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 TF/01 Failure to get a lock on a hints database can have serious
10 consequences so log it to the panic log.
12 TF/02 Log LMTP confirmation messages in the same way as SMTP,
13 controlled using the smtp_confirmation log selector.
15 TF/03 Include the error message when we fail to unlink a spool file.
17 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
18 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
19 for maintaining out-of-tree patches for some time.
21 PP/01 Bugzilla 139: Documentation and portability issues.
22 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
23 Handle per-OS dynamic-module compilation flags.
25 PP/02 Let /dev/null have normal permissions.
26 The 4.73 fixes were a little too stringent and complained about the
27 permissions on /dev/null. Exempt it from some checks.
28 Reported by Andreas M. Kirchwitz.
30 PP/03 Report version information for many libraries, including
31 Exim version information for dynamically loaded libraries. Created
32 version.h, now support a version extension string for distributors
33 who patch heavily. Dynamic module ABI change.
35 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
36 privilege escalation vulnerability whereby the Exim run-time user
37 can cause root to append content of the attacker's choosing to
44 PP/01 Date: & Message-Id: revert to normally being appended to a message,
45 only prepend for the Resent-* case. Fixes regression introduced in
46 Exim 4.70 by NM/22 for Bugzilla 607.
48 PP/02 Include check_rfc2047_length in configure.default because we're seeing
49 increasing numbers of administrators be bitten by this.
51 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
53 PP/03 Bugzilla 994: added openssl_options main configuration option.
55 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
57 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
59 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
61 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
62 without a peer certificate, leading to a segfault because of an
63 assumption that peers always have certificates. Be a little more
64 paranoid. Problem reported by Martin Tscholak.
66 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
67 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
68 NB: ClamAV planning to remove STREAM in "middle of 2010".
69 CL also introduces -bmalware, various -d+acl logging additions and
70 more caution in buffer sizes.
72 PP/09 Implemented reverse_ip expansion operator.
74 PP/10 Bugzilla 937: provide a "debug" ACL control.
76 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
78 PP/12 Bugzilla 973: Implement --version.
80 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
82 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
84 PP/15 Bugzilla 816: support multiple condition rules on Routers.
86 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
87 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
88 ignore trailing whitespace.
90 JJ/02 prevent non-panic DKIM error from being sent to paniclog
92 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
95 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
96 Notification from Dr Andrew Aitchison.
98 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
99 ExtendedDetectionInfo response format.
100 Notification from John Horne.
102 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
105 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
106 XSL and documented dependency on system catalogs, with examples of how
109 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
112 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
113 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
114 configuration file which is writeable by the Exim user or group.
116 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
117 of configuration files to cover files specified with the -C option if
118 they are going to be used with root privileges, not just the default
121 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
122 option (effectively making it always true).
124 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
125 files to be used while preserving root privileges.
127 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
128 that rogue child processes cannot use them.
130 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
131 run-time user, instead of root.
133 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
134 Exim run-time user without dropping privileges.
136 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
137 result string, instead of calling string_vformat() twice with the same
140 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
141 for other users. Others should always drop root privileges if they use
142 -C on the command line, even for a whitelisted configure file.
144 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
146 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
152 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
153 $data_path, and $header_path variables; fixed documentation bugs and
156 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
157 exipick to access non-standard spools, including the "frozen" queue
160 NM/01 Bugzilla 965: Support mysql stored procedures.
161 Patch from Alain Williams
163 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
165 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
166 Patch from Andreas Metzler
168 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
169 Patch from Kirill Miazine
171 NM/05 Bugzilla 671: Added umask to procmail example.
173 JJ/03 installed exipick 20100323.0, fixing doc bug
175 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
176 directory. Notification and patch from Dan Rosenberg.
178 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
180 TK/02 Improve log output when DKIM signing operation fails.
182 MH/01 Treat the transport option dkim_domain as a colon separated
183 list, not as a single string, and sign the message with each element,
184 omitting multiple occurences of the same signer.
186 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
187 Bugzilla 985, 986. Patch by Simon Arlott
189 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
190 Patch by Simon Arlott
192 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
193 MBX locking. Notification from Dan Rosenberg.
199 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
201 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
203 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
205 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
207 NM/04 Bugzilla 915: Flag broken perl installation during build.
213 TK/01 Added patch by Johannes Berg that expands the main option
214 "spamd_address" if it starts with a dollar sign.
216 TK/02 Write list of recipients to X-Envelope-Sender header when building
217 the mbox-format spool file for content scanning (suggested by Jakob
220 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
221 (http://www.dcc-servers.net/) support via dccifd. Activated by
222 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
224 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
225 by Mark Daniel Reidel <mr@df.eu>.
227 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
228 When building exim an external PCRE library is now needed -
229 PCRE is a system library on the majority of modern systems.
230 See entry on PCRE_LIBS in EDITME file.
232 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
233 conversation. Added nologin parameter to request.
234 Patch contributed by Kirill Miazine.
236 TF/01 Do not log submission mode rewrites if they do not change the address.
238 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
240 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
241 log files in place. Contributed by Roberto Lima.
243 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
245 TF/03 Bugzilla 615: When checking the local_parts router precondition
246 after a local_part_suffix or local_part_prefix option, Exim now
247 does not use the address's named list lookup cache, since this
248 contains cached lookups for the whole local part.
250 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
251 Robert Millan. Documentation is in experimental-spec.txt.
253 TF/04 Bugzilla 668: Fix parallel build (make -j).
255 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
257 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
258 Patch provided by Jan Srzednicki.
260 TF/05 Leading white space used to be stripped from $spam_report which
261 wrecked the formatting. Now it is preserved.
263 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
264 that they are available at delivery time.
266 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
268 TF/08 TLS error reporting now respects the incoming_interface and
269 incoming_port log selectors.
271 TF/09 Produce a more useful error message if an SMTP transport's hosts
272 setting expands to an empty string.
274 NM/06 Bugzilla 744: EXPN did not work under TLS.
275 Patch provided by Phil Pennock.
277 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
278 Patch provided by Richard Godbee.
280 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
281 acl_smtp_notquit, added index entry.
283 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
284 Patch provided by Eugene Bujak.
286 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
287 accept(). Patch provided by Maxim Dounin.
289 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
290 Patch provided by Phil Pennock.
292 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
294 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
295 Patch provided by Brad "anomie" Jorsch.
297 NM/14 Bugzilla 622: Added timeout setting to transport filter.
298 Patch provided by Dean Brooks.
300 TK/05 Add native DKIM support (does not depend on external libraries).
302 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
303 Patch provided by Graeme Fowler.
305 NM/16 Bugzilla 851: Documentation example syntax fix.
307 NM/17 Changed NOTICE file to remove references to embedded PCRE.
309 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
312 NM/19 Bugzilla 745: TLS version reporting.
313 Patch provided by Phil Pennock.
315 NM/20 Bugzilla 167: bool: condition support.
316 Patch provided by Phil Pennock.
318 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
319 clients. Patch provided by Phil Pennock.
321 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
322 Patch provided by Brad "anomie" Jorsch.
324 NM/23 Bugzilla 687: Fix misparses in eximstats.
325 Patch provided by Heiko Schlittermann.
327 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
328 Patch provided by Heiko Schlittermann.
330 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
331 plus update to original patch.
333 NM/26 Bugzilla 799: Documentation correction for ratelimit.
335 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
336 Patch provided by David Brownlee.
338 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
340 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
342 NM/30 Bugzilla 888: TLS documentation bugfixes.
344 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
346 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
347 Unlike the original bugzilla I have changed all shell scripts in src tree.
349 NM/33 Bugzilla 898: Transport filter timeout fix.
350 Patch by Todd Rinaldo.
352 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
353 Patch by Serge Demonchaux.
355 NM/35 Bugzilla 39: Base64 decode bug fixes.
356 Patch by Jakob Hirsch.
358 NM/36 Bugzilla 909: Correct connect() call in dcc code.
360 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
362 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
364 NM/39 Bugzilla 911: Fixed MakeLinks build script.
370 TK/01 Add preliminary DKIM support. Currently requires a forked version of
371 ALT-N's libdkim that I have put here:
372 http://duncanthrax.net/exim-experimental/
374 Note to Michael Haardt: I had to rename some vars in sieve.c. They
375 were called 'true' and it seems that C99 defines that as a reserved
376 keyword to be used with 'bool' variable types. That means you could
377 not include C99-style headers which use bools without triggering
378 build errors in sieve.c.
380 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
381 as mailq or other aliases. Changed the --help handling significantly
382 to do whats expected. exim_usage() emits usage/help information.
384 SC/01 Added the -bylocaldomain option to eximstats.
386 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
388 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
390 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
396 PH/01 Another patch from the Sieve maintainer.
398 PH/02 When an IPv6 address is converted to a string for single-key lookup
399 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
400 dots are used instead of colons so that keys in lsearch files need not
401 contain colons. This was done some time before quoting was made available
402 in lsearch files. However, iplsearch files do require colons in IPv6 keys
403 (notated using the quote facility) so as to distinguish them from IPv4
404 keys. This meant that lookups for IP addresses in host lists did not work
405 for iplsearch lookups.
407 This has been fixed by arranging for IPv6 addresses to be expressed with
408 colons if the lookup type is iplsearch. This is not incompatible, because
409 previously such lookups could never work.
411 The situation is now rather anomolous, since one *can* have colons in
412 ordinary lsearch keys. However, making the change in all cases is
413 incompatible and would probably break a number of configurations.
415 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
418 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
419 conversion specification without a maximum field width, thereby enabling
420 a rogue spamd server to cause a buffer overflow. While nobody in their
421 right mind would setup Exim to query an untrusted spamd server, an
422 attacker that gains access to a server running spamd could potentially
423 exploit this vulnerability to run arbitrary code as the Exim user.
425 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
426 $primary_hostname instead of what libspf2 thinks the hosts name is.
428 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
429 a directory entry by the name of the lookup key. Previously, if a
430 symlink pointed to a non-existing file or a file in a directory that
431 Exim lacked permissions to read, a lookup for a key matching that
432 symlink would fail. Now it is enough that a matching directory entry
433 exists, symlink or not. (Bugzilla 503.)
435 PH/03 The body_linecount and body_zerocount variables are now exported in the
438 PH/04 Added the $dnslist_matched variable.
440 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
441 This means they are set thereafter only if the connection becomes
444 PH/06 Added the client_condition to authenticators so that some can be skipped
445 by clients under certain conditions.
447 PH/07 The error message for a badly-placed control=no_multiline_responses left
448 "_responses" off the end of the name.
450 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
452 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
453 (without spaces) instead of just copying the configuration text.
455 PH/10 Added the /noupdate option to the ratelimit ACL condition.
457 PH/11 Added $max_received_linelength.
459 PH/12 Added +ignore_defer and +include_defer to host lists.
461 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
462 way in which PCRE > 7.0 is built.
464 PH/14 Implemented queue_only_load_latch.
466 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
467 MAIL command. The effect was to mangle the value on 64-bit systems.
469 PH/16 Another patch from the Sieve maintainer.
471 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
473 PH/18 If a system quota error occurred while trying to create the file for
474 a maildir delivery, the message "Mailbox is full" was not appended to the
475 bounce if the delivery eventually timed out. Change 4.67/27 below applied
476 only to a quota excession during the actual writing of the file.
478 PH/19 It seems that peer DN values may contain newlines (and other non-printing
479 characters?) which causes problems in log lines. The DN values are now
480 passed through string_printing() before being added to log lines.
482 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
483 and InterBase are left for another time.)
485 PH/21 Added message_body_newlines option.
487 PH/22 Guard against possible overflow in moan_check_errorcopy().
489 PH/23 POSIX allows open() to be a macro; guard against that.
491 PH/24 If the recipient of an error message contained an @ in the local part
492 (suitably quoted, of course), incorrect values were put in $domain and
493 $local_part during the evaluation of errors_copy.
499 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
500 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
503 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
504 issue a MAIL command.
506 PH/02 In an ACL statement such as
508 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
510 if a client was not listed at all, or was listed with a value other than
511 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
512 the condition was not true (as it should be), so access was not denied.
513 The bug was that the ! inversion was incorrectly passed on to the second
514 item. This has been fixed.
516 PH/03 Added additional dnslists conditions == and =& which are different from
517 = and & when the dns lookup returns more than one IP address.
519 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
520 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
522 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
523 FSYNC, which compiles an option called disable_fsync that allows for
524 bypassing fsync(). The documentation is heavily laced with warnings.
526 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
528 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
529 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
530 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
531 including adding "make clean"; (3) Added -fPIC when compiling the test
532 dynamically loaded module, to get rid of a warning.
534 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
535 message fails, move_frozen_messages = true and ignore_bounce_errors_after
536 = 0s. The bug is otherwise harmless.
538 PH/07 There was a bug in the dovecot authenticator such that the value of
539 $auth1 could be overwritten, and so not correctly preserved, after a
540 successful authentication. This usually meant that the value preserved by
541 the server_setid option was incorrect.
543 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
545 PH/09 Installed PCRE release 7.0.
547 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
548 run for batched SMTP input. It is now run at the start of every message
549 in the batch. While fixing this I discovered that the process information
550 (output by running exiwhat) was not always getting set for -bs and -bS
551 input. This is fixed, and it now also says "batched" for BSMTP.
553 PH/11 Added control=no_pipelining.
555 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
556 patch, slightly modified), and move the expansion of helo_data till after
557 the connection is made in the smtp transport (so it can use these
560 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
562 PH/14 Added log_selector = +pid.
564 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
566 PH/16 Add ${if forany and ${if forall.
568 PH/17 Added dsn_from option to vary the From: line in DSNs.
570 PH/18 Flush SMTP output before performing a callout, unless control =
571 no_callout_flush is set.
573 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
574 was true (the default) a successful delivery failed to delete the retry
575 item, thus causing premature timeout of the address. The bug is now
578 PH/20 Added hosts_avoid_pipelining to the smtp transport.
580 PH/21 Long custom messages for fakedefer and fakereject are now split up
581 into multiline reponses in the same way that messages for "deny" and
582 other ACL rejections are.
584 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
585 with slight modification.
587 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
588 draft, changing the syntax and factoring some duplicate code".
590 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
591 for deliveries of the second and subsequent messages over the same SMTP
594 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
595 ${reduce, with only minor "tidies".
597 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
599 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
600 expansion side effects.
602 PH/27 When a message times out after an over-quota error from an Exim-imposed
603 quota, the bounce message says "mailbox is full". This message was not
604 being given when it was a system quota that was exceeded. It now should
607 MH/03 Made $recipients available in local_scan(). local_scan() already has
608 better access to the recipient list through recipients_list[], but
609 $recipients can be useful in postmaster-provided expansion strings.
611 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
612 in the case of a MAIL command with additional options following the
613 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
614 were accidentally chopped off.
616 PH/29 SMTP synchronization checks are implemented when a command is read -
617 there is a check that no more input is waiting when there shouldn't be
618 any. However, for some commands, a delay in an ACL can mean that it is
619 some time before the response is written. In this time, more input might
620 arrive, invalidly. So now there are extra checks after an ACL has run for
621 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
622 pipelining has not been advertised.
624 PH/30 MH's patch to allow iscntrl() characters to be list separators.
626 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
627 returned in the SMTP response when smtp_return_error_details was false.
630 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
631 instead of the C I/O that was originally supplied, because problems were
634 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
635 Exim which did not show up earlier: it was assuming that a call to
636 SSL_CTX_set_info_callback() might give an error value. In fact, there is
637 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
638 was a macro that became an assignment, so it seemed to work. This has
639 changed to a proper function call with a void return, hence the compile
640 error. Exim's code has been fixed.
642 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
645 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
647 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
649 PH/37 If a message is not accepted after it has had an id assigned (e.g.
650 because it turns out to be too big or there is a timeout) there is no
651 "Completed" line in the log. When some messages of this type were
652 selected by exigrep, they were listed as "not completed". Others were
653 picked up by some special patterns. I have improved the selection
654 criteria to be more general.
656 PH/38 The host_find_failed option in the manualroute router can now be set
657 to "ignore", to completely ignore a host whose IP address cannot be
658 found. If all hosts are ignored, the behaviour is controlled by the new
659 host_all_ignored option.
661 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
662 homing or because of multiple MX records with /mx) generated more than
663 one IP address, and the following item turned out to be the local host,
664 all the secondary addresses of the first item were incorrectly removed
665 from the list, along with the local host and any following hosts (which
666 is what is supposed to happen).
668 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
669 whoever called Exim into the -H file. In the case of the daemon it was
670 behaving confusingly. When first started, it used values for whoever
671 started the daemon, but after a SIGHUP it used the Exim user (because it
672 calls itself on a restart). I have changed the code so that it now always
675 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
676 message are rejected with the same error (e.g. no authentication or bad
677 sender address), and a DATA command is nevertheless sent (as can happen
678 with PIPELINING or a stupid MUA), the error message that was given to the
679 RCPT commands is included in the rejection of the DATA command. This is
680 intended to be helpful for MUAs that show only the final error to their
683 PH/42 Another patch from the Sieve maintainer.
685 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
686 Eximstats - Fixed some broken HTML links and added missing column headers
688 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
689 columns for Rejects, Temp Rejects, Ham, and Spam rows.
691 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
693 PH/43 Yet another patch from the Sieve maintainer.
695 PH/44 I found a way to check for a TCP/IP connection going away before sending
696 the response to the final '.' that terminates a message, but only in the
697 case where the client has not sent further data following the '.'
698 (unfortunately, this is allowed). However, in many cases there won't be
699 any further data because there won't be any more messages to send. A call
700 to select() can be used: if it shows that the input is "ready", there is
701 either input waiting, or the socket has been closed. An attempt to read
702 the next input character can distinguish the two cases. Previously, Exim
703 would have sent an OK response which the client would never have see.
704 This could lead to message repetition. This fix should cure that, at
705 least in a lot of common cases.
707 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
708 advertised in response to EHLO.
714 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
715 fixed by 4.65/MH/01 (is this a record?) are fixed:
717 (i) An empty string was always treated as zero by the numeric comparison
718 operators. This behaviour has been restored.
720 (ii) It is documented that the numeric comparison operators always treat
721 their arguments as decimal numbers. This was broken in that numbers
722 starting with 0 were being interpreted as octal.
724 While fixing these problems I realized that there was another issue that
725 hadn't been noticed. Values of message_size_limit (both the global option
726 and the transport option) were treated as octal if they started with 0.
727 The documentation was vague. These values are now always treated as
728 decimal, and I will make that clear in the documentation.
734 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
735 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
738 MH/01 Don't check that the operands of numeric comparison operators are
739 integers when their expansion is in "skipping" mode (fixes bug
740 introduced by 4.64-PH/07).
742 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
743 child addresses, Exim now panics and dies. Previously, because the count
744 is held in a short int, deliveries were likely to be lost. As such a
745 large number of recipients for a single message is ridiculous
746 (performance will be very, very poor), I have chosen to impose a limit
747 rather than extend the field.
753 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
754 leftover -K file (the existence of which was triggered by #402).
755 While we were at it, introduced process PID as part of the -K
756 filename. This should rule out race conditions when creating
759 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
760 processing considerably. Previous code took too long for large mails,
761 triggering a timeout which in turn triggers #401.
763 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
764 in the DK code in transports.c. sendfile() is not really portable,
765 hence the _LINUX specificness.
767 TF/01 In the add_headers option to the mail command in an Exim filter,
768 there was a bug that Exim would claim a syntax error in any
769 header after the first one which had an odd number of characters
772 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
773 callout verification, Exim cached a "reject" for the entire domain. This
774 is correct for most verifications, but it is not correct for a recipient
775 verification with use_sender or use_postmaster set, because in that case
776 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
777 case of MAIL FROM:<> rejection from other early rejections (e.g.
778 rejection of HELO). When verifying a recipient using a non-null MAIL
779 address, the cache is ignored if it shows MAIL FROM:<> rejection.
780 Whatever the result of the callout, the value of the domain cache is
781 left unchanged (for any other kind of callout, getting as far as trying
782 RCPT means that the domain itself is ok).
784 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
787 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
788 manifest itself as EPIPE rather than ECONNECT. When tidying away a
789 session, the daemon ignores ECONNECT errors and logs others; it now
790 ignores EPIPE as well.
792 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
793 (quoted-printable decoding).
795 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
796 later the small subsequent patch to fix an introduced bug.
798 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
800 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
802 PH/08 An error is now given if message_size_limit is specified negative.
804 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
805 to be given (somewhat) arbitrary names.
807 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
810 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
811 miscellaneous code fixes
813 PH/10 Added the log_reject_target ACL modifier to specify where to log
816 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
817 hostname. This is wrong, because it relates to the incoming message (and
818 probably the interface on which it is arriving) and not to the outgoing
819 callout (which could be using a different interface). This has been
820 changed to use the value of the helo_data option from the smtp transport
821 instead - this is what is used when a message is actually being sent. If
822 there is no remote transport (possible with a router that sets up host
823 addresses), $smtp_active_hostname is used.
825 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
826 tweaks were necessary in order to get it to work (see also 21 below):
827 (a) The code assumed that strncpy() returns a negative number on buffer
828 overflow, which isn't the case. Replaced with Exim's string_format()
830 (b) There were several signed/unsigned issues. I just did the minimum
831 hacking in of casts. There is scope for a larger refactoring.
832 (c) The code used strcasecmp() which is not a standard C function.
833 Replaced with Exim's strcmpic() function.
834 (d) The code set only $1; it now sets $auth1 as well.
835 (e) A simple test gave the error "authentication client didn't specify
836 service in request". It would seem that Dovecot has changed its
837 interface. Fortunately there's a specification; I followed it and
838 changed what the client sends and it appears to be working now.
840 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
843 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
844 address A is aliased to B and C, where B exists and C does not. Without
845 -v the output is "A verified" because verification stops after a
846 successful redirection if more than one address is generated. However,
847 with -v the child addresses are also verified. Exim was outputting "A
848 failed to verify" and then showing the successful verification for C,
849 with its parentage. It now outputs "B failed to verify", showing B's
850 parentage before showing the successful verification of C.
852 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
853 look up a TXT record in a specific list after matching in a combined
856 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
857 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
858 they consult the DNS. I had assumed they would set it the way they
859 wanted; and indeed my experiments on Linux seem to show that in some
860 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
861 To be on the safe side, however, I have now made the interface to
862 host_find_byname() similar to host_find_bydns(), with an argument
863 containing the DNS resolver options. The host_find_byname() function now
864 sets these options at its start, just as host_find_bydns() does. The smtp
865 transport options dns_qualify_single and dns_search_parents are passed to
866 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
867 of host_find_byname() use the default settings of RES_DEFNAMES
868 (qualify_single) but not RES_DNSRCH (search_parents).
870 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
871 spool_read_header() do less string testing, by means of a preliminary
872 switch on the second character of optional "-foo" lines. (This is
873 overdue, caused by the large number of possibilities that now exist.
874 Originally there were few.) While I was there, I also converted the
875 str(n)cmp tests so they don't re-test the leading "-" and the first
876 character, in the hope this might squeeze out yet more improvement.
878 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
879 flag allowing group syntax was set by the header_syntax check but not
880 turned off, possible causing trouble later; (2) The flag was not being
881 set at all for the header_verify test, causing "group"-style headers to
882 be rejected. I have now set it in this case, and also caused header_
883 verify to ignore an empty address taken from a group. While doing this, I
884 came across some other cases where the code for allowing group syntax
885 while scanning a header line wasn't quite right (mostly, not resetting
886 the flag correctly in the right place). These bugs could have caused
887 trouble for malformed header lines. I hope it is now all correct.
889 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
890 with the "reply" argument non-NULL. The code, however (which originally
891 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
892 but it didn't always do it. This confused somebody who was copying the
893 code for some other use. I have removed all the tests.
895 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
896 feature that was used to support insecure browsers during the U.S. crypto
897 embargo. It requires special client support, and Exim is probably the
898 only MTA that supported it -- and would never use it because real RSA is
899 always available. This code has been removed, because it had the bad
900 effect of slowing Exim down by computing (never used) parameters for the
901 RSA_EXPORT functionality.
903 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
904 authenticator to fail if there's a tab character in the incoming data
905 (there should never be unless someone is messing about, as it's supposed
906 to be base64-encoded). Also added, on Timo's advice, the "secured" option
907 if the connection is using TLS or if the remote IP is the same as the
908 local IP, and the "valid-client-cert option" if a client certificate has
911 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
912 authenticators. This can be used for authorization after authentication
913 succeeds. (In the case of plaintext, it servers for both authentication
916 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
917 if any retry times were supplied.
919 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
920 connection, where there is no client IP address to check. In this
921 situation, the verify now always succeeds.
923 PH/25 Applied John Jetmore's -Mset patch.
925 PH/26 Added -bem to be like -Mset, but loading a message from a file.
927 PH/27 In a string expansion for a processed (not raw) header when multiple
928 headers of the same name were present, leading whitespace was being
929 removed from all of them, but trailing whitespace was being removed only
930 from the last one. Now trailing whitespace is removed from each header
931 before concatenation. Completely empty headers in a concatenation (as
934 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
935 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
937 PH/29 [Removed. This was a change that I later backed out, and forgot to
938 correct the ChangeLog entry (that I had efficiently created) before
939 committing the later change.]
941 PH/30 Exim was sometimes attempting to deliver messages that had suffered
942 address errors (4xx response to RCPT) over the same connection as other
943 messages routed to the same hosts. Such deliveries are always "forced",
944 so retry times are not inspected. This resulted in far too many retries
945 for the affected addresses. The effect occurred only when there were more
946 hosts than the hosts_max_try setting in the smtp transport when it had
947 the 4xx errors. Those hosts that it had tried were not added to the list
948 of hosts for which the message was waiting, so if all were tried, there
949 was no problem. Two fixes have been applied:
951 (i) If there are any address or message errors in an SMTP delivery, none
952 of the hosts (tried or untried) are now added to the list of hosts
953 for which the message is waiting, so the message should not be a
954 candidate for sending over the same connection that was used for a
955 successful delivery of some other message. This seems entirely
956 reasonable: after all the message is NOT "waiting for some host".
957 This is so "obvious" that I'm not sure why it wasn't done
958 previously. Hope I haven't missed anything, but it can't do any
959 harm, as the worst effect is to miss an optimization.
961 (ii) If, despite (i), such a delivery is accidentally attempted, the
962 routing retry time is respected, so at least it doesn't keep
963 hammering the server.
965 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
966 in ${readsocket because some servers need this prod.
968 PH/32 Added some extra debug output when updating a wait-xxx database.
970 PH/33 The hint "could be header name not terminated by colon", which has been
971 given for certain expansion errors for a long time, was not being given
972 for the ${if def:h_colon_omitted{... case.
974 PH/34 The spec says: "With one important exception, whenever a domain list is
975 being scanned, $domain contains the subject domain." There was at least
976 one case where this was not true.
978 PH/35 The error "getsockname() failed: connection reset by peer" was being
979 written to the panic log as well as the main log, but it isn't really
980 panic-worthy as it just means the connection died rather early on. I have
981 removed the panic log writing for the ECONNRESET error when getsockname()
984 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
985 runs only) independently of the message's sender address. This meant
986 that, if the 4xx error was in fact related to the sender, a different
987 message to the same recipient with a different sender could confuse
988 things. In particualar, this can happen when sending to a greylisting
989 server, but other circumstances could also provoke similar problems.
990 I have changed the default so that the retry time for these errors is now
991 based a combination of the sender and recipient addresses. This change
992 can be overridden by setting address_retry_include_sender=false in the
995 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
996 remote server are returned as part of bounce messages. This was not
997 happening for LMTP over a pipe (the lmtp transport), but now it is the
998 same for both kinds of LMTP.
1000 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1001 in header lines that were in fact CNAMEs. This is no longer the case.
1003 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1004 and queue runs started by the daemon processed all messages. This has
1005 been fixed so that -R and -S can now usefully be given with -q<time>.
1007 PH/40 Import PCRE release 6.7 (fixes some bugs).
1009 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1011 PH/42 Give an error if -q is specified more than once.
1013 PH/43 Renamed the variables $interface_address and $interface_port as
1014 $received_ip_address and $received_port, to make it clear that these
1015 values apply to message reception, and not to the outgoing interface when
1016 a message is delivered. (The old names remain recognized, of course.)
1018 PH/44 There was no timeout on the connect() call when using a Unix domain
1019 socket in the ${readsocket expansion. There now is.
1021 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1022 be meaningful with "accept".
1024 SC/01 Eximstats V1.43
1025 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1027 SC/02 Eximstats V1.44
1028 Use a glob alias rather than an array ref in the generated
1029 parser. This improves both readability and performance.
1031 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1032 Collect SpamAssassin and rejection statistics.
1033 Don't display local sender or destination tables unless
1034 there is data to show.
1035 Added average volumes into the top table text output.
1037 SC/04 Eximstats V1.46
1038 Collect data on the number of addresses (recipients)
1039 as well as the number of messages.
1041 SC/05 Eximstats V1.47
1042 Added 'Message too big' to the list of mail rejection
1043 reasons (thanks to Marco Gaiarin).
1045 SC/06 Eximstats V1.48
1046 Mainlog lines which have GMT offsets and are too short to
1047 have a flag are now skipped.
1049 SC/07 Eximstats V1.49 (Alain Williams)
1050 Added the -emptyok flag.
1052 SC/08 Eximstats V1.50
1053 Fixes for obtaining the IP address from reject messages.
1055 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1056 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1057 whitesspace changes from 4.64-PH/27
1059 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1062 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1063 are found, allow negative numbers in numeric criteria)
1065 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1067 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1068 to match changes made in 4.64-PH/43
1070 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1072 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1073 "domain" is actually an IP address, and force a failure. This locks out
1074 those revolvers/nameservers that support "A-for-A" lookups, in
1075 contravention of the specifications.
1077 PH/48 When a host name was looked up from an IP address, and the subsequent
1078 forward lookup of the name timed out, the host name was left in
1079 $sender_host_name, contrary to the specification.
1081 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1082 restricted to single-key lookups, Exim was not diagnosing an error if
1083 * or *@ was used with a query-style lookup.
1085 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1087 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1088 long ago, but noone interested enough thought of it. Let's just say that
1089 the "1.1" means that there are some new functions that weren't there at
1090 some point in the past.
1092 PH/51 Error processing for expansion failure of helo_data from an smtp
1093 transport during callout processing was broken.
1095 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1096 tested/used via the -bh/-bhc/-bs options.
1098 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1099 bug, fixed in subsequent PCRE releases).
1101 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1102 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1104 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1110 SC/01 Use a glob alias rather than an array ref in eximstats generated
1111 parser. This improves both readability and performance.
1113 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1114 Don't display local sender or destination tables in eximstats unless
1115 there is data to show.
1116 Added average volumes into the eximstats top table text output.
1118 SC/03 Collect data on the number of addresses (recipients) as well
1119 as the number of messages in eximstats.
1121 TF/01 Correct an error in the documentation for the redirect router. Exim
1122 does (usually) call initgroups() when daemonizing.
1124 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1125 with consistent privilege compared to when running as a daemon.
1127 TF/03 Note in the spec that $authenticated_id is not set for local
1128 submissions from trusted users.
1130 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1131 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1133 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1134 by adding some example configuration directives to the default
1135 configuration file. A little bit of work is required to uncomment the
1136 directives and define how usernames and passwords are checked, but
1137 there is now a framework to start from.
1139 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1140 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1141 without this. I don't know how relevant this is to other LDAP libraries.
1143 PH/02 Add the verb name to the "unknown ACL verb" error.
1145 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1147 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1149 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1150 directory not expanded when it should be if an expanded home directory
1151 was set for the address (which is overridden by the transport).
1153 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1156 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1157 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1158 because it is too late at that time, and has no effect.
1160 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1161 security issue with \' (bugzilla #107). I could not use the
1162 PQescapeStringConn() function, because it needs a PGconn value as one of
1165 PH/08 When testing addresses using -bt, indicate those final addresses that
1166 are duplicates that would not cause an additional delivery. At least one
1167 person was confused, thinking that -bt output corresponded to deliveries.
1168 (Suppressing duplicates isn't a good idea as you lose the information
1169 about possibly different redirections that led to the duplicates.)
1171 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1172 systems where poll() doesn't work, in particular OS X.
1174 PH/10 Added more information to debugging output for retry time not reached.
1176 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1177 operations in malware.c.
1179 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1182 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1183 syslog_duplication set false, log lines that would normally be written
1184 both the the main log and to the reject log were not written to syslog at
1187 PH/14 In the default configuration, change the use of "message" in ACL warn
1188 statements to "add_header".
1190 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1191 not followed by a command (e.g. "seen endif").
1193 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1194 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1197 PH/17 Added extra conditions to the default value of delay_warning_condition
1201 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1202 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1203 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1206 The Auto-Submitted: and various List- headers are standardised, whereas I
1207 don't think Precedence: ever was.
1209 PH/18 Refactored debugging code in route_finduser() to show more information,
1210 in particular, the error code if getpwnam() issues one.
1212 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1213 This is apparently needed in addition to the PH/07 change above to avoid
1214 any possible encoding problems.
1216 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1217 but not after initializing Perl.
1219 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1220 output them only if debugging. By default they are written stderr,
1221 apparently, which is not desirable.
1223 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1226 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1229 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1231 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1232 authenticated or an ident call has been made. Suppress the default
1233 values for $authenticated_id and $authenticated_sender (but permit -oMai
1234 and -oMas) when testing with -bh.
1236 PH/24 Re-jigged the order of the tests in the default configuration so that the
1237 tests for valid domains and recipients precede the DNS black list and CSA
1238 tests, on the grounds that those ones are more expensive.
1240 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1241 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1242 This bug exists in every version of Exim that I still have, right back to
1245 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1246 However, an attempt to turn on case-sensitivity in a regex key by
1247 including (?-i) didn't work because the subject string was already
1248 lowercased, and the effects were non-intuitive. It turns out that a
1249 one-line patch can be used to allow (?-i) to work as expected.
1255 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1256 other effects) broke the use of negated acl sub-conditions.
1258 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1261 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1262 "Deny" causes Exim to reject the incoming connection with a 554 error.
1263 Unfortunately, if there is a major crisis, such as a disk failure,
1264 tcp-wrappers gives "deny", whereas what one would like would be some
1265 kind of temporary error. A kludge has been added to help with this.
1266 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1267 554 error is used if errno is still zero or contains ENOENT (which occurs
1268 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1271 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1273 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1274 errors. Otherwise a message that provokes a temporary error (when other
1275 messages do not) can cause a whole host to time out.
1277 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1278 addresses were routed directly to files or pipes from a redirect router.
1279 File deliveries just didn't batch; pipe deliveries might have suffered
1282 PH/06 A failure to get a lock for a hints database would erroneously always say
1283 "Failed to get write lock", even when it was really a read lock.
1285 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1286 of 0600. This has been changed to use the value of the lockfile_mode
1287 option (which defaults to 0600).
1289 PH/08 Applied small patch from the Sieve maintainer.
1291 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1292 folder from quota calculations, a direct delivery into this folder messed
1293 up the contents of the maildirsize file. This was because the regex was
1294 used only to exclude .Trash (or whatever) when the size of the mailbox
1295 was calculated. There was no check that a delivery was happening into an
1296 excluded directory. This bug has been fixed by ignoring all quota
1297 processing for deliveries into excluded directories.
1299 PH/10 Added the maildirfolder_create_regex option to appendfile.
1305 PH/01 The code for finding all the local interface addresses on a FreeBSD
1306 system running IPv6 was broken. This may well have applied to all BSD
1307 systems, as well as to others that have similar system calls. The broken
1308 code found IPv4 interfaces correctly, but gave incorrect values for the
1309 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1310 that it would not match correctly against @[] and not recognize the IPv6
1313 PH/02 The ipliteral router was not recognizing addresses of the form user@
1314 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1316 PH/03 Added disable_ipv6.
1318 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1319 decoded form, because it is most often used to construct To: headers
1320 lines in autoreplies, and the decoded form may well be syntactically
1321 invalid. However, $reply_address has leading white space removed, and all
1322 newlines turned into spaces so that the autoreply transport does not
1325 PH/05 If group was specified without a user on a router, and no group or user
1326 was specified on a transport, the group from the router was ignored.
1328 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1329 for visible compile-time settings that can be used to change these
1330 numbers, for those that want even more. Backwards compatibility with old
1331 spool files has been maintained. However, going back to a previous Exim
1332 release will lost any variables that are in spool files.
1334 PH/07 Two small changes when running in the test harness: increase delay when
1335 passing a TCP/IP connection to a new process, in case the original
1336 process has to generate a bounce, and remove special handling of
1337 127.0.0.2 (sic), which is no longer necessary.
1339 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1340 be the same on different OS.
1342 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1345 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1346 whether --show-vars was specified or not
1348 JJ/02 exipick: Added support for new ACL variable spool format introduced
1351 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1352 syntactically invalid From: or Reply-to: line, and a filter used this to
1353 generate an autoreply, and therefore failed to obtain an address for the
1354 autoreply, Exim could try to deliver to a non-existent relative file
1355 name, causing unrelated and misleading errors. What now happens is that
1356 it logs this as a hard delivery error, but does not attempt to create a
1359 PH/11 The exinext utility has a -C option for testing purposes, but although
1360 the given file was scanned by exinext itself; it wasn't being passed on
1361 when Exim was called.
1363 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1364 an end-of-file indication when reading a command response.
1366 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1367 compiled. In many other places in Exim, IPv6 addresses are always
1368 recognized, so I have changed this. It also means that IPv4 domain
1369 literals of the form [IPV4:n.n.n.n] are now always recognized.
1371 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1372 used if the router is not running as root, for example, when verifying at
1373 ACL time, or when using -bh. The debugging output from this situation was
1374 non-existent - all you got was a failure to exec. I have made two
1377 (a) Failures to set uid/gid, the current directory, or a process leader
1378 in a subprocess such as that created by queryprogram now generate
1379 suitable debugging ouput when -d is set.
1381 (b) The queryprogram router detects when it is not running as root,
1382 outputs suitable debugging information if -d is set, and then runs
1383 the subprocess without attempting to change uid/gid.
1385 PH/15 Minor change to Makefile for building test_host (undocumented testing
1388 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1389 additional section of a DNS packet that returns MX or SRV records.
1390 Instead, it always explicitly searches for A/AAAA records. This avoids
1391 major problems that occur when a DNS server includes only records of one
1392 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1393 fixed another bug: if SRV records were looked up and the corresponding
1394 address records were *not* found in the additional section, the port
1395 values from the SRV records were lost.
1397 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1398 using the correct key (the original address) when searching the retry
1399 rules in order to find which one to use for generating the retry hint.
1401 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1402 adding the default one. Similarly, if it contains a Reply-To: header, the
1403 errors_reply_to option, if set, is not used.
1405 PH/19 When calculating a retry time, Exim used to measure the "time since
1406 failure" by looking at the "first failed" field in the retry record. Now
1407 it does not use this if it is later than than the arrival time of the
1408 message. Instead it uses the arrival time. This makes for better
1409 behaviour in cases where some deliveries succeed, thus re-setting the
1410 "first failed" field. An example is a quota failure for a huge message
1411 when small messages continue to be delivered. Without this change, the
1412 "time since failure" will always be short, possible causing more frequent
1413 delivery attempts for the huge message than are intended.
1414 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1416 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1417 $1, $2, $3) because the numerical variables can be reset during some
1418 expansion items (e.g. "match"), thereby losing the authentication data.
1420 PH/21 Make -bV show the size of off_t variables so that the test suite can
1421 decide whether to run tests for quotas > 2G.
1423 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1424 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1425 filecount value is greater than 2G or if a quota value is greater than 2G
1426 on a system where the size of off_t is not greater than 4, a panic error
1429 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1430 never match. The debug and -bh output now contains an explicit error
1431 message indicating a malformed IPv4 address or mask.
1433 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1434 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1435 PH/23 above applies.
1437 PH/25 Do not write to syslog when running in the test harness. The only
1438 occasion when this arises is a failure to open the main or panic logs
1439 (for which there is an explicit test).
1441 PH/26 Added the /no_tell option to "control=freeze".
1443 PH/27 If a host name lookup failed very early in a connection, for example, if
1444 the IP address matched host_lookup and the reverse lookup yielded a name
1445 that did not have a forward lookup, an error message of the form "no IP
1446 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1447 could be logged. Now it outputs the IP address instead of "NULL".
1449 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1450 allows the sender and the authenticated sender to be set when
1451 submitting a message from within Exim. Since child_open_exim() is
1452 documented for local_scan(), the new function should be too.
1454 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1455 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1456 results in an empty string is now treated as unset.
1458 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1460 PH/31 Added sender_verify_fail logging option.
1462 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1463 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1464 tidied the source and removed it altogether.
1466 PH/33 When a queue run was abandoned because the load average was too high, a
1467 log line was always written; now it is written only if the queue_run log
1468 selector is set. In addition, the log line for abandonment now contains
1469 information about the queue run such as the pid. This is always present
1470 in "start" and "stop" lines but was omitted from the "abandon" line.
1472 PH/34 Omit spaces between a header name and the colon in the error message that
1473 is given when verify = headers_syntax fails (if there are lots of them,
1474 the message gets confusing).
1476 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1477 names, as there are now some PTR records that contain slashes. This check
1478 is only to protect against broken name servers that fall over on strange
1479 characters, so the fact that it applies to all lookups doesn't matter.
1481 PH/36 Now that the new test suite is complete, we can remove some of the
1482 special code in Exim that was needed for the old test suite. For example,
1483 sorting DNS records because real resolvers return them in an arbitrary
1484 order. The new test suite's fake resolver always returns records in the
1487 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1488 bounces) except when queue_only is set, to avoid logging races between
1489 the different processes.
1491 PH/38 Panic-die if .include specifies a non-absolute path.
1493 PH/39 A tweak to the "H" retry rule from its user.
1495 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1496 a label. They prevented compilation on older perls.
1498 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1499 a warning to be raised on newish perls.
1501 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1502 on queue. Changes to match documented behaviour of showing count of
1503 messages matching specified criteria.
1505 PH/40 Changed the default ident timeout from 30s to 5s.
1507 PH/41 Added support for the use of login_cap features, on those BSD systems
1508 that have them, for controlling the resources used by pipe deliveries.
1510 PH/42 The content-scanning code uses fopen() to create files in which to put
1511 message data. Previously it was not paying any attention to the mode of
1512 the files. Exim runs with umask(0) because the rest of the code creates
1513 files with open(), and sets the required mode explicitly. Thus, these
1514 files were ending up world-writeable. This was not a big issue, because,
1515 being within the spool directory, they were not world-accessible. I have
1516 created a function called modefopen, which takes an additional mode
1517 argument. It sets umask(777), creates the file, chmods it to the required
1518 mode, then resets the umask. All the relevant calls to fopen() in the
1519 content scanning code have been changed to use this function.
1521 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1522 to 24 hours. This avoids potential overflow problems when processing G
1523 and H retry rules. I suspect nobody ever tinkers with this value.
1525 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1527 PH/45 When the plaintext authenticator is running as a client, the server's
1528 challenges are checked to ensure they are valid base64 strings. By
1529 default, the authentication attempt is cancelled if an invalid string is
1530 received. Setting client_ignore_invalid_base64 true ignores these errors.
1531 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1532 they are received. Thus, the responses can be made to depend on the
1533 challenges. If an invalid string is ignored, an empty string is placed in
1536 PH/46 Messages that are created by the autoreply transport now contains a
1537 References: header, in accordance with RFCs 2822 and 3834.
1539 PH/47 Added authenticated_sender_force to the smtp transport.
1541 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1543 PH/49 Installed latest patch from the Sieve maintainer.
1545 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1546 was also set, the appendfile transport was unnecessarily scanning a
1547 directory of message files (e.g. for maildir delivery) to find the count
1548 of files (along with the size), even though it did not need this
1549 information. It now does the scan only if it needs to find either the
1550 size of the count of files.
1552 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1554 PH/52 Two bugs concerned with error handling when the smtp transport is
1557 (i) Exim was not creating retry information for temporary errors given
1558 for individual recipients after the DATA command when the smtp transport
1559 was used in LMTP mode. This meant that they could be retried too
1560 frequently, and not timed out correctly.
1562 (ii) Exim was setting the flag that allows error details to be returned
1563 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1564 recipients that were returned after the DATA command.
1566 PH/53 This is related to PH/52, but is more general: for any failing address,
1567 when detailed error information was permitted to be returned to the
1568 sender, but the error was temporary, then after the final timeout, only
1569 "retry timeout exceeded" was returned. Now it returns the full error as
1570 well as "retry timeout exceeded".
1572 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1573 do this, and (what is worse) MTAs that accept it.
1575 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1576 will now be deprecated.
1578 PH/56 New os.c-cygwin from the Cygwin maintainer.
1580 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1581 formats (previously only available in exim formats via -bpr, -bpru,
1582 and -bpra. Now also available in native and exiqgrep formats)
1584 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1585 with very large, slow to parse queues
1587 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1589 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1591 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1592 responses to authentication challenges, though it was showing the
1593 challenges; (ii) I've removed the CR characters from the debug output for
1596 PH/58 Allow for the insertion of a newline as well as a space when a string
1597 is turned into more than one encoded-word during RFC 2047 encoding. The
1598 Sieve code now uses this.
1600 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1601 data_4xx, lost_connection, tls_required.
1603 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1604 message was being sent as an SMTP response.
1606 PH/61 Add -l and -k options to exicyclog.
1608 PH/62 When verifying, if an address was redirected to one new address, so that
1609 verification continued, and the new address failed or deferred after
1610 having set something in $address_data, the value of $address_data was not
1611 passed back to the ACL. This was different to the case when no
1612 redirection occurred. The value is now passed back in both cases.
1614 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1615 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1616 use login_cap.h, so on its own it isn't the distinguishing feature. The
1617 new name refers directly to the setclassresources() function.
1619 PH/65 Added configuration files for NetBSD3.
1621 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1623 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1624 is preferred over IPv4.
1626 PH/68 The bounce_return_message and bounce_return_body options were not being
1627 honoured for bounces generated during the reception of non-SMTP messages.
1628 In particular, this applied to messages rejected by the ACL. This bug has
1629 been fixed. However, if bounce_return_message is true and bounce_return_
1630 body is false, the headers that are returned for a non-SMTP message
1631 include only those that have been read before the error was detected.
1632 (In the case of an ACL rejection, they have all been read.)
1634 PH/69 The HTML version of the specification is now built in a directory called
1635 spec_html instead of spec.html, because the latter looks like a path with
1636 a MIME-type, and this confuses some software.
1638 PH/70 Catch two compiler warnings in sieve.c.
1640 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1641 function verify_get_ident() calls ip_connect() to connect a socket, but
1642 if the "connect()" function timed out, ip_connect() used to close the
1643 socket. However, verify_get_ident() also closes the socket later, and in
1644 between Exim writes to the log, which may get opened at this point. When
1645 the socket was closed in ip_connect(), the log could get the same file
1646 descriptor number as the socket. This naturally causes chaos. The fix is
1647 not to close the socket in ip_connect(); the socket should be closed by
1648 the function that creates it. There was only one place in the code where
1649 this was missing, in the iplookup router, which I don't think anybody now
1650 uses, but I've fixed it anyway.
1652 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1653 well as to direct DNS lookups. Otherwise the handling of names in host
1654 lists is inconsistent and therefore confusing.
1660 PH/01 Two changes to the default runtime configuration:
1662 (1) Move the checks for relay_from_hosts and authenticated clients from
1663 after to before the (commented out) DNS black list checks.
1665 (2) Add control=submission to the relay_from_hosts and authenticated
1666 clients checks, on the grounds that messages accepted by these
1667 statements are most likely to be submissions.
1669 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1671 (1) Generate an error if the third argument for the ${prvs expansion is
1674 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1677 (3) Reset the variables that are obtained from the first argument of
1678 ${prvscheck and used in the second argument before leaving the code,
1679 because their memory is reclaimed, so using them afterwards may do
1682 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1683 one (it's much easier than Tom thought :-).
1685 (5) Because of (4), we can now allow for the use of $prvscheck_result
1686 inside the third argument.
1688 PH/03 For some reason, the default setting of PATH when running a command from
1689 a pipe transport was just "/usr/bin". I have changed it to
1692 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1693 anything to be listed in the output from -bV.
1695 PH/05 When a filter generated an autoreply, the entire To: header line was
1696 quoted in the delivery log line, like this:
1698 => >A.N.Other <ano@some.domain> <original@ddress> ...
1700 This has been changed so that it extracts the operative address. There
1701 may be more than one such address. If so, they are comma-separated, like
1704 => >ano@some.domain,ona@other.domain <original@ddress> ...
1706 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1707 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1708 not being looked up in the rDNS to get a host name, Exim was showing the
1709 IP address twice in Received: lines, even though the IP addresses were
1710 identical. For example:
1712 Received: from [1.2.3.4] (helo=[1.2.3.4])
1714 However, if the real host name was known, it was omitting the HELO data
1715 if it matched the actual IP address. This has been tidied up so that it
1716 doesn't show the same IP address twice.
1718 PH/07 When both +timestamp and +memory debugging was on, the value given by
1719 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1720 called by the debug printing, thereby overwriting the timestamp buffer.
1721 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1723 PH/08 When the original message was included in an autoreply transport, it
1724 always said "this is a copy of the message, including all the headers",
1725 even if body_only or headers_only was set. It now gives an appropriate
1728 PH/09 Applied a patch from the Sieve maintainer which:
1730 o fixes some comments
1731 o adds the (disabled) notify extension core
1732 o adds some debug output for the result of if/elsif tests
1733 o points to the current vacation draft in the documentation
1734 and documents the missing references header update
1738 o fixes a bug in processing the envelope test (when testing
1739 multiple envelope elements, the last element determinted the
1742 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1743 Electronic Mail") by including:
1745 Auto-submitted: auto-generated
1747 in the messages that it generates (bounce messages and others, such as
1748 warnings). In the case of bounce messages for non-SMTP mesages, there was
1749 also a typo: it was using "Auto_submitted" (underscore instead of
1750 hyphen). Since every message generated by Exim is necessarily in response
1751 to another message, thes have all been changed to:
1753 Auto-Submitted: auto-replied
1755 in accordance with these statements in the RFC:
1757 The auto-replied keyword:
1759 - SHOULD be used on messages sent in direct response to another
1760 message by an automatic process,
1762 - MUST NOT be used on manually-generated messages,
1764 - MAY be used on Delivery Status Notifications (DSNs) and Message
1765 Disposition Notifications (MDNs),
1767 - MUST NOT be used on messages generated by automatic or periodic
1768 processes, except for messages which are automatic responses to
1771 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1772 to the default Received: header definition.
1774 PH/12 Added log selector acl_warn_skipped (default on).
1776 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1777 variables because (a) they are in the wrong storage pool and (b) even if
1778 they were copied, it wouldn't work properly because of the caching.
1780 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1781 checking when decoding. Apparently there are clients that generate
1782 overlong encoded strings. Why am I not surprised?
1784 PH/15 If the first argument of "${if match_address" was not empty, but did not
1785 contain an "@" character, Exim crashed. Now it writes a panic log message
1786 and treats the condition as false.
1788 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1790 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1791 extension "envelope-auth". The code is finished and in agreement with
1792 other implementations, but there is no documentation so far and in fact,
1793 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1794 not changing the active code.
1796 Print executed "if" and "elsif" statements when debugging is used. This
1797 helps a great deal to understand what a filter does.
1799 Document more things not specified clearly in RFC3028. I had all this
1800 sorted out, when out of a sudden new issues came to my mind. Oops."
1802 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1805 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1806 canonical form (as documented). However, after a host name lookup from
1807 the IP address, check_host() was doing a simple string comparison with
1808 addresses acquired from the DNS when checking that the found name did
1809 have the original IP as one of its addresses. Since any found IPv6
1810 addresses are likely to be in abbreviated form, the comparison could
1811 fail. Luckily, there already exists a function for doing the comparison
1812 by converting both addresses to binary, so now that is used instead of
1813 the text comparison.
1815 PH/20 There was another similar case to PH/19, when a complete host name was
1816 given in a host list; looking up its IP address could give an abbreviated
1817 form, whereas the current host's name might or might not be abbreviated.
1818 The same fix has been applied.
1824 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1825 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1828 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1829 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1831 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1833 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1834 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1835 Also, for "Auto-Submitted", treat anything other than "no" as
1836 non-personal, in accordance with RFC 3834. (Previously it treated
1837 anything starting "auto-" as non-personal.)
1839 TF/01 The control=submission/name=... option had a problem with syntax
1840 errors if the name included a slash character. The /name= option
1841 now slurps the rest of the string, so it can include any characters
1842 but it must come last in the list of options (after /sender_retain
1845 PH/05 Some modifications to the interface to the fake nameserver for the new
1853 TK/01 Added the "success_on_redirect" address verification option. See
1854 NewStuff for rationale and an example.
1856 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1858 PH/02 Patch to exigrep to allow it to work on syslog lines.
1860 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1861 fread() to skip over the body file's header line, because in Cygwin the
1862 header line is locked and is inaccessible.
1864 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1865 co-exist for some time) to make it clear that it is the Exim ID that is
1866 referenced, not the Message-ID: header line.
1868 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1869 string_format() function, because snprintf() does not exist on all
1872 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1873 ${stat: expansion item.
1875 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1876 protocol synchronization error", to keep the pedants happy.
1878 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1879 well as for IRIX systems, when gcc is being used. See the host.c source
1882 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1884 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1887 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1888 transport and to the smtp transport in LMTP mode.
1890 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1892 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1893 run a filter in a subprocess. This could lead to confusion in subsequent
1894 lookups in the parent process. There should also be a search_tidyup() at
1895 the end of the subprocess.
1897 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1898 only if the host matched helo_try_verify_hosts, which caused the
1899 verification to occur when the EHLO/HELO command was issued. The ACL just
1900 tested the remembered result. Now, if a previous verification attempt has
1901 not happened, "verify = helo" does it there and then.
1903 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1905 TK/03 Fix log output including CR from clamd.
1907 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1908 exist provoked a memory error which could cause a segfault.
1910 PH/15 Installed PCRE 6.2
1912 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1914 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1915 of the problem. Specifically, suggested +O2 rather than +O1 for the
1918 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1920 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1921 record, the port was not correctly logged when the outgoing_port log
1922 selector was set (it logged the transort's default port).
1924 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1925 fallback_hosts, and "hosts" in the smtp transport.
1927 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1928 host errors such as "Connection refused".
1930 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1931 authentication with radiusclient 0.4.9:
1933 - Error returned from rc_read_config was caught wrongly
1934 - Username/password not passed on to radius server due to wrong length.
1936 The presumption is that some radiusclient API changes for 4.51/PH/17
1937 were not taken care of correctly. The code is still untested by me (my
1938 Linux distribution still has 0.3.2 of radiusclient), but it was
1939 contributed by a Radius user.
1941 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1942 expanding the "port" option of the smtp transport.
1944 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1945 while reading a MIME header. Thanks to Tom Hughes for a patch.
1947 PH/24 Include config.h inside local_scan.h so that configuration settings are
1950 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1951 that in an ACL for RCPT (for example), you can examine exactly what was
1954 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1955 commands, but it was not correctly comparing the address with the actual
1956 client host address. Thus, it would show the EHLO address in Received:
1957 header lines when this was not necessary.
1959 PH/27 Added the % operator to ${eval:}.
1961 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1962 it should be ignoring failures (because with -C, for example, it has lost
1963 privilege). It wasn't ignoring creation failures other than "already
1966 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1969 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1970 because an input line was too long, either on its own, or by virtue of
1971 too many continuations, the temporary file was not being removed, and the
1972 return code was incorrect.
1974 PH/31 Missing "BOOL" in function definition in filtertest.c.
1976 PH/32 Applied Sieve patches from the maintainer.
1978 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1980 PH/33 Added "verify = not_blind".
1982 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1983 Local/Makefile (with some defaults set). These are used in built scripts
1984 such as exicyclog, but they have never been used in the exim_install
1985 script (though there are many overriding facilities there). I have
1986 arranged that the exim_install script now takes note of these two
1989 PH/35 Installed configuration files for Dragonfly.
1991 PH/36 When a locally submitted message by a trusted user did not contain a
1992 From: header, and the sender address was obtained from -f or from an SMTP
1993 MAIL command, and the trusted user did not use -F to supply a sender
1994 name, $originator_name was incorrectly used when constructing a From:
1995 header. Furthermore, $originator_name was used for submission mode
1996 messages from external hosts without From: headers in a similar way,
1997 which is clearly wrong.
1999 PH/37 Added control=suppress_local_fixups.
2001 PH/38 When log_selector = +received_sender was set, and the addition of the
2002 sender made the log line's construction buffer exactly full, or one byte
2003 less than full, an overflow happened when the terminating "\n" was
2006 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2007 when the result of a list match is failure because a DNS lookup failed.
2009 PH/40 RM_COMMAND is now used in the building process.
2011 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2012 the "build-* directories that it finds.
2014 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2015 address in a domain literal was a prefix of an interface address.
2017 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2018 when verifying a sender address, unless rewrite_headers is false.
2020 PH/44 Wrote a long comment about why errors_to addresses are verified as
2021 recipients, not senders.
2023 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2024 the ratelimit ACL was added.
2026 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2028 PH/46 Added extra information about PostgreSQL errors to the error string.
2030 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2031 avoiding the need to install special zones in a real server. This is
2032 backwards compatible; if it can't find the fake resolver, it drops back.
2033 Thus, both old and new test suites can be run.
2035 TF/02 Added util/ratelimit.pl
2037 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2038 clock is set back in time.
2040 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2041 Candler <B.Candler@pobox.com>.
2043 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2044 OK for addresses that are the result of redirections.
2046 PH/48 A number of further additions for the benefit of the new test suite,
2047 including a fake gethostbyname() that interfaces to the fake DNS resolver
2050 TF/06 The fix for widen_domains has also been applied to qualify_single and
2051 search_parents which are the other dnslookup options that can cause
2054 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2057 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2059 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2060 service" if it encountered an MX record with an empty target hostname.
2061 The message is now "an MX or SRV record indicated no SMTP service".
2063 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2064 if the DNS of the sending site is misconfigured. This is quite a
2065 common situation. This change restores the behaviour of treating a
2066 helo verification defer as a failure.
2068 PH/51 If self=fail was set on a router, the bounce message did not include the
2069 actual error message.
2075 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2077 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2078 command itself ended in error, the underlying message about the transport
2079 filter timeout was being overwritten with the pipe command error. Now the
2080 underlying error message should be appended to the second error message.
2082 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2084 PH/02 Reduce the amount of output that "make" produces by default. Full output
2085 can still be requested.
2087 PH/03 The warning log line about a condition test deferring for a "warn" verb
2088 was being output only once per connection, rather than after each
2089 occurrence (because it was using the same function as for successful
2090 "warn" verbs). This seems wrong, so I have changed it.
2092 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2093 it should not have, which might have caused a crash in the right
2094 circumstances, but probably never did.
2096 PH/04 Installed a modified version of Tony Finch's patch to make submission
2097 mode fix the return path as well as the Sender: header line, and to
2098 add a /name= option so that you can make the user's friendly name appear
2101 TF/03 Added the control = fakedefer ACL modifier.
2103 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2104 Mark Lowes for thorough testing.
2106 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2108 TK/03 Merged latest SRS patch from Miles Wilton.
2110 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2111 with the definition in sysexits.h (which is #included earlier).
2112 Fortunately, Exim does not actually use EX_OK. The code used to try to
2113 preserve the sysexits.h value, by assumimg that macro definitions were
2114 scanned for macro replacements. I have been disabused of this notion,
2115 so now the code just undefines EX_OK before #including unistd.h.
2117 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2118 in the smtp transport. When a block could not be written in a single
2119 write() function, the timeout was being re-applied to each part-write.
2120 This seems wrong - if the receiver was accepting one byte at a time it
2121 would take for ever. The timeout is now adjusted when this happens. It
2122 doesn't have to be particularly precise.
2124 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2125 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2127 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2128 without a domain if the check to <postmaster@domain> fails.
2130 SC/01 Eximstats: added -xls and the ability to specify output files
2131 (patch written by Frank Heydlauf).
2133 SC/02 Eximstats: use FileHandles for outputing results.
2135 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2137 SC/04 Eximstats: fixed display of large numbers with -nvr option
2139 SC/05 Eximstats: fixed merging of reports with empty tables.
2141 SC/06 Eximstats: added the -include_original_destination flag
2143 SC/07 Eximstats: removed tabs and trailing whitespace.
2145 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2147 TK/06 MBOX spool code: Add real "From " MBOX separator line
2148 so the .eml file is really in mbox format (even though
2149 most programs do not really care). Patch from Alex Miller.
2151 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2152 The latter is generated from $received_to and is only set if the
2153 message has one envelope recipient. SA can use these headers,
2154 obviously out-of-the-box. Patch from Alex Miller.
2156 PH/08 The ${def test on a variable was returning false if the variable's
2157 value was "0", contrary to what the specification has always said!
2158 The result should be true unless the variable is empty.
2160 PH/09 The syntax error of a character other than { following "${if
2161 def:variable_name" (after optional whitespace) was not being diagnosed.
2162 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2163 accidental colon was present, for example, could give incorrect results.
2165 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2166 result is used (not including appendfile, where other changes are about
2169 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2170 This involved changing a lot of size variables from int to off_t. It
2171 should work with maildirs and everything.
2173 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2174 spamd dying while we are connected to it.
2176 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2179 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2180 function for BDB 4.3.
2182 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2184 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2185 the sources for PCRE have been re-arranged and more files are now
2188 PH/15 The code I had for printing potentially long long variables in PH/11
2189 above was not the best (it lost precision). The length of off_t variables
2190 is now inspected at build time, and an appropriate printing format (%ld
2191 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2192 to be "long long int" or "long int". This is needed for the internal
2193 formatting function string_vformat().
2195 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2196 the configuration file to be ":syslog", then the script "guesses" where
2197 the logs files are, rather than using the compiled in default. In our
2198 case the guess is not the same as the compiled default, so the script
2199 suddenly stopped working when I started to use syslog. The patch checks
2200 to see if log_file_path is "". If so, it attempts to read it from exim
2201 with no configuration file to get the compiled in version, before it
2202 falls back to the previous guessing code."
2204 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2205 implementing BATV in an Exim configuration. See NewStuff for the gory
2208 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2209 Makefile that are specific to HP-UX.
2211 PH/18 If the "use_postmaster" option was set for a recipient callout together
2212 with the "random" option, the postmaster address was used as the MAIL
2213 FROM address for the random test, but not for the subsequent recipient
2214 test. It is now used for both.
2216 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2217 patch removes a few documentation additions to RFC 3028, because the
2218 latest draft now contains them. It adds the new en;ascii-case comparator
2219 and a new error check for 8bit text in MIME parts. Comparator and
2220 require names are now matched exactly. I enabled the subaddress
2221 extension, but it is not well tested yet (read: it works for me)."
2223 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2224 rework some of the code of TK/09 above to avoid the hardwired use of
2225 "%lld" and "long long". Replaced the call to snprintf() with a call to
2228 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2229 records point to non-existent hosts", "retry timeout exceeded", and
2230 "retry time not reached for any host after a long failure period".
2232 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2233 experimental DomainKeys support:
2235 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2236 (2) On an error such as an illegally used "control", the wrong name for
2237 the control was given.
2239 These problems did NOT occur unless DomainKeys support was compiled.
2241 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2243 PH/24 Added ${if match_ip condition.
2245 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2246 databases so that it will be absolutely obvious if a crash occurs in the
2247 DB library. This is a regular occurrence (often caused by mis-matched
2250 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2251 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2252 picked up on a user's system that detects such things. There doesn't seem
2253 to be a gcc warning option for this - only an attribute that has to be
2254 put on the function's prototype. It seems that in Fedora Core 4 they have
2255 set this on a number of new functions. No doubt there will be more in due
2258 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2259 specify a transport. However, if an address that was verified by such a
2260 router was the subject of a callout, Exim crashed because it tried to
2261 read the rcpt_include_affixes from the non-existent transport. Now it
2262 just assumes that the setting of that option is false. This bug was
2263 introduced by 4.51/PH/31.
2265 PH/28 Changed -d+all to exclude +memory, because that information is very
2266 rarely of interest, but it makes the output a lot bigger. People tend to
2267 do -d+all out of habit.
2269 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2270 code in os-type was giving problems when libc.so lives in lib64, like on
2273 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2274 aren't the modern standard, and it seems that some systems' include files
2275 don't always have them. Exim was already checking for some of the newer
2276 ones like T_AAAA, and defining it itself. I've added checks for all the
2277 record types that Exim uses.
2279 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2280 not automatically generating a new one, as it is supposed to. This
2281 prevented TLS from working. If the file did exist, but contained invalid
2282 data, a new version was generated, as expected. It was only the case of a
2283 non-existent file that was broken.
2285 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2286 with a change in libdomainkeys > 0.64.
2288 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2289 from DNS. If the selector record carries the flag, it now has
2290 precedence over the domain-wide flag.
2292 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2294 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2295 the use of a port name that isn't defined in /etc/services) occurred, the
2296 message was deferred as in a normal delivery, and thus remained on the
2297 spool, instead of being failed because of the mua_wrapper setting. This
2298 is now fixed, and I tidied up some of the mua_wrapper messages at the
2301 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2302 the messages in a hash of arrays rather than using individual hashes.
2303 This is a bit cleaner and results in dramatic memory savings, albeit
2304 at a slight CPU cost.
2306 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2307 as requested by Marc Sherman.
2309 SC/10 Eximstats: added histograms for user specified patterns as requested
2312 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2314 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2315 fopen() in the content-scanning modules that did not already have it.
2321 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2322 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2324 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2326 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2328 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2329 to test Sieve filters that use "vacation".
2331 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2332 that changes the way the GnuTLS parameters are stored in the cache file.
2333 The new format can be generated externally. For backward compatibility,
2334 if the data in the cache doesn't make sense, Exim assumes it has read an
2335 old-format file, and it generates new data and writes a new file. This
2336 means that you can't go back to an older release without removing the
2339 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2340 work if there are any delivery delays because "one_time" forces the
2341 parent to be marked "delivered", so its unseen clone is never tried
2342 again. For this reason, Exim now forbids the simultaneous setting of
2345 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2346 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2347 entry if you want to know the details. The fix, however, neglected to
2348 consider the case when local delivery batching is involved. The test for
2349 "previously delivered" was not happening when checking to see if an
2350 address could be batched with a previous (undelivered) one; under
2351 certain circumstances this could lead to multiple deliveries to the same
2354 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2355 in its include files, and this causes problems building Exim.
2357 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2358 header_syntax) but Exim was just ignoring anything given after a slash.
2359 In particular, this caused confusion with an attempt to use "verify =
2360 reverse_host_lookup/defer_ok". An error is now given when options are
2361 supplied for verify items that do not have them. (Maybe reverse_host_
2362 lookup should have a defer_ok option, but that's a different point.)
2364 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2365 defined by RFC 821) to 2048, because there were problems with some AUTH
2366 commands, and RFC 1869 says the size should be increased for extended
2367 SMTP commands that take arguments.
2369 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2372 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2373 "unknown" error; now it says that the functionality isn't in the binary.
2375 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2376 an address' error message when a string expansion fails (syntax or
2377 whatever). Otherwise the password may appear in the log. Following change
2378 PH/42 below, there is no longer a chance of it appearing in a bounce
2381 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2383 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2384 of its addresses was ever tried. (Bugzilla bug #2.)
2386 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2387 the result incorrectly in the debug output. (It correctly added a newline
2388 to what was transported.)
2390 TF/01 Added $received_time.
2392 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2393 commented out examples of how to interface to a virus scanner and to
2394 SpamAssassin. Also added commented examples of av_scanner and
2395 spamd_address settings.
2397 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2398 and controls are allowed in which ACLs. There were a couple of minor
2399 errors. Some of the entries in the conditions table (which is a table of
2400 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2401 negation of where the condition IS allowed.
2403 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2405 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2406 header file does not have a version number, so I've had to invent a new
2407 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2408 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2409 radiusclient), but it was contributed by a Radius user.
2411 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2412 files or directories, for OpenSSL.
2414 PH/19 When an Exim process that is running as root has to create an Exim log
2415 file, it does so in a subprocess that runs as exim:exim so as to get the
2416 ownership right at creation (otherwise, other Exim processes might see
2417 the file with the wrong ownership). There was no test for failure of this
2418 fork() call, which would lead to the process getting stuck as it waited
2419 for a non-existent subprocess. Forks do occasionally fail when resources
2420 run out. I reviewed all the other calls to fork(); they all seem to check
2423 PH/20 When checking for unexpected SMTP input at connect time (before writing
2424 the banner), Exim was not dealing correctly with a non-positive return
2425 from the read() function. If the client had disconnected by this time,
2426 the result was a log entry for a synchronization error with an empty
2427 string after "input=" when read() returned zero. If read() returned -1
2428 (an event I could not check), uninitialized data bytes were printed.
2429 There were reports of junk text (parts of files, etc) appearing after
2432 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2434 PH/22 Added support for macro redefinition, and (re)definition in between
2435 driver and ACL definitions.
2437 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2438 forgetting to use the resulting value; it was using the unexpanded value.
2440 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2441 hadn't been configured. The fix is from Juergen Kreileder, who
2442 understands it better than I do:
2444 "Here's what I see happening with three configured cyrus_sasl
2445 authenticators configured (plain, login, cram-md5):
2447 On startup auth_cyrus_sasl_init() gets called for each of these.
2448 This means three calls to sasl_listmech() without a specified mech_list.
2449 => SASL tests which mechs of all available mechs actually work
2450 => three warnings about OTP not working
2451 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2453 With the patch, sasl_listmech() also gets called three times. But now
2454 SASL's mech_list option is set to the server_mech specified in the the
2455 authenticator. Or in other words, the answer from sasl_listmech()
2456 gets limited to just the mech you're testing for (which is different
2458 => the return list contains just 'plain' or 'login', 'cram-md5' or
2459 nothing depending on the value of ob->server_mech.
2461 I've just tested the patch: Authentication still works fine,
2462 unavailable mechs specified in the exim configuration are still
2463 caught, and the auth.log warnings about OTP are gone."
2465 PH/25 When debugging is enabled, the contents of the command line are added
2466 to the debugging output, even when log_selector=+arguments is not
2469 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2470 answer is "GNU", and only if the return is "GNU/something" is the answer
2473 PH/27 $acl_verify_message is now set immediately after the failure of a
2474 verification in an ACL, and so is available in subsequent modifiers. In
2475 particular, the message can be preserved by coding like this:
2477 warn !verify = sender
2478 set acl_m0 = $acl_verify_message
2480 Previously, $acl_verify_message was set only while expanding "message"
2481 and "log_message" when a very denied access.
2483 PH/28 Modified OS/os.c-Linux with
2485 -#ifndef OS_LOAD_AVERAGE
2486 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2488 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2489 nomenclature these days.)
2491 PH/29 Installed patch from the Sieve maintainer that adds the options
2492 sieve_useraddress and sieve_subaddress to the redirect router.
2494 PH/30 In these circumstances:
2495 . Two addresses routed to the same list of hosts;
2496 . First host does not offer TLS;
2497 . First host accepts first address;
2498 . First host gives temporary error to second address;
2499 . Second host offers TLS and a TLS session is established;
2500 . Second host accepts second address.
2501 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2502 and peerdn, if requested) that were in fact used only for the second
2505 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2506 attention to any local part prefix or suffix that was matched by the
2507 router that accepted the address. It now behaves in the same way as it
2508 does for delivery: the affixes are removed from the local part unless
2509 rcpt_include_affixes is set on the transport.
2511 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2512 timeout during the DATA phase of an incoming message.
2514 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2515 applied a patch sanctioned by the Sieve maintainer.
2517 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2518 the uid or gid is negative. A case of a negative gid caused this to be
2519 noticed. The fix allows for either to be negative.
2521 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2522 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2525 PH/36 The change PH/12 above was broken. Fixed it.
2527 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2528 the grounds that routing the same address twice would always produce the
2529 same answer. This might have been true once, but it is certainly no
2530 longer true now. Routing a child address may depend on the previous
2531 routing that produced that child. Some complicated redirection strategies
2532 went wrong when messages had multiple recipients, and made Exim's
2533 behaviour dependent on the order in which the addresses were given.
2535 I have moved the duplicate checking until after the routing is complete.
2536 Exim scans the addresses that are assigned to local and remote
2537 transports, and removes any duplicates. This means that more work will be
2538 done, as duplicates will always all be routed, but duplicates are
2539 presumably rare, so I don't expect this is of any significance.
2541 For deliveries to pipes, files, and autoreplies, the duplicate checking
2542 still happens during the routing process, since they are not going to be
2545 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2546 It corrects a timeout issue with spamd. This is Ian's comment: "The
2547 background is that sometimes spamd either never reads data from a
2548 connection it has accepted, or it never writes response data. The exiscan
2549 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2550 blindly assumes that writes won't block so it may never time out."
2552 PH/39 Allow G after quota size as well as K and M.
2554 PH/40 The value set for $authenticated_id in an authenticator may not contain
2555 binary zeroes or newlines because the value is written to log lines and
2556 to spool files. There was no check on this. Now the value is run through
2557 the string_printing() function so that such characters are converted to
2558 printable escape sequences.
2560 PH/41 $message_linecount is a new variable that contains the total number of
2561 lines in the message. Compare $body_linecount, which is the count for the
2564 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2565 bounce and delay warning messages, except in certain special cases, which
2568 (a) An SMTP error message from a remote host;
2569 (b) A message specified in a :fail: redirection;
2570 (c) A message specified in a "fail" command in a system filter;
2571 (d) A message specified in a FAIL return from the queryprogram router;
2572 (e) A message specified by the cannot_route_message router option.
2574 In these cases only, Exim does include the error details in bounce and
2575 warning messages. There are also a few cases where bland messages such
2576 as "unrouteable address" or "local delivery error" are given.
2578 PH/43 $value is now also set for the "else" part of a ${run expansion.
2580 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2581 being worked on, but at least Exim now implements the latest version to
2584 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2585 process to complete was treated as a delivery failure, a timeout while
2586 writing the message to the pipe was logged, but erroneously treated as a
2587 successful delivery. Such timeouts include transport filter timeouts. For
2588 consistency with the overall process timeout, these timeouts are now
2589 treated as errors, giving rise to delivery failures by default. However,
2590 there is now a new Boolean option for the pipe transport called
2591 timeout_defer, which, if set TRUE, converts the failures into defers for
2592 both kinds of timeout. A transport filter timeout is now identified in
2595 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2596 systems where "make" and "gmake" are different, calling "gmake" at top
2597 level broke things. I've arranged for the value of $(MAKE) to be passed
2598 from the Makefile to this script so that it can call the same version of
2602 A note about Exim versions 4.44 and 4.50
2603 ----------------------------------------
2605 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2606 changes of various kinds. As a consequence, a big documentation update was
2607 needed. This delayed the release for rather longer than seemed good, especially
2608 in the light of a couple of (minor) security issues. Therefore, the changes
2609 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2610 release. So 4.44 and 4.50 are in effect two different branches that both start
2613 I have left the 4.50 change log unchanged; it contains all the changes since
2614 4.43. The change log for 4.44 is below; many of its items are identical to
2615 those for 4.50. This seems to be the most sensible way to preserve the
2616 historical information.
2622 1. Minor wording change to the doc/README.SIEVE file.
2624 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2625 computation of the current number of files was incorrect.
2627 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2628 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2631 4. Give more explanation in the error message when the command for a transport
2632 filter fails to execute.
2634 5. There are several places where Exim runs a non-Exim command in a
2635 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2636 was being done only for the command run by the queryprogram router. It is
2637 now done for all such subprocesses. The other cases are: ${run, transport
2638 filters, and the commands run by the lmtp and pipe transports.
2640 6. Added CONFIGURE_GROUP build-time option.
2642 7. Some older OS have a limit of 256 on the maximum number of file
2643 descriptors. Exim was using setrlimit() to set 1000 as a large value
2644 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2645 systems. I've change it so that if it can't get 1000, it tries for 256.
2647 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2648 was an oversight, and furthermore, ever since the addition of extra
2649 controls (e.g. 4.43/32), the checks on when to allow different forms of
2650 "control" were broken. There should now be diagnostics for all cases when a
2651 control that does not make sense is encountered.
2653 9. Added the /retain_sender option to "control=submission".
2655 10. $recipients is now available in the predata ACL (oversight).
2657 11. Tidy the search cache before the fork to do a delivery from a message
2658 received from the command line. Otherwise the child will trigger a lookup
2659 failure and thereby defer the delivery if it tries to use (for example) a
2660 cached ldap connection that the parent has called unbind on.
2662 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2663 of $address_data from the recipient verification was clobbered by the
2664 sender verification.
2666 13. The value of address_data from a sender verification is now available in
2667 $sender_address_data in subsequent conditions in the ACL statement.
2669 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2671 15. Added a new option "connect=<time>" to callout options, to set a different
2674 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2675 was its contents. (It was OK if the option was not defined at all.)
2677 17. A "Completed" log line is now written for messages that are removed from
2678 the spool by the -Mrm option.
2680 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2681 information about exactly what failed.
2683 19. Added -dd to debug only the daemon process.
2685 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2686 handles timeouts, both on the server side and network timeouts. Renamed the
2687 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2689 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2690 It is now set to "smtps".
2692 22. $host_address is now set to the target address during the checking of
2693 ignore_target_hosts.
2695 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2696 being passed; this would have caused $sender_host_name to have been used if
2697 matching the list had actually called for a host name (not very likely,
2698 since this list is usually IP addresses). A host name is now passed as
2701 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2702 code that specifies a non-restarting handler (typically sigaction() in
2703 modern systems) in an attempt to fix a rare and obscure crash bug.
2705 25. Narrowed the window for a race in the daemon that could cause it to ignore
2706 SIGCHLD signals. This is not a major problem, because they are used only to
2707 wake it up if nothing else does.
2709 26. A malformed maildirsize file could cause Exim to calculate negative values
2710 for the mailbox size or file count. Odd effects could occur as a result.
2711 The maildirsize information is now recalculated if the size or filecount
2714 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2715 support for a long time. Removed HAVE_SYS_VFS_H.
2717 28. Installed the latest version of exipick from John Jetmore.
2719 29. In an address list, if the pattern was not a regular expression, an empty
2720 subject address (from a bounce message) matched only if the pattern was an
2721 empty string. Non-empty patterns were not even tested. This was the wrong
2722 because it is perfectly reasonable to use an empty address as part of a
2723 database query. An empty address is now tested by patterns that are
2724 lookups. However, all the other forms of pattern expect the subject to
2725 contain a local part and a domain, and therefore, for them, an empty
2726 address still always fails if the pattern is not itself empty.
2728 30. Exim went into a mad DNS loop when attempting to do a callout where the
2729 host was specified on an smtp transport, and looking it up yielded more
2730 than one IP address.
2732 31. Re-factored the code for checking spool and log partition space into a
2733 function that finds that data and another that does the check. The former
2734 is then used to implement four new variables: $spool_space, $log_space,
2735 $spool_inodes, and $log_inodes.
2737 32. The RFC2047 encoding function was originally intended for short strings
2738 such as real names; it was not keeping to the 75-character limit for
2739 encoded words that the RFC imposes. It now respects the limit, and
2740 generates multiple encoded words if necessary. To be on the safe side, I
2741 have increased the buffer size for the ${rfc2047: expansion operator from
2744 33. It is now permitted to omit both strings after an "if" condition; if the
2745 condition is true, the result is "true". As before, when the second string
2746 is omitted, a false condition yields an empty string. This makes it less
2747 cumbersome to write custom ACL and router conditions.
2749 34. Failure to deliver a bounce message always caused it to be frozen, even if
2750 there was an errors_to setting on the router. The errors_to setting is now
2753 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2754 canonical form (fully expanded) before being placed in
2755 $sender_host_address.
2757 36. The table in the code that translates DNS record types into text (T_A to
2758 "A" for instance) was missing entries for NS and CNAME. It is just possible
2759 that this could have caused confusion if both these types were looked up
2760 for the same domain, because the text type is used as part of Exim's
2761 per-process caching. But the chance of anyone hitting this buglet seems
2764 37. The dnsdb lookup has been extended in a number of ways.
2766 (1) There is a new type, "zns", which walks up the domain tree until it
2767 finds some nameserver records. It should be used with care.
2769 (2) There is a new type, "mxh", which is like "mx" except that it returns
2770 just the host names, not the priorities.
2772 (3) It is now possible to give a list of domains (or IP addresses) to be
2773 looked up. The behaviour when one of the lookups defers can be
2774 controlled by a keyword.
2776 (4) It is now possible to specify the separator character for use when
2777 multiple records are returned.
2779 38. The dnslists ACL condition has been extended: it is now possible to supply
2780 a list of IP addresses and/or domains to be looked up in a particular DNS
2783 39. Added log_selector=+queue_time_overall.
2785 40. When running the queue in the test harness, wait just a tad after forking a
2786 delivery process, to get repeatability of debugging output.
2788 41. Include certificate and key file names in error message when GnuTLS fails
2789 to set them up, because the GnuTLS error message doesn't include the name
2790 of the failing file when there is a problem reading it.
2792 42. Allow both -bf and -bF in the same test run.
2794 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2796 44. The "Exiscan patch" is now merged into the mainline Exim source.
2798 45. Sometimes the final signoff response after QUIT could fail to get
2799 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2800 < 0 before doing a fflush(). This bug looks as though it goes back to the
2801 introduction of TLS in release 3.20, but "sometimes" must have been rare
2802 because the tests only now provoked it.
2804 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2805 (this can affect the format of dates).
2807 47. exim_tidydb, when checking for the continued existence of a message for
2808 which it has found a message-specific retry record, was not finding
2809 messages that were in split spool directories. Consequently, it was
2810 deleting retry records that should have stayed in existence.
2812 48. Steve fixed some bugs in eximstats.
2814 49. The SPA authentication driver was not abandoning authentication and moving
2815 on to the next authenticator when an expansion was forced to fail,
2816 contradicting the general specification for all authenticators. Instead it
2817 was generating a temporary error. It now behaves as specified.
2819 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2820 (the order specifies the preference for clients). The order is now AES256,
2821 AES128, 3DES, ARCFOUR128.
2823 51. Small patch to Sieve code - explicitly set From: when generating an
2826 52. Exim crashed if a remote delivery caused a very long error message to be
2827 recorded - for instance if somebody sent an entire SpamAssassin report back
2828 as a large number of 550 error lines. This bug was coincidentally fixed by
2829 increasing the size of one of Exim's internal buffers (big_buffer) that
2830 happened as part of the Exiscan merge. However, to be on the safe side, I
2831 have made the code more robust (and fixed the comments that describe what
2834 53. Now that there can be additional text after "Completed" in log lines (if
2835 the queue_time_overall log selector is set), a one-byte patch to exigrep
2836 was needed to allow it to recognize "Completed" as not the last thing in
2839 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2840 patch that reportedly fixes this has been added. I am not expert enough to
2841 create a test for it. This is what the patch creator wrote:
2843 "I found a little strange behaviour of ldap code when working with
2844 Windows 2003 AD Domain, where users was placed in more than one
2845 Organization Units. When I tried to give exim partial DN, the exit code
2846 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2847 But simultaneously result of request was absolutely normal ldap result,
2848 so I produce this patch..."
2850 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2851 REFERENCE, so I have modified the code to exclude the patch when that macro
2854 55. Some experimental protocols are using DNS PTR records for new purposes. The
2855 keys for these records are domain names, not reversed IP addresses. The
2856 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2857 leaves it alone. Component reversal etc. now happens only for IP addresses.
2860 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2862 57. Double the size of the debug message buffer (to 2048) so that more of very
2863 long debug lines gets shown.
2865 58. The exicyclog utility now does better if the number of log files to keep
2866 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2868 59. Two changes related to the smtp_active_hostname option:
2870 (1) $smtp_active_hostname is now available as a variable.
2871 (2) The default for smtp_banner uses $smtp_active_hostname instead
2872 of $primary_hostname.
2874 60. The host_aton() function is supposed to be passed a string that is known
2875 to be a valid IP address. However, in the case of IPv6 addresses, it was
2876 not checking this. This is a hostage to fortune. Exim now panics and dies
2877 if the condition is not met. A case was found where this could be provoked
2878 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2879 components; fortuitously, this particular loophole had already been fixed
2880 by change 4.50/55 above.
2882 If there are any other similar loopholes, the new check in host_aton()
2883 itself should stop them being exploited. The report I received stated that
2884 data on the command line could provoke the exploit when Exim was running as
2885 exim, but did not say which command line option was involved. All I could
2886 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2887 running as the user.
2890 61. There was a buffer overflow vulnerability in the SPA authentication code
2891 (which came originally from the Samba project). I have added a test to the
2892 spa_base64_to_bits() function which I hope fixes it.
2895 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2896 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2898 63. The daemon start-up calls getloadavg() while still root for those OS that
2899 need the first call to be done as root, but it missed one case: when
2900 deliver_queue_load_max is set with deliver_drop_privilege. This is
2901 necessary for the benefit of the queue runner, because there is no re-exec
2902 when deliver_drop_privilege is set.
2904 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2905 This has been fixed.
2907 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2908 was in use, was not putting the data itself into the right store pool;
2909 consequently, it could be overwritten for a subsequent message in the same
2910 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2913 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2915 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2916 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2917 the function were treating the return as a boolean value, which happened to
2918 work because 0=false and not-0=true, but is not correct code.
2920 68. The host_aton() function was not handling scoped IPv6 addresses (those
2921 with, for example, "%eth0" on the end) correctly.
2923 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2924 negated items (that is, ~something) in unsigned ints. Some compilers
2925 apparently mutter when there is no cast.
2927 70. If an address verification called from an ACL failed, and did not produce a
2928 user-specific message (i.e. there was only a "system" message), nothing was
2929 put in $acl_verify_message. In this situation, it now puts the system
2932 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2933 session; change 4.31/43 added the unwanted input to the log line - except
2934 that it did not do this in the start of session case. It now does.
2936 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2937 This is wrong and can cause the other end to generate a synchronization
2938 error if it is another Exim or anything else that does the synchronization
2939 check. A QUIT command is no longer sent after a timeout.
2941 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2942 during host lookups.
2944 74. The defer_ok option of callout verification was not working if it was used
2945 when verifying addresses in header lines, that is, for this case:
2947 verify = header_sender/callout=defer_ok
2949 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2950 those file descriptors could be used for SMTP connections. If anything
2951 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2952 could be sent to the SMTP client, causing chaos. The daemon now opens
2953 stdin, stdout, and stderr to /dev/null when it puts itself into the
2956 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2957 log by default. The user can override this with suitable Perl magic.
2959 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2960 the log message when discard triggers, was not working for the DATA ACL or
2961 for the non-SMTP ACL.
2963 78. Error message wording change in sieve.c.
2965 79. If smtp_accept_max_per_host was set, the number of connections could be
2966 restricted to fewer than expected, because the daemon was trying to set up
2967 a new connection before checking whether the processes handling previous
2968 connections had finished. The check for completed processes is now done
2969 earlier. On busy systems, this bug wouldn't be noticed because something
2970 else would have woken the daemon, and it would have reaped the completed
2973 80. If a message was submitted locally by a user whose login name contained one
2974 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2975 It caused a spool format error. I have fixed the spool reading code. A
2976 related problem was that the "from" clause in the Received: line became
2977 illegal because of the space(s). It is now covered by ${quote_local_part.
2979 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2982 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2984 83. Patch from Sieve maintainer to fix unterminated string problem in
2985 "vacation" handling.
2987 84. Some minor changes to the Linux configuration files to help with other
2988 OS variants using glibc.
2990 85. One more patch for Sieve to update vacation handling to latest spec.
2993 ----------------------------------------------------
2994 See the note above about the 4.44 and 4.50 releases.
2995 ----------------------------------------------------
3001 1. Change 4.43/35 introduced a bug that caused file counts to be
3002 incorrectly computed when quota_filecount was set in an appendfile
3005 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3006 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3009 3. Give more explanation in the error message when the command for a transport
3010 filter fails to execute.
3012 4. There are several places where Exim runs a non-Exim command in a
3013 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3014 was being done only for the command run by the queryprogram router. It is
3015 now done for all such subprocesses. The other cases are: ${run, transport
3016 filters, and the commands run by the lmtp and pipe transports.
3018 5. Some older OS have a limit of 256 on the maximum number of file
3019 descriptors. Exim was using setrlimit() to set 1000 as a large value
3020 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3021 systems. I've change it so that if it can't get 1000, it tries for 256.
3023 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3024 was an oversight, and furthermore, ever since the addition of extra
3025 controls (e.g. 4.43/32), the checks on when to allow different forms of
3026 "control" were broken. There should now be diagnostics for all cases when a
3027 control that does not make sense is encountered.
3029 7. $recipients is now available in the predata ACL (oversight).
3031 8. Tidy the search cache before the fork to do a delivery from a message
3032 received from the command line. Otherwise the child will trigger a lookup
3033 failure and thereby defer the delivery if it tries to use (for example) a
3034 cached ldap connection that the parent has called unbind on.
3036 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3037 of $address_data from the recipient verification was clobbered by the
3038 sender verification.
3040 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3041 was its contents. (It was OK if the option was not defined at all.)
3043 11. A "Completed" log line is now written for messages that are removed from
3044 the spool by the -Mrm option.
3046 12. $host_address is now set to the target address during the checking of
3047 ignore_target_hosts.
3049 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3050 being passed; this would have caused $sender_host_name to have been used if
3051 matching the list had actually called for a host name (not very likely,
3052 since this list is usually IP addresses). A host name is now passed as
3055 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3056 code that specifies a non-restarting handler (typically sigaction() in
3057 modern systems) in an attempt to fix a rare and obscure crash bug.
3059 15. Narrowed the window for a race in the daemon that could cause it to ignore
3060 SIGCHLD signals. This is not a major problem, because they are used only to
3061 wake it up if nothing else does.
3063 16. A malformed maildirsize file could cause Exim to calculate negative values
3064 for the mailbox size or file count. Odd effects could occur as a result.
3065 The maildirsize information is now recalculated if the size or filecount
3068 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3069 support for a long time. Removed HAVE_SYS_VFS_H.
3071 18. Updated exipick to current release from John Jetmore.
3073 19. Allow an empty sender to be matched against a lookup in an address list.
3074 Previously the only cases considered were a regular expression, or an
3077 20. Exim went into a mad DNS lookup loop when doing a callout where the
3078 host was specified on the transport, if the DNS lookup yielded more than
3081 21. The RFC2047 encoding function was originally intended for short strings
3082 such as real names; it was not keeping to the 75-character limit for
3083 encoded words that the RFC imposes. It now respects the limit, and
3084 generates multiple encoded words if necessary. To be on the safe side, I
3085 have increased the buffer size for the ${rfc2047: expansion operator from
3088 22. Failure to deliver a bounce message always caused it to be frozen, even if
3089 there was an errors_to setting on the router. The errors_to setting is now
3092 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3093 canonical form (fully expanded) before being placed in
3094 $sender_host_address.
3096 24. Updated eximstats to version 1.33
3098 25. Include certificate and key file names in error message when GnuTLS fails
3099 to set them up, because the GnuTLS error message doesn't include the name
3100 of the failing file when there is a problem reading it.
3102 26. Expand error message when OpenSSL has problems setting up cert/key files.
3105 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3106 (this can affect the format of dates).
3108 28. exim_tidydb, when checking for the continued existence of a message for
3109 which it has found a message-specific retry record, was not finding
3110 messages that were in split spool directories. Consequently, it was
3111 deleting retry records that should have stayed in existence.
3113 29. eximstats updated to version 1.35
3114 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3115 1.35 - bugfix such that pie charts by volume are generated correctly
3117 30. The SPA authentication driver was not abandoning authentication and moving
3118 on to the next authenticator when an expansion was forced to fail,
3119 contradicting the general specification for all authenticators. Instead it
3120 was generating a temporary error. It now behaves as specified.
3122 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3123 (the order specifies the preference for clients). The order is now AES256,
3124 AES128, 3DES, ARCFOUR128.
3126 31. Small patch to Sieve code - explicitly set From: when generating an
3129 32. Exim crashed if a remote delivery caused a very long error message to be
3130 recorded - for instance if somebody sent an entire SpamAssassin report back
3131 as a large number of 550 error lines. This bug was coincidentally fixed by
3132 increasing the size of one of Exim's internal buffers (big_buffer) that
3133 happened as part of the Exiscan merge. However, to be on the safe side, I
3134 have made the code more robust (and fixed the comments that describe what
3137 33. Some experimental protocols are using DNS PTR records for new purposes. The
3138 keys for these records are domain names, not reversed IP addresses. The
3139 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3140 leaves it alone. Component reversal etc. now happens only for IP addresses.
3143 34. The host_aton() function is supposed to be passed a string that is known
3144 to be a valid IP address. However, in the case of IPv6 addresses, it was
3145 not checking this. This is a hostage to fortune. Exim now panics and dies
3146 if the condition is not met. A case was found where this could be provoked
3147 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3148 components; fortuitously, this particular loophole had already been fixed
3149 by change 4.50/55 or 4.44/33 above.
3151 If there are any other similar loopholes, the new check in host_aton()
3152 itself should stop them being exploited. The report I received stated that
3153 data on the command line could provoke the exploit when Exim was running as
3154 exim, but did not say which command line option was involved. All I could
3155 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3156 running as the user.
3159 35. There was a buffer overflow vulnerability in the SPA authentication code
3160 (which came originally from the Samba project). I have added a test to the
3161 spa_base64_to_bits() function which I hope fixes it.
3164 36. The daemon start-up calls getloadavg() while still root for those OS that
3165 need the first call to be done as root, but it missed one case: when
3166 deliver_queue_load_max is set with deliver_drop_privilege. This is
3167 necessary for the benefit of the queue runner, because there is no re-exec
3168 when deliver_drop_privilege is set.
3170 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3171 was in use, was not putting the data itself into the right store pool;
3172 consequently, it could be overwritten for a subsequent message in the same
3173 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3176 38. Sometimes the final signoff response after QUIT could fail to get
3177 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3178 < 0 before doing a fflush(). This bug looks as though it goes back to the
3179 introduction of TLS in release 3.20, but "sometimes" must have been rare
3180 because the tests only now provoked it.
3186 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3187 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3188 Now it returns an integer. A number of calls were still expecting a T/F
3189 return. Fortuitously, in all cases, the tests worked in OK situations,
3190 which is the norm. However, things would have gone wrong on any write
3191 failures on the smtp file descriptor. This function is used when sending
3192 messages over SMTP and also when doing verify callouts.
3194 2. When Exim is called to do synchronous delivery of a locally submitted
3195 message (the -odf or -odi options), it no longer closes stderr before doing
3198 3. Implemented the mua_wrapper option.
3200 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3202 5. Implemented the functions header_remove(), header_testname(),
3203 header_add_at_position(), and receive_remove_recipient(), and exported them
3206 6. If an ACL "warn" statement specified the addition of headers, Exim already
3207 inserted X-ACL-Warn: at the start if there was no header name. However, it
3208 was not making this test for the second and subsequent header lines if
3209 there were newlines in the string. This meant that an invalid header could
3210 be inserted if Exim was badly configured.
3212 7. Allow an ACL "warn" statement to add header lines at the start or after all
3213 the Received: headers, as well as at the end.
3215 8. Added the rcpt_4xx retry error code.
3217 9. Added postmaster_mailfrom=xxx to callout verification option.
3219 10. Added mailfrom=xxxx to the callout verification option, for verify=
3222 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3223 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3225 12. Inserted some casts to stop certain compilers warning when using pointer
3226 differences as field lengths or precisions in printf-type calls (mostly
3227 affecting debugging statements).
3229 13. Added optional readline() support for -be (dynamically loaded).
3231 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3232 same clock tick as a message's arrival, so that its received time was the
3233 same as the "first fail" time on the retry record, and that message
3234 remained on the queue past the ultimate address timeout, every queue runner
3235 would try a delivery (because it was past the ultimate address timeout) but
3236 after another failure, the ultimate address timeout, which should have then
3237 bounced the address, did not kick in. This was a "< instead of <=" error;
3238 in most cases the first failure would have been in the next clock tick
3239 after the received time, and all would be well.
3241 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3242 being recognized when the domain list was tested by the match_domain
3243 condition in an expansion string.
3245 16. Added the ${str2b64: operator.
3247 17. Exim was always calling setrlimit() to set a large limit for the number of
3248 processes, without checking whether the existing limit was already
3249 adequate. (It did check for the limit on file descriptors.) Furthermore,
3250 errors from getrlimit() and setrlimit() were being ignored. Now they are
3251 logged to the main and panic logs, but Exim does carry on, to try to do its
3252 job under whatever limits there are.
3254 18. Imported PCRE 5.0.
3256 19. Trivial typo in log message " temporarily refused connection" (the leading
3259 20. If the log selector return_path_on_delivery was set and an address was
3260 redirected to /dev/null, the delivery process crashed because it assumed
3261 that a return path would always be set for a "successful" delivery. In this
3262 case, the whole delivery is bypassed as an optimization, and therefore no
3265 21. Internal re-arrangement: the function for sending a challenge and reading
3266 a response while authentication was assuming a zero-terminated challenge
3267 string. It's now changed to take a pointer and a length, to allow for
3268 binary data in such strings.
3270 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3272 23. Exim was not respecting finduser_retries when seeking the login of the
3273 uid under which it was called; it was always trying 10 times. (The default
3274 setting of finduser_retries is zero.) Also, it was sleeping after the final
3275 failure, which is pointless.
3277 24. Implemented tls_on_connect_ports.
3279 25. Implemented acl_smtp_predata.
3281 26. If the domain in control=submission is set empty, Exim assumes that the
3282 authenticated id is a complete email address when it generates From: or
3283 Sender: header lines.
3285 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3286 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3287 chown and chgrp in /bin and hostname in /usr/bin.
3289 28. Exim was keeping the "process log" file open after each use, just as it
3290 does for the main log. This opens the possibility of it remaining open for
3291 long periods when the USR1 signal hits a daemon. Occasional processlog
3292 errors were reported, that could have been caused by this. Anyway, it seems
3293 much more sensible not to leave this file open at all, so that is what now
3296 29. The long-running daemon process does not normally write to the log once it
3297 has entered its main loop, and it closes the log before doing so. This is
3298 so that log files can straightforwardly be renamed and moved. However,
3299 there are a couple of unusual error situations where the daemon does write
3300 log entries, and I had neglected to close the log afterwards.
3302 30. The text of an SMTP error response that was received during a remote
3303 delivery was being truncated at 512 bytes. This is too short for some of
3304 the long messages that one sometimes sees. I've increased the limit to
3307 31. It is now possible to make retry rules that apply only when a message has a
3308 specific sender, in particular, an empty sender.
3310 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3311 it possible to be selective about when SMTP synchronization is enforced.
3313 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3315 32. Implemented hosts_connection_nolog.
3317 33. Added an ACL for QUIT.
3319 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3322 35. Added mailbox_size and mailbox_filecount to appendfile.
3324 36. Added control = no_multiline_responses to ACLs.
3326 37. There was a bug in the logic of the code that waits for the clock to tick
3327 in the case where the clock went backwards by a substantial amount such
3328 that the microsecond fraction of "now" was more than the microsecond
3329 fraction of "then" (but the whole seconds number was less).
3331 38. Added support for the libradius Radius client library this is found on
3332 FreeBSD (previously only the radiusclient library was supported).
3338 1. When certain lookups returned multiple values in the form name=value, the
3339 quoting of the values was not always being done properly. Specifically:
3340 (a) If the value started with a double quote, but contained no whitespace,
3342 (b) If the value contained whitespace other than a space character (i.e.
3343 tabs or newlines or carriage returns) it was not quoted.
3344 This fix has been applied to the mysql and pgsql lookups by writing a
3345 separate quoting function and calling it from the lookup code. The fix
3346 should probably also be applied to nisplus, ibase and oracle lookups, but
3347 since I cannot test any of those, I have not disturbed their existing code.
3349 2. A hit in the callout cache for a specific address caused a log line with no
3350 reason for rejecting RCPT. Now it says "Previous (cached) callout
3351 verification failure".
3353 3. There was an off-by-one bug in the queryprogram router. An over-long
3354 return line was truncated at 256 instead of 255 characters, thereby
3355 overflowing its buffer with the terminating zero. As well as fixing this, I
3356 have increased the buffer size to 1024 (and made a note to document this).
3358 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3359 when Exim is waiting for an SMTP response from a remote server, Exim
3360 restarts its select() call on the socket, thereby resetting its timeout.
3361 This is not a problem when such interrupts are rare. Somebody set up a cron
3362 job to run exiwhat every 2 minutes, which is less than the normal select()
3363 timeout (5 or 10 minutes). This meant that the select() timeout never
3364 kicked in because it was always reset. I have fixed this by comparing the
3365 time when an interrupt arrives with the time at the start of the first call
3366 to select(). If more time than the timeout has elapsed, the interrupt is
3367 treated as a timeout.
3369 5. Some internal re-factoring in preparation for the addition of Sieve
3370 extensions (by MH). In particular, the "personal" test is moved to a
3371 separate function, and given an option for scanning Cc: and Bcc: (which is
3372 not set for Exim filters).
3374 6. When Exim created an email address using the login of the caller as the
3375 local part (e.g. when creating a From: or Sender: header line), it was not
3376 quoting the local part when it contained special characters such as @.
3378 7. Installed new OpenBSD configuration files.
3380 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3381 try to make them clearer.
3383 9. Callout options, other than the timeout value, were being ignored when
3384 verifying sender addresses in header lines. For example, when using
3386 verify = header_sender/callout=no_cache
3388 the cache was (incorrectly) being used.
3390 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3391 only the Cygwin environment.
3393 11. When return_path_on_delivery was set as a log selector, if different remote
3394 addresses in the same message used different return paths and parallel
3395 remote delivery occurred, the wrong values would sometimes be logged.
3396 (Whenever a remote delivery process finished, the return path value from
3397 the most recently started remote delivery process was logged.)
3399 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3400 lines when AUTH and/or TLS are in use. This is the "received protocol"
3401 field. Exim used to use "asmtp" for authenticated SMTP, without any
3402 indication (in the protocol name) for TLS use. Now it follows the RFC and
3403 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3404 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3405 names appear in log lines as well as in Received: header lines.
3407 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3408 extensions, and comparison tests, and to fix some bugs.
3410 14. Changes to the "personal" filter test:
3412 (1) The test was buggy in that it was just doing the equivalent of
3413 "contains" tests on header lines. For example, if a user's address was
3414 anne@some.where, the "personal" test would incorrectly be true for
3416 To: susanne@some.where
3418 This test is now done by extracting each address from the header in turn,
3419 and checking the entire address. Other tests that are part of "personal"
3420 are now done using regular expressions (for example, to check local parts
3421 of addresses in From: header lines).
3423 (2) The list of non-personal local parts in From: addresses has been
3424 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3425 taken from the Sieve specification recommendations.
3427 (3) If the message contains any header line starting with "List-" it is
3428 treated as non-personal.
3430 (4) The test for "circular" in the Subject: header line has been removed
3431 because it now seems ill-conceived.
3433 15. Minor typos in src/EDITME comments corrected.
3435 16. Installed latest exipick from John Jetmore.
3437 17. If headers_add on a router specified a text string that was too long for
3438 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3439 of string_sprintf() is now avoided.
3441 18. $message_body_size was not set (it was always zero) when running the DATA
3442 ACL and the local_scan() function.
3444 19. For the "mail" command in an Exim filter, no default was being set for
3445 the once_repeat time, causing a random time value to be used if "once" was
3446 specified. (If the value happened to be <= 0, no repeat happened.) The
3447 default is now 0s, meaning "never repeat". The "vacation" command was OK
3448 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3449 (I found it when inspecting the code).
3451 20. There is now an overall timeout for performing a callout verification. It
3452 defaults to 4 times the callout timeout, which applies to individual SMTP
3453 commands during the callout. The overall timeout applies when there is more
3454 than one host that can be tried. The timeout is checked before trying the
3455 next host. This prevents very long delays if there are a large number of
3456 hosts and all are timing out (e.g. when the network connections are timing
3457 out). The value of the overall timeout can be changed by specifying an
3458 additional sub-option for "callout", called "maxwait". For example:
3460 verify = sender/callout=5s,maxwait=20s
3462 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3463 to the end before writing, but this should make it even safer).
3465 22. Exim was forgetting that it had advertised PIPELINING for the second and
3466 subsequent messages on an SMTP connection. It was also not resetting its
3467 memory on STARTTLS and an internal HELO.
3469 23. When Exim logs an SMTP synchronization error within a session, it now
3470 records whether PIPELINING has been advertised or not.
3472 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3473 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3474 rather than long int.
3476 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3478 26. Added the never_mail option to autoreply.
3484 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3485 crash if the getsockname() call failed; this can happen if a connection is
3486 closed very soon after it is established. The problem was simply in the
3487 order in which certain operations were done, causing Exim to try to write
3488 to the SMTP stream before it had set up the file descriptor. The bug has
3489 been fixed by making things happen in the correct order.
3495 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3496 before the connection was closed, thus losing the rejection response.
3498 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3499 some early Solaris releases, but causes trouble in current releases where
3500 socklen_t is defined.
3502 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3505 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3508 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3509 that releasing the top of it at the end releases what was used for sub-
3510 expansions (unless the block got too big). However, discard this block if
3511 the first thing is a variable or header, so that we can use its block when
3512 it is dynamic (useful for very large $message_headers, for example).
3514 6. Lookups now cache *every* query, not just the most recent. A new, separate
3515 store pool is used for this. It can be recovered when all lookup caches are
3516 flushed. Lookups now release memory at the end of their result strings.
3517 This has involved some general refactoring of the lookup sources.
3519 7. Some code has been added to the store_xxx() functions to reduce the amount
3520 of flapping under certain conditions.
3522 8. log_incoming_interface used to affect only the <= reception log lines. Now
3523 it causes the local interface and port to be added to several more SMTP log
3524 lines, for example "SMTP connection from", and rejection lines.
3526 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3528 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3530 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3531 could be overwritten at the end of the current message (or the start of a
3532 new message if it was set in a HELO ACL). The value is now preserved for
3533 the duration of the SMTP connection.
3535 12. If a transport had a headers_rewrite setting, and a matching header line
3536 contained an unqualified address, that address was qualified, even if it
3537 did not match any rewriting rules. The underlying bug was that the values
3538 of the flags that permit the existence of unqualified sender and recipient
3539 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3540 non-local messages, and by -bnq for local messages) were not being
3541 preserved with the message after it was received.
3543 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3544 "next input=" as part of the text comprising the host identity instead of
3545 the correct text. The code was using the same buffer for two different
3546 strings. However, depending on which order the printing function evaluated
3547 its arguments, the bug did not always show up. Under Linux, for example, my
3548 test suite worked just fine.
3550 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3551 doesn't work with some older versions of Perl. It has been changed to "my",
3552 which in any case is probably the better facility to use.
3554 15. A really picky compiler found some instances of statements for creating
3555 error messages that either had too many or two few arguments for the format
3558 16. The size of the buffer for calls to the DNS resolver has been increased
3559 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3560 for addresses that have a lot of PTR records. This alleviates a problem; it
3561 does not fully solve it.
3563 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3564 buffer now truncates the list and logs the incident, which is the same
3565 action as happens when Exim is looking up a host name and its aliases.
3566 Previously in this situation something unpredictable would happen;
3567 sometimes it was "internal error: store_reset failed".
3569 18. If a server dropped the connection unexpectedly when an Exim client was
3570 using GnuTLS and trying to read a response, the client delivery process
3571 crashed while trying to generate an error log message.
3573 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3574 string, for example:
3576 warn message = H1: something\nH2: something
3578 the text was added as a single header line from Exim's point of view
3579 though it ended up OK in the delivered message. However, searching for the
3580 second and subsequent header lines using $h_h2: did not work. This has been
3581 fixed. Similarly, if a system filter added multiple headers in this way,
3582 the routers could not see them.
3584 20. Expanded the error message when iplsearch is called with an invalid key to
3585 suggest using net-iplsearch in a host list.
3587 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3588 ACLs are no longer actually imposed (and a message to that effect is
3591 22. If a "gecos" field in a passwd entry contained escaped characters, in
3592 particular, if it contained a \" sequence, Exim got it wrong when building
3593 a From: or a Sender: header from that name. A second bug also caused
3594 incorrect handling when an unquoted " was present following a character
3595 that needed quoting.
3597 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3598 was not being matched caselessly.
3600 24. Arranged for all hyphens in the exim.8 source to be escaped with
3603 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3604 properly. Recipient callout cache records were still being keyed to include
3605 the sender, even when use_sender was set false. This led to far more
3606 callouts that were necessary. The sender is no longer included in the key
3607 when use_sender is false.
3609 26. Added "control = submission" modifier to ACLs.
3611 27. Added the ${base62d: operator to decode base 62 numbers.
3613 28. dnsdb lookups can now access SRV records.
3615 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3616 the configuration file.
3618 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3619 (-v) mode. This makes the output for a verbose queue run more intelligible.
3621 31. Added a use_postmaster feature to recipient callouts.
3623 32. Added the $body_zerocount variable, containing the number of binary zero
3624 bytes in the message body.
3626 33. The time of last modification of the "new" subdirectory is now used as the
3627 "mailbox time last read" when there is a quota error for a maildir
3630 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3632 35. Added +ignore_unknown as a special item in host lists.
3634 36. Code for decoding IPv6 addresses in host lists is now included, even if
3635 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3636 address was recognized as an IP address, but was then not correctly decoded
3637 into binary, causing unexpected and incorrect effects when compared with
3644 1. Very minor rewording of debugging text in manualroute to say "list of
3645 hosts" instead of "hostlist".
3647 2. If verify=header_syntax was set, and a header line with an unqualified
3648 address (no domain) and a large number of spaces between the end of the
3649 name and the colon was received, the reception process suffered a buffer
3650 overflow, and (when I tested it) crashed. This was caused by some obsolete
3651 code that should have been removed. The fix is to remove it!
3653 3. When running in the test harness, delay a bit after writing a bounce
3654 message to get a bit more predictability in the log output.
3656 4. Added a call to search_tidyup() just before forking a reception process. In
3657 theory, someone could use a lookup in the expansion of smtp_accept_max_
3658 per_host which, without the tidyup, could leave open a database connection.
3660 5. Added the variables $recipient_data and $sender_data which get set from a
3661 lookup success in an ACL "recipients" or "senders" condition, or a router
3662 "senders" option, similar to $domain_data and $local_part_data.
3664 6. Moved the writing of debug_print from before to after the "senders" test
3667 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3668 problems for message scanning, either using a data ACL, or using
3669 local_scan() because the Received: header was not generated till after they
3670 were called (in order to set the time as the time of reception completion).
3671 I have revised the way this works. The header is now generated after the
3672 body is received, but before the ACL or local_scan() are called. After they
3673 are run, the timestamp in the header is updated.
3679 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3680 before starting a queue runner without re-exec. This happened only when
3681 deliver_drop_privilege was set or when the Exim user was set to root. The
3682 effect of the bug was that timeouts during subsequent deliveries caused
3683 crashes instead of being properly handled. The handler is now left at its
3684 default (and expected) setting.
3686 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3687 message, again when deliver_drop_privilege is set or Exim is run as root.
3688 The bug described in (1) was not present in this case, but the tidying up
3689 of the other signals was missing. I have made the two cases consistent.
3691 3. The ignore_target_hosts setting on a manualroute router was being ignored
3692 for hosts that were looked up using the /MX notation.
3694 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3697 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3698 operated on the sender address. After changing the $sender_address to <>
3699 for the sender address verify, Exim was re-instated it as the original
3700 (before rewriting) address, but remembering that it had rewritten it, so it
3701 wasn't rewriting it again. This bug also had the effect of breaking the
3702 sender address verification caching when the sender address was rewritten.
3704 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3705 This has been changed so that if the ip literal address matches
3706 ignore_target_hosts, the router declines.
3708 7. Added expansion conditions match_domain, match_address, and match_local_
3709 part (NOT match_host).
3711 8. The placeholder for the Received: header didn't have a length field set.
3713 9. Added code to Exim itself and to exim_lock to test for a specific race
3714 condition that could lead to file corruption when using MBX delivery. The
3715 issue is with the lockfile that is created in /tmp. If this file is removed
3716 after a process has opened it but before that process has acquired a lock,
3717 there is the potential for a second process to recreate the file and also
3718 acquire a lock. This could lead to two Exim processes writing to the file
3719 at the same time. The added code performs the same test as UW imapd; it
3720 checks after acquiring the lock that its file descriptor still refers to
3721 the same named file.
3723 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3724 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3727 11. Added the smtp_active_hostname option. If used, this will typically be made
3728 to depend on the incoming interface address. Because $interface_address is
3729 not set up until the daemon has forked a reception process, error responses
3730 that can happen earlier (such as "too many connections") no longer contain
3733 12. If an expansion in a condition on a "warn" statement fails because a lookup
3734 defers, the "warn" statement is abandoned, and the next ACL statement is
3735 processed. Previously this caused the whole ACL to be aborted.
3737 13. Added the iplsearch lookup type.
3739 14. Added ident_timeout as a log selector.
3741 15. Added tls_certificate_verified as a log selector.
3743 16. Added a global option tls_require_ciphers (compare the smtp transport
3744 option of the same name). This controls incoming TLS connections.
3746 17. I finally figured out how to make tls_require_ciphers do a similar thing
3747 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3748 before starting the TLS session.
3750 18. Tabs are now shown as \t in -bP output.
3752 19. If the log selector return_path_on_delivery was set, Exim crashed when
3753 bouncing a message because it had too many Received: header lines.
3755 20. If two routers both had headers_remove settings, and the first one included
3756 a superfluous trailing colon, the final name in the first list and the
3757 first name in the second list were incorrectly joined into one item (with a
3758 colon in the middle).
3764 1. Added -C and -D options to the exinext utility, mainly to make it easier
3765 to include in the automated testing, but these could be helpful when
3766 multiple configurations are in use.
3768 2. The exinext utility was not formatting the output nicely when there was
3769 an alternate port involved in the retry record key, nor when there was a
3770 message id as well (for retries that were specific to a specific message
3771 and a specific host). It was also confused by IPv6 addresses, because of
3772 the additional colons they contain. I have fixed the IPv4 problem, and
3773 patched it up to do a reasonable job for IPv6.
3775 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3776 delivery, the log line now contains "pipelined" if PIPELINING was used.
3778 4. An SMTP transport process used to panic and die if the bind() call to set
3779 an explicit outgoing interface failed. This has been changed; it is now
3780 treated in the same way as a connect() failure.
3782 5. A reference to $sender_host_name in the part of a conditional expansion
3783 that was being skipped was still causing a DNS lookup. This no longer
3786 6. The def: expansion condition was not recognizing references to header lines
3787 that used bh_ and bheader_.
3789 7. Added the _cache feature to named lists.
3791 8. The code for checking quota_filecount in the appendfile transport was
3792 allowing one more file than it should have been.
3794 9. For compatibility with Sendmail, the command line option
3802 and sets the incoming protocol and host name (for trusted callers). The
3803 host name and its colon can be omitted when only the protocol is to be set.
3804 Note the Exim already has two private options, -pd and -ps, that refer to
3805 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3806 "s", but I don't think that's a major issue.
3808 10. A number of refactoring changes to the code, none of which should affect
3811 (a) The number of logging options was getting close to filling up the
3812 32-bit word that was used as a bit map. I have split them into two classes:
3813 those that are passed in the argument to log_write(), and those that are
3814 only ever tested independently outside of that function. These are now in
3815 separate 32-bit words, so there is plenty of room for expansion again.
3816 There is no change in the user interface or the logging behaviour.
3818 (b) When building, for example, log lines, the code previously used a
3819 macro that called string_cat() twice, in order to add two strings. This is
3820 not really sufficiently general. Furthermore, there was one instance where
3821 it was actually wrong because one of the argument was used twice, and in
3822 one call a function was used. (As it happened, calling the function twice
3823 did not affect the overall behaviour.) The macro has been replaced by a
3824 function that can join an arbitrary number of extra strings onto a growing
3827 (c) The code for expansion conditions now uses a table and a binary chop
3828 instead of a serial search (which was left over from when there were very
3829 few conditions). Also, it now recognizes conditions like "pam" even when
3830 the relevant support is not compiled in: a suitably worded error message is
3831 given if an attempt is made to use such a condition.
3833 11. Added ${time_interval:xxxxx}.
3835 12. A bug was causing one of the ddress fields not to be passed back correctly
3836 from remote delivery subprocesses. The field in question was not being
3837 subsequently used, so this caused to problems in practice.
3839 13. Added new log selectors queue_time and deliver_time.
3841 14. Might have fixed a bug in maildirsizefile handling that threw up
3842 "unexpected character" debug warnings, and recalculated the data
3843 unnecessarily. In any case, I expanded the warning message to give more
3846 15. Added the message "Restricted characters in address" to the statements in
3847 the default ACL that block characters like @ and % in local parts.
3849 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3850 Three changes have been made:
3852 (a) There was a serious bug; a negative response to MAIL caused the whole
3853 recipient domain to be cached as invalid, thereby blocking all messages
3854 to all local parts at the same domain, from all senders. This bug has
3855 been fixed. The domain is no longer cached after a negative response to
3856 MAIL if the sender used is not empty.
3858 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3861 (c) A new callout option, "use_sender" has been added for people who want
3862 the modified behaviour.
3868 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3871 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3872 indeed breaks things for older releases.
3874 3. Added additional logging to the case where there is a problem reading data
3875 from a filter that is running in a subprocess using a pipe, in order to
3876 try to track down a specific problem.
3878 4. Testing facility fudge: when running in the test harness and attempting
3879 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3880 getting "No route to host". Convert this to a timeout.
3882 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3885 6. Some OS don't have socklen_t but use size_t instead. This affects the
3886 fifth argument of getsockopt() amongst other things. This is now
3887 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3888 can be set for individual OS. I have set it for SunOS5, OSF1, and
3889 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3890 some earlier ones do not.
3892 7. Change 4.30/15 was not doing the test caselessly.
3894 8. The standard form for an IPv6 address literal was being rejected by address
3895 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3896 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3897 this, as well as the form without the "IPv6" on the front (but only when
3898 address literals are enabled, of course).
3900 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3902 10. Exim crashed if a message with an empty sender address specified by -f
3903 encountered a router with an errors_to setting. This could be provoked only
3904 by a command such as
3908 where an empty string was supplied; "<>" did not hit this bug.
3910 11. Installed PCRE release 4.5.
3912 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3913 remained set. It is now erased.
3915 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3916 times from message ids (which are base 36 rather than the normal 62).
3918 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3919 were being counted as actual protocol errors, and logged if the log
3920 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3921 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3922 and DATA following a set of rejected RCPTs do not count as protocol errors.
3923 In other words, Exim assumes they were pipelined, though this may not
3924 actually be the case. Of course, in all cases the client gets an
3925 appropriate error code.
3927 15. If a lookup fails in an ACL condition, a message about the failure may
3928 be available; it is used if testing the ACL cannot continue, because most
3929 such messages specify what the cause of the deferral is. However, some
3930 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3931 that caused an old message to be retained and used if a later statement
3932 caused a defer, replacing the real cause of the deferral.
3934 16. If an IP address had so many PTR records that the DNS lookup buffer
3935 was not large enough to hold them, Exim could crash while trying to process
3936 the truncated data. It now detects and logs this case.
3938 17. Further to 4.21/58, another change has been made: if (and only if) the
3939 first line of a message (the first header line) ends with CRLF, a bare LF
3940 in a subsequent header line has a space inserted after it, so as not to
3941 terminate the header.
3943 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3944 unnecessarily, used atoi() instead of strtol(), and didn't check the
3945 termination when getting file sizes from file names by regex.
3947 19. Completely re-implemented the support for maildirsize files, in the light
3948 of a number of problems with the previous contributed implementation
3949 (4.30/29). In particular:
3951 . If the quota is zero, the maildirsize file is maintained, but no quota is
3954 . If the maildir directory does not exist, it is created before any attempt
3955 to write a maildirsize file.
3957 . The quota value in the file is just a cache; if the quota is changed in
3958 the transport, the new value overrides.
3960 . A regular expression is available for excluding directories from the
3963 20. The autoreply transport checks the characters in options that define the
3964 message's headers; it allows continued headers, but it was checking with
3965 isspace() after an embedded newline instead of explicitly looking for a
3968 21. If all the "regular" hosts to which an address was routed had passed their
3969 expiry times, and had not reached their retry times, the address was
3970 bounced, even if fallback hosts were defined. Now Exim should go on to try
3973 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3974 equivalent code in the SMTP transport. Some hosts send humungous responses
3975 to HELO/EHLO, more than 1024 it seems.
3977 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3978 gives compiler warnings in some environments. I've now done it "properly",
3981 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3982 (because of problems with the built-in one) was declared to return uschar *
3983 instead of char *, causing compiler failure.
3985 25. Fixed a file descriptor leak when processing alias/forward files.
3987 26. Fixed a minor format string issue in dbfn.c.
3989 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3991 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3992 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3995 29. When a sender address is verified, it is cached, to save repeating the test
3996 when there is more than one recipient in a message. However, when the
3997 verification involves a callout, it is possible for different callout
3998 options to be set for different recipients. It is too complicated to keep
3999 track of this in the cache, so now Exim always runs a verification when a
4000 callout is required, relying on the callout cache for the optimization.
4001 The overhead is duplication of the address routing, but this should not be
4004 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4005 to be verified with callout=postmaster, and the main callout worked but the
4006 postmaster check failed, the verification correctly failed. However, if a
4007 subsequent RCPT command asked for sender verification *without* the
4008 postmaster check, incorrect caching caused this verification also to fail,
4011 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4012 it was not caching the DNS options (qualify_single, search_parents) that
4013 were used when the lookup failed. A subsequent lookup with different
4014 options therefore always gave the same answer, though there were cases
4015 where it should not have. (Example: a "domains = !$mx_any" option on a
4016 dnslookup router: the "domains" option is always processed without any
4017 widening, but the router might have qualify_single set.) Now Exim uses the
4018 cached value only when the same options are set.
4020 32. Added John Jetmore's "exipick" utility to the distribution.
4022 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4023 than a timeout (e.g. a certificate is required, and is not provided), an
4024 Exim server now closes the connection immediately. Previously it waited for
4025 the client to close - but if the client is SSL, it seems that they each
4026 wait for each other, leading to a delay before one of them times out.
4028 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4029 maintained 0.8.x compatibility because I don't think many are using it, and
4030 it is clearly obsolete.
4032 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4035 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4036 client certificate was expired. A simple patch fixes this, though I don't
4037 understand the full logic of why the verify callback is called multiple
4040 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4041 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4042 which causes problems with some clients (such as the Certicom SSL Plus
4043 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4044 disables the coutermeasure allowing Eudora to connect."
4046 38. Exim was not checking that a write() to a log file succeeded. This could
4047 lead to Bad Things if a log got too big, in particular if it hit a file
4048 size limit. Exim now panics and dies if it cannot write to a log file, just
4049 as it does if it cannot open a log file.
4051 39. Modified OS/Makefile-Linux so that it now contains
4053 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4055 The two -D definitions ensure that Exim is compiled with large file
4056 support, which makes it possible to handle log files that are bigger than
4059 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4060 instance) a domain was checked against a named list that involved a lookup,
4061 causing $domain_data to be set, then another domain was checked against the
4062 same list, then the first domain was re-checked, the value of $domain_data
4063 after the final check could be wrong. In particular, if the second check
4064 failed, it could be set empty. This bug probably also applied to
4067 41. The strip_trailing_dot option was not being applied to the address given
4068 with the -f command-line option.
4070 42. The code for reading a message's header from the spool was incrementing
4071 $received_count, but never initializing it. This meant that the value was
4072 incorrect (doubled) while delivering a message in the same process in which
4073 it was received. In the most common configuration of Exim, this never
4074 happens - a fresh exec is done - but it can happen when
4075 deliver_drop_privilege is set.
4077 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4078 it now includes up to 150 characters of the unexpected data in the log
4081 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4082 and building data strings. The size of both of these buffers was 10 000
4083 bytes - far larger than anybody would *ever* want, thought I. Needless to
4084 say, somebody hit the limit. I have increased the maximum line length to
4085 20 000 and the maximum data length of concatenated lines to 100 000. I have
4086 also fixed two bugs, because there was no checking on these buffers. Tsk,
4087 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4088 buffer is too small.
4090 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4091 lsearch lookups. Now it does.
4093 46. When parsing a route_list item in a manualroute router, a fixed-length
4094 buffer was used for the list of hosts. I made this 1024 bytes long,
4095 thinking that nobody would ever have a list of hosts that long. Wrong.
4096 Somebody had a whole pile of complicated expansion conditions, and the
4097 string was silently truncated, leading to an expansion error. It turns out
4098 that it is easier to change to an unlimited length (owing to other changes
4099 that have happened since this code was originally written) than to build
4100 structure for giving a limitation error. The length of the item that
4101 expands into the list of hosts is now unlimited.
4103 47. The lsearch lookup could not handle data where the length of text line was
4104 more than 4095 characters. Such lines were truncated, leading to shortened
4105 data being returned. It should now handle lines of any length.
4107 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4108 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4111 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4112 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4113 no longer happens in comment lines. A list of replacements is now placed
4114 at the head of all of the source files, except those whose only change is
4115 to replace PERL_COMMAND in the very first #! line.
4117 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4118 messages on the queue, with a bottom-up merge sort, using code contributed
4119 by Michael Haardt. This should make operations like -bp somewhat faster on
4120 large queues. It won't affect queue runners, except when queue_run_in_order
4123 51. Installed eximstats 1.31 in the distribution.
4125 52. Added support for SRV lookups to the dnslookup router.
4127 53. If an ACL referred to $message_body or $message_body_end, the value was not
4128 reset for any messages that followed in the same SMTP session.
4130 54. The store-handling optimization for building very long strings was not
4131 differentiating between the different store pools. I don't think this
4132 actually made any difference in practice, but I've tidied it.
4134 55. While running the routers to verify a sender address, $sender_address
4135 was still set to the sender address. This is wrong, because when routing to
4136 send a bounce to the sender, it would be empty. Therefore, I have changed
4137 it so that, while verifying a sender address, $sender_address is set to <>.
4138 (There is no change to what happens when verifying a recipient address.)
4140 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4141 target A or AAAA records (if not already returned) without resetting the
4142 qualify_single or search_parents options of the DNS resolver. These are
4143 inappropriate in this case because the targets of MX and SRV records must
4144 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4145 target that, when qualified, matched something in the local domain. These
4146 two options are now turned off when doing these lookups.
4148 57. It seems that at least some releases of Reiserfs (which does not have the
4149 concept of a fixed number of inodes) returns zero and not -1 for the
4150 number of available inodes. This interacted badly with check_spool_inodes,
4151 which assumed that -1 was the "no such thing" setting. What I have done is
4152 to check that the total number of inodes is greater than zero before doing
4153 the test of how many are available.
4155 58. When a "warn" ACL statement has a log_message modifier, the message is
4156 remembered, and not repeated. This is to avoid a lot of repetition when a
4157 message has many recipients that cause the same warning to be written.
4158 Howewer, Exim was preserving the list of already written lines for an
4159 entire SMTP session, which doesn't seem right. The memory is now reset if a
4160 new message is started.
4162 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4163 debugging output unless log_rewrite was also set.
4165 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4166 of (int)(handle) when we know that handle contains (void *)(-1).
4168 61. The Exim daemon panic-logs an error return when it closes the incoming
4169 connection. However "connection reset by peer" seems to be common, and
4170 isn't really an error worthy of noting specially, so that particular error
4173 62. When Exim is trying to find all the local interfaces, it used to panic and
4174 die if the ioctl to get the interface flags failed. However, it seems that
4175 on at least one OS (Solaris 9) it is possible to have an interface that is
4176 included in the list of interfaces, but for which you get a failure error
4177 for this call. This happens when the interface is not "plumbed" into a
4178 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4179 failure of the "get flags" call assumes that the interface is down.
4181 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4182 makes life easier for people who are doing arithmetic on fields extracted
4183 from dates, where you often get leading zeros that should not be
4184 interpreted as octal.
4186 64. Added qualify_domain to the redirect router, to override the global
4189 65. If a pathologically long header line contained very many addresses (the
4190 report of this problem mentioned 10 000) and each of them was rewritten,
4191 Exim could use up a very large amount of memory. (It kept on making new
4192 copies of the header line as it rewrote, and never released the old ones.)
4193 At the expense of a bit more processing, the header rewriting function has
4194 been changed so that it no longer eats memory in this way.
4196 66. The generation of the Received: header has been moved from the time that a
4197 message starts to be received, to the time that it finishes. The timestamp
4198 in the Received: header should now be very close to that of the <= log
4199 line. There are two side-effects of this change:
4201 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4202 logged header lines no longer include the local Received: line, because
4203 it has not yet been created. The same applies to a copy of the message
4204 that is returned to a non-SMTP sender when a message is rejected.
4206 (b) When a filter file is tested using -bf, no additional Received: header
4207 is added to the test message. After some thought, I decided that this
4210 This change does not affect the value of $received_for. It is still set
4211 after address rewriting, but before local_scan() is called.
4213 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4215 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4216 gave an unhelpful panic error message, and a defer error. I have managed to
4217 change this behaviour so that it now rejects any supplied certificate,
4218 which seems right, as the list of acceptable certificates is empty.
4220 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4221 gave an unhelpful defer error. I have not managed to make this reject any
4222 supplied certificates, but the error message it gives is "no certificate
4223 supplied", which is not helpful.
4225 70. exigrep's output now also includes lines that are not associated with any
4226 message, but which match the given pattern. Implemented by a patch from
4227 Martin Sluka, which also tidied up the Perl a bit.
4229 71. Recipient callout verification, like sender verification, was using <> in
4230 the MAIL FROM command. This isn't really the right thing, since the actual
4231 sender may affect whether the remote host accepts the recipient or not. I
4232 have changed it to use the actual sender in the callout; this means that
4233 the cache record is now keyed on a recipient/sender pair, not just the
4234 recipient address. There doesn't seem to be a real danger of callout loops,
4235 since a callout by the remote host to check the sender would use <>.
4236 [SEE ABOVE: changed after hitting problems.]
4238 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4239 temporary errors. However, in the case of such a code being given after
4240 the end of a data transmission (i.e. after ".") Exim was failing to write
4241 a retry record for the message. (Yes, there was some broken host that was
4242 actually sending 8xx at this point.)
4244 73. An unknown lookup type in a host list could cause Exim to panic-die when
4245 the list was checked. (An example that provoked this was putting <; in the
4246 middle of a list instead of at the start.) If this happened during a DATA
4247 ACL check, a -D file could be left lying around. This kind of configuration
4248 error no longer causes Exim to die; instead it causes a defer errror. The
4249 incident is still logged to the main and panic logs.
4251 74. Buglet left over from Exim 3 conversion. The message "too many messages
4252 in one connection" was written to the rejectlog but not the mainlog, except
4253 when address rewriting (yes!) was being logged.
4255 75. Added write_rejectlog option.
4257 76. When a system filter was run not as root (that is, when system_filter_user
4258 was set), the values of the $n variables were not being returned to the
4259 main process; thus, they were not subsequently available in the $sn
4262 77. Added +return_path_on_delivery log selector.
4264 78. A connection timeout was being treated differently from recipients deferred
4265 when testing hosts_max_try with a message that was older than the host's
4266 retry timeout. (The host should not be counted, thus allowing all hosts to
4267 be tried at least once before bouncing.) This may have been the cause of an
4268 occasionally reported bug whereby a message would remain on the queue
4269 longer than the retry timeout, but would be bounced if a delivery was
4270 forced. I say "may" because I never totally pinned down the problem;
4271 setting up timeout/retry tests is difficult. See also the next item.
4273 79. The ultimate address timeout was not being applied to errors that involved
4274 a combination of host plus message (for example, a timeout on a MAIL
4275 command). When an address resolved to a number of possible hosts, and they
4276 were not all tried for each delivery (e.g. because of hosts_max_try), a
4277 message could remain on the queue longer than the retry timeout.
4279 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4282 81. Fixed an obscure SMTP outgoing bug which required at least the following
4283 conditions: (a) there was another message waiting for the same server;
4284 (b) the server returned 5xx to all RCPT commands in the first message so
4285 that the message was not completed; (c) the server dropped the connection
4286 or gave a negative response to the RSET that Exim sends to abort the
4287 transaction. The observed case was a dropped connection after DATA that had
4288 been sent in pipelining mode. That is, the server had advertised PIPELINING
4289 but was not implementing it correctly. The effect of the bug was incorrect
4290 behaviour, such as trying another host, and this could lead to a crash.
4296 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4297 and daemon.c were passed as pointers to ints; they should have been
4298 pointers to socklen_t variables (which are typically unsigned ints).
4300 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4303 3. Fixed a really odd bug that affected only the testing scheme; patching a
4304 certain fixed string in the binary changed the value of another string that
4305 happened to be identical to the end of the original first string.
4307 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4308 name", it returns that address as the IP address. On some operating
4309 systems (e.g. Solaris), it also passes back the IP address string as the
4310 "host name". However, on others (e.g. Linux), it passes back an empty
4311 string. Exim wasn't checking for this, and was changing the host name to an
4312 empty string, assuming it had been canonicized.
4314 5. Although rare, it is permitted to have more than one PTR record for a given
4315 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4316 all the names associated with an address, because they do in Solaris.
4317 However, it seems that they do not in Linux for data that comes from the
4318 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4319 I found this out when I moved to a new Linux workstation and tried to run
4320 the Exim test suite.
4322 To get round this problem I have changed the code so that it now does its
4323 own call to the DNS to look up PTR records when searching for a host name.
4324 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4325 addresses that are only in /etc/hosts are still found.
4327 This behaviour is, however, controlled by an option called host_lookup_
4328 order, which defaults to "bydns:byaddr". If people want to use the other
4329 order, or indeed, just use one or the other means of lookup, they can
4330 specify it in this variable.
4332 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4333 some operating systems, this comes back from gethostbyaddr() as an empty
4334 string, and this is what Exim used to test for. However, it seems that in
4335 other systems, "." is yielded. Exim now tests for this case too.
4337 7. The values of check_spool_space and check_log_space are now held internally
4338 as a number of kilobytes instead of an absolute number of bytes. If a
4339 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4340 kilobyte. This means that much larger values can be stored.
4342 8. Exim monitor: an attempt to get the action menu when not actually pointing
4343 at a message produces an empty menu entitled "No message selected". This
4344 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4345 no entries in it ("Shell widget menu has zero width and/or height"). So I
4346 have added a single, blank menu entry in this case.
4348 9. Added ${quote_local_part.
4350 10. MIME decoding is now applied to the contents of Subject: header lines when
4353 11. Now that a reference to $sender_host_address automatically causes a reverse
4354 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4355 host lookup before query-style lookups in lists that might use this
4356 variable. This has therefore been abolished, and the "net-" prefix is no
4357 longer necessary for query-style lookups.
4359 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4360 have been a typo for LFLAGS, so it has been changed.
4362 13. The install script calls Exim with "-C /dev/null" in order to find the
4363 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4364 to be output. Howeve, since Exim outputs its version number before the
4365 error, it didn't break the script. It just looked ugly. I fixed this by
4366 always allowing "-C /dev/null" if the caller is root.
4368 14. Ignore overlarge ACL variable number when reading spool file - insurance
4369 against a later release with more variables having written the file.
4371 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4372 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4373 this, as well as the form without the "IPv6" on the front.
4375 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4376 OS/Makefile-Darwin file.
4378 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4379 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4381 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4382 variable. However, some OS use ints for inodes. I've added cast to long int
4383 to get rid of the compiler warning.
4385 19. I had forgotten to lock out "/../" in configuration file names when
4386 ALT_CONFIG_PREFIX was set.
4388 20. Routers used for verification do not need to specify transports. However,
4389 if such a router generated a host list, and callout was configured, Exim
4390 crashed, because it could not find a port number from the (non-existent)
4391 transport. It now assumes port 25 in this circumstance.
4393 21. Added the -t option to exigrep.
4395 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4396 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4397 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4398 them is set, LOOKUP_LSEARCH is forced.
4400 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4402 24. Added sender and host information to the "rejected by local_scan()" log
4403 line; previously there was no indication of these.
4405 25. Added .include_if_exists.
4407 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4408 Linux. It turns out that not all file systems support this. Apparently some
4409 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4410 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4411 not supported on the file descriptor, is now ignored when Exim is trying to
4412 sync a directory. This applies only to Linux.
4414 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4416 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4417 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4420 29. Added support for maildirsize files from supplied patch (modified a bit).
4422 30. The use of :fail: followed by an empty string could lead Exim to respond to
4423 sender verification failures with (e.g.):
4425 550 Verification failed for <xxx>
4426 550 Sender verify failed
4428 where the first response line was missing the '-' that indicates it is not
4429 the final line of the response.
4431 31. The loop for finding the name of the user that called Exim had a hardwired
4432 limit of 10; it now uses the value of finduser_retries, which is used for
4433 all other user lookups.
4435 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4438 33. Exim was neglecting to zero errno before one call of strtol() when
4439 expanding a string and expecting an integer value. On some systems this
4440 resulted in spurious "integer overflow" errors. Also, it was casting the
4441 result into an int without checking.
4443 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4444 did not work. The code looks as if it has *never* worked, though it appears
4445 to have been documented since at least releast 1.62. I have made it work.
4447 35. The "timeout_DNS" error in retry rules, also documented since at least
4448 1.62, also never worked. As it isn't clear exactly what this means, and
4449 clearly it isn't a major issue, I have abolished the feature by treating it
4450 as "timeout", and writing a warning to the main and panic logs.
4452 36. The display of retry rules for -brt wasn't always showing the error code
4455 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4456 timeout_connect_A, timeout_connect_MX.
4458 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4459 to the empty sender.
4461 39. The daemon was not analysing the content of -oX till after it had closed
4462 stderr and disconnected from the controlling terminal. This meant that any
4463 syntax errors were only noted on the panic log, and the return code from
4464 the command was 0. By re-arranging the code a little, I've made the
4465 decoding happen first, so such errors now appear on stderr, and the return
4466 code is 1. However, the actual setting up of the sockets still happens in
4467 the disconnected process, so errors there are still only recorded on the
4470 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4471 connections (as happens on some IP stacks) was logged at start up time as
4472 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4473 it from "IPv6 and IPv4", which means that two separate sockets are being
4476 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4477 says whether AF_INET or AF_INET6 was passed as an argument.
4479 42. Exiwhat output was messed up when time zones were included in log
4482 43. Exiwhat now gives more information about the daemon's listening ports,
4483 and whether -tls-on-connect was used.
4485 44. The "port" option of the smtp transport is now expanded.
4487 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4488 silently ignored. Now an error message is written to the main and panic
4491 46. There's a new ACL modifier called "logwrite" which writes to a log file
4492 as soon as it is encountered.
4494 47. Added $local_user_uid and $local_user_gid at routing time.
4496 48. Exim crashed when trying to verify a sender address that was being
4499 49. Exim was recognizing only a space character after ".include". It now also
4500 recognizes a tab character.
4502 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4503 extracting the relevant information from the specification. The man page no
4504 longer contains scrambled data for the -d option, and I've added a section
4505 at the front about calling Exim under different names.
4507 51. Added "extra_headers" argument to the "mail" command in filter files.
4509 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4512 53. Installed eximstats 1.29.
4514 54. Added transport_filter_timeout as a generic transport option.
4516 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4517 Cc: header lines. This was required by RFC 822, but it not required by RFC
4520 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4521 incoming messages that did not have them. Now it does so only if the
4522 message originates locally, that is, if there is no associated remote host
4523 address. When Resent- header lines are present, this applies to the Resent-
4524 lines rather than the non-Resent- lines.
4526 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4527 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4529 58. Messages for configuration errors now include the name of the main
4530 configuration file - useful now that there may be more than one file in a
4531 list (.included file names were always shown).
4533 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4534 for those rare installations that do not start the daemon as root or run it
4535 setuid root. I've cut out the call to initgroups() if the daemon is not
4538 60. The Exim user and group can now be bound into the binary as text strings
4539 that are looked up at the start of Exim's processing.
4541 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4543 62. Added $mailstore_basename variable.
4545 63. Installed patch to sieve.c from Michael Haardt.
4547 64. When Exim failed to open the panic log after failing to open the main log,
4548 the original message it was trying to log was written to stderr and debug
4549 output, but if they were not available (the usual case in production), it
4550 was lost. Now it is written to syslog before the two lines that record the
4551 failures to open the logs.
4553 65. Users' Exim filters run in subprocesses under the user's uid. It is
4554 possible for a "deliver" command or an alias in a "personal" command to
4555 provoke an address rewrite. If logging of address rewriting is configured,
4556 this fails because the process is not running as root or exim. There may be
4557 a better way of dealing with this, but for the moment (because 4.30 needs
4558 to be released), I have disabled address rewrite logging when running a
4559 filter in a non-root, non-exim process.
4565 1. The buildconfig auxiliary program wasn't quoting the value set for
4566 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4567 not defined. This bug was masked in 4.22 by the effect that was fixed in
4570 2. Some messages that were rejected after a message id was allocated were
4571 shown as "incomplete" by exigrep. It no longer does this for messages that
4572 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4574 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4575 have allow_domain_literals set, the ID did not get logged in the <= line.
4576 Domain literals are now always recognized in Message-ID: header lines.
4578 4. The first argument for a ${extract expansion item is the key name or field
4579 number. Leading and trailing spaces in this item were not being ignored,
4580 causing some misleading effects.
4582 5. When deliver_drop_privilege was set, single queue runner processes started
4583 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4584 same command in the process it spins off) were not dropping privilege.
4586 6. When the daemon running as "exim" started a queue runner, it always
4587 re-executed Exim in the spun-off process. This is a waste of effort when
4588 deliver_drop_privilege is set. The new process now just calls the
4589 queue-runner function directly.
4595 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4598 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4599 ignored. Though the use of -oP was forcing the writing of a pid file, it
4600 was always written to the default place.
4602 3. If the message "no IP address found for host xxxx" is generated during
4603 incoming verification, it is now followed by identification of the incoming
4604 connection (so you can more easily find what provoked it).
4606 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4608 5. Added some features to "harden" Exim a bit more against certain attacks:
4610 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4611 be put in Local/Makefile. This is like the never_users runtime option,
4612 but it cannot be overridden. The default setting is "root".
4614 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4615 prefix string with which any file named in a -C command line option
4618 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4619 is retained for -C and -D only if the caller of Exim is root. Without
4620 it, the exim user may also use -C and -D and retain privilege.
4622 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4623 command line option is disabled.
4625 6. Macro names set by the -D option must start with an upper case letter, just
4626 like macro names defined in the configuration file.
4628 7. Added "dereference=" facility to LDAP.
4630 8. Two instances of the typo "uknown" in the source files are fixed.
4632 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4633 the Configure-Makefile script screwed up while processing it.
4635 10. Incorporated PCRE 4.4.
4637 11. The SMTP synchronization check was not operating right at the start of an
4638 SMTP session. For example, it could not catch a HELO sent before the client
4639 waited for the greeting. There is now a check for outstanding input at the
4640 point when the greeting is written. Because of the duplex, asynchronous
4641 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4642 way, but not yet received, when the check is performed.
4644 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4645 on TCP/IP sockets, because this apparently causes some broken clients to
4648 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4649 unchanged) from the Cygwin maintainer.
4651 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4652 was supported instead of testing for mbx. Effectively a typo.
4654 15. The spa authenticator server code was not checking that the input it
4655 received was valid base64.
4657 16. The debug output line for the "set" modifier in ACLs was not showing the
4658 name of the variable that was being set.
4660 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4662 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4663 lookup on its own. Something else was needed to trigger the lookup. For
4664 example, a match in host_lookup or the need for a host name in a host list.
4665 Now, if $sender_host_name is referenced and the host name has not yet been
4666 looked up, a lookup is performed. If the lookup fails, the variable remains
4667 empty, and $host_lookup_failed is set to "1".
4669 19. Added "eqi" as a case-independent comparison operator.
4671 20. The saslauthd authentication condition could segfault if neither service
4672 nor realm was specified.
4674 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4675 error message that was logged was misleading, and incoming SMTP
4676 connections were dropped. The message is now more accurate, and temporary
4677 errors are given to SMTP connections.
4679 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4680 (except RSET) with a 421 error, until QUIT is received. However, it was
4681 failing to send a response to QUIT.
4683 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4684 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4685 ACL is now run after the helo_try_verify_hosts code.
4687 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4688 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4689 case-independent, so other case variants are also recognized). Apparently
4690 some systems use these upper case variants.
4692 25. If more than two messages were waiting for the same host, and a transport
4693 filter was specified for the transport, Exim sent two messages over the
4694 same TCP/IP connection, and then failed with "socket operation on non-
4695 socket" when it tried to send the third.
4697 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4699 27. The extern definition of crypt16() in expand.c was not being excluded when
4700 the OS had its own crypt16() function.
4702 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4703 as a preferred synonym for return_size_limit, both as an option and as an
4706 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4708 30. Changed the default configuration ACL to relax the local part checking rule
4709 for addresses that are not in any local domains. For these addresses,
4710 slashes and pipe symbols are allowed within local parts, but the sequence
4711 /../ is explicitly forbidden.
4713 31. SPA server authentication was not clearing the challenge buffer before
4716 32. log_message in a "warn" ACL statement was writing to the reject log as
4717 well as to the main log, which contradicts the documentation and doesn't
4718 seem right (because no rejection is happening). So I have stopped it.
4720 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4721 However, I am unable to do any testing of this.
4723 34. Fixed an infelicity in the appendfile transport. When checking directories
4724 for a mailbox, to see if any needed to be created, it was accidentally
4725 using path names with one or more superfluous leading slashes; tracing
4726 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4728 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4729 contents are added to the log line that is written for every discarded
4730 recipient. (Previously a log_message setting was ignored.)
4732 36. The ${quote: operator now quotes the string if it is empty.
4734 37. The install script runs exim in order to find its version number. If for
4735 some reason other than non-existence or emptiness, which it checks, it
4736 could not run './exim', it was installing it with an empty version number,
4737 i.e. as "exim-". This error state is now caught, and the installation is
4740 38. An argument was missing from the function that creates an error message
4741 when Exim fails to connect to the socket for saslauthd authentication.
4742 This could cause Exim to crash, or give a corrupted message.
4744 39. Added isip, isip4, and isip6 to ${if conditions.
4746 40. The ACL variables $acl_xx are now saved with the message, and can be
4747 accessed later in routers, transports, and filters.
4749 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4750 strings in the file are not string-expanded.
4752 42. If a MAIL command specified a SIZE value that was too large to fit into an
4753 int variable, the check against message_size_limit failed. Such values are
4754 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4755 day this will have to be increased, but I don't think I want to be around
4756 when emails are that large.
4763 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4764 iconv() is not standard in FreeBSD.
4766 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4767 IPv6 enabled. The observed symptom was a segmentation fault on return from
4768 the function os_common_find_running_interfaces() in src/os.c.
4770 3. In the check_special_case() function in daemon.c I had used "errno" as an
4771 argument name, which causes warnings on some systems. This was basically a
4772 typo, since it was named "eno" in the comments!
4774 4. The code that waits for the clock to tick (at a resolution of some fraction
4775 of a second) so as to ensure message-id uniqueness was always waiting for
4776 at least one whole tick, when it could have waited for less. [This is
4777 almost certainly not relevant at current processor speeds, where it is
4778 unlikely to ever wait at all. But we try to future-proof.]
4780 5. The function that sleeps for a time interval that includes fractions of a
4781 second contained a race. It did not block SIGALRM between setting the
4782 timer, and suspending (a couple of lines later). If the interval was short
4783 and the sigsuspend() was delayed until after it had expired, the suspension
4784 never ended. On busy systems this could lead to processes getting stuck for
4787 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4788 process, before it forks any delivery processes. The open lookup caching
4789 mechanism meant that the open file or database connection was passed into
4790 the delivery process. The problem was that delivery processes always tidy
4791 up cached lookup data. This could cause a problem for the next delivery
4792 process started by the queue runner, because the external queue runner
4793 process does not know about the closure. So the next delivery process
4794 still has data in the lookup cache. In the case of a file lookup, there was
4795 no problem because closing a file descriptor in a subprocess doesn't affect
4796 the parent. However, if the lookup was caching a connection to a database,
4797 the connection was closed, and the second delivery process was likely to
4798 see errors such as "PGSQL: query failed: server closed the connection
4799 unexpectedly". The problem has been fixed by closing all cached lookups
4800 in a queue runner before running a delivery process.
4802 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4803 seem to have the "const" qualifier which it has on other OS. I've
4806 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4807 *of the same type* (client or server) with the same public name that an
4808 error should be diagnosed.
4810 9. When Exim looked up a host name for an IP address, but failed to find the
4811 original IP address when looking up the host name (a safety check), it
4812 output the message "<ip address> does not match any IP for NULL", which was
4813 confusing, to say the least. The bug was that the host name should have
4814 appeared instead of "NULL".
4816 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4817 user that is built into the binary, and the -C or -D options is used, root
4818 privilege is dropped before the configuration file is read. In addition,
4819 logging is switched to stderr instead of the normal log files. If the
4820 configuration then re-defines the Exim user, the unprivileged environment
4821 is probably not what is expected, so Exim logs a panic warning message (but
4824 However, if deliver_drop_privilege is set, the unprivileged state may well
4825 be exactly what is intended, so the warning has been cut out in that case,
4826 and Exim is allowed to try to write to its normal log files.
4832 1. smtp_return_error_details was not giving details for temporary sender
4833 or receiver verification errors.
4835 2. Diagnose a configuration error if two authenticators have the same public
4838 3. Exim used not to create the message log file for a message until the first
4839 delivery attempt. This could be confusing when incoming messages were held
4840 for policy or load reasons. The message log file is now created at the time
4841 the message is received, and an initial "Received" line is written to it.
4843 4. The automatically generated man page for command line options had a minor
4844 bug that caused no ill effects; however, a more serious problem was that
4845 the procedure for building the man page automatically didn't always
4846 operate. Consequently, release 4.20 contains an out-of-date version. This
4847 shouldn't happen again.
4849 5. When building Exim with embedded Perl support, the script that builds the
4850 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4851 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4853 6. The freeze_tell option was not being used for messages that were frozen on
4854 arrival, either by an ACL or by local_scan().
4856 7. Added the smtp_incomplete_transaction log selector.
4858 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4859 was accepting AUTH without a new EHLO.
4861 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4862 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4865 10. Logging of TCP/IP connections (when configured) now happens in the main
4866 daemon process instead of the child process, so that the TCP/IP connection
4867 count is more accurate (but it can never be perfect).
4869 11. The use of "drop" in a nested ACL was not being handled correctly in the
4870 outer ACL. Now, if condition failure induced by the nested "drop" causes
4871 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4872 or "require"), the connection is dropped.
4874 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4875 that yield "discard" can now be used with an "accept" or a "discard" verb,
4876 but an error is generated for any others (because I can't see a useful way
4877 to define what should happen).
4879 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4880 are now processed in the same way as lines in the Exim configuration file.
4881 In particular, continuation lines are supported.
4883 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4885 15. Added -ti meaning -t -i.
4887 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4888 domains, and warn by logging if others are found.
4890 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4891 returned from GIFCONF when Exim is trying to find the list of interfaces on
4892 a host. The code in os.c has been modified to copy each ifreq to an aligned
4893 structure in all cases.
4895 Also, in some cases, the returned ifreq's were being copied to a 'struct
4896 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4897 means the last couple of bytes of an IPv6 address could be chopped if the
4898 ifreq contained only a normal sockaddr (14 bytes storage).
4900 18. Named domain lists were not supported in the hosts_treat_as_local option.
4901 An entry such as +xxxx was not recognized, and was treated as a literal
4904 19. Ensure that header lines added by a DATA ACL are included in the reject log
4905 if the ACL subsequently rejects the message.
4907 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4908 MD5 (which is deprecated).
4910 21. When testing a filter file using -bf, Exim was writing a message when it
4911 took the sender from a "From " line in the message, but it was not doing so
4912 when it took $return_path from a Return-Path: header line. It now does.
4914 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4915 with a valid header line field name (a series of printing characters
4916 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4918 23. Changed "disc" in the source to "disk" to conform to the documentation and
4919 the book and for uniformity.
4921 24. Ignore Sendmail's -Ooption=value command line item.
4923 25. When execve() failed while trying to run a command in a pipe transport,
4924 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4925 could be confused with a return value of 69 from the command itself. This
4926 has been changed to 127, the value the shell returns if it is asked to run
4927 a non-existent command. The wording for the related log line suggests a
4928 non-existent command as the problem.
4930 26. If received_header_text expands to an empty string, do not add a Received:
4931 header line to the message. (Well, it adds a token one on the spool, but
4932 marks it "old" so that it doesn't get used or transmitted.)
4934 27. Installed eximstats 1.28 (addition of -nt option).
4936 28. There was no check for failure on the call to getsockname() in the daemon
4937 code. This can fail if there is a shortage of resources on the system, with
4938 ENOMEM, for example. A temporary error is now given on failure.
4940 29. Contrary to the C standard, it seems that in some environments, the
4941 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4942 program. Exim now does this explicitly; it affects the formatting of
4943 timestamps using strftime().
4945 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4946 complaints. I've now initialized all the variables, to avoid this.
4948 32. Header lines added by a system filter were not being "seen" during
4949 transport-time rewrites.
4951 33. The info_callback() function passed to OpenSSL is set up with type void
4952 (*)(SSL *, int, int), as described somewhere. However, when calling the
4953 function (actually a macro) that sets it up, the type void(*)() is
4954 expected. I've put in a cast to prevent warnings from picky compilers.
4956 34. If a DNS black list lookup found a CNAME record, but there were no A
4957 records associated with the domain it pointed at, Exim crashed.
4959 35. If a DNS black list lookup returned more than one A record, Exim ignored
4960 all but the first. It now scans all returned addresses if a particular IP
4961 value is being sought. In this situation, the contents of the
4962 $dnslist_value variable are a list of all the addresses, separated by a
4965 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4966 to accept a host name and all its aliases if the forward lookup for any of
4967 them yielded the IP address of the incoming connection. Now it accepts only
4968 those names whose forward lookup yields the correct IP address. Any other
4969 names are discarded. This closes a loophole whereby a rogue DNS
4970 administrator could create reverse DNS records to break through a
4971 wildcarded host restriction in an ACL.
4973 37. If a user filter or a system filter that ran in a subprocess used any of
4974 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4975 the wrong values were passed to the pipe command ($thisaddress had the
4976 value of $0, $0 had the value of $1, etc). This bug was introduced by
4977 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4979 38. Improved the line breaking for long SMTP error messages from ACLs.
4980 Previously, if there was no break point between 40 and 75 characters, Exim
4981 left the rest of the message alone. Two changes have been made: (a) I've
4982 reduced the minimum length to 35 characters; (b) if it can't find a break
4983 point between 35 and 75 characters, it looks ahead and uses the first one
4984 that it finds. This may give the occasional overlong line, but at least the
4985 remaining text gets split now.
4987 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4988 file descriptors might be low, and that setting 1000 would always raise it.
4989 It turns out that in some environments, the limit is already over 1000 and
4990 that lowering it causes trouble. So now Exim takes care not to decrease it.
4992 40. When delivering a message, the value of $return_path is set to $sender_
4993 address at the start of routing (routers may change the value). By an
4994 oversight, this default was not being set up when an address was tested by
4995 -bt or -bv, which affected the outcome if any router or filter referred to
4998 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4999 log only when "message" or "log_message" are set. However, if one of the
5000 conditions was an address verification, or a call to a nested ACL, the
5001 messages generated by the underlying test were being passed through. This
5002 no longer happens. The underlying message is available in $acl_verify_
5003 message for both "message" and "log_message" expansions, so it can be
5004 passed through if needed.
5006 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5007 new expansion $bh_ to give the encoded byte string without charset
5008 translation. Translation happens only if iconv() is available; HAVE_ICONV
5009 indicates this at build time. HEADERS_CHARSET gives the charset to
5010 translate to; headers_charset can change it in the configuration, and
5011 "headers charset" can change it in an individual filter file.
5013 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5014 that creates RFC 2047 encoded "words" labels them as that charset instead
5015 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5016 expansion operator; (ii) when Exim creates a From: line for a local
5017 message; (iii) when a header line is rewritten to include a "phrase" part.
5019 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5020 buggy, causing it to skip the first lines of messages whose message ID
5021 ended in 'D'. This would not have bitten before Exim release 4.14, because
5022 message IDs were unlikely to end in 'D' before then. The effect was to have
5023 incorrect size information for certain domains.
5025 45. #include "config.h" was missing at the start of the crypt16.c module. This
5026 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5029 46. If there was a timeout during a "random" callout check, Exim treated it as
5030 a failure of the random address, and carried on sending RSET and the real
5031 address. If the delay was just some slowness somewhere, the response to the
5032 original RCPT would be taken as a response to RSET and so on, causing
5033 mayhem of various kinds.
5035 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5036 when I implemented it. It didn't allow for the fact that some option values
5037 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5038 the right test for positive values.
5040 48. Domain names in DNS records are case-independent. Exim always looks them up
5041 in lower case. Some resolvers return domain names in exactly the case they
5042 appear in the zone file, that is, they may contain uppercase letters. Not
5043 all resolvers do this - some return always lower case. Exim was treating a
5044 change of case by a resolver as a change of domain, similar to a widening
5045 of a domain abbreviation. This triggered its re-routing code and so it was
5046 trying to route what was effectively the same domain again. This normally
5047 caused routing to fail (because the router wouldn't handle the domain
5048 twice). Now Exim checks for this case specially, and just changes the
5049 casing of the domain that it ultimately uses when it transmits the message
5052 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5055 50. If a filter generated a file delivery with a non-absolute name (possible if
5056 no home directory exists for the router), the forbid_file option was not
5059 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5060 the existing equality matching.
5062 52. Exim was using ints instead of ino_t variables in some places where it was
5063 dealing with inode numbers.
5065 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5066 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5067 called TMPDIR, and if it finds it is different, it changes its value.
5069 54. The smtp_printf() function is now made available to local_scan() so
5070 additional output lines can be written before returning. There is also an
5071 smtp_fflush() function to enable the detection of a dropped connection.
5072 The variables smtp_input and smtp_batched_input are exported to
5075 55. Changed the default runtime configuration: the message "Unknown user"
5076 has been removed from the ACL, and instead placed on the localuser router,
5077 using the cannot_route_message feature. This means that any verification
5078 failures that generate their own messages won't get overridden. Similarly,
5079 the "Unrouteable address" message that was in the ACL for unverifiable
5080 relay addresses has also been removed.
5082 56. Added hosts_avoid_esmtp to the smtp transport.
5084 57. The exicyclog script was not checking for the esoteric option
5085 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5086 will work only if exicyclog is run under the appropriate euid.
5088 58. Following a discussion on the list, the rules by which Exim recognises line
5089 endings on incoming messages have been changed. The -dropcr and drop_cr
5090 options are now no-ops, retained only for backwards compatibility. The
5091 following line terminators are recognized: LF CRLF CR. However, special
5092 processing applies to CR:
5094 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5095 nor a local message in the state where . is a terminator.
5097 (ii) If a bare CR is encountered in a header line, an extra space is added
5098 after the line terminator so as not to end the header. The reasoning
5099 behind this is that bare CRs in header lines are most likely either
5100 to be mistakes, or people trying to play silly games.
5102 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5103 was being incorrectly given as 18 bytes larger than it should have been.
5104 This is a VOB (very old bug).
5106 60. This may never have affected anything current, but just in case it has:
5107 When the local host is found other than at the start of a list of hosts,
5108 the local host, those with the same MX, and any that follow, are discarded.
5109 When the list in question was part of a longer list of hosts, the following
5110 hosts (not currently being processed) were also being discarded. This no
5111 longer happens. I'm not sure if this situation could ever has previously
5114 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5117 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5118 header. This is something that is recommended in a new Internet Draft, and
5119 is something that is documented as being done by Sendmail. There are two
5120 possible values. For messages generated by the autoreply transport, Exim
5123 Auto-Submitted: auto-replied
5125 whereas for all other generated messages (e.g. bounces) it adds
5127 Auto-Submitted: auto-generated
5129 63. The "personal" condition in filters now includes a test for the
5130 Auto-Submitted: header. If it contains the string "auto-" the message it
5131 not considered personal.
5133 64. Added rcpt_include_affixes as a generic transport option.
5135 65. Added queue_only_override (default true).
5137 66. Added the syslog_duplication option.
5139 67. If what should have been the first header line of a message consisted of
5140 a space followed by a colon, Exim was mis-interpreting it as a header line.
5141 It isn't of course - it is syntactically invalid and should therefore be
5142 treated as the start of the message body. The misbehaviour could have
5143 caused a number of strange effects, including loss of data in subsequent
5144 header lines, and spool format errors.
5146 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5147 client host had authenticated. This control can now be exercised by an ACL
5148 for more flexibility.
5150 69. By default, callouts do not happen when testing with -bh. There is now a
5151 variant, -bhc, which does actually run the callout code, including
5152 consulting and updating the callout cache.
5154 70. Added support for saslauthd authentication, courtesy of Alexander
5157 71. If statvfs() failed on the spool or log directories while checking their
5158 size for availability, Exim confusingly gave the error "space shortage".
5159 Furthermore, in debugging mode it crashed with a floating point exception.
5160 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5161 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5162 serious problem, Exim now writes to the main and panic logs when this
5163 happens, with details of the failure. It then refuses to accept the
5164 incoming message, giving the message "spool directory problem" or "log
5165 directory problem" with a 421 code for SMTP messages.
5167 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5168 0, 1, and 2 exist, because some OS complain for execs without them (see
5169 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5170 descriptors. However, the code omitted to check that the open succeeded,
5171 causing mysterious errors if for some reason the permissions on /dev/null
5172 got screwed. Now Exim writes a message to the main and panic logs, and
5173 bombs out if it can't open /dev/null.
5175 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5176 interact so that it is all more flexible. It is supposed to remain
5177 backwards compatible. Also added extra_local_interfaces.
5179 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5180 to bomb out with an assertion failure - to the client this appears as a
5181 connection drop. This problem occurs in the part of the code that was taken
5182 from the Samba project. Fortunately, the assertion is in a very simple
5183 function, so I have fixed this by reproducing the function inline in the
5184 one place where it is called, and arranging for authentication to fail
5185 instead of killing the process with assert().
5187 75. The SPA client code was not working when the server requested OEM rather
5188 than Unicode encoding.
5190 76. Added code to make require_files with a specific uid setting more usable in
5191 the case where statting the file as root fails - usually a non-root-mounted
5192 NFS file system. When this happens and the failure is EACCES, Exim now
5193 forks a subprocess and does the per-uid checking as the relevant uid.
5195 77. Added process_log_path.
5197 78. If log_file_path was not explicitly set, a setting of check_log_space or
5198 check_log_inodes was ignored.
5200 79. If a space check for the spool or log partitions fails, the incident is now
5201 logged. Of course, in the latter case the data may get lost...
5203 80. Added the %p formatting code to string_format() so that it can be used to
5204 print addresses in debug_print(). Adjusted all the address printing in the
5205 debugging in store.c to use %p rather than %d.
5207 81. There was a concern that a line of code in smtp_in.c could overflow a
5208 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5209 initially expressed, the concern was not well-founded, because trailing
5210 spaces are removed early. However, if the trailing spaces were followed by
5211 a NULL, they did not get removed, so the overflow was possible. Two fixes
5214 (a) I re-wrote the offending code in a cleaner fashion.
5215 (b) If an incoming SMTP command contains a NULL character, it is rejected
5218 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5219 runs initgroups(), so that if the Exim user is in any additional groups,
5220 they will be used during message reception.
5226 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob