing a simple string comparison with
1907 addresses acquired from the DNS when checking that the found name did
1908 have the original IP as one of its addresses. Since any found IPv6
1909 addresses are likely to be in abbreviated form, the comparison could
1910 fail. Luckily, there already exists a function for doing the comparison
1911 by converting both addresses to binary, so now that is used instead of
1912 the text comparison.
1914 PH/20 There was another similar case to PH/19, when a complete host name was
1915 given in a host list; looking up its IP address could give an abbreviated
1916 form, whereas the current host's name might or might not be abbreviated.
1917 The same fix has been applied.
1923 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1924 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1927 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1928 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1930 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1932 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1933 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1934 Also, for "Auto-Submitted", treat anything other than "no" as
1935 non-personal, in accordance with RFC 3834. (Previously it treated
1936 anything starting "auto-" as non-personal.)
1938 TF/01 The control=submission/name=... option had a problem with syntax
1939 errors if the name included a slash character. The /name= option
1940 now slurps the rest of the string, so it can include any characters
1941 but it must come last in the list of options (after /sender_retain
1944 PH/05 Some modifications to the interface to the fake nameserver for the new
1952 TK/01 Added the "success_on_redirect" address verification option. See
1953 NewStuff for rationale and an example.
1955 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1957 PH/02 Patch to exigrep to allow it to work on syslog lines.
1959 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1960 fread() to skip over the body file's header line, because in Cygwin the
1961 header line is locked and is inaccessible.
1963 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1964 co-exist for some time) to make it clear that it is the Exim ID that is
1965 referenced, not the Message-ID: header line.
1967 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1968 string_format() function, because snprintf() does not exist on all
1971 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1972 ${stat: expansion item.
1974 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1975 protocol synchronization error", to keep the pedants happy.
1977 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1978 well as for IRIX systems, when gcc is being used. See the host.c source
1981 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1983 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1986 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1987 transport and to the smtp transport in LMTP mode.
1989 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1991 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1992 run a filter in a subprocess. This could lead to confusion in subsequent
1993 lookups in the parent process. There should also be a search_tidyup() at
1994 the end of the subprocess.
1996 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1997 only if the host matched helo_try_verify_hosts, which caused the
1998 verification to occur when the EHLO/HELO command was issued. The ACL just
1999 tested the remembered result. Now, if a previous verification attempt has
2000 not happened, "verify = helo" does it there and then.
2002 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2004 TK/03 Fix log output including CR from clamd.
2006 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2007 exist provoked a memory error which could cause a segfault.
2009 PH/15 Installed PCRE 6.2
2011 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2013 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2014 of the problem. Specifically, suggested +O2 rather than +O1 for the
2017 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2019 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2020 record, the port was not correctly logged when the outgoing_port log
2021 selector was set (it logged the transort's default port).
2023 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2024 fallback_hosts, and "hosts" in the smtp transport.
2026 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2027 host errors such as "Connection refused".
2029 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2030 authentication with radiusclient 0.4.9:
2032 - Error returned from rc_read_config was caught wrongly
2033 - Username/password not passed on to radius server due to wrong length.
2035 The presumption is that some radiusclient API changes for 4.51/PH/17
2036 were not taken care of correctly. The code is still untested by me (my
2037 Linux distribution still has 0.3.2 of radiusclient), but it was
2038 contributed by a Radius user.
2040 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2041 expanding the "port" option of the smtp transport.
2043 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2044 while reading a MIME header. Thanks to Tom Hughes for a patch.
2046 PH/24 Include config.h inside local_scan.h so that configuration settings are
2049 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2050 that in an ACL for RCPT (for example), you can examine exactly what was
2053 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2054 commands, but it was not correctly comparing the address with the actual
2055 client host address. Thus, it would show the EHLO address in Received:
2056 header lines when this was not necessary.
2058 PH/27 Added the % operator to ${eval:}.
2060 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2061 it should be ignoring failures (because with -C, for example, it has lost
2062 privilege). It wasn't ignoring creation failures other than "already
2065 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2068 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2069 because an input line was too long, either on its own, or by virtue of
2070 too many continuations, the temporary file was not being removed, and the
2071 return code was incorrect.
2073 PH/31 Missing "BOOL" in function definition in filtertest.c.
2075 PH/32 Applied Sieve patches from the maintainer.
2077 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2079 PH/33 Added "verify = not_blind".
2081 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2082 Local/Makefile (with some defaults set). These are used in built scripts
2083 such as exicyclog, but they have never been used in the exim_install
2084 script (though there are many overriding facilities there). I have
2085 arranged that the exim_install script now takes note of these two
2088 PH/35 Installed configuration files for Dragonfly.
2090 PH/36 When a locally submitted message by a trusted user did not contain a
2091 From: header, and the sender address was obtained from -f or from an SMTP
2092 MAIL command, and the trusted user did not use -F to supply a sender
2093 name, $originator_name was incorrectly used when constructing a From:
2094 header. Furthermore, $originator_name was used for submission mode
2095 messages from external hosts without From: headers in a similar way,
2096 which is clearly wrong.
2098 PH/37 Added control=suppress_local_fixups.
2100 PH/38 When log_selector = +received_sender was set, and the addition of the
2101 sender made the log line's construction buffer exactly full, or one byte
2102 less than full, an overflow happened when the terminating "\n" was
2105 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2106 when the result of a list match is failure because a DNS lookup failed.
2108 PH/40 RM_COMMAND is now used in the building process.
2110 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2111 the "build-* directories that it finds.
2113 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2114 address in a domain literal was a prefix of an interface address.
2116 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2117 when verifying a sender address, unless rewrite_headers is false.
2119 PH/44 Wrote a long comment about why errors_to addresses are verified as
2120 recipients, not senders.
2122 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2123 the ratelimit ACL was added.
2125 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2127 PH/46 Added extra information about PostgreSQL errors to the error string.
2129 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2130 avoiding the need to install special zones in a real server. This is
2131 backwards compatible; if it can't find the fake resolver, it drops back.
2132 Thus, both old and new test suites can be run.
2134 TF/02 Added util/ratelimit.pl
2136 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2137 clock is set back in time.
2139 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2140 Candler <B.Candler@pobox.com>.
2142 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2143 OK for addresses that are the result of redirections.
2145 PH/48 A number of further additions for the benefit of the new test suite,
2146 including a fake gethostbyname() that interfaces to the fake DNS resolver
2149 TF/06 The fix for widen_domains has also been applied to qualify_single and
2150 search_parents which are the other dnslookup options that can cause
2153 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2156 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2158 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2159 service" if it encountered an MX record with an empty target hostname.
2160 The message is now "an MX or SRV record indicated no SMTP service".
2162 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2163 if the DNS of the sending site is misconfigured. This is quite a
2164 common situation. This change restores the behaviour of treating a
2165 helo verification defer as a failure.
2167 PH/51 If self=fail was set on a router, the bounce message did not include the
2168 actual error message.
2174 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2176 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2177 command itself ended in error, the underlying message about the transport
2178 filter timeout was being overwritten with the pipe command error. Now the
2179 underlying error message should be appended to the second error message.
2181 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2183 PH/02 Reduce the amount of output that "make" produces by default. Full output
2184 can still be requested.
2186 PH/03 The warning log line about a condition test deferring for a "warn" verb
2187 was being output only once per connection, rather than after each
2188 occurrence (because it was using the same function as for successful
2189 "warn" verbs). This seems wrong, so I have changed it.
2191 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2192 it should not have, which might have caused a crash in the right
2193 circumstances, but probably never did.
2195 PH/04 Installed a modified version of Tony Finch's patch to make submission
2196 mode fix the return path as well as the Sender: header line, and to
2197 add a /name= option so that you can make the user's friendly name appear
2200 TF/03 Added the control = fakedefer ACL modifier.
2202 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2203 Mark Lowes for thorough testing.
2205 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2207 TK/03 Merged latest SRS patch from Miles Wilton.
2209 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2210 with the definition in sysexits.h (which is #included earlier).
2211 Fortunately, Exim does not actually use EX_OK. The code used to try to
2212 preserve the sysexits.h value, by assumimg that macro definitions were
2213 scanned for macro replacements. I have been disabused of this notion,
2214 so now the code just undefines EX_OK before #including unistd.h.
2216 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2217 in the smtp transport. When a block could not be written in a single
2218 write() function, the timeout was being re-applied to each part-write.
2219 This seems wrong - if the receiver was accepting one byte at a time it
2220 would take for ever. The timeout is now adjusted when this happens. It
2221 doesn't have to be particularly precise.
2223 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2224 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2226 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2227 without a domain if the check to <postmaster@domain> fails.
2229 SC/01 Eximstats: added -xls and the ability to specify output files
2230 (patch written by Frank Heydlauf).
2232 SC/02 Eximstats: use FileHandles for outputing results.
2234 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2236 SC/04 Eximstats: fixed display of large numbers with -nvr option
2238 SC/05 Eximstats: fixed merging of reports with empty tables.
2240 SC/06 Eximstats: added the -include_original_destination flag
2242 SC/07 Eximstats: removed tabs and trailing whitespace.
2244 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2246 TK/06 MBOX spool code: Add real "From " MBOX separator line
2247 so the .eml file is really in mbox format (even though
2248 most programs do not really care). Patch from Alex Miller.
2250 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2251 The latter is generated from $received_to and is only set if the
2252 message has one envelope recipient. SA can use these headers,
2253 obviously out-of-the-box. Patch from Alex Miller.
2255 PH/08 The ${def test on a variable was returning false if the variable's
2256 value was "0", contrary to what the specification has always said!
2257 The result should be true unless the variable is empty.
2259 PH/09 The syntax error of a character other than { following "${if
2260 def:variable_name" (after optional whitespace) was not being diagnosed.
2261 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2262 accidental colon was present, for example, could give incorrect results.
2264 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2265 result is used (not including appendfile, where other changes are about
2268 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2269 This involved changing a lot of size variables from int to off_t. It
2270 should work with maildirs and everything.
2272 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2273 spamd dying while we are connected to it.
2275 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2278 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2279 function for BDB 4.3.
2281 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2283 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2284 the sources for PCRE have been re-arranged and more files are now
2287 PH/15 The code I had for printing potentially long long variables in PH/11
2288 above was not the best (it lost precision). The length of off_t variables
2289 is now inspected at build time, and an appropriate printing format (%ld
2290 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2291 to be "long long int" or "long int". This is needed for the internal
2292 formatting function string_vformat().
2294 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2295 the configuration file to be ":syslog", then the script "guesses" where
2296 the logs files are, rather than using the compiled in default. In our
2297 case the guess is not the same as the compiled default, so the script
2298 suddenly stopped working when I started to use syslog. The patch checks
2299 to see if log_file_path is "". If so, it attempts to read it from exim
2300 with no configuration file to get the compiled in version, before it
2301 falls back to the previous guessing code."
2303 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2304 implementing BATV in an Exim configuration. See NewStuff for the gory
2307 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2308 Makefile that are specific to HP-UX.
2310 PH/18 If the "use_postmaster" option was set for a recipient callout together
2311 with the "random" option, the postmaster address was used as the MAIL
2312 FROM address for the random test, but not for the subsequent recipient
2313 test. It is now used for both.
2315 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2316 patch removes a few documentation additions to RFC 3028, because the
2317 latest draft now contains them. It adds the new en;ascii-case comparator
2318 and a new error check for 8bit text in MIME parts. Comparator and
2319 require names are now matched exactly. I enabled the subaddress
2320 extension, but it is not well tested yet (read: it works for me)."
2322 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2323 rework some of the code of TK/09 above to avoid the hardwired use of
2324 "%lld" and "long long". Replaced the call to snprintf() with a call to
2327 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2328 records point to non-existent hosts", "retry timeout exceeded", and
2329 "retry time not reached for any host after a long failure period".
2331 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2332 experimental DomainKeys support:
2334 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2335 (2) On an error such as an illegally used "control", the wrong name for
2336 the control was given.
2338 These problems did NOT occur unless DomainKeys support was compiled.
2340 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2342 PH/24 Added ${if match_ip condition.
2344 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2345 databases so that it will be absolutely obvious if a crash occurs in the
2346 DB library. This is a regular occurrence (often caused by mis-matched
2349 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2350 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2351 picked up on a user's system that detects such things. There doesn't seem
2352 to be a gcc warning option for this - only an attribute that has to be
2353 put on the function's prototype. It seems that in Fedora Core 4 they have
2354 set this on a number of new functions. No doubt there will be more in due
2357 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2358 specify a transport. However, if an address that was verified by such a
2359 router was the subject of a callout, Exim crashed because it tried to
2360 read the rcpt_include_affixes from the non-existent transport. Now it
2361 just assumes that the setting of that option is false. This bug was
2362 introduced by 4.51/PH/31.
2364 PH/28 Changed -d+all to exclude +memory, because that information is very
2365 rarely of interest, but it makes the output a lot bigger. People tend to
2366 do -d+all out of habit.
2368 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2369 code in os-type was giving problems when libc.so lives in lib64, like on
2372 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2373 aren't the modern standard, and it seems that some systems' include files
2374 don't always have them. Exim was already checking for some of the newer
2375 ones like T_AAAA, and defining it itself. I've added checks for all the
2376 record types that Exim uses.
2378 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2379 not automatically generating a new one, as it is supposed to. This
2380 prevented TLS from working. If the file did exist, but contained invalid
2381 data, a new version was generated, as expected. It was only the case of a
2382 non-existent file that was broken.
2384 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2385 with a change in libdomainkeys > 0.64.
2387 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2388 from DNS. If the selector record carries the flag, it now has
2389 precedence over the domain-wide flag.
2391 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2393 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2394 the use of a port name that isn't defined in /etc/services) occurred, the
2395 message was deferred as in a normal delivery, and thus remained on the
2396 spool, instead of being failed because of the mua_wrapper setting. This
2397 is now fixed, and I tidied up some of the mua_wrapper messages at the
2400 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2401 the messages in a hash of arrays rather than using individual hashes.
2402 This is a bit cleaner and results in dramatic memory savings, albeit
2403 at a slight CPU cost.
2405 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2406 as requested by Marc Sherman.
2408 SC/10 Eximstats: added histograms for user specified patterns as requested
2411 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2413 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2414 fopen() in the content-scanning modules that did not already have it.
2420 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2421 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2423 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2425 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2427 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2428 to test Sieve filters that use "vacation".
2430 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2431 that changes the way the GnuTLS parameters are stored in the cache file.
2432 The new format can be generated externally. For backward compatibility,
2433 if the data in the cache doesn't make sense, Exim assumes it has read an
2434 old-format file, and it generates new data and writes a new file. This
2435 means that you can't go back to an older release without removing the
2438 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2439 work if there are any delivery delays because "one_time" forces the
2440 parent to be marked "delivered", so its unseen clone is never tried
2441 again. For this reason, Exim now forbids the simultaneous setting of
2444 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2445 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2446 entry if you want to know the details. The fix, however, neglected to
2447 consider the case when local delivery batching is involved. The test for
2448 "previously delivered" was not happening when checking to see if an
2449 address could be batched with a previous (undelivered) one; under
2450 certain circumstances this could lead to multiple deliveries to the same
2453 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2454 in its include files, and this causes problems building Exim.
2456 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2457 header_syntax) but Exim was just ignoring anything given after a slash.
2458 In particular, this caused confusion with an attempt to use "verify =
2459 reverse_host_lookup/defer_ok". An error is now given when options are
2460 supplied for verify items that do not have them. (Maybe reverse_host_
2461 lookup should have a defer_ok option, but that's a different point.)
2463 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2464 defined by RFC 821) to 2048, because there were problems with some AUTH
2465 commands, and RFC 1869 says the size should be increased for extended
2466 SMTP commands that take arguments.
2468 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2471 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2472 "unknown" error; now it says that the functionality isn't in the binary.
2474 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2475 an address' error message when a string expansion fails (syntax or
2476 whatever). Otherwise the password may appear in the log. Following change
2477 PH/42 below, there is no longer a chance of it appearing in a bounce
2480 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2482 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2483 of its addresses was ever tried. (Bugzilla bug #2.)
2485 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2486 the result incorrectly in the debug output. (It correctly added a newline
2487 to what was transported.)
2489 TF/01 Added $received_time.
2491 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2492 commented out examples of how to interface to a virus scanner and to
2493 SpamAssassin. Also added commented examples of av_scanner and
2494 spamd_address settings.
2496 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2497 and controls are allowed in which ACLs. There were a couple of minor
2498 errors. Some of the entries in the conditions table (which is a table of
2499 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2500 negation of where the condition IS allowed.
2502 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2504 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2505 header file does not have a version number, so I've had to invent a new
2506 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2507 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2508 radiusclient), but it was contributed by a Radius user.
2510 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2511 files or directories, for OpenSSL.
2513 PH/19 When an Exim process that is running as root has to create an Exim log
2514 file, it does so in a subprocess that runs as exim:exim so as to get the
2515 ownership right at creation (otherwise, other Exim processes might see
2516 the file with the wrong ownership). There was no test for failure of this
2517 fork() call, which would lead to the process getting stuck as it waited
2518 for a non-existent subprocess. Forks do occasionally fail when resources
2519 run out. I reviewed all the other calls to fork(); they all seem to check
2522 PH/20 When checking for unexpected SMTP input at connect time (before writing
2523 the banner), Exim was not dealing correctly with a non-positive return
2524 from the read() function. If the client had disconnected by this time,
2525 the result was a log entry for a synchronization error with an empty
2526 string after "input=" when read() returned zero. If read() returned -1
2527 (an event I could not check), uninitialized data bytes were printed.
2528 There were reports of junk text (parts of files, etc) appearing after
2531 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2533 PH/22 Added support for macro redefinition, and (re)definition in between
2534 driver and ACL definitions.
2536 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2537 forgetting to use the resulting value; it was using the unexpanded value.
2539 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2540 hadn't been configured. The fix is from Juergen Kreileder, who
2541 understands it better than I do:
2543 "Here's what I see happening with three configured cyrus_sasl
2544 authenticators configured (plain, login, cram-md5):
2546 On startup auth_cyrus_sasl_init() gets called for each of these.
2547 This means three calls to sasl_listmech() without a specified mech_list.
2548 => SASL tests which mechs of all available mechs actually work
2549 => three warnings about OTP not working
2550 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2552 With the patch, sasl_listmech() also gets called three times. But now
2553 SASL's mech_list option is set to the server_mech specified in the the
2554 authenticator. Or in other words, the answer from sasl_listmech()
2555 gets limited to just the mech you're testing for (which is different
2557 => the return list contains just 'plain' or 'login', 'cram-md5' or
2558 nothing depending on the value of ob->server_mech.
2560 I've just tested the patch: Authentication still works fine,
2561 unavailable mechs specified in the exim configuration are still
2562 caught, and the auth.log warnings about OTP are gone."
2564 PH/25 When debugging is enabled, the contents of the command line are added
2565 to the debugging output, even when log_selector=+arguments is not
2568 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2569 answer is "GNU", and only if the return is "GNU/something" is the answer
2572 PH/27 $acl_verify_message is now set immediately after the failure of a
2573 verification in an ACL, and so is available in subsequent modifiers. In
2574 particular, the message can be preserved by coding like this:
2576 warn !verify = sender
2577 set acl_m0 = $acl_verify_message
2579 Previously, $acl_verify_message was set only while expanding "message"
2580 and "log_message" when a very denied access.
2582 PH/28 Modified OS/os.c-Linux with
2584 -#ifndef OS_LOAD_AVERAGE
2585 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2587 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2588 nomenclature these days.)
2590 PH/29 Installed patch from the Sieve maintainer that adds the options
2591 sieve_useraddress and sieve_subaddress to the redirect router.
2593 PH/30 In these circumstances:
2594 . Two addresses routed to the same list of hosts;
2595 . First host does not offer TLS;
2596 . First host accepts first address;
2597 . First host gives temporary error to second address;
2598 . Second host offers TLS and a TLS session is established;
2599 . Second host accepts second address.
2600 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2601 and peerdn, if requested) that were in fact used only for the second
2604 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2605 attention to any local part prefix or suffix that was matched by the
2606 router that accepted the address. It now behaves in the same way as it
2607 does for delivery: the affixes are removed from the local part unless
2608 rcpt_include_affixes is set on the transport.
2610 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2611 timeout during the DATA phase of an incoming message.
2613 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2614 applied a patch sanctioned by the Sieve maintainer.
2616 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2617 the uid or gid is negative. A case of a negative gid caused this to be
2618 noticed. The fix allows for either to be negative.
2620 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2621 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2624 PH/36 The change PH/12 above was broken. Fixed it.
2626 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2627 the grounds that routing the same address twice would always produce the
2628 same answer. This might have been true once, but it is certainly no
2629 longer true now. Routing a child address may depend on the previous
2630 routing that produced that child. Some complicated redirection strategies
2631 went wrong when messages had multiple recipients, and made Exim's
2632 behaviour dependent on the order in which the addresses were given.
2634 I have moved the duplicate checking until after the routing is complete.
2635 Exim scans the addresses that are assigned to local and remote
2636 transports, and removes any duplicates. This means that more work will be
2637 done, as duplicates will always all be routed, but duplicates are
2638 presumably rare, so I don't expect this is of any significance.
2640 For deliveries to pipes, files, and autoreplies, the duplicate checking
2641 still happens during the routing process, since they are not going to be
2644 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2645 It corrects a timeout issue with spamd. This is Ian's comment: "The
2646 background is that sometimes spamd either never reads data from a
2647 connection it has accepted, or it never writes response data. The exiscan
2648 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2649 blindly assumes that writes won't block so it may never time out."
2651 PH/39 Allow G after quota size as well as K and M.
2653 PH/40 The value set for $authenticated_id in an authenticator may not contain
2654 binary zeroes or newlines because the value is written to log lines and
2655 to spool files. There was no check on this. Now the value is run through
2656 the string_printing() function so that such characters are converted to
2657 printable escape sequences.
2659 PH/41 $message_linecount is a new variable that contains the total number of
2660 lines in the message. Compare $body_linecount, which is the count for the
2663 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2664 bounce and delay warning messages, except in certain special cases, which
2667 (a) An SMTP error message from a remote host;
2668 (b) A message specified in a :fail: redirection;
2669 (c) A message specified in a "fail" command in a system filter;
2670 (d) A message specified in a FAIL return from the queryprogram router;
2671 (e) A message specified by the cannot_route_message router option.
2673 In these cases only, Exim does include the error details in bounce and
2674 warning messages. There are also a few cases where bland messages such
2675 as "unrouteable address" or "local delivery error" are given.
2677 PH/43 $value is now also set for the "else" part of a ${run expansion.
2679 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2680 being worked on, but at least Exim now implements the latest version to
2683 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2684 process to complete was treated as a delivery failure, a timeout while
2685 writing the message to the pipe was logged, but erroneously treated as a
2686 successful delivery. Such timeouts include transport filter timeouts. For
2687 consistency with the overall process timeout, these timeouts are now
2688 treated as errors, giving rise to delivery failures by default. However,
2689 there is now a new Boolean option for the pipe transport called
2690 timeout_defer, which, if set TRUE, converts the failures into defers for
2691 both kinds of timeout. A transport filter timeout is now identified in
2694 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2695 systems where "make" and "gmake" are different, calling "gmake" at top
2696 level broke things. I've arranged for the value of $(MAKE) to be passed
2697 from the Makefile to this script so that it can call the same version of
2701 A note about Exim versions 4.44 and 4.50
2702 ----------------------------------------
2704 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2705 changes of various kinds. As a consequence, a big documentation update was
2706 needed. This delayed the release for rather longer than seemed good, especially
2707 in the light of a couple of (minor) security issues. Therefore, the changes
2708 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2709 release. So 4.44 and 4.50 are in effect two different branches that both start
2712 I have left the 4.50 change log unchanged; it contains all the changes since
2713 4.43. The change log for 4.44 is below; many of its items are identical to
2714 those for 4.50. This seems to be the most sensible way to preserve the
2715 historical information.
2721 1. Minor wording change to the doc/README.SIEVE file.
2723 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2724 computation of the current number of files was incorrect.
2726 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2727 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2730 4. Give more explanation in the error message when the command for a transport
2731 filter fails to execute.
2733 5. There are several places where Exim runs a non-Exim command in a
2734 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2735 was being done only for the command run by the queryprogram router. It is
2736 now done for all such subprocesses. The other cases are: ${run, transport
2737 filters, and the commands run by the lmtp and pipe transports.
2739 6. Added CONFIGURE_GROUP build-time option.
2741 7. Some older OS have a limit of 256 on the maximum number of file
2742 descriptors. Exim was using setrlimit() to set 1000 as a large value
2743 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2744 systems. I've change it so that if it can't get 1000, it tries for 256.
2746 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2747 was an oversight, and furthermore, ever since the addition of extra
2748 controls (e.g. 4.43/32), the checks on when to allow different forms of
2749 "control" were broken. There should now be diagnostics for all cases when a
2750 control that does not make sense is encountered.
2752 9. Added the /retain_sender option to "control=submission".
2754 10. $recipients is now available in the predata ACL (oversight).
2756 11. Tidy the search cache before the fork to do a delivery from a message
2757 received from the command line. Otherwise the child will trigger a lookup
2758 failure and thereby defer the delivery if it tries to use (for example) a
2759 cached ldap connection that the parent has called unbind on.
2761 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2762 of $address_data from the recipient verification was clobbered by the
2763 sender verification.
2765 13. The value of address_data from a sender verification is now available in
2766 $sender_address_data in subsequent conditions in the ACL statement.
2768 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2770 15. Added a new option "connect=<time>" to callout options, to set a different
2773 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2774 was its contents. (It was OK if the option was not defined at all.)
2776 17. A "Completed" log line is now written for messages that are removed from
2777 the spool by the -Mrm option.
2779 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2780 information about exactly what failed.
2782 19. Added -dd to debug only the daemon process.
2784 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2785 handles timeouts, both on the server side and network timeouts. Renamed the
2786 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2788 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2789 It is now set to "smtps".
2791 22. $host_address is now set to the target address during the checking of
2792 ignore_target_hosts.
2794 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2795 being passed; this would have caused $sender_host_name to have been used if
2796 matching the list had actually called for a host name (not very likely,
2797 since this list is usually IP addresses). A host name is now passed as
2800 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2801 code that specifies a non-restarting handler (typically sigaction() in
2802 modern systems) in an attempt to fix a rare and obscure crash bug.
2804 25. Narrowed the window for a race in the daemon that could cause it to ignore
2805 SIGCHLD signals. This is not a major problem, because they are used only to
2806 wake it up if nothing else does.
2808 26. A malformed maildirsize file could cause Exim to calculate negative values
2809 for the mailbox size or file count. Odd effects could occur as a result.
2810 The maildirsize information is now recalculated if the size or filecount
2813 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2814 support for a long time. Removed HAVE_SYS_VFS_H.
2816 28. Installed the latest version of exipick from John Jetmore.
2818 29. In an address list, if the pattern was not a regular expression, an empty
2819 subject address (from a bounce message) matched only if the pattern was an
2820 empty string. Non-empty patterns were not even tested. This was the wrong
2821 because it is perfectly reasonable to use an empty address as part of a
2822 database query. An empty address is now tested by patterns that are
2823 lookups. However, all the other forms of pattern expect the subject to
2824 contain a local part and a domain, and therefore, for them, an empty
2825 address still always fails if the pattern is not itself empty.
2827 30. Exim went into a mad DNS loop when attempting to do a callout where the
2828 host was specified on an smtp transport, and looking it up yielded more
2829 than one IP address.
2831 31. Re-factored the code for checking spool and log partition space into a
2832 function that finds that data and another that does the check. The former
2833 is then used to implement four new variables: $spool_space, $log_space,
2834 $spool_inodes, and $log_inodes.
2836 32. The RFC2047 encoding function was originally intended for short strings
2837 such as real names; it was not keeping to the 75-character limit for
2838 encoded words that the RFC imposes. It now respects the limit, and
2839 generates multiple encoded words if necessary. To be on the safe side, I
2840 have increased the buffer size for the ${rfc2047: expansion operator from
2843 33. It is now permitted to omit both strings after an "if" condition; if the
2844 condition is true, the result is "true". As before, when the second string
2845 is omitted, a false condition yields an empty string. This makes it less
2846 cumbersome to write custom ACL and router conditions.
2848 34. Failure to deliver a bounce message always caused it to be frozen, even if
2849 there was an errors_to setting on the router. The errors_to setting is now
2852 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2853 canonical form (fully expanded) before being placed in
2854 $sender_host_address.
2856 36. The table in the code that translates DNS record types into text (T_A to
2857 "A" for instance) was missing entries for NS and CNAME. It is just possible
2858 that this could have caused confusion if both these types were looked up
2859 for the same domain, because the text type is used as part of Exim's
2860 per-process caching. But the chance of anyone hitting this buglet seems
2863 37. The dnsdb lookup has been extended in a number of ways.
2865 (1) There is a new type, "zns", which walks up the domain tree until it
2866 finds some nameserver records. It should be used with care.
2868 (2) There is a new type, "mxh", which is like "mx" except that it returns
2869 just the host names, not the priorities.
2871 (3) It is now possible to give a list of domains (or IP addresses) to be
2872 looked up. The behaviour when one of the lookups defers can be
2873 controlled by a keyword.
2875 (4) It is now possible to specify the separator character for use when
2876 multiple records are returned.
2878 38. The dnslists ACL condition has been extended: it is now possible to supply
2879 a list of IP addresses and/or domains to be looked up in a particular DNS
2882 39. Added log_selector=+queue_time_overall.
2884 40. When running the queue in the test harness, wait just a tad after forking a
2885 delivery process, to get repeatability of debugging output.
2887 41. Include certificate and key file names in error message when GnuTLS fails
2888 to set them up, because the GnuTLS error message doesn't include the name
2889 of the failing file when there is a problem reading it.
2891 42. Allow both -bf and -bF in the same test run.
2893 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2895 44. The "Exiscan patch" is now merged into the mainline Exim source.
2897 45. Sometimes the final signoff response after QUIT could fail to get
2898 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2899 < 0 before doing a fflush(). This bug looks as though it goes back to the
2900 introduction of TLS in release 3.20, but "sometimes" must have been rare
2901 because the tests only now provoked it.
2903 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2904 (this can affect the format of dates).
2906 47. exim_tidydb, when checking for the continued existence of a message for
2907 which it has found a message-specific retry record, was not finding
2908 messages that were in split spool directories. Consequently, it was
2909 deleting retry records that should have stayed in existence.
2911 48. Steve fixed some bugs in eximstats.
2913 49. The SPA authentication driver was not abandoning authentication and moving
2914 on to the next authenticator when an expansion was forced to fail,
2915 contradicting the general specification for all authenticators. Instead it
2916 was generating a temporary error. It now behaves as specified.
2918 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2919 (the order specifies the preference for clients). The order is now AES256,
2920 AES128, 3DES, ARCFOUR128.
2922 51. Small patch to Sieve code - explicitly set From: when generating an
2925 52. Exim crashed if a remote delivery caused a very long error message to be
2926 recorded - for instance if somebody sent an entire SpamAssassin report back
2927 as a large number of 550 error lines. This bug was coincidentally fixed by
2928 increasing the size of one of Exim's internal buffers (big_buffer) that
2929 happened as part of the Exiscan merge. However, to be on the safe side, I
2930 have made the code more robust (and fixed the comments that describe what
2933 53. Now that there can be additional text after "Completed" in log lines (if
2934 the queue_time_overall log selector is set), a one-byte patch to exigrep
2935 was needed to allow it to recognize "Completed" as not the last thing in
2938 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2939 patch that reportedly fixes this has been added. I am not expert enough to
2940 create a test for it. This is what the patch creator wrote:
2942 "I found a little strange behaviour of ldap code when working with
2943 Windows 2003 AD Domain, where users was placed in more than one
2944 Organization Units. When I tried to give exim partial DN, the exit code
2945 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2946 But simultaneously result of request was absolutely normal ldap result,
2947 so I produce this patch..."
2949 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2950 REFERENCE, so I have modified the code to exclude the patch when that macro
2953 55. Some experimental protocols are using DNS PTR records for new purposes. The
2954 keys for these records are domain names, not reversed IP addresses. The
2955 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2956 leaves it alone. Component reversal etc. now happens only for IP addresses.
2959 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2961 57. Double the size of the debug message buffer (to 2048) so that more of very
2962 long debug lines gets shown.
2964 58. The exicyclog utility now does better if the number of log files to keep
2965 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2967 59. Two changes related to the smtp_active_hostname option:
2969 (1) $smtp_active_hostname is now available as a variable.
2970 (2) The default for smtp_banner uses $smtp_active_hostname instead
2971 of $primary_hostname.
2973 60. The host_aton() function is supposed to be passed a string that is known
2974 to be a valid IP address. However, in the case of IPv6 addresses, it was
2975 not checking this. This is a hostage to fortune. Exim now panics and dies
2976 if the condition is not met. A case was found where this could be provoked
2977 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2978 components; fortuitously, this particular loophole had already been fixed
2979 by change 4.50/55 above.
2981 If there are any other similar loopholes, the new check in host_aton()
2982 itself should stop them being exploited. The report I received stated that
2983 data on the command line could provoke the exploit when Exim was running as
2984 exim, but did not say which command line option was involved. All I could
2985 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2986 running as the user.
2989 61. There was a buffer overflow vulnerability in the SPA authentication code
2990 (which came originally from the Samba project). I have added a test to the
2991 spa_base64_to_bits() function which I hope fixes it.
2994 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2995 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2997 63. The daemon start-up calls getloadavg() while still root for those OS that
2998 need the first call to be done as root, but it missed one case: when
2999 deliver_queue_load_max is set with deliver_drop_privilege. This is
3000 necessary for the benefit of the queue runner, because there is no re-exec
3001 when deliver_drop_privilege is set.
3003 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3004 This has been fixed.
3006 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3007 was in use, was not putting the data itself into the right store pool;
3008 consequently, it could be overwritten for a subsequent message in the same
3009 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3012 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3014 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3015 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3016 the function were treating the return as a boolean value, which happened to
3017 work because 0=false and not-0=true, but is not correct code.
3019 68. The host_aton() function was not handling scoped IPv6 addresses (those
3020 with, for example, "%eth0" on the end) correctly.
3022 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3023 negated items (that is, ~something) in unsigned ints. Some compilers
3024 apparently mutter when there is no cast.
3026 70. If an address verification called from an ACL failed, and did not produce a
3027 user-specific message (i.e. there was only a "system" message), nothing was
3028 put in $acl_verify_message. In this situation, it now puts the system
3031 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3032 session; change 4.31/43 added the unwanted input to the log line - except
3033 that it did not do this in the start of session case. It now does.
3035 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3036 This is wrong and can cause the other end to generate a synchronization
3037 error if it is another Exim or anything else that does the synchronization
3038 check. A QUIT command is no longer sent after a timeout.
3040 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3041 during host lookups.
3043 74. The defer_ok option of callout verification was not working if it was used
3044 when verifying addresses in header lines, that is, for this case:
3046 verify = header_sender/callout=defer_ok
3048 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3049 those file descriptors could be used for SMTP connections. If anything
3050 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3051 could be sent to the SMTP client, causing chaos. The daemon now opens
3052 stdin, stdout, and stderr to /dev/null when it puts itself into the
3055 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3056 log by default. The user can override this with suitable Perl magic.
3058 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3059 the log message when discard triggers, was not working for the DATA ACL or
3060 for the non-SMTP ACL.
3062 78. Error message wording change in sieve.c.
3064 79. If smtp_accept_max_per_host was set, the number of connections could be
3065 restricted to fewer than expected, because the daemon was trying to set up
3066 a new connection before checking whether the processes handling previous
3067 connections had finished. The check for completed processes is now done
3068 earlier. On busy systems, this bug wouldn't be noticed because something
3069 else would have woken the daemon, and it would have reaped the completed
3072 80. If a message was submitted locally by a user whose login name contained one
3073 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3074 It caused a spool format error. I have fixed the spool reading code. A
3075 related problem was that the "from" clause in the Received: line became
3076 illegal because of the space(s). It is now covered by ${quote_local_part.
3078 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3081 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3083 83. Patch from Sieve maintainer to fix unterminated string problem in
3084 "vacation" handling.
3086 84. Some minor changes to the Linux configuration files to help with other
3087 OS variants using glibc.
3089 85. One more patch for Sieve to update vacation handling to latest spec.
3092 ----------------------------------------------------
3093 See the note above about the 4.44 and 4.50 releases.
3094 ----------------------------------------------------
3100 1. Change 4.43/35 introduced a bug that caused file counts to be
3101 incorrectly computed when quota_filecount was set in an appendfile
3104 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3105 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3108 3. Give more explanation in the error message when the command for a transport
3109 filter fails to execute.
3111 4. There are several places where Exim runs a non-Exim command in a
3112 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3113 was being done only for the command run by the queryprogram router. It is
3114 now done for all such subprocesses. The other cases are: ${run, transport
3115 filters, and the commands run by the lmtp and pipe transports.
3117 5. Some older OS have a limit of 256 on the maximum number of file
3118 descriptors. Exim was using setrlimit() to set 1000 as a large value
3119 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3120 systems. I've change it so that if it can't get 1000, it tries for 256.
3122 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3123 was an oversight, and furthermore, ever since the addition of extra
3124 controls (e.g. 4.43/32), the checks on when to allow different forms of
3125 "control" were broken. There should now be diagnostics for all cases when a
3126 control that does not make sense is encountered.
3128 7. $recipients is now available in the predata ACL (oversight).
3130 8. Tidy the search cache before the fork to do a delivery from a message
3131 received from the command line. Otherwise the child will trigger a lookup
3132 failure and thereby defer the delivery if it tries to use (for example) a
3133 cached ldap connection that the parent has called unbind on.
3135 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3136 of $address_data from the recipient verification was clobbered by the
3137 sender verification.
3139 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3140 was its contents. (It was OK if the option was not defined at all.)
3142 11. A "Completed" log line is now written for messages that are removed from
3143 the spool by the -Mrm option.
3145 12. $host_address is now set to the target address during the checking of
3146 ignore_target_hosts.
3148 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3149 being passed; this would have caused $sender_host_name to have been used if
3150 matching the list had actually called for a host name (not very likely,
3151 since this list is usually IP addresses). A host name is now passed as
3154 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3155 code that specifies a non-restarting handler (typically sigaction() in
3156 modern systems) in an attempt to fix a rare and obscure crash bug.
3158 15. Narrowed the window for a race in the daemon that could cause it to ignore
3159 SIGCHLD signals. This is not a major problem, because they are used only to
3160 wake it up if nothing else does.
3162 16. A malformed maildirsize file could cause Exim to calculate negative values
3163 for the mailbox size or file count. Odd effects could occur as a result.
3164 The maildirsize information is now recalculated if the size or filecount
3167 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3168 support for a long time. Removed HAVE_SYS_VFS_H.
3170 18. Updated exipick to current release from John Jetmore.
3172 19. Allow an empty sender to be matched against a lookup in an address list.
3173 Previously the only cases considered were a regular expression, or an
3176 20. Exim went into a mad DNS lookup loop when doing a callout where the
3177 host was specified on the transport, if the DNS lookup yielded more than
3180 21. The RFC2047 encoding function was originally intended for short strings
3181 such as real names; it was not keeping to the 75-character limit for
3182 encoded words that the RFC imposes. It now respects the limit, and
3183 generates multiple encoded words if necessary. To be on the safe side, I
3184 have increased the buffer size for the ${rfc2047: expansion operator from
3187 22. Failure to deliver a bounce message always caused it to be frozen, even if
3188 there was an errors_to setting on the router. The errors_to setting is now
3191 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3192 canonical form (fully expanded) before being placed in
3193 $sender_host_address.
3195 24. Updated eximstats to version 1.33
3197 25. Include certificate and key file names in error message when GnuTLS fails
3198 to set them up, because the GnuTLS error message doesn't include the name
3199 of the failing file when there is a problem reading it.
3201 26. Expand error message when OpenSSL has problems setting up cert/key files.
3204 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3205 (this can affect the format of dates).
3207 28. exim_tidydb, when checking for the continued existence of a message for
3208 which it has found a message-specific retry record, was not finding
3209 messages that were in split spool directories. Consequently, it was
3210 deleting retry records that should have stayed in existence.
3212 29. eximstats updated to version 1.35
3213 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3214 1.35 - bugfix such that pie charts by volume are generated correctly
3216 30. The SPA authentication driver was not abandoning authentication and moving
3217 on to the next authenticator when an expansion was forced to fail,
3218 contradicting the general specification for all authenticators. Instead it
3219 was generating a temporary error. It now behaves as specified.
3221 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3222 (the order specifies the preference for clients). The order is now AES256,
3223 AES128, 3DES, ARCFOUR128.
3225 31. Small patch to Sieve code - explicitly set From: when generating an
3228 32. Exim crashed if a remote delivery caused a very long error message to be
3229 recorded - for instance if somebody sent an entire SpamAssassin report back
3230 as a large number of 550 error lines. This bug was coincidentally fixed by
3231 increasing the size of one of Exim's internal buffers (big_buffer) that
3232 happened as part of the Exiscan merge. However, to be on the safe side, I
3233 have made the code more robust (and fixed the comments that describe what
3236 33. Some experimental protocols are using DNS PTR records for new purposes. The
3237 keys for these records are domain names, not reversed IP addresses. The
3238 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3239 leaves it alone. Component reversal etc. now happens only for IP addresses.
3242 34. The host_aton() function is supposed to be passed a string that is known
3243 to be a valid IP address. However, in the case of IPv6 addresses, it was
3244 not checking this. This is a hostage to fortune. Exim now panics and dies
3245 if the condition is not met. A case was found where this could be provoked
3246 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3247 components; fortuitously, this particular loophole had already been fixed
3248 by change 4.50/55 or 4.44/33 above.
3250 If there are any other similar loopholes, the new check in host_aton()
3251 itself should stop them being exploited. The report I received stated that
3252 data on the command line could provoke the exploit when Exim was running as
3253 exim, but did not say which command line option was involved. All I could
3254 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3255 running as the user.
3258 35. There was a buffer overflow vulnerability in the SPA authentication code
3259 (which came originally from the Samba project). I have added a test to the
3260 spa_base64_to_bits() function which I hope fixes it.
3263 36. The daemon start-up calls getloadavg() while still root for those OS that
3264 need the first call to be done as root, but it missed one case: when
3265 deliver_queue_load_max is set with deliver_drop_privilege. This is
3266 necessary for the benefit of the queue runner, because there is no re-exec
3267 when deliver_drop_privilege is set.
3269 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3270 was in use, was not putting the data itself into the right store pool;
3271 consequently, it could be overwritten for a subsequent message in the same
3272 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3275 38. Sometimes the final signoff response after QUIT could fail to get
3276 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3277 < 0 before doing a fflush(). This bug looks as though it goes back to the
3278 introduction of TLS in release 3.20, but "sometimes" must have been rare
3279 because the tests only now provoked it.
3285 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3286 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3287 Now it returns an integer. A number of calls were still expecting a T/F
3288 return. Fortuitously, in all cases, the tests worked in OK situations,
3289 which is the norm. However, things would have gone wrong on any write
3290 failures on the smtp file descriptor. This function is used when sending
3291 messages over SMTP and also when doing verify callouts.
3293 2. When Exim is called to do synchronous delivery of a locally submitted
3294 message (the -odf or -odi options), it no longer closes stderr before doing
3297 3. Implemented the mua_wrapper option.
3299 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3301 5. Implemented the functions header_remove(), header_testname(),
3302 header_add_at_position(), and receive_remove_recipient(), and exported them
3305 6. If an ACL "warn" statement specified the addition of headers, Exim already
3306 inserted X-ACL-Warn: at the start if there was no header name. However, it
3307 was not making this test for the second and subsequent header lines if
3308 there were newlines in the string. This meant that an invalid header could
3309 be inserted if Exim was badly configured.
3311 7. Allow an ACL "warn" statement to add header lines at the start or after all
3312 the Received: headers, as well as at the end.
3314 8. Added the rcpt_4xx retry error code.
3316 9. Added postmaster_mailfrom=xxx to callout verification option.
3318 10. Added mailfrom=xxxx to the callout verification option, for verify=
3321 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3322 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3324 12. Inserted some casts to stop certain compilers warning when using pointer
3325 differences as field lengths or precisions in printf-type calls (mostly
3326 affecting debugging statements).
3328 13. Added optional readline() support for -be (dynamically loaded).
3330 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3331 same clock tick as a message's arrival, so that its received time was the
3332 same as the "first fail" time on the retry record, and that message
3333 remained on the queue past the ultimate address timeout, every queue runner
3334 would try a delivery (because it was past the ultimate address timeout) but
3335 after another failure, the ultimate address timeout, which should have then
3336 bounced the address, did not kick in. This was a "< instead of <=" error;
3337 in most cases the first failure would have been in the next clock tick
3338 after the received time, and all would be well.
3340 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3341 being recognized when the domain list was tested by the match_domain
3342 condition in an expansion string.
3344 16. Added the ${str2b64: operator.
3346 17. Exim was always calling setrlimit() to set a large limit for the number of
3347 processes, without checking whether the existing limit was already
3348 adequate. (It did check for the limit on file descriptors.) Furthermore,
3349 errors from getrlimit() and setrlimit() were being ignored. Now they are
3350 logged to the main and panic logs, but Exim does carry on, to try to do its
3351 job under whatever limits there are.
3353 18. Imported PCRE 5.0.
3355 19. Trivial typo in log message " temporarily refused connection" (the leading
3358 20. If the log selector return_path_on_delivery was set and an address was
3359 redirected to /dev/null, the delivery process crashed because it assumed
3360 that a return path would always be set for a "successful" delivery. In this
3361 case, the whole delivery is bypassed as an optimization, and therefore no
3364 21. Internal re-arrangement: the function for sending a challenge and reading
3365 a response while authentication was assuming a zero-terminated challenge
3366 string. It's now changed to take a pointer and a length, to allow for
3367 binary data in such strings.
3369 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3371 23. Exim was not respecting finduser_retries when seeking the login of the
3372 uid under which it was called; it was always trying 10 times. (The default
3373 setting of finduser_retries is zero.) Also, it was sleeping after the final
3374 failure, which is pointless.
3376 24. Implemented tls_on_connect_ports.
3378 25. Implemented acl_smtp_predata.
3380 26. If the domain in control=submission is set empty, Exim assumes that the
3381 authenticated id is a complete email address when it generates From: or
3382 Sender: header lines.
3384 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3385 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3386 chown and chgrp in /bin and hostname in /usr/bin.
3388 28. Exim was keeping the "process log" file open after each use, just as it
3389 does for the main log. This opens the possibility of it remaining open for
3390 long periods when the USR1 signal hits a daemon. Occasional processlog
3391 errors were reported, that could have been caused by this. Anyway, it seems
3392 much more sensible not to leave this file open at all, so that is what now
3395 29. The long-running daemon process does not normally write to the log once it
3396 has entered its main loop, and it closes the log before doing so. This is
3397 so that log files can straightforwardly be renamed and moved. However,
3398 there are a couple of unusual error situations where the daemon does write
3399 log entries, and I had neglected to close the log afterwards.
3401 30. The text of an SMTP error response that was received during a remote
3402 delivery was being truncated at 512 bytes. This is too short for some of
3403 the long messages that one sometimes sees. I've increased the limit to
3406 31. It is now possible to make retry rules that apply only when a message has a
3407 specific sender, in particular, an empty sender.
3409 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3410 it possible to be selective about when SMTP synchronization is enforced.
3412 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3414 32. Implemented hosts_connection_nolog.
3416 33. Added an ACL for QUIT.
3418 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3421 35. Added mailbox_size and mailbox_filecount to appendfile.
3423 36. Added control = no_multiline_responses to ACLs.
3425 37. There was a bug in the logic of the code that waits for the clock to tick
3426 in the case where the clock went backwards by a substantial amount such
3427 that the microsecond fraction of "now" was more than the microsecond
3428 fraction of "then" (but the whole seconds number was less).
3430 38. Added support for the libradius Radius client library this is found on
3431 FreeBSD (previously only the radiusclient library was supported).
3437 1. When certain lookups returned multiple values in the form name=value, the
3438 quoting of the values was not always being done properly. Specifically:
3439 (a) If the value started with a double quote, but contained no whitespace,
3441 (b) If the value contained whitespace other than a space character (i.e.
3442 tabs or newlines or carriage returns) it was not quoted.
3443 This fix has been applied to the mysql and pgsql lookups by writing a
3444 separate quoting function and calling it from the lookup code. The fix
3445 should probably also be applied to nisplus, ibase and oracle lookups, but
3446 since I cannot test any of those, I have not disturbed their existing code.
3448 2. A hit in the callout cache for a specific address caused a log line with no
3449 reason for rejecting RCPT. Now it says "Previous (cached) callout
3450 verification failure".
3452 3. There was an off-by-one bug in the queryprogram router. An over-long
3453 return line was truncated at 256 instead of 255 characters, thereby
3454 overflowing its buffer with the terminating zero. As well as fixing this, I
3455 have increased the buffer size to 1024 (and made a note to document this).
3457 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3458 when Exim is waiting for an SMTP response from a remote server, Exim
3459 restarts its select() call on the socket, thereby resetting its timeout.
3460 This is not a problem when such interrupts are rare. Somebody set up a cron
3461 job to run exiwhat every 2 minutes, which is less than the normal select()
3462 timeout (5 or 10 minutes). This meant that the select() timeout never
3463 kicked in because it was always reset. I have fixed this by comparing the
3464 time when an interrupt arrives with the time at the start of the first call
3465 to select(). If more time than the timeout has elapsed, the interrupt is
3466 treated as a timeout.
3468 5. Some internal re-factoring in preparation for the addition of Sieve
3469 extensions (by MH). In particular, the "personal" test is moved to a
3470 separate function, and given an option for scanning Cc: and Bcc: (which is
3471 not set for Exim filters).
3473 6. When Exim created an email address using the login of the caller as the
3474 local part (e.g. when creating a From: or Sender: header line), it was not
3475 quoting the local part when it contained special characters such as @.
3477 7. Installed new OpenBSD configuration files.
3479 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3480 try to make them clearer.
3482 9. Callout options, other than the timeout value, were being ignored when
3483 verifying sender addresses in header lines. For example, when using
3485 verify = header_sender/callout=no_cache
3487 the cache was (incorrectly) being used.
3489 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3490 only the Cygwin environment.
3492 11. When return_path_on_delivery was set as a log selector, if different remote
3493 addresses in the same message used different return paths and parallel
3494 remote delivery occurred, the wrong values would sometimes be logged.
3495 (Whenever a remote delivery process finished, the return path value from
3496 the most recently started remote delivery process was logged.)
3498 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3499 lines when AUTH and/or TLS are in use. This is the "received protocol"
3500 field. Exim used to use "asmtp" for authenticated SMTP, without any
3501 indication (in the protocol name) for TLS use. Now it follows the RFC and
3502 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3503 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3504 names appear in log lines as well as in Received: header lines.
3506 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3507 extensions, and comparison tests, and to fix some bugs.
3509 14. Changes to the "personal" filter test:
3511 (1) The test was buggy in that it was just doing the equivalent of
3512 "contains" tests on header lines. For example, if a user's address was
3513 anne@some.where, the "personal" test would incorrectly be true for
3515 To: susanne@some.where
3517 This test is now done by extracting each address from the header in turn,
3518 and checking the entire address. Other tests that are part of "personal"
3519 are now done using regular expressions (for example, to check local parts
3520 of addresses in From: header lines).
3522 (2) The list of non-personal local parts in From: addresses has been
3523 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3524 taken from the Sieve specification recommendations.
3526 (3) If the message contains any header line starting with "List-" it is
3527 treated as non-personal.
3529 (4) The test for "circular" in the Subject: header line has been removed
3530 because it now seems ill-conceived.
3532 15. Minor typos in src/EDITME comments corrected.
3534 16. Installed latest exipick from John Jetmore.
3536 17. If headers_add on a router specified a text string that was too long for
3537 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3538 of string_sprintf() is now avoided.
3540 18. $message_body_size was not set (it was always zero) when running the DATA
3541 ACL and the local_scan() function.
3543 19. For the "mail" command in an Exim filter, no default was being set for
3544 the once_repeat time, causing a random time value to be used if "once" was
3545 specified. (If the value happened to be <= 0, no repeat happened.) The
3546 default is now 0s, meaning "never repeat". The "vacation" command was OK
3547 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3548 (I found it when inspecting the code).
3550 20. There is now an overall timeout for performing a callout verification. It
3551 defaults to 4 times the callout timeout, which applies to individual SMTP
3552 commands during the callout. The overall timeout applies when there is more
3553 than one host that can be tried. The timeout is checked before trying the
3554 next host. This prevents very long delays if there are a large number of
3555 hosts and all are timing out (e.g. when the network connections are timing
3556 out). The value of the overall timeout can be changed by specifying an
3557 additional sub-option for "callout", called "maxwait". For example:
3559 verify = sender/callout=5s,maxwait=20s
3561 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3562 to the end before writing, but this should make it even safer).
3564 22. Exim was forgetting that it had advertised PIPELINING for the second and
3565 subsequent messages on an SMTP connection. It was also not resetting its
3566 memory on STARTTLS and an internal HELO.
3568 23. When Exim logs an SMTP synchronization error within a session, it now
3569 records whether PIPELINING has been advertised or not.
3571 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3572 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3573 rather than long int.
3575 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3577 26. Added the never_mail option to autoreply.
3583 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3584 crash if the getsockname() call failed; this can happen if a connection is
3585 closed very soon after it is established. The problem was simply in the
3586 order in which certain operations were done, causing Exim to try to write
3587 to the SMTP stream before it had set up the file descriptor. The bug has
3588 been fixed by making things happen in the correct order.
3594 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3595 before the connection was closed, thus losing the rejection response.
3597 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3598 some early Solaris releases, but causes trouble in current releases where
3599 socklen_t is defined.
3601 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3604 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3607 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3608 that releasing the top of it at the end releases what was used for sub-
3609 expansions (unless the block got too big). However, discard this block if
3610 the first thing is a variable or header, so that we can use its block when
3611 it is dynamic (useful for very large $message_headers, for example).
3613 6. Lookups now cache *every* query, not just the most recent. A new, separate
3614 store pool is used for this. It can be recovered when all lookup caches are
3615 flushed. Lookups now release memory at the end of their result strings.
3616 This has involved some general refactoring of the lookup sources.
3618 7. Some code has been added to the store_xxx() functions to reduce the amount
3619 of flapping under certain conditions.
3621 8. log_incoming_interface used to affect only the <= reception log lines. Now
3622 it causes the local interface and port to be added to several more SMTP log
3623 lines, for example "SMTP connection from", and rejection lines.
3625 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3627 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3629 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3630 could be overwritten at the end of the current message (or the start of a
3631 new message if it was set in a HELO ACL). The value is now preserved for
3632 the duration of the SMTP connection.
3634 12. If a transport had a headers_rewrite setting, and a matching header line
3635 contained an unqualified address, that address was qualified, even if it
3636 did not match any rewriting rules. The underlying bug was that the values
3637 of the flags that permit the existence of unqualified sender and recipient
3638 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3639 non-local messages, and by -bnq for local messages) were not being
3640 preserved with the message after it was received.
3642 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3643 "next input=" as part of the text comprising the host identity instead of
3644 the correct text. The code was using the same buffer for two different
3645 strings. However, depending on which order the printing function evaluated
3646 its arguments, the bug did not always show up. Under Linux, for example, my
3647 test suite worked just fine.
3649 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3650 doesn't work with some older versions of Perl. It has been changed to "my",
3651 which in any case is probably the better facility to use.
3653 15. A really picky compiler found some instances of statements for creating
3654 error messages that either had too many or two few arguments for the format
3657 16. The size of the buffer for calls to the DNS resolver has been increased
3658 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3659 for addresses that have a lot of PTR records. This alleviates a problem; it
3660 does not fully solve it.
3662 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3663 buffer now truncates the list and logs the incident, which is the same
3664 action as happens when Exim is looking up a host name and its aliases.
3665 Previously in this situation something unpredictable would happen;
3666 sometimes it was "internal error: store_reset failed".
3668 18. If a server dropped the connection unexpectedly when an Exim client was
3669 using GnuTLS and trying to read a response, the client delivery process
3670 crashed while trying to generate an error log message.
3672 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3673 string, for example:
3675 warn message = H1: something\nH2: something
3677 the text was added as a single header line from Exim's point of view
3678 though it ended up OK in the delivered message. However, searching for the
3679 second and subsequent header lines using $h_h2: did not work. This has been
3680 fixed. Similarly, if a system filter added multiple headers in this way,
3681 the routers could not see them.
3683 20. Expanded the error message when iplsearch is called with an invalid key to
3684 suggest using net-iplsearch in a host list.
3686 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3687 ACLs are no longer actually imposed (and a message to that effect is
3690 22. If a "gecos" field in a passwd entry contained escaped characters, in
3691 particular, if it contained a \" sequence, Exim got it wrong when building
3692 a From: or a Sender: header from that name. A second bug also caused
3693 incorrect handling when an unquoted " was present following a character
3694 that needed quoting.
3696 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3697 was not being matched caselessly.
3699 24. Arranged for all hyphens in the exim.8 source to be escaped with
3702 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3703 properly. Recipient callout cache records were still being keyed to include
3704 the sender, even when use_sender was set false. This led to far more
3705 callouts that were necessary. The sender is no longer included in the key
3706 when use_sender is false.
3708 26. Added "control = submission" modifier to ACLs.
3710 27. Added the ${base62d: operator to decode base 62 numbers.
3712 28. dnsdb lookups can now access SRV records.
3714 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3715 the configuration file.
3717 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3718 (-v) mode. This makes the output for a verbose queue run more intelligible.
3720 31. Added a use_postmaster feature to recipient callouts.
3722 32. Added the $body_zerocount variable, containing the number of binary zero
3723 bytes in the message body.
3725 33. The time of last modification of the "new" subdirectory is now used as the
3726 "mailbox time last read" when there is a quota error for a maildir
3729 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3731 35. Added +ignore_unknown as a special item in host lists.
3733 36. Code for decoding IPv6 addresses in host lists is now included, even if
3734 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3735 address was recognized as an IP address, but was then not correctly decoded
3736 into binary, causing unexpected and incorrect effects when compared with
3743 1. Very minor rewording of debugging text in manualroute to say "list of
3744 hosts" instead of "hostlist".
3746 2. If verify=header_syntax was set, and a header line with an unqualified
3747 address (no domain) and a large number of spaces between the end of the
3748 name and the colon was received, the reception process suffered a buffer
3749 overflow, and (when I tested it) crashed. This was caused by some obsolete
3750 code that should have been removed. The fix is to remove it!
3752 3. When running in the test harness, delay a bit after writing a bounce
3753 message to get a bit more predictability in the log output.
3755 4. Added a call to search_tidyup() just before forking a reception process. In
3756 theory, someone could use a lookup in the expansion of smtp_accept_max_
3757 per_host which, without the tidyup, could leave open a database connection.
3759 5. Added the variables $recipient_data and $sender_data which get set from a
3760 lookup success in an ACL "recipients" or "senders" condition, or a router
3761 "senders" option, similar to $domain_data and $local_part_data.
3763 6. Moved the writing of debug_print from before to after the "senders" test
3766 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3767 problems for message scanning, either using a data ACL, or using
3768 local_scan() because the Received: header was not generated till after they
3769 were called (in order to set the time as the time of reception completion).
3770 I have revised the way this works. The header is now generated after the
3771 body is received, but before the ACL or local_scan() are called. After they
3772 are run, the timestamp in the header is updated.
3778 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3779 before starting a queue runner without re-exec. This happened only when
3780 deliver_drop_privilege was set or when the Exim user was set to root. The
3781 effect of the bug was that timeouts during subsequent deliveries caused
3782 crashes instead of being properly handled. The handler is now left at its
3783 default (and expected) setting.
3785 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3786 message, again when deliver_drop_privilege is set or Exim is run as root.
3787 The bug described in (1) was not present in this case, but the tidying up
3788 of the other signals was missing. I have made the two cases consistent.
3790 3. The ignore_target_hosts setting on a manualroute router was being ignored
3791 for hosts that were looked up using the /MX notation.
3793 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3796 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3797 operated on the sender address. After changing the $sender_address to <>
3798 for the sender address verify, Exim was re-instated it as the original
3799 (before rewriting) address, but remembering that it had rewritten it, so it
3800 wasn't rewriting it again. This bug also had the effect of breaking the
3801 sender address verification caching when the sender address was rewritten.
3803 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3804 This has been changed so that if the ip literal address matches
3805 ignore_target_hosts, the router declines.
3807 7. Added expansion conditions match_domain, match_address, and match_local_
3808 part (NOT match_host).
3810 8. The placeholder for the Received: header didn't have a length field set.
3812 9. Added code to Exim itself and to exim_lock to test for a specific race
3813 condition that could lead to file corruption when using MBX delivery. The
3814 issue is with the lockfile that is created in /tmp. If this file is removed
3815 after a process has opened it but before that process has acquired a lock,
3816 there is the potential for a second process to recreate the file and also
3817 acquire a lock. This could lead to two Exim processes writing to the file
3818 at the same time. The added code performs the same test as UW imapd; it
3819 checks after acquiring the lock that its file descriptor still refers to
3820 the same named file.
3822 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3823 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3826 11. Added the smtp_active_hostname option. If used, this will typically be made
3827 to depend on the incoming interface address. Because $interface_address is
3828 not set up until the daemon has forked a reception process, error responses
3829 that can happen earlier (such as "too many connections") no longer contain
3832 12. If an expansion in a condition on a "warn" statement fails because a lookup
3833 defers, the "warn" statement is abandoned, and the next ACL statement is
3834 processed. Previously this caused the whole ACL to be aborted.
3836 13. Added the iplsearch lookup type.
3838 14. Added ident_timeout as a log selector.
3840 15. Added tls_certificate_verified as a log selector.
3842 16. Added a global option tls_require_ciphers (compare the smtp transport
3843 option of the same name). This controls incoming TLS connections.
3845 17. I finally figured out how to make tls_require_ciphers do a similar thing
3846 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3847 before starting the TLS session.
3849 18. Tabs are now shown as \t in -bP output.
3851 19. If the log selector return_path_on_delivery was set, Exim crashed when
3852 bouncing a message because it had too many Received: header lines.
3854 20. If two routers both had headers_remove settings, and the first one included
3855 a superfluous trailing colon, the final name in the first list and the
3856 first name in the second list were incorrectly joined into one item (with a
3857 colon in the middle).
3863 1. Added -C and -D options to the exinext utility, mainly to make it easier
3864 to include in the automated testing, but these could be helpful when
3865 multiple configurations are in use.
3867 2. The exinext utility was not formatting the output nicely when there was
3868 an alternate port involved in the retry record key, nor when there was a
3869 message id as well (for retries that were specific to a specific message
3870 and a specific host). It was also confused by IPv6 addresses, because of
3871 the additional colons they contain. I have fixed the IPv4 problem, and
3872 patched it up to do a reasonable job for IPv6.
3874 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3875 delivery, the log line now contains "pipelined" if PIPELINING was used.
3877 4. An SMTP transport process used to panic and die if the bind() call to set
3878 an explicit outgoing interface failed. This has been changed; it is now
3879 treated in the same way as a connect() failure.
3881 5. A reference to $sender_host_name in the part of a conditional expansion
3882 that was being skipped was still causing a DNS lookup. This no longer
3885 6. The def: expansion condition was not recognizing references to header lines
3886 that used bh_ and bheader_.
3888 7. Added the _cache feature to named lists.
3890 8. The code for checking quota_filecount in the appendfile transport was
3891 allowing one more file than it should have been.
3893 9. For compatibility with Sendmail, the command line option
3901 and sets the incoming protocol and host name (for trusted callers). The
3902 host name and its colon can be omitted when only the protocol is to be set.
3903 Note the Exim already has two private options, -pd and -ps, that refer to
3904 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3905 "s", but I don't think that's a major issue.
3907 10. A number of refactoring changes to the code, none of which should affect
3910 (a) The number of logging options was getting close to filling up the
3911 32-bit word that was used as a bit map. I have split them into two classes:
3912 those that are passed in the argument to log_write(), and those that are
3913 only ever tested independently outside of that function. These are now in
3914 separate 32-bit words, so there is plenty of room for expansion again.
3915 There is no change in the user interface or the logging behaviour.
3917 (b) When building, for example, log lines, the code previously used a
3918 macro that called string_cat() twice, in order to add two strings. This is
3919 not really sufficiently general. Furthermore, there was one instance where
3920 it was actually wrong because one of the argument was used twice, and in
3921 one call a function was used. (As it happened, calling the function twice
3922 did not affect the overall behaviour.) The macro has been replaced by a
3923 function that can join an arbitrary number of extra strings onto a growing
3926 (c) The code for expansion conditions now uses a table and a binary chop
3927 instead of a serial search (which was left over from when there were very
3928 few conditions). Also, it now recognizes conditions like "pam" even when
3929 the relevant support is not compiled in: a suitably worded error message is
3930 given if an attempt is made to use such a condition.
3932 11. Added ${time_interval:xxxxx}.
3934 12. A bug was causing one of the ddress fields not to be passed back correctly
3935 from remote delivery subprocesses. The field in question was not being
3936 subsequently used, so this caused to problems in practice.
3938 13. Added new log selectors queue_time and deliver_time.
3940 14. Might have fixed a bug in maildirsizefile handling that threw up
3941 "unexpected character" debug warnings, and recalculated the data
3942 unnecessarily. In any case, I expanded the warning message to give more
3945 15. Added the message "Restricted characters in address" to the statements in
3946 the default ACL that block characters like @ and % in local parts.
3948 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3949 Three changes have been made:
3951 (a) There was a serious bug; a negative response to MAIL caused the whole
3952 recipient domain to be cached as invalid, thereby blocking all messages
3953 to all local parts at the same domain, from all senders. This bug has
3954 been fixed. The domain is no longer cached after a negative response to
3955 MAIL if the sender used is not empty.
3957 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3960 (c) A new callout option, "use_sender" has been added for people who want
3961 the modified behaviour.
3967 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3970 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3971 indeed breaks things for older releases.
3973 3. Added additional logging to the case where there is a problem reading data
3974 from a filter that is running in a subprocess using a pipe, in order to
3975 try to track down a specific problem.
3977 4. Testing facility fudge: when running in the test harness and attempting
3978 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3979 getting "No route to host". Convert this to a timeout.
3981 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3984 6. Some OS don't have socklen_t but use size_t instead. This affects the
3985 fifth argument of getsockopt() amongst other things. This is now
3986 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3987 can be set for individual OS. I have set it for SunOS5, OSF1, and
3988 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3989 some earlier ones do not.
3991 7. Change 4.30/15 was not doing the test caselessly.
3993 8. The standard form for an IPv6 address literal was being rejected by address
3994 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3995 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3996 this, as well as the form without the "IPv6" on the front (but only when
3997 address literals are enabled, of course).
3999 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4001 10. Exim crashed if a message with an empty sender address specified by -f
4002 encountered a router with an errors_to setting. This could be provoked only
4003 by a command such as
4007 where an empty string was supplied; "<>" did not hit this bug.
4009 11. Installed PCRE release 4.5.
4011 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4012 remained set. It is now erased.
4014 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4015 times from message ids (which are base 36 rather than the normal 62).
4017 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4018 were being counted as actual protocol errors, and logged if the log
4019 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4020 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4021 and DATA following a set of rejected RCPTs do not count as protocol errors.
4022 In other words, Exim assumes they were pipelined, though this may not
4023 actually be the case. Of course, in all cases the client gets an
4024 appropriate error code.
4026 15. If a lookup fails in an ACL condition, a message about the failure may
4027 be available; it is used if testing the ACL cannot continue, because most
4028 such messages specify what the cause of the deferral is. However, some
4029 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4030 that caused an old message to be retained and used if a later statement
4031 caused a defer, replacing the real cause of the deferral.
4033 16. If an IP address had so many PTR records that the DNS lookup buffer
4034 was not large enough to hold them, Exim could crash while trying to process
4035 the truncated data. It now detects and logs this case.
4037 17. Further to 4.21/58, another change has been made: if (and only if) the
4038 first line of a message (the first header line) ends with CRLF, a bare LF
4039 in a subsequent header line has a space inserted after it, so as not to
4040 terminate the header.
4042 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4043 unnecessarily, used atoi() instead of strtol(), and didn't check the
4044 termination when getting file sizes from file names by regex.
4046 19. Completely re-implemented the support for maildirsize files, in the light
4047 of a number of problems with the previous contributed implementation
4048 (4.30/29). In particular:
4050 . If the quota is zero, the maildirsize file is maintained, but no quota is
4053 . If the maildir directory does not exist, it is created before any attempt
4054 to write a maildirsize file.
4056 . The quota value in the file is just a cache; if the quota is changed in
4057 the transport, the new value overrides.
4059 . A regular expression is available for excluding directories from the
4062 20. The autoreply transport checks the characters in options that define the
4063 message's headers; it allows continued headers, but it was checking with
4064 isspace() after an embedded newline instead of explicitly looking for a
4067 21. If all the "regular" hosts to which an address was routed had passed their
4068 expiry times, and had not reached their retry times, the address was
4069 bounced, even if fallback hosts were defined. Now Exim should go on to try
4072 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4073 equivalent code in the SMTP transport. Some hosts send humungous responses
4074 to HELO/EHLO, more than 1024 it seems.
4076 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4077 gives compiler warnings in some environments. I've now done it "properly",
4080 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4081 (because of problems with the built-in one) was declared to return uschar *
4082 instead of char *, causing compiler failure.
4084 25. Fixed a file descriptor leak when processing alias/forward files.
4086 26. Fixed a minor format string issue in dbfn.c.
4088 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4090 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4091 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4094 29. When a sender address is verified, it is cached, to save repeating the test
4095 when there is more than one recipient in a message. However, when the
4096 verification involves a callout, it is possible for different callout
4097 options to be set for different recipients. It is too complicated to keep
4098 track of this in the cache, so now Exim always runs a verification when a
4099 callout is required, relying on the callout cache for the optimization.
4100 The overhead is duplication of the address routing, but this should not be
4103 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4104 to be verified with callout=postmaster, and the main callout worked but the
4105 postmaster check failed, the verification correctly failed. However, if a
4106 subsequent RCPT command asked for sender verification *without* the
4107 postmaster check, incorrect caching caused this verification also to fail,
4110 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4111 it was not caching the DNS options (qualify_single, search_parents) that
4112 were used when the lookup failed. A subsequent lookup with different
4113 options therefore always gave the same answer, though there were cases
4114 where it should not have. (Example: a "domains = !$mx_any" option on a
4115 dnslookup router: the "domains" option is always processed without any
4116 widening, but the router might have qualify_single set.) Now Exim uses the
4117 cached value only when the same options are set.
4119 32. Added John Jetmore's "exipick" utility to the distribution.
4121 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4122 than a timeout (e.g. a certificate is required, and is not provided), an
4123 Exim server now closes the connection immediately. Previously it waited for
4124 the client to close - but if the client is SSL, it seems that they each
4125 wait for each other, leading to a delay before one of them times out.
4127 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4128 maintained 0.8.x compatibility because I don't think many are using it, and
4129 it is clearly obsolete.
4131 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4134 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4135 client certificate was expired. A simple patch fixes this, though I don't
4136 understand the full logic of why the verify callback is called multiple
4139 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4140 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4141 which causes problems with some clients (such as the Certicom SSL Plus
4142 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4143 disables the coutermeasure allowing Eudora to connect."
4145 38. Exim was not checking that a write() to a log file succeeded. This could
4146 lead to Bad Things if a log got too big, in particular if it hit a file
4147 size limit. Exim now panics and dies if it cannot write to a log file, just
4148 as it does if it cannot open a log file.
4150 39. Modified OS/Makefile-Linux so that it now contains
4152 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4154 The two -D definitions ensure that Exim is compiled with large file
4155 support, which makes it possible to handle log files that are bigger than
4158 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4159 instance) a domain was checked against a named list that involved a lookup,
4160 causing $domain_data to be set, then another domain was checked against the
4161 same list, then the first domain was re-checked, the value of $domain_data
4162 after the final check could be wrong. In particular, if the second check
4163 failed, it could be set empty. This bug probably also applied to
4166 41. The strip_trailing_dot option was not being applied to the address given
4167 with the -f command-line option.
4169 42. The code for reading a message's header from the spool was incrementing
4170 $received_count, but never initializing it. This meant that the value was
4171 incorrect (doubled) while delivering a message in the same process in which
4172 it was received. In the most common configuration of Exim, this never
4173 happens - a fresh exec is done - but it can happen when
4174 deliver_drop_privilege is set.
4176 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4177 it now includes up to 150 characters of the unexpected data in the log
4180 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4181 and building data strings. The size of both of these buffers was 10 000
4182 bytes - far larger than anybody would *ever* want, thought I. Needless to
4183 say, somebody hit the limit. I have increased the maximum line length to
4184 20 000 and the maximum data length of concatenated lines to 100 000. I have
4185 also fixed two bugs, because there was no checking on these buffers. Tsk,
4186 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4187 buffer is too small.
4189 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4190 lsearch lookups. Now it does.
4192 46. When parsing a route_list item in a manualroute router, a fixed-length
4193 buffer was used for the list of hosts. I made this 1024 bytes long,
4194 thinking that nobody would ever have a list of hosts that long. Wrong.
4195 Somebody had a whole pile of complicated expansion conditions, and the
4196 string was silently truncated, leading to an expansion error. It turns out
4197 that it is easier to change to an unlimited length (owing to other changes
4198 that have happened since this code was originally written) than to build
4199 structure for giving a limitation error. The length of the item that
4200 expands into the list of hosts is now unlimited.
4202 47. The lsearch lookup could not handle data where the length of text line was
4203 more than 4095 characters. Such lines were truncated, leading to shortened
4204 data being returned. It should now handle lines of any length.
4206 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4207 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4210 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4211 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4212 no longer happens in comment lines. A list of replacements is now placed
4213 at the head of all of the source files, except those whose only change is
4214 to replace PERL_COMMAND in the very first #! line.
4216 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4217 messages on the queue, with a bottom-up merge sort, using code contributed
4218 by Michael Haardt. This should make operations like -bp somewhat faster on
4219 large queues. It won't affect queue runners, except when queue_run_in_order
4222 51. Installed eximstats 1.31 in the distribution.
4224 52. Added support for SRV lookups to the dnslookup router.
4226 53. If an ACL referred to $message_body or $message_body_end, the value was not
4227 reset for any messages that followed in the same SMTP session.
4229 54. The store-handling optimization for building very long strings was not
4230 differentiating between the different store pools. I don't think this
4231 actually made any difference in practice, but I've tidied it.
4233 55. While running the routers to verify a sender address, $sender_address
4234 was still set to the sender address. This is wrong, because when routing to
4235 send a bounce to the sender, it would be empty. Therefore, I have changed
4236 it so that, while verifying a sender address, $sender_address is set to <>.
4237 (There is no change to what happens when verifying a recipient address.)
4239 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4240 target A or AAAA records (if not already returned) without resetting the
4241 qualify_single or search_parents options of the DNS resolver. These are
4242 inappropriate in this case because the targets of MX and SRV records must
4243 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4244 target that, when qualified, matched something in the local domain. These
4245 two options are now turned off when doing these lookups.
4247 57. It seems that at least some releases of Reiserfs (which does not have the
4248 concept of a fixed number of inodes) returns zero and not -1 for the
4249 number of available inodes. This interacted badly with check_spool_inodes,
4250 which assumed that -1 was the "no such thing" setting. What I have done is
4251 to check that the total number of inodes is greater than zero before doing
4252 the test of how many are available.
4254 58. When a "warn" ACL statement has a log_message modifier, the message is
4255 remembered, and not repeated. This is to avoid a lot of repetition when a
4256 message has many recipients that cause the same warning to be written.
4257 Howewer, Exim was preserving the list of already written lines for an
4258 entire SMTP session, which doesn't seem right. The memory is now reset if a
4259 new message is started.
4261 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4262 debugging output unless log_rewrite was also set.
4264 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4265 of (int)(handle) when we know that handle contains (void *)(-1).
4267 61. The Exim daemon panic-logs an error return when it closes the incoming
4268 connection. However "connection reset by peer" seems to be common, and
4269 isn't really an error worthy of noting specially, so that particular error
4272 62. When Exim is trying to find all the local interfaces, it used to panic and
4273 die if the ioctl to get the interface flags failed. However, it seems that
4274 on at least one OS (Solaris 9) it is possible to have an interface that is
4275 included in the list of interfaces, but for which you get a failure error
4276 for this call. This happens when the interface is not "plumbed" into a
4277 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4278 failure of the "get flags" call assumes that the interface is down.
4280 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4281 makes life easier for people who are doing arithmetic on fields extracted
4282 from dates, where you often get leading zeros that should not be
4283 interpreted as octal.
4285 64. Added qualify_domain to the redirect router, to override the global
4288 65. If a pathologically long header line contained very many addresses (the
4289 report of this problem mentioned 10 000) and each of them was rewritten,
4290 Exim could use up a very large amount of memory. (It kept on making new
4291 copies of the header line as it rewrote, and never released the old ones.)
4292 At the expense of a bit more processing, the header rewriting function has
4293 been changed so that it no longer eats memory in this way.
4295 66. The generation of the Received: header has been moved from the time that a
4296 message starts to be received, to the time that it finishes. The timestamp
4297 in the Received: header should now be very close to that of the <= log
4298 line. There are two side-effects of this change:
4300 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4301 logged header lines no longer include the local Received: line, because
4302 it has not yet been created. The same applies to a copy of the message
4303 that is returned to a non-SMTP sender when a message is rejected.
4305 (b) When a filter file is tested using -bf, no additional Received: header
4306 is added to the test message. After some thought, I decided that this
4309 This change does not affect the value of $received_for. It is still set
4310 after address rewriting, but before local_scan() is called.
4312 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4314 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4315 gave an unhelpful panic error message, and a defer error. I have managed to
4316 change this behaviour so that it now rejects any supplied certificate,
4317 which seems right, as the list of acceptable certificates is empty.
4319 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4320 gave an unhelpful defer error. I have not managed to make this reject any
4321 supplied certificates, but the error message it gives is "no certificate
4322 supplied", which is not helpful.
4324 70. exigrep's output now also includes lines that are not associated with any
4325 message, but which match the given pattern. Implemented by a patch from
4326 Martin Sluka, which also tidied up the Perl a bit.
4328 71. Recipient callout verification, like sender verification, was using <> in
4329 the MAIL FROM command. This isn't really the right thing, since the actual
4330 sender may affect whether the remote host accepts the recipient or not. I
4331 have changed it to use the actual sender in the callout; this means that
4332 the cache record is now keyed on a recipient/sender pair, not just the
4333 recipient address. There doesn't seem to be a real danger of callout loops,
4334 since a callout by the remote host to check the sender would use <>.
4335 [SEE ABOVE: changed after hitting problems.]
4337 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4338 temporary errors. However, in the case of such a code being given after
4339 the end of a data transmission (i.e. after ".") Exim was failing to write
4340 a retry record for the message. (Yes, there was some broken host that was
4341 actually sending 8xx at this point.)
4343 73. An unknown lookup type in a host list could cause Exim to panic-die when
4344 the list was checked. (An example that provoked this was putting <; in the
4345 middle of a list instead of at the start.) If this happened during a DATA
4346 ACL check, a -D file could be left lying around. This kind of configuration
4347 error no longer causes Exim to die; instead it causes a defer errror. The
4348 incident is still logged to the main and panic logs.
4350 74. Buglet left over from Exim 3 conversion. The message "too many messages
4351 in one connection" was written to the rejectlog but not the mainlog, except
4352 when address rewriting (yes!) was being logged.
4354 75. Added write_rejectlog option.
4356 76. When a system filter was run not as root (that is, when system_filter_user
4357 was set), the values of the $n variables were not being returned to the
4358 main process; thus, they were not subsequently available in the $sn
4361 77. Added +return_path_on_delivery log selector.
4363 78. A connection timeout was being treated differently from recipients deferred
4364 when testing hosts_max_try with a message that was older than the host's
4365 retry timeout. (The host should not be counted, thus allowing all hosts to
4366 be tried at least once before bouncing.) This may have been the cause of an
4367 occasionally reported bug whereby a message would remain on the queue
4368 longer than the retry timeout, but would be bounced if a delivery was
4369 forced. I say "may" because I never totally pinned down the problem;
4370 setting up timeout/retry tests is difficult. See also the next item.
4372 79. The ultimate address timeout was not being applied to errors that involved
4373 a combination of host plus message (for example, a timeout on a MAIL
4374 command). When an address resolved to a number of possible hosts, and they
4375 were not all tried for each delivery (e.g. because of hosts_max_try), a
4376 message could remain on the queue longer than the retry timeout.
4378 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4381 81. Fixed an obscure SMTP outgoing bug which required at least the following
4382 conditions: (a) there was another message waiting for the same server;
4383 (b) the server returned 5xx to all RCPT commands in the first message so
4384 that the message was not completed; (c) the server dropped the connection
4385 or gave a negative response to the RSET that Exim sends to abort the
4386 transaction. The observed case was a dropped connection after DATA that had
4387 been sent in pipelining mode. That is, the server had advertised PIPELINING
4388 but was not implementing it correctly. The effect of the bug was incorrect
4389 behaviour, such as trying another host, and this could lead to a crash.
4395 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4396 and daemon.c were passed as pointers to ints; they should have been
4397 pointers to socklen_t variables (which are typically unsigned ints).
4399 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4402 3. Fixed a really odd bug that affected only the testing scheme; patching a
4403 certain fixed string in the binary changed the value of another string that
4404 happened to be identical to the end of the original first string.
4406 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4407 name", it returns that address as the IP address. On some operating
4408 systems (e.g. Solaris), it also passes back the IP address string as the
4409 "host name". However, on others (e.g. Linux), it passes back an empty
4410 string. Exim wasn't checking for this, and was changing the host name to an
4411 empty string, assuming it had been canonicized.
4413 5. Although rare, it is permitted to have more than one PTR record for a given
4414 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4415 all the names associated with an address, because they do in Solaris.
4416 However, it seems that they do not in Linux for data that comes from the
4417 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4418 I found this out when I moved to a new Linux workstation and tried to run
4419 the Exim test suite.
4421 To get round this problem I have changed the code so that it now does its
4422 own call to the DNS to look up PTR records when searching for a host name.
4423 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4424 addresses that are only in /etc/hosts are still found.
4426 This behaviour is, however, controlled by an option called host_lookup_
4427 order, which defaults to "bydns:byaddr". If people want to use the other
4428 order, or indeed, just use one or the other means of lookup, they can
4429 specify it in this variable.
4431 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4432 some operating systems, this comes back from gethostbyaddr() as an empty
4433 string, and this is what Exim used to test for. However, it seems that in
4434 other systems, "." is yielded. Exim now tests for this case too.
4436 7. The values of check_spool_space and check_log_space are now held internally
4437 as a number of kilobytes instead of an absolute number of bytes. If a
4438 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4439 kilobyte. This means that much larger values can be stored.
4441 8. Exim monitor: an attempt to get the action menu when not actually pointing
4442 at a message produces an empty menu entitled "No message selected". This
4443 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4444 no entries in it ("Shell widget menu has zero width and/or height"). So I
4445 have added a single, blank menu entry in this case.
4447 9. Added ${quote_local_part.
4449 10. MIME decoding is now applied to the contents of Subject: header lines when
4452 11. Now that a reference to $sender_host_address automatically causes a reverse
4453 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4454 host lookup before query-style lookups in lists that might use this
4455 variable. This has therefore been abolished, and the "net-" prefix is no
4456 longer necessary for query-style lookups.
4458 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4459 have been a typo for LFLAGS, so it has been changed.
4461 13. The install script calls Exim with "-C /dev/null" in order to find the
4462 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4463 to be output. Howeve, since Exim outputs its version number before the
4464 error, it didn't break the script. It just looked ugly. I fixed this by
4465 always allowing "-C /dev/null" if the caller is root.
4467 14. Ignore overlarge ACL variable number when reading spool file - insurance
4468 against a later release with more variables having written the file.
4470 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4471 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4472 this, as well as the form without the "IPv6" on the front.
4474 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4475 OS/Makefile-Darwin file.
4477 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4478 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4480 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4481 variable. However, some OS use ints for inodes. I've added cast to long int
4482 to get rid of the compiler warning.
4484 19. I had forgotten to lock out "/../" in configuration file names when
4485 ALT_CONFIG_PREFIX was set.
4487 20. Routers used for verification do not need to specify transports. However,
4488 if such a router generated a host list, and callout was configured, Exim
4489 crashed, because it could not find a port number from the (non-existent)
4490 transport. It now assumes port 25 in this circumstance.
4492 21. Added the -t option to exigrep.
4494 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4495 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4496 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4497 them is set, LOOKUP_LSEARCH is forced.
4499 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4501 24. Added sender and host information to the "rejected by local_scan()" log
4502 line; previously there was no indication of these.
4504 25. Added .include_if_exists.
4506 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4507 Linux. It turns out that not all file systems support this. Apparently some
4508 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4509 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4510 not supported on the file descriptor, is now ignored when Exim is trying to
4511 sync a directory. This applies only to Linux.
4513 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4515 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4516 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4519 29. Added support for maildirsize files from supplied patch (modified a bit).
4521 30. The use of :fail: followed by an empty string could lead Exim to respond to
4522 sender verification failures with (e.g.):
4524 550 Verification failed for <xxx>
4525 550 Sender verify failed
4527 where the first response line was missing the '-' that indicates it is not
4528 the final line of the response.
4530 31. The loop for finding the name of the user that called Exim had a hardwired
4531 limit of 10; it now uses the value of finduser_retries, which is used for
4532 all other user lookups.
4534 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4537 33. Exim was neglecting to zero errno before one call of strtol() when
4538 expanding a string and expecting an integer value. On some systems this
4539 resulted in spurious "integer overflow" errors. Also, it was casting the
4540 result into an int without checking.
4542 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4543 did not work. The code looks as if it has *never* worked, though it appears
4544 to have been documented since at least releast 1.62. I have made it work.
4546 35. The "timeout_DNS" error in retry rules, also documented since at least
4547 1.62, also never worked. As it isn't clear exactly what this means, and
4548 clearly it isn't a major issue, I have abolished the feature by treating it
4549 as "timeout", and writing a warning to the main and panic logs.
4551 36. The display of retry rules for -brt wasn't always showing the error code
4554 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4555 timeout_connect_A, timeout_connect_MX.
4557 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4558 to the empty sender.
4560 39. The daemon was not analysing the content of -oX till after it had closed
4561 stderr and disconnected from the controlling terminal. This meant that any
4562 syntax errors were only noted on the panic log, and the return code from
4563 the command was 0. By re-arranging the code a little, I've made the
4564 decoding happen first, so such errors now appear on stderr, and the return
4565 code is 1. However, the actual setting up of the sockets still happens in
4566 the disconnected process, so errors there are still only recorded on the
4569 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4570 connections (as happens on some IP stacks) was logged at start up time as
4571 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4572 it from "IPv6 and IPv4", which means that two separate sockets are being
4575 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4576 says whether AF_INET or AF_INET6 was passed as an argument.
4578 42. Exiwhat output was messed up when time zones were included in log
4581 43. Exiwhat now gives more information about the daemon's listening ports,
4582 and whether -tls-on-connect was used.
4584 44. The "port" option of the smtp transport is now expanded.
4586 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4587 silently ignored. Now an error message is written to the main and panic
4590 46. There's a new ACL modifier called "logwrite" which writes to a log file
4591 as soon as it is encountered.
4593 47. Added $local_user_uid and $local_user_gid at routing time.
4595 48. Exim crashed when trying to verify a sender address that was being
4598 49. Exim was recognizing only a space character after ".include". It now also
4599 recognizes a tab character.
4601 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4602 extracting the relevant information from the specification. The man page no
4603 longer contains scrambled data for the -d option, and I've added a section
4604 at the front about calling Exim under different names.
4606 51. Added "extra_headers" argument to the "mail" command in filter files.
4608 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4611 53. Installed eximstats 1.29.
4613 54. Added transport_filter_timeout as a generic transport option.
4615 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4616 Cc: header lines. This was required by RFC 822, but it not required by RFC
4619 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4620 incoming messages that did not have them. Now it does so only if the
4621 message originates locally, that is, if there is no associated remote host
4622 address. When Resent- header lines are present, this applies to the Resent-
4623 lines rather than the non-Resent- lines.
4625 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4626 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4628 58. Messages for configuration errors now include the name of the main
4629 configuration file - useful now that there may be more than one file in a
4630 list (.included file names were always shown).
4632 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4633 for those rare installations that do not start the daemon as root or run it
4634 setuid root. I've cut out the call to initgroups() if the daemon is not
4637 60. The Exim user and group can now be bound into the binary as text strings
4638 that are looked up at the start of Exim's processing.
4640 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4642 62. Added $mailstore_basename variable.
4644 63. Installed patch to sieve.c from Michael Haardt.
4646 64. When Exim failed to open the panic log after failing to open the main log,
4647 the original message it was trying to log was written to stderr and debug
4648 output, but if they were not available (the usual case in production), it
4649 was lost. Now it is written to syslog before the two lines that record the
4650 failures to open the logs.
4652 65. Users' Exim filters run in subprocesses under the user's uid. It is
4653 possible for a "deliver" command or an alias in a "personal" command to
4654 provoke an address rewrite. If logging of address rewriting is configured,
4655 this fails because the process is not running as root or exim. There may be
4656 a better way of dealing with this, but for the moment (because 4.30 needs
4657 to be released), I have disabled address rewrite logging when running a
4658 filter in a non-root, non-exim process.
4664 1. The buildconfig auxiliary program wasn't quoting the value set for
4665 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4666 not defined. This bug was masked in 4.22 by the effect that was fixed in
4669 2. Some messages that were rejected after a message id was allocated were
4670 shown as "incomplete" by exigrep. It no longer does this for messages that
4671 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4673 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4674 have allow_domain_literals set, the ID did not get logged in the <= line.
4675 Domain literals are now always recognized in Message-ID: header lines.
4677 4. The first argument for a ${extract expansion item is the key name or field
4678 number. Leading and trailing spaces in this item were not being ignored,
4679 causing some misleading effects.
4681 5. When deliver_drop_privilege was set, single queue runner processes started
4682 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4683 same command in the process it spins off) were not dropping privilege.
4685 6. When the daemon running as "exim" started a queue runner, it always
4686 re-executed Exim in the spun-off process. This is a waste of effort when
4687 deliver_drop_privilege is set. The new process now just calls the
4688 queue-runner function directly.
4694 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4697 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4698 ignored. Though the use of -oP was forcing the writing of a pid file, it
4699 was always written to the default place.
4701 3. If the message "no IP address found for host xxxx" is generated during
4702 incoming verification, it is now followed by identification of the incoming
4703 connection (so you can more easily find what provoked it).
4705 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4707 5. Added some features to "harden" Exim a bit more against certain attacks:
4709 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4710 be put in Local/Makefile. This is like the never_users runtime option,
4711 but it cannot be overridden. The default setting is "root".
4713 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4714 prefix string with which any file named in a -C command line option
4717 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4718 is retained for -C and -D only if the caller of Exim is root. Without
4719 it, the exim user may also use -C and -D and retain privilege.
4721 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4722 command line option is disabled.
4724 6. Macro names set by the -D option must start with an upper case letter, just
4725 like macro names defined in the configuration file.
4727 7. Added "dereference=" facility to LDAP.
4729 8. Two instances of the typo "uknown" in the source files are fixed.
4731 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4732 the Configure-Makefile script screwed up while processing it.
4734 10. Incorporated PCRE 4.4.
4736 11. The SMTP synchronization check was not operating right at the start of an
4737 SMTP session. For example, it could not catch a HELO sent before the client
4738 waited for the greeting. There is now a check for outstanding input at the
4739 point when the greeting is written. Because of the duplex, asynchronous
4740 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4741 way, but not yet received, when the check is performed.
4743 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4744 on TCP/IP sockets, because this apparently causes some broken clients to
4747 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4748 unchanged) from the Cygwin maintainer.
4750 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4751 was supported instead of testing for mbx. Effectively a typo.
4753 15. The spa authenticator server code was not checking that the input it
4754 received was valid base64.
4756 16. The debug output line for the "set" modifier in ACLs was not showing the
4757 name of the variable that was being set.
4759 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4761 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4762 lookup on its own. Something else was needed to trigger the lookup. For
4763 example, a match in host_lookup or the need for a host name in a host list.
4764 Now, if $sender_host_name is referenced and the host name has not yet been
4765 looked up, a lookup is performed. If the lookup fails, the variable remains
4766 empty, and $host_lookup_failed is set to "1".
4768 19. Added "eqi" as a case-independent comparison operator.
4770 20. The saslauthd authentication condition could segfault if neither service
4771 nor realm was specified.
4773 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4774 error message that was logged was misleading, and incoming SMTP
4775 connections were dropped. The message is now more accurate, and temporary
4776 errors are given to SMTP connections.
4778 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4779 (except RSET) with a 421 error, until QUIT is received. However, it was
4780 failing to send a response to QUIT.
4782 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4783 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4784 ACL is now run after the helo_try_verify_hosts code.
4786 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4787 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4788 case-independent, so other case variants are also recognized). Apparently
4789 some systems use these upper case variants.
4791 25. If more than two messages were waiting for the same host, and a transport
4792 filter was specified for the transport, Exim sent two messages over the
4793 same TCP/IP connection, and then failed with "socket operation on non-
4794 socket" when it tried to send the third.
4796 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4798 27. The extern definition of crypt16() in expand.c was not being excluded when
4799 the OS had its own crypt16() function.
4801 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4802 as a preferred synonym for return_size_limit, both as an option and as an
4805 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4807 30. Changed the default configuration ACL to relax the local part checking rule
4808 for addresses that are not in any local domains. For these addresses,
4809 slashes and pipe symbols are allowed within local parts, but the sequence
4810 /../ is explicitly forbidden.
4812 31. SPA server authentication was not clearing the challenge buffer before
4815 32. log_message in a "warn" ACL statement was writing to the reject log as
4816 well as to the main log, which contradicts the documentation and doesn't
4817 seem right (because no rejection is happening). So I have stopped it.
4819 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4820 However, I am unable to do any testing of this.
4822 34. Fixed an infelicity in the appendfile transport. When checking directories
4823 for a mailbox, to see if any needed to be created, it was accidentally
4824 using path names with one or more superfluous leading slashes; tracing
4825 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4827 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4828 contents are added to the log line that is written for every discarded
4829 recipient. (Previously a log_message setting was ignored.)
4831 36. The ${quote: operator now quotes the string if it is empty.
4833 37. The install script runs exim in order to find its version number. If for
4834 some reason other than non-existence or emptiness, which it checks, it
4835 could not run './exim', it was installing it with an empty version number,
4836 i.e. as "exim-". This error state is now caught, and the installation is
4839 38. An argument was missing from the function that creates an error message
4840 when Exim fails to connect to the socket for saslauthd authentication.
4841 This could cause Exim to crash, or give a corrupted message.
4843 39. Added isip, isip4, and isip6 to ${if conditions.
4845 40. The ACL variables $acl_xx are now saved with the message, and can be
4846 accessed later in routers, transports, and filters.
4848 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4849 strings in the file are not string-expanded.
4851 42. If a MAIL command specified a SIZE value that was too large to fit into an
4852 int variable, the check against message_size_limit failed. Such values are
4853 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4854 day this will have to be increased, but I don't think I want to be around
4855 when emails are that large.
4862 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4863 iconv() is not standard in FreeBSD.
4865 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4866 IPv6 enabled. The observed symptom was a segmentation fault on return from
4867 the function os_common_find_running_interfaces() in src/os.c.
4869 3. In the check_special_case() function in daemon.c I had used "errno" as an
4870 argument name, which causes warnings on some systems. This was basically a
4871 typo, since it was named "eno" in the comments!
4873 4. The code that waits for the clock to tick (at a resolution of some fraction
4874 of a second) so as to ensure message-id uniqueness was always waiting for
4875 at least one whole tick, when it could have waited for less. [This is
4876 almost certainly not relevant at current processor speeds, where it is
4877 unlikely to ever wait at all. But we try to future-proof.]
4879 5. The function that sleeps for a time interval that includes fractions of a
4880 second contained a race. It did not block SIGALRM between setting the
4881 timer, and suspending (a couple of lines later). If the interval was short
4882 and the sigsuspend() was delayed until after it had expired, the suspension
4883 never ended. On busy systems this could lead to processes getting stuck for
4886 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4887 process, before it forks any delivery processes. The open lookup caching
4888 mechanism meant that the open file or database connection was passed into
4889 the delivery process. The problem was that delivery processes always tidy
4890 up cached lookup data. This could cause a problem for the next delivery
4891 process started by the queue runner, because the external queue runner
4892 process does not know about the closure. So the next delivery process
4893 still has data in the lookup cache. In the case of a file lookup, there was
4894 no problem because closing a file descriptor in a subprocess doesn't affect
4895 the parent. However, if the lookup was caching a connection to a database,
4896 the connection was closed, and the second delivery process was likely to
4897 see errors such as "PGSQL: query failed: server closed the connection
4898 unexpectedly". The problem has been fixed by closing all cached lookups
4899 in a queue runner before running a delivery process.
4901 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4902 seem to have the "const" qualifier which it has on other OS. I've
4905 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4906 *of the same type* (client or server) with the same public name that an
4907 error should be diagnosed.
4909 9. When Exim looked up a host name for an IP address, but failed to find the
4910 original IP address when looking up the host name (a safety check), it
4911 output the message "<ip address> does not match any IP for NULL", which was
4912 confusing, to say the least. The bug was that the host name should have
4913 appeared instead of "NULL".
4915 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4916 user that is built into the binary, and the -C or -D options is used, root
4917 privilege is dropped before the configuration file is read. In addition,
4918 logging is switched to stderr instead of the normal log files. If the
4919 configuration then re-defines the Exim user, the unprivileged environment
4920 is probably not what is expected, so Exim logs a panic warning message (but
4923 However, if deliver_drop_privilege is set, the unprivileged state may well
4924 be exactly what is intended, so the warning has been cut out in that case,
4925 and Exim is allowed to try to write to its normal log files.
4931 1. smtp_return_error_details was not giving details for temporary sender
4932 or receiver verification errors.
4934 2. Diagnose a configuration error if two authenticators have the same public
4937 3. Exim used not to create the message log file for a message until the first
4938 delivery attempt. This could be confusing when incoming messages were held
4939 for policy or load reasons. The message log file is now created at the time
4940 the message is received, and an initial "Received" line is written to it.
4942 4. The automatically generated man page for command line options had a minor
4943 bug that caused no ill effects; however, a more serious problem was that
4944 the procedure for building the man page automatically didn't always
4945 operate. Consequently, release 4.20 contains an out-of-date version. This
4946 shouldn't happen again.
4948 5. When building Exim with embedded Perl support, the script that builds the
4949 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4950 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4952 6. The freeze_tell option was not being used for messages that were frozen on
4953 arrival, either by an ACL or by local_scan().
4955 7. Added the smtp_incomplete_transaction log selector.
4957 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4958 was accepting AUTH without a new EHLO.
4960 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4961 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4964 10. Logging of TCP/IP connections (when configured) now happens in the main
4965 daemon process instead of the child process, so that the TCP/IP connection
4966 count is more accurate (but it can never be perfect).
4968 11. The use of "drop" in a nested ACL was not being handled correctly in the
4969 outer ACL. Now, if condition failure induced by the nested "drop" causes
4970 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4971 or "require"), the connection is dropped.
4973 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4974 that yield "discard" can now be used with an "accept" or a "discard" verb,
4975 but an error is generated for any others (because I can't see a useful way
4976 to define what should happen).
4978 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4979 are now processed in the same way as lines in the Exim configuration file.
4980 In particular, continuation lines are supported.
4982 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4984 15. Added -ti meaning -t -i.
4986 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4987 domains, and warn by logging if others are found.
4989 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4990 returned from GIFCONF when Exim is trying to find the list of interfaces on
4991 a host. The code in os.c has been modified to copy each ifreq to an aligned
4992 structure in all cases.
4994 Also, in some cases, the returned ifreq's were being copied to a 'struct
4995 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4996 means the last couple of bytes of an IPv6 address could be chopped if the
4997 ifreq contained only a normal sockaddr (14 bytes storage).
4999 18. Named domain lists were not supported in the hosts_treat_as_local option.
5000 An entry such as +xxxx was not recognized, and was treated as a literal
5003 19. Ensure that header lines added by a DATA ACL are included in the reject log
5004 if the ACL subsequently rejects the message.
5006 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5007 MD5 (which is deprecated).
5009 21. When testing a filter file using -bf, Exim was writing a message when it
5010 took the sender from a "From " line in the message, but it was not doing so
5011 when it took $return_path from a Return-Path: header line. It now does.
5013 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5014 with a valid header line field name (a series of printing characters
5015 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5017 23. Changed "disc" in the source to "disk" to conform to the documentation and
5018 the book and for uniformity.
5020 24. Ignore Sendmail's -Ooption=value command line item.
5022 25. When execve() failed while trying to run a command in a pipe transport,
5023 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5024 could be confused with a return value of 69 from the command itself. This
5025 has been changed to 127, the value the shell returns if it is asked to run
5026 a non-existent command. The wording for the related log line suggests a
5027 non-existent command as the problem.
5029 26. If received_header_text expands to an empty string, do not add a Received:
5030 header line to the message. (Well, it adds a token one on the spool, but
5031 marks it "old" so that it doesn't get used or transmitted.)
5033 27. Installed eximstats 1.28 (addition of -nt option).
5035 28. There was no check for failure on the call to getsockname() in the daemon
5036 code. This can fail if there is a shortage of resources on the system, with
5037 ENOMEM, for example. A temporary error is now given on failure.
5039 29. Contrary to the C standard, it seems that in some environments, the
5040 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5041 program. Exim now does this explicitly; it affects the formatting of
5042 timestamps using strftime().
5044 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5045 complaints. I've now initialized all the variables, to avoid this.
5047 32. Header lines added by a system filter were not being "seen" during
5048 transport-time rewrites.
5050 33. The info_callback() function passed to OpenSSL is set up with type void
5051 (*)(SSL *, int, int), as described somewhere. However, when calling the
5052 function (actually a macro) that sets it up, the type void(*)() is
5053 expected. I've put in a cast to prevent warnings from picky compilers.
5055 34. If a DNS black list lookup found a CNAME record, but there were no A
5056 records associated with the domain it pointed at, Exim crashed.
5058 35. If a DNS black list lookup returned more than one A record, Exim ignored
5059 all but the first. It now scans all returned addresses if a particular IP
5060 value is being sought. In this situation, the contents of the
5061 $dnslist_value variable are a list of all the addresses, separated by a
5064 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5065 to accept a host name and all its aliases if the forward lookup for any of
5066 them yielded the IP address of the incoming connection. Now it accepts only
5067 those names whose forward lookup yields the correct IP address. Any other
5068 names are discarded. This closes a loophole whereby a rogue DNS
5069 administrator could create reverse DNS records to break through a
5070 wildcarded host restriction in an ACL.
5072 37. If a user filter or a system filter that ran in a subprocess used any of
5073 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5074 the wrong values were passed to the pipe command ($thisaddress had the
5075 value of $0, $0 had the value of $1, etc). This bug was introduced by
5076 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5078 38. Improved the line breaking for long SMTP error messages from ACLs.
5079 Previously, if there was no break point between 40 and 75 characters, Exim
5080 left the rest of the message alone. Two changes have been made: (a) I've
5081 reduced the minimum length to 35 characters; (b) if it can't find a break
5082 point between 35 and 75 characters, it looks ahead and uses the first one
5083 that it finds. This may give the occasional overlong line, but at least the
5084 remaining text gets split now.
5086 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5087 file descriptors might be low, and that setting 1000 would always raise it.
5088 It turns out that in some environments, the limit is already over 1000 and
5089 that lowering it causes trouble. So now Exim takes care not to decrease it.
5091 40. When delivering a message, the value of $return_path is set to $sender_
5092 address at the start of routing (routers may change the value). By an
5093 oversight, this default was not being set up when an address was tested by
5094 -bt or -bv, which affected the outcome if any router or filter referred to
5097 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5098 log only when "message" or "log_message" are set. However, if one of the
5099 conditions was an address verification, or a call to a nested ACL, the
5100 messages generated by the underlying test were being passed through. This
5101 no longer happens. The underlying message is available in $acl_verify_
5102 message for both "message" and "log_message" expansions, so it can be
5103 passed through if needed.
5105 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5106 new expansion $bh_ to give the encoded byte string without charset
5107 translation. Translation happens only if iconv() is available; HAVE_ICONV
5108 indicates this at build time. HEADERS_CHARSET gives the charset to
5109 translate to; headers_charset can change it in the configuration, and
5110 "headers charset" can change it in an individual filter file.
5112 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5113 that creates RFC 2047 encoded "words" labels them as that charset instead
5114 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5115 expansion operator; (ii) when Exim creates a From: line for a local
5116 message; (iii) when a header line is rewritten to include a "phrase" part.
5118 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5119 buggy, causing it to skip the first lines of messages whose message ID
5120 ended in 'D'. This would not have bitten before Exim release 4.14, because
5121 message IDs were unlikely to end in 'D' before then. The effect was to have
5122 incorrect size information for certain domains.
5124 45. #include "config.h" was missing at the start of the crypt16.c module. This
5125 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5128 46. If there was a timeout during a "random" callout check, Exim treated it as
5129 a failure of the random address, and carried on sending RSET and the real
5130 address. If the delay was just some slowness somewhere, the response to the
5131 original RCPT would be taken as a response to RSET and so on, causing
5132 mayhem of various kinds.
5134 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5135 when I implemented it. It didn't allow for the fact that some option values
5136 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5137 the right test for positive values.
5139 48. Domain names in DNS records are case-independent. Exim always looks them up
5140 in lower case. Some resolvers return domain names in exactly the case they
5141 appear in the zone file, that is, they may contain uppercase letters. Not
5142 all resolvers do this - some return always lower case. Exim was treating a
5143 change of case by a resolver as a change of domain, similar to a widening
5144 of a domain abbreviation. This triggered its re-routing code and so it was
5145 trying to route what was effectively the same domain again. This normally
5146 caused routing to fail (because the router wouldn't handle the domain
5147 twice). Now Exim checks for this case specially, and just changes the
5148 casing of the domain that it ultimately uses when it transmits the message
5151 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5154 50. If a filter generated a file delivery with a non-absolute name (possible if
5155 no home directory exists for the router), the forbid_file option was not
5158 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5159 the existing equality matching.
5161 52. Exim was using ints instead of ino_t variables in some places where it was
5162 dealing with inode numbers.
5164 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5165 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5166 called TMPDIR, and if it finds it is different, it changes its value.
5168 54. The smtp_printf() function is now made available to local_scan() so
5169 additional output lines can be written before returning. There is also an
5170 smtp_fflush() function to enable the detection of a dropped connection.
5171 The variables smtp_input and smtp_batched_input are exported to
5174 55. Changed the default runtime configuration: the message "Unknown user"
5175 has been removed from the ACL, and instead placed on the localuser router,
5176 using the cannot_route_message feature. This means that any verification
5177 failures that generate their own messages won't get overridden. Similarly,
5178 the "Unrouteable address" message that was in the ACL for unverifiable
5179 relay addresses has also been removed.
5181 56. Added hosts_avoid_esmtp to the smtp transport.
5183 57. The exicyclog script was not checking for the esoteric option
5184 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5185 will work only if exicyclog is run under the appropriate euid.
5187 58. Following a discussion on the list, the rules by which Exim recognises line
5188 endings on incoming messages have been changed. The -dropcr and drop_cr
5189 options are now no-ops, retained only for backwards compatibility. The
5190 following line terminators are recognized: LF CRLF CR. However, special
5191 processing applies to CR:
5193 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5194 nor a local message in the state where . is a terminator.
5196 (ii) If a bare CR is encountered in a header line, an extra space is added
5197 after the line terminator so as not to end the header. The reasoning
5198 behind this is that bare CRs in header lines are most likely either
5199 to be mistakes, or people trying to play silly games.
5201 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5202 was being incorrectly given as 18 bytes larger than it should have been.
5203 This is a VOB (very old bug).
5205 60. This may never have affected anything current, but just in case it has:
5206 When the local host is found other than at the start of a list of hosts,
5207 the local host, those with the same MX, and any that follow, are discarded.
5208 When the list in question was part of a longer list of hosts, the following
5209 hosts (not currently being processed) were also being discarded. This no
5210 longer happens. I'm not sure if this situation could ever has previously
5213 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5216 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5217 header. This is something that is recommended in a new Internet Draft, and
5218 is something that is documented as being done by Sendmail. There are two
5219 possible values. For messages generated by the autoreply transport, Exim
5222 Auto-Submitted: auto-replied
5224 whereas for all other generated messages (e.g. bounces) it adds
5226 Auto-Submitted: auto-generated
5228 63. The "personal" condition in filters now includes a test for the
5229 Auto-Submitted: header. If it contains the string "auto-" the message it
5230 not considered personal.
5232 64. Added rcpt_include_affixes as a generic transport option.
5234 65. Added queue_only_override (default true).
5236 66. Added the syslog_duplication option.
5238 67. If what should have been the first header line of a message consisted of
5239 a space followed by a colon, Exim was mis-interpreting it as a header line.
5240 It isn't of course - it is syntactically invalid and should therefore be
5241 treated as the start of the message body. The misbehaviour could have
5242 caused a number of strange effects, including loss of data in subsequent
5243 header lines, and spool format errors.
5245 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5246 client host had authenticated. This control can now be exercised by an ACL
5247 for more flexibility.
5249 69. By default, callouts do not happen when testing with -bh. There is now a
5250 variant, -bhc, which does actually run the callout code, including
5251 consulting and updating the callout cache.
5253 70. Added support for saslauthd authentication, courtesy of Alexander
5256 71. If statvfs() failed on the spool or log directories while checking their
5257 size for availability, Exim confusingly gave the error "space shortage".
5258 Furthermore, in debugging mode it crashed with a floating point exception.
5259 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5260 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5261 serious problem, Exim now writes to the main and panic logs when this
5262 happens, with details of the failure. It then refuses to accept the
5263 incoming message, giving the message "spool directory problem" or "log
5264 directory problem" with a 421 code for SMTP messages.
5266 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5267 0, 1, and 2 exist, because some OS complain for execs without them (see
5268 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5269 descriptors. However, the code omitted to check that the open succeeded,
5270 causing mysterious errors if for some reason the permissions on /dev/null
5271 got screwed. Now Exim writes a message to the main and panic logs, and
5272 bombs out if it can't open /dev/null.
5274 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5275 interact so that it is all more flexible. It is supposed to remain
5276 backwards compatible. Also added extra_local_interfaces.
5278 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5279 to bomb out with an assertion failure - to the client this appears as a
5280 connection drop. This problem occurs in the part of the code that was taken
5281 from the Samba project. Fortunately, the assertion is in a very simple
5282 function, so I have fixed this by reproducing the function inline in the
5283 one place where it is called, and arranging for authentication to fail
5284 instead of killing the process with assert().
5286 75. The SPA client code was not working when the server requested OEM rather
5287 than Unicode encoding.
5289 76. Added code to make require_files with a specific uid setting more usable in
5290 the case where statting the file as root fails - usually a non-root-mounted
5291 NFS file system. When this happens and the failure is EACCES, Exim now
5292 forks a subprocess and does the per-uid checking as the relevant uid.
5294 77. Added process_log_path.
5296 78. If log_file_path was not explicitly set, a setting of check_log_space or
5297 check_log_inodes was ignored.
5299 79. If a space check for the spool or log partitions fails, the incident is now
5300 logged. Of course, in the latter case the data may get lost...
5302 80. Added the %p formatting code to string_format() so that it can be used to
5303 print addresses in debug_print(). Adjusted all the address printing in the
5304 debugging in store.c to use %p rather than %d.
5306 81. There was a concern that a line of code in smtp_in.c could overflow a
5307 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5308 initially expressed, the concern was not well-founded, because trailing
5309 spaces are removed early. However, if the trailing spaces were followed by
5310 a NULL, they did not get removed, so the overflow was possible. Two fixes
5313 (a) I re-wrote the offending code in a cleaner fashion.
5314 (b) If an incoming SMTP command contains a NULL character, it is rejected
5317 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5318 runs initgroups(), so that if the Exim user is in any additional groups,
5319 they will be used during message reception.
5325 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob