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
40 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
43 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
44 If dropping privileges for untrusted macros, we disabled normal logging
45 on the basis that it would fail; for the Exim run-time user, this is not
46 the case, and it resulted in successful deliveries going unlogged.
47 Fixed. Reported by Andreas Metzler.
53 PP/01 Date: & Message-Id: revert to normally being appended to a message,
54 only prepend for the Resent-* case. Fixes regression introduced in
55 Exim 4.70 by NM/22 for Bugzilla 607.
57 PP/02 Include check_rfc2047_length in configure.default because we're seeing
58 increasing numbers of administrators be bitten by this.
60 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
62 PP/03 Bugzilla 994: added openssl_options main configuration option.
64 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
66 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
68 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
70 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
71 without a peer certificate, leading to a segfault because of an
72 assumption that peers always have certificates. Be a little more
73 paranoid. Problem reported by Martin Tscholak.
75 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
76 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
77 NB: ClamAV planning to remove STREAM in "middle of 2010".
78 CL also introduces -bmalware, various -d+acl logging additions and
79 more caution in buffer sizes.
81 PP/09 Implemented reverse_ip expansion operator.
83 PP/10 Bugzilla 937: provide a "debug" ACL control.
85 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
87 PP/12 Bugzilla 973: Implement --version.
89 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
91 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
93 PP/15 Bugzilla 816: support multiple condition rules on Routers.
95 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
96 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
97 ignore trailing whitespace.
99 JJ/02 prevent non-panic DKIM error from being sent to paniclog
101 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
104 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
105 Notification from Dr Andrew Aitchison.
107 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
108 ExtendedDetectionInfo response format.
109 Notification from John Horne.
111 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
114 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
115 XSL and documented dependency on system catalogs, with examples of how
118 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
121 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
122 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
123 configuration file which is writeable by the Exim user or group.
125 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
126 of configuration files to cover files specified with the -C option if
127 they are going to be used with root privileges, not just the default
130 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
131 option (effectively making it always true).
133 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
134 files to be used while preserving root privileges.
136 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
137 that rogue child processes cannot use them.
139 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
140 run-time user, instead of root.
142 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
143 Exim run-time user without dropping privileges.
145 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
146 result string, instead of calling string_vformat() twice with the same
149 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
150 for other users. Others should always drop root privileges if they use
151 -C on the command line, even for a whitelisted configure file.
153 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
155 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
161 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
162 $data_path, and $header_path variables; fixed documentation bugs and
165 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
166 exipick to access non-standard spools, including the "frozen" queue
169 NM/01 Bugzilla 965: Support mysql stored procedures.
170 Patch from Alain Williams
172 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
174 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
175 Patch from Andreas Metzler
177 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
178 Patch from Kirill Miazine
180 NM/05 Bugzilla 671: Added umask to procmail example.
182 JJ/03 installed exipick 20100323.0, fixing doc bug
184 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
185 directory. Notification and patch from Dan Rosenberg.
187 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
189 TK/02 Improve log output when DKIM signing operation fails.
191 MH/01 Treat the transport option dkim_domain as a colon separated
192 list, not as a single string, and sign the message with each element,
193 omitting multiple occurences of the same signer.
195 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
196 Bugzilla 985, 986. Patch by Simon Arlott
198 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
199 Patch by Simon Arlott
201 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
202 MBX locking. Notification from Dan Rosenberg.
208 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
210 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
212 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
214 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
216 NM/04 Bugzilla 915: Flag broken perl installation during build.
222 TK/01 Added patch by Johannes Berg that expands the main option
223 "spamd_address" if it starts with a dollar sign.
225 TK/02 Write list of recipients to X-Envelope-Sender header when building
226 the mbox-format spool file for content scanning (suggested by Jakob
229 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
230 (http://www.dcc-servers.net/) support via dccifd. Activated by
231 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
233 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
234 by Mark Daniel Reidel <mr@df.eu>.
236 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
237 When building exim an external PCRE library is now needed -
238 PCRE is a system library on the majority of modern systems.
239 See entry on PCRE_LIBS in EDITME file.
241 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
242 conversation. Added nologin parameter to request.
243 Patch contributed by Kirill Miazine.
245 TF/01 Do not log submission mode rewrites if they do not change the address.
247 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
249 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
250 log files in place. Contributed by Roberto Lima.
252 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
254 TF/03 Bugzilla 615: When checking the local_parts router precondition
255 after a local_part_suffix or local_part_prefix option, Exim now
256 does not use the address's named list lookup cache, since this
257 contains cached lookups for the whole local part.
259 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
260 Robert Millan. Documentation is in experimental-spec.txt.
262 TF/04 Bugzilla 668: Fix parallel build (make -j).
264 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
266 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
267 Patch provided by Jan Srzednicki.
269 TF/05 Leading white space used to be stripped from $spam_report which
270 wrecked the formatting. Now it is preserved.
272 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
273 that they are available at delivery time.
275 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
277 TF/08 TLS error reporting now respects the incoming_interface and
278 incoming_port log selectors.
280 TF/09 Produce a more useful error message if an SMTP transport's hosts
281 setting expands to an empty string.
283 NM/06 Bugzilla 744: EXPN did not work under TLS.
284 Patch provided by Phil Pennock.
286 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
287 Patch provided by Richard Godbee.
289 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
290 acl_smtp_notquit, added index entry.
292 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
293 Patch provided by Eugene Bujak.
295 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
296 accept(). Patch provided by Maxim Dounin.
298 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
299 Patch provided by Phil Pennock.
301 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
303 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
304 Patch provided by Brad "anomie" Jorsch.
306 NM/14 Bugzilla 622: Added timeout setting to transport filter.
307 Patch provided by Dean Brooks.
309 TK/05 Add native DKIM support (does not depend on external libraries).
311 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
312 Patch provided by Graeme Fowler.
314 NM/16 Bugzilla 851: Documentation example syntax fix.
316 NM/17 Changed NOTICE file to remove references to embedded PCRE.
318 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
321 NM/19 Bugzilla 745: TLS version reporting.
322 Patch provided by Phil Pennock.
324 NM/20 Bugzilla 167: bool: condition support.
325 Patch provided by Phil Pennock.
327 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
328 clients. Patch provided by Phil Pennock.
330 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
331 Patch provided by Brad "anomie" Jorsch.
333 NM/23 Bugzilla 687: Fix misparses in eximstats.
334 Patch provided by Heiko Schlittermann.
336 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
337 Patch provided by Heiko Schlittermann.
339 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
340 plus update to original patch.
342 NM/26 Bugzilla 799: Documentation correction for ratelimit.
344 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
345 Patch provided by David Brownlee.
347 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
349 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
351 NM/30 Bugzilla 888: TLS documentation bugfixes.
353 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
355 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
356 Unlike the original bugzilla I have changed all shell scripts in src tree.
358 NM/33 Bugzilla 898: Transport filter timeout fix.
359 Patch by Todd Rinaldo.
361 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
362 Patch by Serge Demonchaux.
364 NM/35 Bugzilla 39: Base64 decode bug fixes.
365 Patch by Jakob Hirsch.
367 NM/36 Bugzilla 909: Correct connect() call in dcc code.
369 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
371 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
373 NM/39 Bugzilla 911: Fixed MakeLinks build script.
379 TK/01 Add preliminary DKIM support. Currently requires a forked version of
380 ALT-N's libdkim that I have put here:
381 http://duncanthrax.net/exim-experimental/
383 Note to Michael Haardt: I had to rename some vars in sieve.c. They
384 were called 'true' and it seems that C99 defines that as a reserved
385 keyword to be used with 'bool' variable types. That means you could
386 not include C99-style headers which use bools without triggering
387 build errors in sieve.c.
389 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
390 as mailq or other aliases. Changed the --help handling significantly
391 to do whats expected. exim_usage() emits usage/help information.
393 SC/01 Added the -bylocaldomain option to eximstats.
395 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
397 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
399 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
405 PH/01 Another patch from the Sieve maintainer.
407 PH/02 When an IPv6 address is converted to a string for single-key lookup
408 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
409 dots are used instead of colons so that keys in lsearch files need not
410 contain colons. This was done some time before quoting was made available
411 in lsearch files. However, iplsearch files do require colons in IPv6 keys
412 (notated using the quote facility) so as to distinguish them from IPv4
413 keys. This meant that lookups for IP addresses in host lists did not work
414 for iplsearch lookups.
416 This has been fixed by arranging for IPv6 addresses to be expressed with
417 colons if the lookup type is iplsearch. This is not incompatible, because
418 previously such lookups could never work.
420 The situation is now rather anomolous, since one *can* have colons in
421 ordinary lsearch keys. However, making the change in all cases is
422 incompatible and would probably break a number of configurations.
424 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
427 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
428 conversion specification without a maximum field width, thereby enabling
429 a rogue spamd server to cause a buffer overflow. While nobody in their
430 right mind would setup Exim to query an untrusted spamd server, an
431 attacker that gains access to a server running spamd could potentially
432 exploit this vulnerability to run arbitrary code as the Exim user.
434 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
435 $primary_hostname instead of what libspf2 thinks the hosts name is.
437 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
438 a directory entry by the name of the lookup key. Previously, if a
439 symlink pointed to a non-existing file or a file in a directory that
440 Exim lacked permissions to read, a lookup for a key matching that
441 symlink would fail. Now it is enough that a matching directory entry
442 exists, symlink or not. (Bugzilla 503.)
444 PH/03 The body_linecount and body_zerocount variables are now exported in the
447 PH/04 Added the $dnslist_matched variable.
449 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
450 This means they are set thereafter only if the connection becomes
453 PH/06 Added the client_condition to authenticators so that some can be skipped
454 by clients under certain conditions.
456 PH/07 The error message for a badly-placed control=no_multiline_responses left
457 "_responses" off the end of the name.
459 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
461 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
462 (without spaces) instead of just copying the configuration text.
464 PH/10 Added the /noupdate option to the ratelimit ACL condition.
466 PH/11 Added $max_received_linelength.
468 PH/12 Added +ignore_defer and +include_defer to host lists.
470 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
471 way in which PCRE > 7.0 is built.
473 PH/14 Implemented queue_only_load_latch.
475 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
476 MAIL command. The effect was to mangle the value on 64-bit systems.
478 PH/16 Another patch from the Sieve maintainer.
480 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
482 PH/18 If a system quota error occurred while trying to create the file for
483 a maildir delivery, the message "Mailbox is full" was not appended to the
484 bounce if the delivery eventually timed out. Change 4.67/27 below applied
485 only to a quota excession during the actual writing of the file.
487 PH/19 It seems that peer DN values may contain newlines (and other non-printing
488 characters?) which causes problems in log lines. The DN values are now
489 passed through string_printing() before being added to log lines.
491 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
492 and InterBase are left for another time.)
494 PH/21 Added message_body_newlines option.
496 PH/22 Guard against possible overflow in moan_check_errorcopy().
498 PH/23 POSIX allows open() to be a macro; guard against that.
500 PH/24 If the recipient of an error message contained an @ in the local part
501 (suitably quoted, of course), incorrect values were put in $domain and
502 $local_part during the evaluation of errors_copy.
508 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
509 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
512 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
513 issue a MAIL command.
515 PH/02 In an ACL statement such as
517 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
519 if a client was not listed at all, or was listed with a value other than
520 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
521 the condition was not true (as it should be), so access was not denied.
522 The bug was that the ! inversion was incorrectly passed on to the second
523 item. This has been fixed.
525 PH/03 Added additional dnslists conditions == and =& which are different from
526 = and & when the dns lookup returns more than one IP address.
528 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
529 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
531 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
532 FSYNC, which compiles an option called disable_fsync that allows for
533 bypassing fsync(). The documentation is heavily laced with warnings.
535 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
537 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
538 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
539 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
540 including adding "make clean"; (3) Added -fPIC when compiling the test
541 dynamically loaded module, to get rid of a warning.
543 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
544 message fails, move_frozen_messages = true and ignore_bounce_errors_after
545 = 0s. The bug is otherwise harmless.
547 PH/07 There was a bug in the dovecot authenticator such that the value of
548 $auth1 could be overwritten, and so not correctly preserved, after a
549 successful authentication. This usually meant that the value preserved by
550 the server_setid option was incorrect.
552 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
554 PH/09 Installed PCRE release 7.0.
556 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
557 run for batched SMTP input. It is now run at the start of every message
558 in the batch. While fixing this I discovered that the process information
559 (output by running exiwhat) was not always getting set for -bs and -bS
560 input. This is fixed, and it now also says "batched" for BSMTP.
562 PH/11 Added control=no_pipelining.
564 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
565 patch, slightly modified), and move the expansion of helo_data till after
566 the connection is made in the smtp transport (so it can use these
569 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
571 PH/14 Added log_selector = +pid.
573 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
575 PH/16 Add ${if forany and ${if forall.
577 PH/17 Added dsn_from option to vary the From: line in DSNs.
579 PH/18 Flush SMTP output before performing a callout, unless control =
580 no_callout_flush is set.
582 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
583 was true (the default) a successful delivery failed to delete the retry
584 item, thus causing premature timeout of the address. The bug is now
587 PH/20 Added hosts_avoid_pipelining to the smtp transport.
589 PH/21 Long custom messages for fakedefer and fakereject are now split up
590 into multiline reponses in the same way that messages for "deny" and
591 other ACL rejections are.
593 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
594 with slight modification.
596 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
597 draft, changing the syntax and factoring some duplicate code".
599 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
600 for deliveries of the second and subsequent messages over the same SMTP
603 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
604 ${reduce, with only minor "tidies".
606 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
608 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
609 expansion side effects.
611 PH/27 When a message times out after an over-quota error from an Exim-imposed
612 quota, the bounce message says "mailbox is full". This message was not
613 being given when it was a system quota that was exceeded. It now should
616 MH/03 Made $recipients available in local_scan(). local_scan() already has
617 better access to the recipient list through recipients_list[], but
618 $recipients can be useful in postmaster-provided expansion strings.
620 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
621 in the case of a MAIL command with additional options following the
622 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
623 were accidentally chopped off.
625 PH/29 SMTP synchronization checks are implemented when a command is read -
626 there is a check that no more input is waiting when there shouldn't be
627 any. However, for some commands, a delay in an ACL can mean that it is
628 some time before the response is written. In this time, more input might
629 arrive, invalidly. So now there are extra checks after an ACL has run for
630 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
631 pipelining has not been advertised.
633 PH/30 MH's patch to allow iscntrl() characters to be list separators.
635 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
636 returned in the SMTP response when smtp_return_error_details was false.
639 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
640 instead of the C I/O that was originally supplied, because problems were
643 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
644 Exim which did not show up earlier: it was assuming that a call to
645 SSL_CTX_set_info_callback() might give an error value. In fact, there is
646 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
647 was a macro that became an assignment, so it seemed to work. This has
648 changed to a proper function call with a void return, hence the compile
649 error. Exim's code has been fixed.
651 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
654 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
656 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
658 PH/37 If a message is not accepted after it has had an id assigned (e.g.
659 because it turns out to be too big or there is a timeout) there is no
660 "Completed" line in the log. When some messages of this type were
661 selected by exigrep, they were listed as "not completed". Others were
662 picked up by some special patterns. I have improved the selection
663 criteria to be more general.
665 PH/38 The host_find_failed option in the manualroute router can now be set
666 to "ignore", to completely ignore a host whose IP address cannot be
667 found. If all hosts are ignored, the behaviour is controlled by the new
668 host_all_ignored option.
670 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
671 homing or because of multiple MX records with /mx) generated more than
672 one IP address, and the following item turned out to be the local host,
673 all the secondary addresses of the first item were incorrectly removed
674 from the list, along with the local host and any following hosts (which
675 is what is supposed to happen).
677 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
678 whoever called Exim into the -H file. In the case of the daemon it was
679 behaving confusingly. When first started, it used values for whoever
680 started the daemon, but after a SIGHUP it used the Exim user (because it
681 calls itself on a restart). I have changed the code so that it now always
684 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
685 message are rejected with the same error (e.g. no authentication or bad
686 sender address), and a DATA command is nevertheless sent (as can happen
687 with PIPELINING or a stupid MUA), the error message that was given to the
688 RCPT commands is included in the rejection of the DATA command. This is
689 intended to be helpful for MUAs that show only the final error to their
692 PH/42 Another patch from the Sieve maintainer.
694 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
695 Eximstats - Fixed some broken HTML links and added missing column headers
697 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
698 columns for Rejects, Temp Rejects, Ham, and Spam rows.
700 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
702 PH/43 Yet another patch from the Sieve maintainer.
704 PH/44 I found a way to check for a TCP/IP connection going away before sending
705 the response to the final '.' that terminates a message, but only in the
706 case where the client has not sent further data following the '.'
707 (unfortunately, this is allowed). However, in many cases there won't be
708 any further data because there won't be any more messages to send. A call
709 to select() can be used: if it shows that the input is "ready", there is
710 either input waiting, or the socket has been closed. An attempt to read
711 the next input character can distinguish the two cases. Previously, Exim
712 would have sent an OK response which the client would never have see.
713 This could lead to message repetition. This fix should cure that, at
714 least in a lot of common cases.
716 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
717 advertised in response to EHLO.
723 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
724 fixed by 4.65/MH/01 (is this a record?) are fixed:
726 (i) An empty string was always treated as zero by the numeric comparison
727 operators. This behaviour has been restored.
729 (ii) It is documented that the numeric comparison operators always treat
730 their arguments as decimal numbers. This was broken in that numbers
731 starting with 0 were being interpreted as octal.
733 While fixing these problems I realized that there was another issue that
734 hadn't been noticed. Values of message_size_limit (both the global option
735 and the transport option) were treated as octal if they started with 0.
736 The documentation was vague. These values are now always treated as
737 decimal, and I will make that clear in the documentation.
743 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
744 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
747 MH/01 Don't check that the operands of numeric comparison operators are
748 integers when their expansion is in "skipping" mode (fixes bug
749 introduced by 4.64-PH/07).
751 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
752 child addresses, Exim now panics and dies. Previously, because the count
753 is held in a short int, deliveries were likely to be lost. As such a
754 large number of recipients for a single message is ridiculous
755 (performance will be very, very poor), I have chosen to impose a limit
756 rather than extend the field.
762 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
763 leftover -K file (the existence of which was triggered by #402).
764 While we were at it, introduced process PID as part of the -K
765 filename. This should rule out race conditions when creating
768 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
769 processing considerably. Previous code took too long for large mails,
770 triggering a timeout which in turn triggers #401.
772 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
773 in the DK code in transports.c. sendfile() is not really portable,
774 hence the _LINUX specificness.
776 TF/01 In the add_headers option to the mail command in an Exim filter,
777 there was a bug that Exim would claim a syntax error in any
778 header after the first one which had an odd number of characters
781 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
782 callout verification, Exim cached a "reject" for the entire domain. This
783 is correct for most verifications, but it is not correct for a recipient
784 verification with use_sender or use_postmaster set, because in that case
785 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
786 case of MAIL FROM:<> rejection from other early rejections (e.g.
787 rejection of HELO). When verifying a recipient using a non-null MAIL
788 address, the cache is ignored if it shows MAIL FROM:<> rejection.
789 Whatever the result of the callout, the value of the domain cache is
790 left unchanged (for any other kind of callout, getting as far as trying
791 RCPT means that the domain itself is ok).
793 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
796 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
797 manifest itself as EPIPE rather than ECONNECT. When tidying away a
798 session, the daemon ignores ECONNECT errors and logs others; it now
799 ignores EPIPE as well.
801 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
802 (quoted-printable decoding).
804 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
805 later the small subsequent patch to fix an introduced bug.
807 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
809 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
811 PH/08 An error is now given if message_size_limit is specified negative.
813 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
814 to be given (somewhat) arbitrary names.
816 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
819 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
820 miscellaneous code fixes
822 PH/10 Added the log_reject_target ACL modifier to specify where to log
825 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
826 hostname. This is wrong, because it relates to the incoming message (and
827 probably the interface on which it is arriving) and not to the outgoing
828 callout (which could be using a different interface). This has been
829 changed to use the value of the helo_data option from the smtp transport
830 instead - this is what is used when a message is actually being sent. If
831 there is no remote transport (possible with a router that sets up host
832 addresses), $smtp_active_hostname is used.
834 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
835 tweaks were necessary in order to get it to work (see also 21 below):
836 (a) The code assumed that strncpy() returns a negative number on buffer
837 overflow, which isn't the case. Replaced with Exim's string_format()
839 (b) There were several signed/unsigned issues. I just did the minimum
840 hacking in of casts. There is scope for a larger refactoring.
841 (c) The code used strcasecmp() which is not a standard C function.
842 Replaced with Exim's strcmpic() function.
843 (d) The code set only $1; it now sets $auth1 as well.
844 (e) A simple test gave the error "authentication client didn't specify
845 service in request". It would seem that Dovecot has changed its
846 interface. Fortunately there's a specification; I followed it and
847 changed what the client sends and it appears to be working now.
849 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
852 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
853 address A is aliased to B and C, where B exists and C does not. Without
854 -v the output is "A verified" because verification stops after a
855 successful redirection if more than one address is generated. However,
856 with -v the child addresses are also verified. Exim was outputting "A
857 failed to verify" and then showing the successful verification for C,
858 with its parentage. It now outputs "B failed to verify", showing B's
859 parentage before showing the successful verification of C.
861 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
862 look up a TXT record in a specific list after matching in a combined
865 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
866 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
867 they consult the DNS. I had assumed they would set it the way they
868 wanted; and indeed my experiments on Linux seem to show that in some
869 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
870 To be on the safe side, however, I have now made the interface to
871 host_find_byname() similar to host_find_bydns(), with an argument
872 containing the DNS resolver options. The host_find_byname() function now
873 sets these options at its start, just as host_find_bydns() does. The smtp
874 transport options dns_qualify_single and dns_search_parents are passed to
875 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
876 of host_find_byname() use the default settings of RES_DEFNAMES
877 (qualify_single) but not RES_DNSRCH (search_parents).
879 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
880 spool_read_header() do less string testing, by means of a preliminary
881 switch on the second character of optional "-foo" lines. (This is
882 overdue, caused by the large number of possibilities that now exist.
883 Originally there were few.) While I was there, I also converted the
884 str(n)cmp tests so they don't re-test the leading "-" and the first
885 character, in the hope this might squeeze out yet more improvement.
887 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
888 flag allowing group syntax was set by the header_syntax check but not
889 turned off, possible causing trouble later; (2) The flag was not being
890 set at all for the header_verify test, causing "group"-style headers to
891 be rejected. I have now set it in this case, and also caused header_
892 verify to ignore an empty address taken from a group. While doing this, I
893 came across some other cases where the code for allowing group syntax
894 while scanning a header line wasn't quite right (mostly, not resetting
895 the flag correctly in the right place). These bugs could have caused
896 trouble for malformed header lines. I hope it is now all correct.
898 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
899 with the "reply" argument non-NULL. The code, however (which originally
900 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
901 but it didn't always do it. This confused somebody who was copying the
902 code for some other use. I have removed all the tests.
904 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
905 feature that was used to support insecure browsers during the U.S. crypto
906 embargo. It requires special client support, and Exim is probably the
907 only MTA that supported it -- and would never use it because real RSA is
908 always available. This code has been removed, because it had the bad
909 effect of slowing Exim down by computing (never used) parameters for the
910 RSA_EXPORT functionality.
912 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
913 authenticator to fail if there's a tab character in the incoming data
914 (there should never be unless someone is messing about, as it's supposed
915 to be base64-encoded). Also added, on Timo's advice, the "secured" option
916 if the connection is using TLS or if the remote IP is the same as the
917 local IP, and the "valid-client-cert option" if a client certificate has
920 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
921 authenticators. This can be used for authorization after authentication
922 succeeds. (In the case of plaintext, it servers for both authentication
925 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
926 if any retry times were supplied.
928 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
929 connection, where there is no client IP address to check. In this
930 situation, the verify now always succeeds.
932 PH/25 Applied John Jetmore's -Mset patch.
934 PH/26 Added -bem to be like -Mset, but loading a message from a file.
936 PH/27 In a string expansion for a processed (not raw) header when multiple
937 headers of the same name were present, leading whitespace was being
938 removed from all of them, but trailing whitespace was being removed only
939 from the last one. Now trailing whitespace is removed from each header
940 before concatenation. Completely empty headers in a concatenation (as
943 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
944 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
946 PH/29 [Removed. This was a change that I later backed out, and forgot to
947 correct the ChangeLog entry (that I had efficiently created) before
948 committing the later change.]
950 PH/30 Exim was sometimes attempting to deliver messages that had suffered
951 address errors (4xx response to RCPT) over the same connection as other
952 messages routed to the same hosts. Such deliveries are always "forced",
953 so retry times are not inspected. This resulted in far too many retries
954 for the affected addresses. The effect occurred only when there were more
955 hosts than the hosts_max_try setting in the smtp transport when it had
956 the 4xx errors. Those hosts that it had tried were not added to the list
957 of hosts for which the message was waiting, so if all were tried, there
958 was no problem. Two fixes have been applied:
960 (i) If there are any address or message errors in an SMTP delivery, none
961 of the hosts (tried or untried) are now added to the list of hosts
962 for which the message is waiting, so the message should not be a
963 candidate for sending over the same connection that was used for a
964 successful delivery of some other message. This seems entirely
965 reasonable: after all the message is NOT "waiting for some host".
966 This is so "obvious" that I'm not sure why it wasn't done
967 previously. Hope I haven't missed anything, but it can't do any
968 harm, as the worst effect is to miss an optimization.
970 (ii) If, despite (i), such a delivery is accidentally attempted, the
971 routing retry time is respected, so at least it doesn't keep
972 hammering the server.
974 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
975 in ${readsocket because some servers need this prod.
977 PH/32 Added some extra debug output when updating a wait-xxx database.
979 PH/33 The hint "could be header name not terminated by colon", which has been
980 given for certain expansion errors for a long time, was not being given
981 for the ${if def:h_colon_omitted{... case.
983 PH/34 The spec says: "With one important exception, whenever a domain list is
984 being scanned, $domain contains the subject domain." There was at least
985 one case where this was not true.
987 PH/35 The error "getsockname() failed: connection reset by peer" was being
988 written to the panic log as well as the main log, but it isn't really
989 panic-worthy as it just means the connection died rather early on. I have
990 removed the panic log writing for the ECONNRESET error when getsockname()
993 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
994 runs only) independently of the message's sender address. This meant
995 that, if the 4xx error was in fact related to the sender, a different
996 message to the same recipient with a different sender could confuse
997 things. In particualar, this can happen when sending to a greylisting
998 server, but other circumstances could also provoke similar problems.
999 I have changed the default so that the retry time for these errors is now
1000 based a combination of the sender and recipient addresses. This change
1001 can be overridden by setting address_retry_include_sender=false in the
1004 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1005 remote server are returned as part of bounce messages. This was not
1006 happening for LMTP over a pipe (the lmtp transport), but now it is the
1007 same for both kinds of LMTP.
1009 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1010 in header lines that were in fact CNAMEs. This is no longer the case.
1012 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1013 and queue runs started by the daemon processed all messages. This has
1014 been fixed so that -R and -S can now usefully be given with -q<time>.
1016 PH/40 Import PCRE release 6.7 (fixes some bugs).
1018 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1020 PH/42 Give an error if -q is specified more than once.
1022 PH/43 Renamed the variables $interface_address and $interface_port as
1023 $received_ip_address and $received_port, to make it clear that these
1024 values apply to message reception, and not to the outgoing interface when
1025 a message is delivered. (The old names remain recognized, of course.)
1027 PH/44 There was no timeout on the connect() call when using a Unix domain
1028 socket in the ${readsocket expansion. There now is.
1030 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1031 be meaningful with "accept".
1033 SC/01 Eximstats V1.43
1034 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1036 SC/02 Eximstats V1.44
1037 Use a glob alias rather than an array ref in the generated
1038 parser. This improves both readability and performance.
1040 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1041 Collect SpamAssassin and rejection statistics.
1042 Don't display local sender or destination tables unless
1043 there is data to show.
1044 Added average volumes into the top table text output.
1046 SC/04 Eximstats V1.46
1047 Collect data on the number of addresses (recipients)
1048 as well as the number of messages.
1050 SC/05 Eximstats V1.47
1051 Added 'Message too big' to the list of mail rejection
1052 reasons (thanks to Marco Gaiarin).
1054 SC/06 Eximstats V1.48
1055 Mainlog lines which have GMT offsets and are too short to
1056 have a flag are now skipped.
1058 SC/07 Eximstats V1.49 (Alain Williams)
1059 Added the -emptyok flag.
1061 SC/08 Eximstats V1.50
1062 Fixes for obtaining the IP address from reject messages.
1064 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1065 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1066 whitesspace changes from 4.64-PH/27
1068 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1071 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1072 are found, allow negative numbers in numeric criteria)
1074 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1076 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1077 to match changes made in 4.64-PH/43
1079 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1081 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1082 "domain" is actually an IP address, and force a failure. This locks out
1083 those revolvers/nameservers that support "A-for-A" lookups, in
1084 contravention of the specifications.
1086 PH/48 When a host name was looked up from an IP address, and the subsequent
1087 forward lookup of the name timed out, the host name was left in
1088 $sender_host_name, contrary to the specification.
1090 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1091 restricted to single-key lookups, Exim was not diagnosing an error if
1092 * or *@ was used with a query-style lookup.
1094 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1096 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1097 long ago, but noone interested enough thought of it. Let's just say that
1098 the "1.1" means that there are some new functions that weren't there at
1099 some point in the past.
1101 PH/51 Error processing for expansion failure of helo_data from an smtp
1102 transport during callout processing was broken.
1104 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1105 tested/used via the -bh/-bhc/-bs options.
1107 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1108 bug, fixed in subsequent PCRE releases).
1110 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1111 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1113 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1119 SC/01 Use a glob alias rather than an array ref in eximstats generated
1120 parser. This improves both readability and performance.
1122 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1123 Don't display local sender or destination tables in eximstats unless
1124 there is data to show.
1125 Added average volumes into the eximstats top table text output.
1127 SC/03 Collect data on the number of addresses (recipients) as well
1128 as the number of messages in eximstats.
1130 TF/01 Correct an error in the documentation for the redirect router. Exim
1131 does (usually) call initgroups() when daemonizing.
1133 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1134 with consistent privilege compared to when running as a daemon.
1136 TF/03 Note in the spec that $authenticated_id is not set for local
1137 submissions from trusted users.
1139 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1140 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1142 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1143 by adding some example configuration directives to the default
1144 configuration file. A little bit of work is required to uncomment the
1145 directives and define how usernames and passwords are checked, but
1146 there is now a framework to start from.
1148 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1149 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1150 without this. I don't know how relevant this is to other LDAP libraries.
1152 PH/02 Add the verb name to the "unknown ACL verb" error.
1154 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1156 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1158 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1159 directory not expanded when it should be if an expanded home directory
1160 was set for the address (which is overridden by the transport).
1162 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1165 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1166 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1167 because it is too late at that time, and has no effect.
1169 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1170 security issue with \' (bugzilla #107). I could not use the
1171 PQescapeStringConn() function, because it needs a PGconn value as one of
1174 PH/08 When testing addresses using -bt, indicate those final addresses that
1175 are duplicates that would not cause an additional delivery. At least one
1176 person was confused, thinking that -bt output corresponded to deliveries.
1177 (Suppressing duplicates isn't a good idea as you lose the information
1178 about possibly different redirections that led to the duplicates.)
1180 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1181 systems where poll() doesn't work, in particular OS X.
1183 PH/10 Added more information to debugging output for retry time not reached.
1185 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1186 operations in malware.c.
1188 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1191 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1192 syslog_duplication set false, log lines that would normally be written
1193 both the the main log and to the reject log were not written to syslog at
1196 PH/14 In the default configuration, change the use of "message" in ACL warn
1197 statements to "add_header".
1199 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1200 not followed by a command (e.g. "seen endif").
1202 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1203 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1206 PH/17 Added extra conditions to the default value of delay_warning_condition
1210 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1211 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1212 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1215 The Auto-Submitted: and various List- headers are standardised, whereas I
1216 don't think Precedence: ever was.
1218 PH/18 Refactored debugging code in route_finduser() to show more information,
1219 in particular, the error code if getpwnam() issues one.
1221 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1222 This is apparently needed in addition to the PH/07 change above to avoid
1223 any possible encoding problems.
1225 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1226 but not after initializing Perl.
1228 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1229 output them only if debugging. By default they are written stderr,
1230 apparently, which is not desirable.
1232 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1235 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1238 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1240 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1241 authenticated or an ident call has been made. Suppress the default
1242 values for $authenticated_id and $authenticated_sender (but permit -oMai
1243 and -oMas) when testing with -bh.
1245 PH/24 Re-jigged the order of the tests in the default configuration so that the
1246 tests for valid domains and recipients precede the DNS black list and CSA
1247 tests, on the grounds that those ones are more expensive.
1249 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1250 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1251 This bug exists in every version of Exim that I still have, right back to
1254 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1255 However, an attempt to turn on case-sensitivity in a regex key by
1256 including (?-i) didn't work because the subject string was already
1257 lowercased, and the effects were non-intuitive. It turns out that a
1258 one-line patch can be used to allow (?-i) to work as expected.
1264 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1265 other effects) broke the use of negated acl sub-conditions.
1267 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1270 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1271 "Deny" causes Exim to reject the incoming connection with a 554 error.
1272 Unfortunately, if there is a major crisis, such as a disk failure,
1273 tcp-wrappers gives "deny", whereas what one would like would be some
1274 kind of temporary error. A kludge has been added to help with this.
1275 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1276 554 error is used if errno is still zero or contains ENOENT (which occurs
1277 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1280 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1282 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1283 errors. Otherwise a message that provokes a temporary error (when other
1284 messages do not) can cause a whole host to time out.
1286 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1287 addresses were routed directly to files or pipes from a redirect router.
1288 File deliveries just didn't batch; pipe deliveries might have suffered
1291 PH/06 A failure to get a lock for a hints database would erroneously always say
1292 "Failed to get write lock", even when it was really a read lock.
1294 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1295 of 0600. This has been changed to use the value of the lockfile_mode
1296 option (which defaults to 0600).
1298 PH/08 Applied small patch from the Sieve maintainer.
1300 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1301 folder from quota calculations, a direct delivery into this folder messed
1302 up the contents of the maildirsize file. This was because the regex was
1303 used only to exclude .Trash (or whatever) when the size of the mailbox
1304 was calculated. There was no check that a delivery was happening into an
1305 excluded directory. This bug has been fixed by ignoring all quota
1306 processing for deliveries into excluded directories.
1308 PH/10 Added the maildirfolder_create_regex option to appendfile.
1314 PH/01 The code for finding all the local interface addresses on a FreeBSD
1315 system running IPv6 was broken. This may well have applied to all BSD
1316 systems, as well as to others that have similar system calls. The broken
1317 code found IPv4 interfaces correctly, but gave incorrect values for the
1318 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1319 that it would not match correctly against @[] and not recognize the IPv6
1322 PH/02 The ipliteral router was not recognizing addresses of the form user@
1323 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1325 PH/03 Added disable_ipv6.
1327 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1328 decoded form, because it is most often used to construct To: headers
1329 lines in autoreplies, and the decoded form may well be syntactically
1330 invalid. However, $reply_address has leading white space removed, and all
1331 newlines turned into spaces so that the autoreply transport does not
1334 PH/05 If group was specified without a user on a router, and no group or user
1335 was specified on a transport, the group from the router was ignored.
1337 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1338 for visible compile-time settings that can be used to change these
1339 numbers, for those that want even more. Backwards compatibility with old
1340 spool files has been maintained. However, going back to a previous Exim
1341 release will lost any variables that are in spool files.
1343 PH/07 Two small changes when running in the test harness: increase delay when
1344 passing a TCP/IP connection to a new process, in case the original
1345 process has to generate a bounce, and remove special handling of
1346 127.0.0.2 (sic), which is no longer necessary.
1348 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1349 be the same on different OS.
1351 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1354 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1355 whether --show-vars was specified or not
1357 JJ/02 exipick: Added support for new ACL variable spool format introduced
1360 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1361 syntactically invalid From: or Reply-to: line, and a filter used this to
1362 generate an autoreply, and therefore failed to obtain an address for the
1363 autoreply, Exim could try to deliver to a non-existent relative file
1364 name, causing unrelated and misleading errors. What now happens is that
1365 it logs this as a hard delivery error, but does not attempt to create a
1368 PH/11 The exinext utility has a -C option for testing purposes, but although
1369 the given file was scanned by exinext itself; it wasn't being passed on
1370 when Exim was called.
1372 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1373 an end-of-file indication when reading a command response.
1375 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1376 compiled. In many other places in Exim, IPv6 addresses are always
1377 recognized, so I have changed this. It also means that IPv4 domain
1378 literals of the form [IPV4:n.n.n.n] are now always recognized.
1380 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1381 used if the router is not running as root, for example, when verifying at
1382 ACL time, or when using -bh. The debugging output from this situation was
1383 non-existent - all you got was a failure to exec. I have made two
1386 (a) Failures to set uid/gid, the current directory, or a process leader
1387 in a subprocess such as that created by queryprogram now generate
1388 suitable debugging ouput when -d is set.
1390 (b) The queryprogram router detects when it is not running as root,
1391 outputs suitable debugging information if -d is set, and then runs
1392 the subprocess without attempting to change uid/gid.
1394 PH/15 Minor change to Makefile for building test_host (undocumented testing
1397 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1398 additional section of a DNS packet that returns MX or SRV records.
1399 Instead, it always explicitly searches for A/AAAA records. This avoids
1400 major problems that occur when a DNS server includes only records of one
1401 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1402 fixed another bug: if SRV records were looked up and the corresponding
1403 address records were *not* found in the additional section, the port
1404 values from the SRV records were lost.
1406 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1407 using the correct key (the original address) when searching the retry
1408 rules in order to find which one to use for generating the retry hint.
1410 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1411 adding the default one. Similarly, if it contains a Reply-To: header, the
1412 errors_reply_to option, if set, is not used.
1414 PH/19 When calculating a retry time, Exim used to measure the "time since
1415 failure" by looking at the "first failed" field in the retry record. Now
1416 it does not use this if it is later than than the arrival time of the
1417 message. Instead it uses the arrival time. This makes for better
1418 behaviour in cases where some deliveries succeed, thus re-setting the
1419 "first failed" field. An example is a quota failure for a huge message
1420 when small messages continue to be delivered. Without this change, the
1421 "time since failure" will always be short, possible causing more frequent
1422 delivery attempts for the huge message than are intended.
1423 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1425 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1426 $1, $2, $3) because the numerical variables can be reset during some
1427 expansion items (e.g. "match"), thereby losing the authentication data.
1429 PH/21 Make -bV show the size of off_t variables so that the test suite can
1430 decide whether to run tests for quotas > 2G.
1432 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1433 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1434 filecount value is greater than 2G or if a quota value is greater than 2G
1435 on a system where the size of off_t is not greater than 4, a panic error
1438 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1439 never match. The debug and -bh output now contains an explicit error
1440 message indicating a malformed IPv4 address or mask.
1442 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1443 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1444 PH/23 above applies.
1446 PH/25 Do not write to syslog when running in the test harness. The only
1447 occasion when this arises is a failure to open the main or panic logs
1448 (for which there is an explicit test).
1450 PH/26 Added the /no_tell option to "control=freeze".
1452 PH/27 If a host name lookup failed very early in a connection, for example, if
1453 the IP address matched host_lookup and the reverse lookup yielded a name
1454 that did not have a forward lookup, an error message of the form "no IP
1455 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1456 could be logged. Now it outputs the IP address instead of "NULL".
1458 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1459 allows the sender and the authenticated sender to be set when
1460 submitting a message from within Exim. Since child_open_exim() is
1461 documented for local_scan(), the new function should be too.
1463 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1464 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1465 results in an empty string is now treated as unset.
1467 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1469 PH/31 Added sender_verify_fail logging option.
1471 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1472 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1473 tidied the source and removed it altogether.
1475 PH/33 When a queue run was abandoned because the load average was too high, a
1476 log line was always written; now it is written only if the queue_run log
1477 selector is set. In addition, the log line for abandonment now contains
1478 information about the queue run such as the pid. This is always present
1479 in "start" and "stop" lines but was omitted from the "abandon" line.
1481 PH/34 Omit spaces between a header name and the colon in the error message that
1482 is given when verify = headers_syntax fails (if there are lots of them,
1483 the message gets confusing).
1485 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1486 names, as there are now some PTR records that contain slashes. This check
1487 is only to protect against broken name servers that fall over on strange
1488 characters, so the fact that it applies to all lookups doesn't matter.
1490 PH/36 Now that the new test suite is complete, we can remove some of the
1491 special code in Exim that was needed for the old test suite. For example,
1492 sorting DNS records because real resolvers return them in an arbitrary
1493 order. The new test suite's fake resolver always returns records in the
1496 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1497 bounces) except when queue_only is set, to avoid logging races between
1498 the different processes.
1500 PH/38 Panic-die if .include specifies a non-absolute path.
1502 PH/39 A tweak to the "H" retry rule from its user.
1504 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1505 a label. They prevented compilation on older perls.
1507 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1508 a warning to be raised on newish perls.
1510 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1511 on queue. Changes to match documented behaviour of showing count of
1512 messages matching specified criteria.
1514 PH/40 Changed the default ident timeout from 30s to 5s.
1516 PH/41 Added support for the use of login_cap features, on those BSD systems
1517 that have them, for controlling the resources used by pipe deliveries.
1519 PH/42 The content-scanning code uses fopen() to create files in which to put
1520 message data. Previously it was not paying any attention to the mode of
1521 the files. Exim runs with umask(0) because the rest of the code creates
1522 files with open(), and sets the required mode explicitly. Thus, these
1523 files were ending up world-writeable. This was not a big issue, because,
1524 being within the spool directory, they were not world-accessible. I have
1525 created a function called modefopen, which takes an additional mode
1526 argument. It sets umask(777), creates the file, chmods it to the required
1527 mode, then resets the umask. All the relevant calls to fopen() in the
1528 content scanning code have been changed to use this function.
1530 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1531 to 24 hours. This avoids potential overflow problems when processing G
1532 and H retry rules. I suspect nobody ever tinkers with this value.
1534 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1536 PH/45 When the plaintext authenticator is running as a client, the server's
1537 challenges are checked to ensure they are valid base64 strings. By
1538 default, the authentication attempt is cancelled if an invalid string is
1539 received. Setting client_ignore_invalid_base64 true ignores these errors.
1540 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1541 they are received. Thus, the responses can be made to depend on the
1542 challenges. If an invalid string is ignored, an empty string is placed in
1545 PH/46 Messages that are created by the autoreply transport now contains a
1546 References: header, in accordance with RFCs 2822 and 3834.
1548 PH/47 Added authenticated_sender_force to the smtp transport.
1550 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1552 PH/49 Installed latest patch from the Sieve maintainer.
1554 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1555 was also set, the appendfile transport was unnecessarily scanning a
1556 directory of message files (e.g. for maildir delivery) to find the count
1557 of files (along with the size), even though it did not need this
1558 information. It now does the scan only if it needs to find either the
1559 size of the count of files.
1561 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1563 PH/52 Two bugs concerned with error handling when the smtp transport is
1566 (i) Exim was not creating retry information for temporary errors given
1567 for individual recipients after the DATA command when the smtp transport
1568 was used in LMTP mode. This meant that they could be retried too
1569 frequently, and not timed out correctly.
1571 (ii) Exim was setting the flag that allows error details to be returned
1572 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1573 recipients that were returned after the DATA command.
1575 PH/53 This is related to PH/52, but is more general: for any failing address,
1576 when detailed error information was permitted to be returned to the
1577 sender, but the error was temporary, then after the final timeout, only
1578 "retry timeout exceeded" was returned. Now it returns the full error as
1579 well as "retry timeout exceeded".
1581 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1582 do this, and (what is worse) MTAs that accept it.
1584 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1585 will now be deprecated.
1587 PH/56 New os.c-cygwin from the Cygwin maintainer.
1589 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1590 formats (previously only available in exim formats via -bpr, -bpru,
1591 and -bpra. Now also available in native and exiqgrep formats)
1593 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1594 with very large, slow to parse queues
1596 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1598 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1600 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1601 responses to authentication challenges, though it was showing the
1602 challenges; (ii) I've removed the CR characters from the debug output for
1605 PH/58 Allow for the insertion of a newline as well as a space when a string
1606 is turned into more than one encoded-word during RFC 2047 encoding. The
1607 Sieve code now uses this.
1609 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1610 data_4xx, lost_connection, tls_required.
1612 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1613 message was being sent as an SMTP response.
1615 PH/61 Add -l and -k options to exicyclog.
1617 PH/62 When verifying, if an address was redirected to one new address, so that
1618 verification continued, and the new address failed or deferred after
1619 having set something in $address_data, the value of $address_data was not
1620 passed back to the ACL. This was different to the case when no
1621 redirection occurred. The value is now passed back in both cases.
1623 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1624 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1625 use login_cap.h, so on its own it isn't the distinguishing feature. The
1626 new name refers directly to the setclassresources() function.
1628 PH/65 Added configuration files for NetBSD3.
1630 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1632 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1633 is preferred over IPv4.
1635 PH/68 The bounce_return_message and bounce_return_body options were not being
1636 honoured for bounces generated during the reception of non-SMTP messages.
1637 In particular, this applied to messages rejected by the ACL. This bug has
1638 been fixed. However, if bounce_return_message is true and bounce_return_
1639 body is false, the headers that are returned for a non-SMTP message
1640 include only those that have been read before the error was detected.
1641 (In the case of an ACL rejection, they have all been read.)
1643 PH/69 The HTML version of the specification is now built in a directory called
1644 spec_html instead of spec.html, because the latter looks like a path with
1645 a MIME-type, and this confuses some software.
1647 PH/70 Catch two compiler warnings in sieve.c.
1649 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1650 function verify_get_ident() calls ip_connect() to connect a socket, but
1651 if the "connect()" function timed out, ip_connect() used to close the
1652 socket. However, verify_get_ident() also closes the socket later, and in
1653 between Exim writes to the log, which may get opened at this point. When
1654 the socket was closed in ip_connect(), the log could get the same file
1655 descriptor number as the socket. This naturally causes chaos. The fix is
1656 not to close the socket in ip_connect(); the socket should be closed by
1657 the function that creates it. There was only one place in the code where
1658 this was missing, in the iplookup router, which I don't think anybody now
1659 uses, but I've fixed it anyway.
1661 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1662 well as to direct DNS lookups. Otherwise the handling of names in host
1663 lists is inconsistent and therefore confusing.
1669 PH/01 Two changes to the default runtime configuration:
1671 (1) Move the checks for relay_from_hosts and authenticated clients from
1672 after to before the (commented out) DNS black list checks.
1674 (2) Add control=submission to the relay_from_hosts and authenticated
1675 clients checks, on the grounds that messages accepted by these
1676 statements are most likely to be submissions.
1678 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1680 (1) Generate an error if the third argument for the ${prvs expansion is
1683 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1686 (3) Reset the variables that are obtained from the first argument of
1687 ${prvscheck and used in the second argument before leaving the code,
1688 because their memory is reclaimed, so using them afterwards may do
1691 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1692 one (it's much easier than Tom thought :-).
1694 (5) Because of (4), we can now allow for the use of $prvscheck_result
1695 inside the third argument.
1697 PH/03 For some reason, the default setting of PATH when running a command from
1698 a pipe transport was just "/usr/bin". I have changed it to
1701 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1702 anything to be listed in the output from -bV.
1704 PH/05 When a filter generated an autoreply, the entire To: header line was
1705 quoted in the delivery log line, like this:
1707 => >A.N.Other <ano@some.domain> <original@ddress> ...
1709 This has been changed so that it extracts the operative address. There
1710 may be more than one such address. If so, they are comma-separated, like
1713 => >ano@some.domain,ona@other.domain <original@ddress> ...
1715 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1716 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1717 not being looked up in the rDNS to get a host name, Exim was showing the
1718 IP address twice in Received: lines, even though the IP addresses were
1719 identical. For example:
1721 Received: from [1.2.3.4] (helo=[1.2.3.4])
1723 However, if the real host name was known, it was omitting the HELO data
1724 if it matched the actual IP address. This has been tidied up so that it
1725 doesn't show the same IP address twice.
1727 PH/07 When both +timestamp and +memory debugging was on, the value given by
1728 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1729 called by the debug printing, thereby overwriting the timestamp buffer.
1730 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1732 PH/08 When the original message was included in an autoreply transport, it
1733 always said "this is a copy of the message, including all the headers",
1734 even if body_only or headers_only was set. It now gives an appropriate
1737 PH/09 Applied a patch from the Sieve maintainer which:
1739 o fixes some comments
1740 o adds the (disabled) notify extension core
1741 o adds some debug output for the result of if/elsif tests
1742 o points to the current vacation draft in the documentation
1743 and documents the missing references header update
1747 o fixes a bug in processing the envelope test (when testing
1748 multiple envelope elements, the last element determinted the
1751 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1752 Electronic Mail") by including:
1754 Auto-submitted: auto-generated
1756 in the messages that it generates (bounce messages and others, such as
1757 warnings). In the case of bounce messages for non-SMTP mesages, there was
1758 also a typo: it was using "Auto_submitted" (underscore instead of
1759 hyphen). Since every message generated by Exim is necessarily in response
1760 to another message, thes have all been changed to:
1762 Auto-Submitted: auto-replied
1764 in accordance with these statements in the RFC:
1766 The auto-replied keyword:
1768 - SHOULD be used on messages sent in direct response to another
1769 message by an automatic process,
1771 - MUST NOT be used on manually-generated messages,
1773 - MAY be used on Delivery Status Notifications (DSNs) and Message
1774 Disposition Notifications (MDNs),
1776 - MUST NOT be used on messages generated by automatic or periodic
1777 processes, except for messages which are automatic responses to
1780 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1781 to the default Received: header definition.
1783 PH/12 Added log selector acl_warn_skipped (default on).
1785 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1786 variables because (a) they are in the wrong storage pool and (b) even if
1787 they were copied, it wouldn't work properly because of the caching.
1789 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1790 checking when decoding. Apparently there are clients that generate
1791 overlong encoded strings. Why am I not surprised?
1793 PH/15 If the first argument of "${if match_address" was not empty, but did not
1794 contain an "@" character, Exim crashed. Now it writes a panic log message
1795 and treats the condition as false.
1797 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1799 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1800 extension "envelope-auth". The code is finished and in agreement with
1801 other implementations, but there is no documentation so far and in fact,
1802 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1803 not changing the active code.
1805 Print executed "if" and "elsif" statements when debugging is used. This
1806 helps a great deal to understand what a filter does.
1808 Document more things not specified clearly in RFC3028. I had all this
1809 sorted out, when out of a sudden new issues came to my mind. Oops."
1811 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1814 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1815 canonical form (as documented). However, after a host name lookup from
1816 the IP address, check_host() was doing a simple string comparison with
1817 addresses acquired from the DNS when checking that the found name did
1818 have the original IP as one of its addresses. Since any found IPv6
1819 addresses are likely to be in abbreviated form, the comparison could
1820 fail. Luckily, there already exists a function for doing the comparison
1821 by converting both addresses to binary, so now that is used instead of
1822 the text comparison.
1824 PH/20 There was another similar case to PH/19, when a complete host name was
1825 given in a host list; looking up its IP address could give an abbreviated
1826 form, whereas the current host's name might or might not be abbreviated.
1827 The same fix has been applied.
1833 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1834 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1837 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1838 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1840 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1842 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1843 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1844 Also, for "Auto-Submitted", treat anything other than "no" as
1845 non-personal, in accordance with RFC 3834. (Previously it treated
1846 anything starting "auto-" as non-personal.)
1848 TF/01 The control=submission/name=... option had a problem with syntax
1849 errors if the name included a slash character. The /name= option
1850 now slurps the rest of the string, so it can include any characters
1851 but it must come last in the list of options (after /sender_retain
1854 PH/05 Some modifications to the interface to the fake nameserver for the new
1862 TK/01 Added the "success_on_redirect" address verification option. See
1863 NewStuff for rationale and an example.
1865 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1867 PH/02 Patch to exigrep to allow it to work on syslog lines.
1869 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1870 fread() to skip over the body file's header line, because in Cygwin the
1871 header line is locked and is inaccessible.
1873 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1874 co-exist for some time) to make it clear that it is the Exim ID that is
1875 referenced, not the Message-ID: header line.
1877 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1878 string_format() function, because snprintf() does not exist on all
1881 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1882 ${stat: expansion item.
1884 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1885 protocol synchronization error", to keep the pedants happy.
1887 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1888 well as for IRIX systems, when gcc is being used. See the host.c source
1891 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1893 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1896 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1897 transport and to the smtp transport in LMTP mode.
1899 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1901 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1902 run a filter in a subprocess. This could lead to confusion in subsequent
1903 lookups in the parent process. There should also be a search_tidyup() at
1904 the end of the subprocess.
1906 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1907 only if the host matched helo_try_verify_hosts, which caused the
1908 verification to occur when the EHLO/HELO command was issued. The ACL just
1909 tested the remembered result. Now, if a previous verification attempt has
1910 not happened, "verify = helo" does it there and then.
1912 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1914 TK/03 Fix log output including CR from clamd.
1916 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1917 exist provoked a memory error which could cause a segfault.
1919 PH/15 Installed PCRE 6.2
1921 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1923 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1924 of the problem. Specifically, suggested +O2 rather than +O1 for the
1927 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1929 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1930 record, the port was not correctly logged when the outgoing_port log
1931 selector was set (it logged the transort's default port).
1933 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1934 fallback_hosts, and "hosts" in the smtp transport.
1936 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1937 host errors such as "Connection refused".
1939 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1940 authentication with radiusclient 0.4.9:
1942 - Error returned from rc_read_config was caught wrongly
1943 - Username/password not passed on to radius server due to wrong length.
1945 The presumption is that some radiusclient API changes for 4.51/PH/17
1946 were not taken care of correctly. The code is still untested by me (my
1947 Linux distribution still has 0.3.2 of radiusclient), but it was
1948 contributed by a Radius user.
1950 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1951 expanding the "port" option of the smtp transport.
1953 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1954 while reading a MIME header. Thanks to Tom Hughes for a patch.
1956 PH/24 Include config.h inside local_scan.h so that configuration settings are
1959 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1960 that in an ACL for RCPT (for example), you can examine exactly what was
1963 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1964 commands, but it was not correctly comparing the address with the actual
1965 client host address. Thus, it would show the EHLO address in Received:
1966 header lines when this was not necessary.
1968 PH/27 Added the % operator to ${eval:}.
1970 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1971 it should be ignoring failures (because with -C, for example, it has lost
1972 privilege). It wasn't ignoring creation failures other than "already
1975 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1978 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1979 because an input line was too long, either on its own, or by virtue of
1980 too many continuations, the temporary file was not being removed, and the
1981 return code was incorrect.
1983 PH/31 Missing "BOOL" in function definition in filtertest.c.
1985 PH/32 Applied Sieve patches from the maintainer.
1987 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1989 PH/33 Added "verify = not_blind".
1991 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1992 Local/Makefile (with some defaults set). These are used in built scripts
1993 such as exicyclog, but they have never been used in the exim_install
1994 script (though there are many overriding facilities there). I have
1995 arranged that the exim_install script now takes note of these two
1998 PH/35 Installed configuration files for Dragonfly.
2000 PH/36 When a locally submitted message by a trusted user did not contain a
2001 From: header, and the sender address was obtained from -f or from an SMTP
2002 MAIL command, and the trusted user did not use -F to supply a sender
2003 name, $originator_name was incorrectly used when constructing a From:
2004 header. Furthermore, $originator_name was used for submission mode
2005 messages from external hosts without From: headers in a similar way,
2006 which is clearly wrong.
2008 PH/37 Added control=suppress_local_fixups.
2010 PH/38 When log_selector = +received_sender was set, and the addition of the
2011 sender made the log line's construction buffer exactly full, or one byte
2012 less than full, an overflow happened when the terminating "\n" was
2015 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2016 when the result of a list match is failure because a DNS lookup failed.
2018 PH/40 RM_COMMAND is now used in the building process.
2020 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2021 the "build-* directories that it finds.
2023 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2024 address in a domain literal was a prefix of an interface address.
2026 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2027 when verifying a sender address, unless rewrite_headers is false.
2029 PH/44 Wrote a long comment about why errors_to addresses are verified as
2030 recipients, not senders.
2032 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2033 the ratelimit ACL was added.
2035 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2037 PH/46 Added extra information about PostgreSQL errors to the error string.
2039 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2040 avoiding the need to install special zones in a real server. This is
2041 backwards compatible; if it can't find the fake resolver, it drops back.
2042 Thus, both old and new test suites can be run.
2044 TF/02 Added util/ratelimit.pl
2046 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2047 clock is set back in time.
2049 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2050 Candler <B.Candler@pobox.com>.
2052 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2053 OK for addresses that are the result of redirections.
2055 PH/48 A number of further additions for the benefit of the new test suite,
2056 including a fake gethostbyname() that interfaces to the fake DNS resolver
2059 TF/06 The fix for widen_domains has also been applied to qualify_single and
2060 search_parents which are the other dnslookup options that can cause
2063 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2066 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2068 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2069 service" if it encountered an MX record with an empty target hostname.
2070 The message is now "an MX or SRV record indicated no SMTP service".
2072 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2073 if the DNS of the sending site is misconfigured. This is quite a
2074 common situation. This change restores the behaviour of treating a
2075 helo verification defer as a failure.
2077 PH/51 If self=fail was set on a router, the bounce message did not include the
2078 actual error message.
2084 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2086 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2087 command itself ended in error, the underlying message about the transport
2088 filter timeout was being overwritten with the pipe command error. Now the
2089 underlying error message should be appended to the second error message.
2091 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2093 PH/02 Reduce the amount of output that "make" produces by default. Full output
2094 can still be requested.
2096 PH/03 The warning log line about a condition test deferring for a "warn" verb
2097 was being output only once per connection, rather than after each
2098 occurrence (because it was using the same function as for successful
2099 "warn" verbs). This seems wrong, so I have changed it.
2101 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2102 it should not have, which might have caused a crash in the right
2103 circumstances, but probably never did.
2105 PH/04 Installed a modified version of Tony Finch's patch to make submission
2106 mode fix the return path as well as the Sender: header line, and to
2107 add a /name= option so that you can make the user's friendly name appear
2110 TF/03 Added the control = fakedefer ACL modifier.
2112 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2113 Mark Lowes for thorough testing.
2115 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2117 TK/03 Merged latest SRS patch from Miles Wilton.
2119 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2120 with the definition in sysexits.h (which is #included earlier).
2121 Fortunately, Exim does not actually use EX_OK. The code used to try to
2122 preserve the sysexits.h value, by assumimg that macro definitions were
2123 scanned for macro replacements. I have been disabused of this notion,
2124 so now the code just undefines EX_OK before #including unistd.h.
2126 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2127 in the smtp transport. When a block could not be written in a single
2128 write() function, the timeout was being re-applied to each part-write.
2129 This seems wrong - if the receiver was accepting one byte at a time it
2130 would take for ever. The timeout is now adjusted when this happens. It
2131 doesn't have to be particularly precise.
2133 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2134 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2136 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2137 without a domain if the check to <postmaster@domain> fails.
2139 SC/01 Eximstats: added -xls and the ability to specify output files
2140 (patch written by Frank Heydlauf).
2142 SC/02 Eximstats: use FileHandles for outputing results.
2144 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2146 SC/04 Eximstats: fixed display of large numbers with -nvr option
2148 SC/05 Eximstats: fixed merging of reports with empty tables.
2150 SC/06 Eximstats: added the -include_original_destination flag
2152 SC/07 Eximstats: removed tabs and trailing whitespace.
2154 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2156 TK/06 MBOX spool code: Add real "From " MBOX separator line
2157 so the .eml file is really in mbox format (even though
2158 most programs do not really care). Patch from Alex Miller.
2160 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2161 The latter is generated from $received_to and is only set if the
2162 message has one envelope recipient. SA can use these headers,
2163 obviously out-of-the-box. Patch from Alex Miller.
2165 PH/08 The ${def test on a variable was returning false if the variable's
2166 value was "0", contrary to what the specification has always said!
2167 The result should be true unless the variable is empty.
2169 PH/09 The syntax error of a character other than { following "${if
2170 def:variable_name" (after optional whitespace) was not being diagnosed.
2171 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2172 accidental colon was present, for example, could give incorrect results.
2174 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2175 result is used (not including appendfile, where other changes are about
2178 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2179 This involved changing a lot of size variables from int to off_t. It
2180 should work with maildirs and everything.
2182 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2183 spamd dying while we are connected to it.
2185 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2188 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2189 function for BDB 4.3.
2191 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2193 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2194 the sources for PCRE have been re-arranged and more files are now
2197 PH/15 The code I had for printing potentially long long variables in PH/11
2198 above was not the best (it lost precision). The length of off_t variables
2199 is now inspected at build time, and an appropriate printing format (%ld
2200 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2201 to be "long long int" or "long int". This is needed for the internal
2202 formatting function string_vformat().
2204 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2205 the configuration file to be ":syslog", then the script "guesses" where
2206 the logs files are, rather than using the compiled in default. In our
2207 case the guess is not the same as the compiled default, so the script
2208 suddenly stopped working when I started to use syslog. The patch checks
2209 to see if log_file_path is "". If so, it attempts to read it from exim
2210 with no configuration file to get the compiled in version, before it
2211 falls back to the previous guessing code."
2213 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2214 implementing BATV in an Exim configuration. See NewStuff for the gory
2217 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2218 Makefile that are specific to HP-UX.
2220 PH/18 If the "use_postmaster" option was set for a recipient callout together
2221 with the "random" option, the postmaster address was used as the MAIL
2222 FROM address for the random test, but not for the subsequent recipient
2223 test. It is now used for both.
2225 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2226 patch removes a few documentation additions to RFC 3028, because the
2227 latest draft now contains them. It adds the new en;ascii-case comparator
2228 and a new error check for 8bit text in MIME parts. Comparator and
2229 require names are now matched exactly. I enabled the subaddress
2230 extension, but it is not well tested yet (read: it works for me)."
2232 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2233 rework some of the code of TK/09 above to avoid the hardwired use of
2234 "%lld" and "long long". Replaced the call to snprintf() with a call to
2237 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2238 records point to non-existent hosts", "retry timeout exceeded", and
2239 "retry time not reached for any host after a long failure period".
2241 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2242 experimental DomainKeys support:
2244 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2245 (2) On an error such as an illegally used "control", the wrong name for
2246 the control was given.
2248 These problems did NOT occur unless DomainKeys support was compiled.
2250 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2252 PH/24 Added ${if match_ip condition.
2254 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2255 databases so that it will be absolutely obvious if a crash occurs in the
2256 DB library. This is a regular occurrence (often caused by mis-matched
2259 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2260 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2261 picked up on a user's system that detects such things. There doesn't seem
2262 to be a gcc warning option for this - only an attribute that has to be
2263 put on the function's prototype. It seems that in Fedora Core 4 they have
2264 set this on a number of new functions. No doubt there will be more in due
2267 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2268 specify a transport. However, if an address that was verified by such a
2269 router was the subject of a callout, Exim crashed because it tried to
2270 read the rcpt_include_affixes from the non-existent transport. Now it
2271 just assumes that the setting of that option is false. This bug was
2272 introduced by 4.51/PH/31.
2274 PH/28 Changed -d+all to exclude +memory, because that information is very
2275 rarely of interest, but it makes the output a lot bigger. People tend to
2276 do -d+all out of habit.
2278 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2279 code in os-type was giving problems when libc.so lives in lib64, like on
2282 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2283 aren't the modern standard, and it seems that some systems' include files
2284 don't always have them. Exim was already checking for some of the newer
2285 ones like T_AAAA, and defining it itself. I've added checks for all the
2286 record types that Exim uses.
2288 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2289 not automatically generating a new one, as it is supposed to. This
2290 prevented TLS from working. If the file did exist, but contained invalid
2291 data, a new version was generated, as expected. It was only the case of a
2292 non-existent file that was broken.
2294 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2295 with a change in libdomainkeys > 0.64.
2297 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2298 from DNS. If the selector record carries the flag, it now has
2299 precedence over the domain-wide flag.
2301 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2303 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2304 the use of a port name that isn't defined in /etc/services) occurred, the
2305 message was deferred as in a normal delivery, and thus remained on the
2306 spool, instead of being failed because of the mua_wrapper setting. This
2307 is now fixed, and I tidied up some of the mua_wrapper messages at the
2310 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2311 the messages in a hash of arrays rather than using individual hashes.
2312 This is a bit cleaner and results in dramatic memory savings, albeit
2313 at a slight CPU cost.
2315 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2316 as requested by Marc Sherman.
2318 SC/10 Eximstats: added histograms for user specified patterns as requested
2321 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2323 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2324 fopen() in the content-scanning modules that did not already have it.
2330 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2331 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2333 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2335 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2337 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2338 to test Sieve filters that use "vacation".
2340 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2341 that changes the way the GnuTLS parameters are stored in the cache file.
2342 The new format can be generated externally. For backward compatibility,
2343 if the data in the cache doesn't make sense, Exim assumes it has read an
2344 old-format file, and it generates new data and writes a new file. This
2345 means that you can't go back to an older release without removing the
2348 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2349 work if there are any delivery delays because "one_time" forces the
2350 parent to be marked "delivered", so its unseen clone is never tried
2351 again. For this reason, Exim now forbids the simultaneous setting of
2354 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2355 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2356 entry if you want to know the details. The fix, however, neglected to
2357 consider the case when local delivery batching is involved. The test for
2358 "previously delivered" was not happening when checking to see if an
2359 address could be batched with a previous (undelivered) one; under
2360 certain circumstances this could lead to multiple deliveries to the same
2363 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2364 in its include files, and this causes problems building Exim.
2366 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2367 header_syntax) but Exim was just ignoring anything given after a slash.
2368 In particular, this caused confusion with an attempt to use "verify =
2369 reverse_host_lookup/defer_ok". An error is now given when options are
2370 supplied for verify items that do not have them. (Maybe reverse_host_
2371 lookup should have a defer_ok option, but that's a different point.)
2373 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2374 defined by RFC 821) to 2048, because there were problems with some AUTH
2375 commands, and RFC 1869 says the size should be increased for extended
2376 SMTP commands that take arguments.
2378 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2381 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2382 "unknown" error; now it says that the functionality isn't in the binary.
2384 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2385 an address' error message when a string expansion fails (syntax or
2386 whatever). Otherwise the password may appear in the log. Following change
2387 PH/42 below, there is no longer a chance of it appearing in a bounce
2390 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2392 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2393 of its addresses was ever tried. (Bugzilla bug #2.)
2395 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2396 the result incorrectly in the debug output. (It correctly added a newline
2397 to what was transported.)
2399 TF/01 Added $received_time.
2401 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2402 commented out examples of how to interface to a virus scanner and to
2403 SpamAssassin. Also added commented examples of av_scanner and
2404 spamd_address settings.
2406 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2407 and controls are allowed in which ACLs. There were a couple of minor
2408 errors. Some of the entries in the conditions table (which is a table of
2409 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2410 negation of where the condition IS allowed.
2412 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2414 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2415 header file does not have a version number, so I've had to invent a new
2416 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2417 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2418 radiusclient), but it was contributed by a Radius user.
2420 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2421 files or directories, for OpenSSL.
2423 PH/19 When an Exim process that is running as root has to create an Exim log
2424 file, it does so in a subprocess that runs as exim:exim so as to get the
2425 ownership right at creation (otherwise, other Exim processes might see
2426 the file with the wrong ownership). There was no test for failure of this
2427 fork() call, which would lead to the process getting stuck as it waited
2428 for a non-existent subprocess. Forks do occasionally fail when resources
2429 run out. I reviewed all the other calls to fork(); they all seem to check
2432 PH/20 When checking for unexpected SMTP input at connect time (before writing
2433 the banner), Exim was not dealing correctly with a non-positive return
2434 from the read() function. If the client had disconnected by this time,
2435 the result was a log entry for a synchronization error with an empty
2436 string after "input=" when read() returned zero. If read() returned -1
2437 (an event I could not check), uninitialized data bytes were printed.
2438 There were reports of junk text (parts of files, etc) appearing after
2441 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2443 PH/22 Added support for macro redefinition, and (re)definition in between
2444 driver and ACL definitions.
2446 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2447 forgetting to use the resulting value; it was using the unexpanded value.
2449 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2450 hadn't been configured. The fix is from Juergen Kreileder, who
2451 understands it better than I do:
2453 "Here's what I see happening with three configured cyrus_sasl
2454 authenticators configured (plain, login, cram-md5):
2456 On startup auth_cyrus_sasl_init() gets called for each of these.
2457 This means three calls to sasl_listmech() without a specified mech_list.
2458 => SASL tests which mechs of all available mechs actually work
2459 => three warnings about OTP not working
2460 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2462 With the patch, sasl_listmech() also gets called three times. But now
2463 SASL's mech_list option is set to the server_mech specified in the the
2464 authenticator. Or in other words, the answer from sasl_listmech()
2465 gets limited to just the mech you're testing for (which is different
2467 => the return list contains just 'plain' or 'login', 'cram-md5' or
2468 nothing depending on the value of ob->server_mech.
2470 I've just tested the patch: Authentication still works fine,
2471 unavailable mechs specified in the exim configuration are still
2472 caught, and the auth.log warnings about OTP are gone."
2474 PH/25 When debugging is enabled, the contents of the command line are added
2475 to the debugging output, even when log_selector=+arguments is not
2478 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2479 answer is "GNU", and only if the return is "GNU/something" is the answer
2482 PH/27 $acl_verify_message is now set immediately after the failure of a
2483 verification in an ACL, and so is available in subsequent modifiers. In
2484 particular, the message can be preserved by coding like this:
2486 warn !verify = sender
2487 set acl_m0 = $acl_verify_message
2489 Previously, $acl_verify_message was set only while expanding "message"
2490 and "log_message" when a very denied access.
2492 PH/28 Modified OS/os.c-Linux with
2494 -#ifndef OS_LOAD_AVERAGE
2495 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2497 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2498 nomenclature these days.)
2500 PH/29 Installed patch from the Sieve maintainer that adds the options
2501 sieve_useraddress and sieve_subaddress to the redirect router.
2503 PH/30 In these circumstances:
2504 . Two addresses routed to the same list of hosts;
2505 . First host does not offer TLS;
2506 . First host accepts first address;
2507 . First host gives temporary error to second address;
2508 . Second host offers TLS and a TLS session is established;
2509 . Second host accepts second address.
2510 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2511 and peerdn, if requested) that were in fact used only for the second
2514 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2515 attention to any local part prefix or suffix that was matched by the
2516 router that accepted the address. It now behaves in the same way as it
2517 does for delivery: the affixes are removed from the local part unless
2518 rcpt_include_affixes is set on the transport.
2520 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2521 timeout during the DATA phase of an incoming message.
2523 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2524 applied a patch sanctioned by the Sieve maintainer.
2526 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2527 the uid or gid is negative. A case of a negative gid caused this to be
2528 noticed. The fix allows for either to be negative.
2530 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2531 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2534 PH/36 The change PH/12 above was broken. Fixed it.
2536 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2537 the grounds that routing the same address twice would always produce the
2538 same answer. This might have been true once, but it is certainly no
2539 longer true now. Routing a child address may depend on the previous
2540 routing that produced that child. Some complicated redirection strategies
2541 went wrong when messages had multiple recipients, and made Exim's
2542 behaviour dependent on the order in which the addresses were given.
2544 I have moved the duplicate checking until after the routing is complete.
2545 Exim scans the addresses that are assigned to local and remote
2546 transports, and removes any duplicates. This means that more work will be
2547 done, as duplicates will always all be routed, but duplicates are
2548 presumably rare, so I don't expect this is of any significance.
2550 For deliveries to pipes, files, and autoreplies, the duplicate checking
2551 still happens during the routing process, since they are not going to be
2554 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2555 It corrects a timeout issue with spamd. This is Ian's comment: "The
2556 background is that sometimes spamd either never reads data from a
2557 connection it has accepted, or it never writes response data. The exiscan
2558 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2559 blindly assumes that writes won't block so it may never time out."
2561 PH/39 Allow G after quota size as well as K and M.
2563 PH/40 The value set for $authenticated_id in an authenticator may not contain
2564 binary zeroes or newlines because the value is written to log lines and
2565 to spool files. There was no check on this. Now the value is run through
2566 the string_printing() function so that such characters are converted to
2567 printable escape sequences.
2569 PH/41 $message_linecount is a new variable that contains the total number of
2570 lines in the message. Compare $body_linecount, which is the count for the
2573 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2574 bounce and delay warning messages, except in certain special cases, which
2577 (a) An SMTP error message from a remote host;
2578 (b) A message specified in a :fail: redirection;
2579 (c) A message specified in a "fail" command in a system filter;
2580 (d) A message specified in a FAIL return from the queryprogram router;
2581 (e) A message specified by the cannot_route_message router option.
2583 In these cases only, Exim does include the error details in bounce and
2584 warning messages. There are also a few cases where bland messages such
2585 as "unrouteable address" or "local delivery error" are given.
2587 PH/43 $value is now also set for the "else" part of a ${run expansion.
2589 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2590 being worked on, but at least Exim now implements the latest version to
2593 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2594 process to complete was treated as a delivery failure, a timeout while
2595 writing the message to the pipe was logged, but erroneously treated as a
2596 successful delivery. Such timeouts include transport filter timeouts. For
2597 consistency with the overall process timeout, these timeouts are now
2598 treated as errors, giving rise to delivery failures by default. However,
2599 there is now a new Boolean option for the pipe transport called
2600 timeout_defer, which, if set TRUE, converts the failures into defers for
2601 both kinds of timeout. A transport filter timeout is now identified in
2604 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2605 systems where "make" and "gmake" are different, calling "gmake" at top
2606 level broke things. I've arranged for the value of $(MAKE) to be passed
2607 from the Makefile to this script so that it can call the same version of
2611 A note about Exim versions 4.44 and 4.50
2612 ----------------------------------------
2614 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2615 changes of various kinds. As a consequence, a big documentation update was
2616 needed. This delayed the release for rather longer than seemed good, especially
2617 in the light of a couple of (minor) security issues. Therefore, the changes
2618 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2619 release. So 4.44 and 4.50 are in effect two different branches that both start
2622 I have left the 4.50 change log unchanged; it contains all the changes since
2623 4.43. The change log for 4.44 is below; many of its items are identical to
2624 those for 4.50. This seems to be the most sensible way to preserve the
2625 historical information.
2631 1. Minor wording change to the doc/README.SIEVE file.
2633 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2634 computation of the current number of files was incorrect.
2636 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2637 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2640 4. Give more explanation in the error message when the command for a transport
2641 filter fails to execute.
2643 5. There are several places where Exim runs a non-Exim command in a
2644 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2645 was being done only for the command run by the queryprogram router. It is
2646 now done for all such subprocesses. The other cases are: ${run, transport
2647 filters, and the commands run by the lmtp and pipe transports.
2649 6. Added CONFIGURE_GROUP build-time option.
2651 7. Some older OS have a limit of 256 on the maximum number of file
2652 descriptors. Exim was using setrlimit() to set 1000 as a large value
2653 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2654 systems. I've change it so that if it can't get 1000, it tries for 256.
2656 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2657 was an oversight, and furthermore, ever since the addition of extra
2658 controls (e.g. 4.43/32), the checks on when to allow different forms of
2659 "control" were broken. There should now be diagnostics for all cases when a
2660 control that does not make sense is encountered.
2662 9. Added the /retain_sender option to "control=submission".
2664 10. $recipients is now available in the predata ACL (oversight).
2666 11. Tidy the search cache before the fork to do a delivery from a message
2667 received from the command line. Otherwise the child will trigger a lookup
2668 failure and thereby defer the delivery if it tries to use (for example) a
2669 cached ldap connection that the parent has called unbind on.
2671 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2672 of $address_data from the recipient verification was clobbered by the
2673 sender verification.
2675 13. The value of address_data from a sender verification is now available in
2676 $sender_address_data in subsequent conditions in the ACL statement.
2678 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2680 15. Added a new option "connect=<time>" to callout options, to set a different
2683 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2684 was its contents. (It was OK if the option was not defined at all.)
2686 17. A "Completed" log line is now written for messages that are removed from
2687 the spool by the -Mrm option.
2689 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2690 information about exactly what failed.
2692 19. Added -dd to debug only the daemon process.
2694 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2695 handles timeouts, both on the server side and network timeouts. Renamed the
2696 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2698 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2699 It is now set to "smtps".
2701 22. $host_address is now set to the target address during the checking of
2702 ignore_target_hosts.
2704 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2705 being passed; this would have caused $sender_host_name to have been used if
2706 matching the list had actually called for a host name (not very likely,
2707 since this list is usually IP addresses). A host name is now passed as
2710 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2711 code that specifies a non-restarting handler (typically sigaction() in
2712 modern systems) in an attempt to fix a rare and obscure crash bug.
2714 25. Narrowed the window for a race in the daemon that could cause it to ignore
2715 SIGCHLD signals. This is not a major problem, because they are used only to
2716 wake it up if nothing else does.
2718 26. A malformed maildirsize file could cause Exim to calculate negative values
2719 for the mailbox size or file count. Odd effects could occur as a result.
2720 The maildirsize information is now recalculated if the size or filecount
2723 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2724 support for a long time. Removed HAVE_SYS_VFS_H.
2726 28. Installed the latest version of exipick from John Jetmore.
2728 29. In an address list, if the pattern was not a regular expression, an empty
2729 subject address (from a bounce message) matched only if the pattern was an
2730 empty string. Non-empty patterns were not even tested. This was the wrong
2731 because it is perfectly reasonable to use an empty address as part of a
2732 database query. An empty address is now tested by patterns that are
2733 lookups. However, all the other forms of pattern expect the subject to
2734 contain a local part and a domain, and therefore, for them, an empty
2735 address still always fails if the pattern is not itself empty.
2737 30. Exim went into a mad DNS loop when attempting to do a callout where the
2738 host was specified on an smtp transport, and looking it up yielded more
2739 than one IP address.
2741 31. Re-factored the code for checking spool and log partition space into a
2742 function that finds that data and another that does the check. The former
2743 is then used to implement four new variables: $spool_space, $log_space,
2744 $spool_inodes, and $log_inodes.
2746 32. The RFC2047 encoding function was originally intended for short strings
2747 such as real names; it was not keeping to the 75-character limit for
2748 encoded words that the RFC imposes. It now respects the limit, and
2749 generates multiple encoded words if necessary. To be on the safe side, I
2750 have increased the buffer size for the ${rfc2047: expansion operator from
2753 33. It is now permitted to omit both strings after an "if" condition; if the
2754 condition is true, the result is "true". As before, when the second string
2755 is omitted, a false condition yields an empty string. This makes it less
2756 cumbersome to write custom ACL and router conditions.
2758 34. Failure to deliver a bounce message always caused it to be frozen, even if
2759 there was an errors_to setting on the router. The errors_to setting is now
2762 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2763 canonical form (fully expanded) before being placed in
2764 $sender_host_address.
2766 36. The table in the code that translates DNS record types into text (T_A to
2767 "A" for instance) was missing entries for NS and CNAME. It is just possible
2768 that this could have caused confusion if both these types were looked up
2769 for the same domain, because the text type is used as part of Exim's
2770 per-process caching. But the chance of anyone hitting this buglet seems
2773 37. The dnsdb lookup has been extended in a number of ways.
2775 (1) There is a new type, "zns", which walks up the domain tree until it
2776 finds some nameserver records. It should be used with care.
2778 (2) There is a new type, "mxh", which is like "mx" except that it returns
2779 just the host names, not the priorities.
2781 (3) It is now possible to give a list of domains (or IP addresses) to be
2782 looked up. The behaviour when one of the lookups defers can be
2783 controlled by a keyword.
2785 (4) It is now possible to specify the separator character for use when
2786 multiple records are returned.
2788 38. The dnslists ACL condition has been extended: it is now possible to supply
2789 a list of IP addresses and/or domains to be looked up in a particular DNS
2792 39. Added log_selector=+queue_time_overall.
2794 40. When running the queue in the test harness, wait just a tad after forking a
2795 delivery process, to get repeatability of debugging output.
2797 41. Include certificate and key file names in error message when GnuTLS fails
2798 to set them up, because the GnuTLS error message doesn't include the name
2799 of the failing file when there is a problem reading it.
2801 42. Allow both -bf and -bF in the same test run.
2803 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2805 44. The "Exiscan patch" is now merged into the mainline Exim source.
2807 45. Sometimes the final signoff response after QUIT could fail to get
2808 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2809 < 0 before doing a fflush(). This bug looks as though it goes back to the
2810 introduction of TLS in release 3.20, but "sometimes" must have been rare
2811 because the tests only now provoked it.
2813 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2814 (this can affect the format of dates).
2816 47. exim_tidydb, when checking for the continued existence of a message for
2817 which it has found a message-specific retry record, was not finding
2818 messages that were in split spool directories. Consequently, it was
2819 deleting retry records that should have stayed in existence.
2821 48. Steve fixed some bugs in eximstats.
2823 49. The SPA authentication driver was not abandoning authentication and moving
2824 on to the next authenticator when an expansion was forced to fail,
2825 contradicting the general specification for all authenticators. Instead it
2826 was generating a temporary error. It now behaves as specified.
2828 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2829 (the order specifies the preference for clients). The order is now AES256,
2830 AES128, 3DES, ARCFOUR128.
2832 51. Small patch to Sieve code - explicitly set From: when generating an
2835 52. Exim crashed if a remote delivery caused a very long error message to be
2836 recorded - for instance if somebody sent an entire SpamAssassin report back
2837 as a large number of 550 error lines. This bug was coincidentally fixed by
2838 increasing the size of one of Exim's internal buffers (big_buffer) that
2839 happened as part of the Exiscan merge. However, to be on the safe side, I
2840 have made the code more robust (and fixed the comments that describe what
2843 53. Now that there can be additional text after "Completed" in log lines (if
2844 the queue_time_overall log selector is set), a one-byte patch to exigrep
2845 was needed to allow it to recognize "Completed" as not the last thing in
2848 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2849 patch that reportedly fixes this has been added. I am not expert enough to
2850 create a test for it. This is what the patch creator wrote:
2852 "I found a little strange behaviour of ldap code when working with
2853 Windows 2003 AD Domain, where users was placed in more than one
2854 Organization Units. When I tried to give exim partial DN, the exit code
2855 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2856 But simultaneously result of request was absolutely normal ldap result,
2857 so I produce this patch..."
2859 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2860 REFERENCE, so I have modified the code to exclude the patch when that macro
2863 55. Some experimental protocols are using DNS PTR records for new purposes. The
2864 keys for these records are domain names, not reversed IP addresses. The
2865 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2866 leaves it alone. Component reversal etc. now happens only for IP addresses.
2869 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2871 57. Double the size of the debug message buffer (to 2048) so that more of very
2872 long debug lines gets shown.
2874 58. The exicyclog utility now does better if the number of log files to keep
2875 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2877 59. Two changes related to the smtp_active_hostname option:
2879 (1) $smtp_active_hostname is now available as a variable.
2880 (2) The default for smtp_banner uses $smtp_active_hostname instead
2881 of $primary_hostname.
2883 60. The host_aton() function is supposed to be passed a string that is known
2884 to be a valid IP address. However, in the case of IPv6 addresses, it was
2885 not checking this. This is a hostage to fortune. Exim now panics and dies
2886 if the condition is not met. A case was found where this could be provoked
2887 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2888 components; fortuitously, this particular loophole had already been fixed
2889 by change 4.50/55 above.
2891 If there are any other similar loopholes, the new check in host_aton()
2892 itself should stop them being exploited. The report I received stated that
2893 data on the command line could provoke the exploit when Exim was running as
2894 exim, but did not say which command line option was involved. All I could
2895 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2896 running as the user.
2899 61. There was a buffer overflow vulnerability in the SPA authentication code
2900 (which came originally from the Samba project). I have added a test to the
2901 spa_base64_to_bits() function which I hope fixes it.
2904 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2905 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2907 63. The daemon start-up calls getloadavg() while still root for those OS that
2908 need the first call to be done as root, but it missed one case: when
2909 deliver_queue_load_max is set with deliver_drop_privilege. This is
2910 necessary for the benefit of the queue runner, because there is no re-exec
2911 when deliver_drop_privilege is set.
2913 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2914 This has been fixed.
2916 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2917 was in use, was not putting the data itself into the right store pool;
2918 consequently, it could be overwritten for a subsequent message in the same
2919 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2922 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2924 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2925 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2926 the function were treating the return as a boolean value, which happened to
2927 work because 0=false and not-0=true, but is not correct code.
2929 68. The host_aton() function was not handling scoped IPv6 addresses (those
2930 with, for example, "%eth0" on the end) correctly.
2932 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2933 negated items (that is, ~something) in unsigned ints. Some compilers
2934 apparently mutter when there is no cast.
2936 70. If an address verification called from an ACL failed, and did not produce a
2937 user-specific message (i.e. there was only a "system" message), nothing was
2938 put in $acl_verify_message. In this situation, it now puts the system
2941 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2942 session; change 4.31/43 added the unwanted input to the log line - except
2943 that it did not do this in the start of session case. It now does.
2945 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2946 This is wrong and can cause the other end to generate a synchronization
2947 error if it is another Exim or anything else that does the synchronization
2948 check. A QUIT command is no longer sent after a timeout.
2950 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2951 during host lookups.
2953 74. The defer_ok option of callout verification was not working if it was used
2954 when verifying addresses in header lines, that is, for this case:
2956 verify = header_sender/callout=defer_ok
2958 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2959 those file descriptors could be used for SMTP connections. If anything
2960 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2961 could be sent to the SMTP client, causing chaos. The daemon now opens
2962 stdin, stdout, and stderr to /dev/null when it puts itself into the
2965 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2966 log by default. The user can override this with suitable Perl magic.
2968 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2969 the log message when discard triggers, was not working for the DATA ACL or
2970 for the non-SMTP ACL.
2972 78. Error message wording change in sieve.c.
2974 79. If smtp_accept_max_per_host was set, the number of connections could be
2975 restricted to fewer than expected, because the daemon was trying to set up
2976 a new connection before checking whether the processes handling previous
2977 connections had finished. The check for completed processes is now done
2978 earlier. On busy systems, this bug wouldn't be noticed because something
2979 else would have woken the daemon, and it would have reaped the completed
2982 80. If a message was submitted locally by a user whose login name contained one
2983 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2984 It caused a spool format error. I have fixed the spool reading code. A
2985 related problem was that the "from" clause in the Received: line became
2986 illegal because of the space(s). It is now covered by ${quote_local_part.
2988 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2991 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2993 83. Patch from Sieve maintainer to fix unterminated string problem in
2994 "vacation" handling.
2996 84. Some minor changes to the Linux configuration files to help with other
2997 OS variants using glibc.
2999 85. One more patch for Sieve to update vacation handling to latest spec.
3002 ----------------------------------------------------
3003 See the note above about the 4.44 and 4.50 releases.
3004 ----------------------------------------------------
3010 1. Change 4.43/35 introduced a bug that caused file counts to be
3011 incorrectly computed when quota_filecount was set in an appendfile
3014 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3015 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3018 3. Give more explanation in the error message when the command for a transport
3019 filter fails to execute.
3021 4. There are several places where Exim runs a non-Exim command in a
3022 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3023 was being done only for the command run by the queryprogram router. It is
3024 now done for all such subprocesses. The other cases are: ${run, transport
3025 filters, and the commands run by the lmtp and pipe transports.
3027 5. Some older OS have a limit of 256 on the maximum number of file
3028 descriptors. Exim was using setrlimit() to set 1000 as a large value
3029 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3030 systems. I've change it so that if it can't get 1000, it tries for 256.
3032 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3033 was an oversight, and furthermore, ever since the addition of extra
3034 controls (e.g. 4.43/32), the checks on when to allow different forms of
3035 "control" were broken. There should now be diagnostics for all cases when a
3036 control that does not make sense is encountered.
3038 7. $recipients is now available in the predata ACL (oversight).
3040 8. Tidy the search cache before the fork to do a delivery from a message
3041 received from the command line. Otherwise the child will trigger a lookup
3042 failure and thereby defer the delivery if it tries to use (for example) a
3043 cached ldap connection that the parent has called unbind on.
3045 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3046 of $address_data from the recipient verification was clobbered by the
3047 sender verification.
3049 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3050 was its contents. (It was OK if the option was not defined at all.)
3052 11. A "Completed" log line is now written for messages that are removed from
3053 the spool by the -Mrm option.
3055 12. $host_address is now set to the target address during the checking of
3056 ignore_target_hosts.
3058 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3059 being passed; this would have caused $sender_host_name to have been used if
3060 matching the list had actually called for a host name (not very likely,
3061 since this list is usually IP addresses). A host name is now passed as
3064 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3065 code that specifies a non-restarting handler (typically sigaction() in
3066 modern systems) in an attempt to fix a rare and obscure crash bug.
3068 15. Narrowed the window for a race in the daemon that could cause it to ignore
3069 SIGCHLD signals. This is not a major problem, because they are used only to
3070 wake it up if nothing else does.
3072 16. A malformed maildirsize file could cause Exim to calculate negative values
3073 for the mailbox size or file count. Odd effects could occur as a result.
3074 The maildirsize information is now recalculated if the size or filecount
3077 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3078 support for a long time. Removed HAVE_SYS_VFS_H.
3080 18. Updated exipick to current release from John Jetmore.
3082 19. Allow an empty sender to be matched against a lookup in an address list.
3083 Previously the only cases considered were a regular expression, or an
3086 20. Exim went into a mad DNS lookup loop when doing a callout where the
3087 host was specified on the transport, if the DNS lookup yielded more than
3090 21. The RFC2047 encoding function was originally intended for short strings
3091 such as real names; it was not keeping to the 75-character limit for
3092 encoded words that the RFC imposes. It now respects the limit, and
3093 generates multiple encoded words if necessary. To be on the safe side, I
3094 have increased the buffer size for the ${rfc2047: expansion operator from
3097 22. Failure to deliver a bounce message always caused it to be frozen, even if
3098 there was an errors_to setting on the router. The errors_to setting is now
3101 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3102 canonical form (fully expanded) before being placed in
3103 $sender_host_address.
3105 24. Updated eximstats to version 1.33
3107 25. Include certificate and key file names in error message when GnuTLS fails
3108 to set them up, because the GnuTLS error message doesn't include the name
3109 of the failing file when there is a problem reading it.
3111 26. Expand error message when OpenSSL has problems setting up cert/key files.
3114 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3115 (this can affect the format of dates).
3117 28. exim_tidydb, when checking for the continued existence of a message for
3118 which it has found a message-specific retry record, was not finding
3119 messages that were in split spool directories. Consequently, it was
3120 deleting retry records that should have stayed in existence.
3122 29. eximstats updated to version 1.35
3123 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3124 1.35 - bugfix such that pie charts by volume are generated correctly
3126 30. The SPA authentication driver was not abandoning authentication and moving
3127 on to the next authenticator when an expansion was forced to fail,
3128 contradicting the general specification for all authenticators. Instead it
3129 was generating a temporary error. It now behaves as specified.
3131 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3132 (the order specifies the preference for clients). The order is now AES256,
3133 AES128, 3DES, ARCFOUR128.
3135 31. Small patch to Sieve code - explicitly set From: when generating an
3138 32. Exim crashed if a remote delivery caused a very long error message to be
3139 recorded - for instance if somebody sent an entire SpamAssassin report back
3140 as a large number of 550 error lines. This bug was coincidentally fixed by
3141 increasing the size of one of Exim's internal buffers (big_buffer) that
3142 happened as part of the Exiscan merge. However, to be on the safe side, I
3143 have made the code more robust (and fixed the comments that describe what
3146 33. Some experimental protocols are using DNS PTR records for new purposes. The
3147 keys for these records are domain names, not reversed IP addresses. The
3148 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3149 leaves it alone. Component reversal etc. now happens only for IP addresses.
3152 34. The host_aton() function is supposed to be passed a string that is known
3153 to be a valid IP address. However, in the case of IPv6 addresses, it was
3154 not checking this. This is a hostage to fortune. Exim now panics and dies
3155 if the condition is not met. A case was found where this could be provoked
3156 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3157 components; fortuitously, this particular loophole had already been fixed
3158 by change 4.50/55 or 4.44/33 above.
3160 If there are any other similar loopholes, the new check in host_aton()
3161 itself should stop them being exploited. The report I received stated that
3162 data on the command line could provoke the exploit when Exim was running as
3163 exim, but did not say which command line option was involved. All I could
3164 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3165 running as the user.
3168 35. There was a buffer overflow vulnerability in the SPA authentication code
3169 (which came originally from the Samba project). I have added a test to the
3170 spa_base64_to_bits() function which I hope fixes it.
3173 36. The daemon start-up calls getloadavg() while still root for those OS that
3174 need the first call to be done as root, but it missed one case: when
3175 deliver_queue_load_max is set with deliver_drop_privilege. This is
3176 necessary for the benefit of the queue runner, because there is no re-exec
3177 when deliver_drop_privilege is set.
3179 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3180 was in use, was not putting the data itself into the right store pool;
3181 consequently, it could be overwritten for a subsequent message in the same
3182 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3185 38. Sometimes the final signoff response after QUIT could fail to get
3186 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3187 < 0 before doing a fflush(). This bug looks as though it goes back to the
3188 introduction of TLS in release 3.20, but "sometimes" must have been rare
3189 because the tests only now provoked it.
3195 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3196 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3197 Now it returns an integer. A number of calls were still expecting a T/F
3198 return. Fortuitously, in all cases, the tests worked in OK situations,
3199 which is the norm. However, things would have gone wrong on any write
3200 failures on the smtp file descriptor. This function is used when sending
3201 messages over SMTP and also when doing verify callouts.
3203 2. When Exim is called to do synchronous delivery of a locally submitted
3204 message (the -odf or -odi options), it no longer closes stderr before doing
3207 3. Implemented the mua_wrapper option.
3209 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3211 5. Implemented the functions header_remove(), header_testname(),
3212 header_add_at_position(), and receive_remove_recipient(), and exported them
3215 6. If an ACL "warn" statement specified the addition of headers, Exim already
3216 inserted X-ACL-Warn: at the start if there was no header name. However, it
3217 was not making this test for the second and subsequent header lines if
3218 there were newlines in the string. This meant that an invalid header could
3219 be inserted if Exim was badly configured.
3221 7. Allow an ACL "warn" statement to add header lines at the start or after all
3222 the Received: headers, as well as at the end.
3224 8. Added the rcpt_4xx retry error code.
3226 9. Added postmaster_mailfrom=xxx to callout verification option.
3228 10. Added mailfrom=xxxx to the callout verification option, for verify=
3231 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3232 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3234 12. Inserted some casts to stop certain compilers warning when using pointer
3235 differences as field lengths or precisions in printf-type calls (mostly
3236 affecting debugging statements).
3238 13. Added optional readline() support for -be (dynamically loaded).
3240 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3241 same clock tick as a message's arrival, so that its received time was the
3242 same as the "first fail" time on the retry record, and that message
3243 remained on the queue past the ultimate address timeout, every queue runner
3244 would try a delivery (because it was past the ultimate address timeout) but
3245 after another failure, the ultimate address timeout, which should have then
3246 bounced the address, did not kick in. This was a "< instead of <=" error;
3247 in most cases the first failure would have been in the next clock tick
3248 after the received time, and all would be well.
3250 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3251 being recognized when the domain list was tested by the match_domain
3252 condition in an expansion string.
3254 16. Added the ${str2b64: operator.
3256 17. Exim was always calling setrlimit() to set a large limit for the number of
3257 processes, without checking whether the existing limit was already
3258 adequate. (It did check for the limit on file descriptors.) Furthermore,
3259 errors from getrlimit() and setrlimit() were being ignored. Now they are
3260 logged to the main and panic logs, but Exim does carry on, to try to do its
3261 job under whatever limits there are.
3263 18. Imported PCRE 5.0.
3265 19. Trivial typo in log message " temporarily refused connection" (the leading
3268 20. If the log selector return_path_on_delivery was set and an address was
3269 redirected to /dev/null, the delivery process crashed because it assumed
3270 that a return path would always be set for a "successful" delivery. In this
3271 case, the whole delivery is bypassed as an optimization, and therefore no
3274 21. Internal re-arrangement: the function for sending a challenge and reading
3275 a response while authentication was assuming a zero-terminated challenge
3276 string. It's now changed to take a pointer and a length, to allow for
3277 binary data in such strings.
3279 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3281 23. Exim was not respecting finduser_retries when seeking the login of the
3282 uid under which it was called; it was always trying 10 times. (The default
3283 setting of finduser_retries is zero.) Also, it was sleeping after the final
3284 failure, which is pointless.
3286 24. Implemented tls_on_connect_ports.
3288 25. Implemented acl_smtp_predata.
3290 26. If the domain in control=submission is set empty, Exim assumes that the
3291 authenticated id is a complete email address when it generates From: or
3292 Sender: header lines.
3294 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3295 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3296 chown and chgrp in /bin and hostname in /usr/bin.
3298 28. Exim was keeping the "process log" file open after each use, just as it
3299 does for the main log. This opens the possibility of it remaining open for
3300 long periods when the USR1 signal hits a daemon. Occasional processlog
3301 errors were reported, that could have been caused by this. Anyway, it seems
3302 much more sensible not to leave this file open at all, so that is what now
3305 29. The long-running daemon process does not normally write to the log once it
3306 has entered its main loop, and it closes the log before doing so. This is
3307 so that log files can straightforwardly be renamed and moved. However,
3308 there are a couple of unusual error situations where the daemon does write
3309 log entries, and I had neglected to close the log afterwards.
3311 30. The text of an SMTP error response that was received during a remote
3312 delivery was being truncated at 512 bytes. This is too short for some of
3313 the long messages that one sometimes sees. I've increased the limit to
3316 31. It is now possible to make retry rules that apply only when a message has a
3317 specific sender, in particular, an empty sender.
3319 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3320 it possible to be selective about when SMTP synchronization is enforced.
3322 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3324 32. Implemented hosts_connection_nolog.
3326 33. Added an ACL for QUIT.
3328 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3331 35. Added mailbox_size and mailbox_filecount to appendfile.
3333 36. Added control = no_multiline_responses to ACLs.
3335 37. There was a bug in the logic of the code that waits for the clock to tick
3336 in the case where the clock went backwards by a substantial amount such
3337 that the microsecond fraction of "now" was more than the microsecond
3338 fraction of "then" (but the whole seconds number was less).
3340 38. Added support for the libradius Radius client library this is found on
3341 FreeBSD (previously only the radiusclient library was supported).
3347 1. When certain lookups returned multiple values in the form name=value, the
3348 quoting of the values was not always being done properly. Specifically:
3349 (a) If the value started with a double quote, but contained no whitespace,
3351 (b) If the value contained whitespace other than a space character (i.e.
3352 tabs or newlines or carriage returns) it was not quoted.
3353 This fix has been applied to the mysql and pgsql lookups by writing a
3354 separate quoting function and calling it from the lookup code. The fix
3355 should probably also be applied to nisplus, ibase and oracle lookups, but
3356 since I cannot test any of those, I have not disturbed their existing code.
3358 2. A hit in the callout cache for a specific address caused a log line with no
3359 reason for rejecting RCPT. Now it says "Previous (cached) callout
3360 verification failure".
3362 3. There was an off-by-one bug in the queryprogram router. An over-long
3363 return line was truncated at 256 instead of 255 characters, thereby
3364 overflowing its buffer with the terminating zero. As well as fixing this, I
3365 have increased the buffer size to 1024 (and made a note to document this).
3367 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3368 when Exim is waiting for an SMTP response from a remote server, Exim
3369 restarts its select() call on the socket, thereby resetting its timeout.
3370 This is not a problem when such interrupts are rare. Somebody set up a cron
3371 job to run exiwhat every 2 minutes, which is less than the normal select()
3372 timeout (5 or 10 minutes). This meant that the select() timeout never
3373 kicked in because it was always reset. I have fixed this by comparing the
3374 time when an interrupt arrives with the time at the start of the first call
3375 to select(). If more time than the timeout has elapsed, the interrupt is
3376 treated as a timeout.
3378 5. Some internal re-factoring in preparation for the addition of Sieve
3379 extensions (by MH). In particular, the "personal" test is moved to a
3380 separate function, and given an option for scanning Cc: and Bcc: (which is
3381 not set for Exim filters).
3383 6. When Exim created an email address using the login of the caller as the
3384 local part (e.g. when creating a From: or Sender: header line), it was not
3385 quoting the local part when it contained special characters such as @.
3387 7. Installed new OpenBSD configuration files.
3389 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3390 try to make them clearer.
3392 9. Callout options, other than the timeout value, were being ignored when
3393 verifying sender addresses in header lines. For example, when using
3395 verify = header_sender/callout=no_cache
3397 the cache was (incorrectly) being used.
3399 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3400 only the Cygwin environment.
3402 11. When return_path_on_delivery was set as a log selector, if different remote
3403 addresses in the same message used different return paths and parallel
3404 remote delivery occurred, the wrong values would sometimes be logged.
3405 (Whenever a remote delivery process finished, the return path value from
3406 the most recently started remote delivery process was logged.)
3408 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3409 lines when AUTH and/or TLS are in use. This is the "received protocol"
3410 field. Exim used to use "asmtp" for authenticated SMTP, without any
3411 indication (in the protocol name) for TLS use. Now it follows the RFC and
3412 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3413 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3414 names appear in log lines as well as in Received: header lines.
3416 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3417 extensions, and comparison tests, and to fix some bugs.
3419 14. Changes to the "personal" filter test:
3421 (1) The test was buggy in that it was just doing the equivalent of
3422 "contains" tests on header lines. For example, if a user's address was
3423 anne@some.where, the "personal" test would incorrectly be true for
3425 To: susanne@some.where
3427 This test is now done by extracting each address from the header in turn,
3428 and checking the entire address. Other tests that are part of "personal"
3429 are now done using regular expressions (for example, to check local parts
3430 of addresses in From: header lines).
3432 (2) The list of non-personal local parts in From: addresses has been
3433 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3434 taken from the Sieve specification recommendations.
3436 (3) If the message contains any header line starting with "List-" it is
3437 treated as non-personal.
3439 (4) The test for "circular" in the Subject: header line has been removed
3440 because it now seems ill-conceived.
3442 15. Minor typos in src/EDITME comments corrected.
3444 16. Installed latest exipick from John Jetmore.
3446 17. If headers_add on a router specified a text string that was too long for
3447 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3448 of string_sprintf() is now avoided.
3450 18. $message_body_size was not set (it was always zero) when running the DATA
3451 ACL and the local_scan() function.
3453 19. For the "mail" command in an Exim filter, no default was being set for
3454 the once_repeat time, causing a random time value to be used if "once" was
3455 specified. (If the value happened to be <= 0, no repeat happened.) The
3456 default is now 0s, meaning "never repeat". The "vacation" command was OK
3457 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3458 (I found it when inspecting the code).
3460 20. There is now an overall timeout for performing a callout verification. It
3461 defaults to 4 times the callout timeout, which applies to individual SMTP
3462 commands during the callout. The overall timeout applies when there is more
3463 than one host that can be tried. The timeout is checked before trying the
3464 next host. This prevents very long delays if there are a large number of
3465 hosts and all are timing out (e.g. when the network connections are timing
3466 out). The value of the overall timeout can be changed by specifying an
3467 additional sub-option for "callout", called "maxwait". For example:
3469 verify = sender/callout=5s,maxwait=20s
3471 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3472 to the end before writing, but this should make it even safer).
3474 22. Exim was forgetting that it had advertised PIPELINING for the second and
3475 subsequent messages on an SMTP connection. It was also not resetting its
3476 memory on STARTTLS and an internal HELO.
3478 23. When Exim logs an SMTP synchronization error within a session, it now
3479 records whether PIPELINING has been advertised or not.
3481 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3482 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3483 rather than long int.
3485 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3487 26. Added the never_mail option to autoreply.
3493 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3494 crash if the getsockname() call failed; this can happen if a connection is
3495 closed very soon after it is established. The problem was simply in the
3496 order in which certain operations were done, causing Exim to try to write
3497 to the SMTP stream before it had set up the file descriptor. The bug has
3498 been fixed by making things happen in the correct order.
3504 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3505 before the connection was closed, thus losing the rejection response.
3507 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3508 some early Solaris releases, but causes trouble in current releases where
3509 socklen_t is defined.
3511 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3514 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3517 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3518 that releasing the top of it at the end releases what was used for sub-
3519 expansions (unless the block got too big). However, discard this block if
3520 the first thing is a variable or header, so that we can use its block when
3521 it is dynamic (useful for very large $message_headers, for example).
3523 6. Lookups now cache *every* query, not just the most recent. A new, separate
3524 store pool is used for this. It can be recovered when all lookup caches are
3525 flushed. Lookups now release memory at the end of their result strings.
3526 This has involved some general refactoring of the lookup sources.
3528 7. Some code has been added to the store_xxx() functions to reduce the amount
3529 of flapping under certain conditions.
3531 8. log_incoming_interface used to affect only the <= reception log lines. Now
3532 it causes the local interface and port to be added to several more SMTP log
3533 lines, for example "SMTP connection from", and rejection lines.
3535 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3537 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3539 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3540 could be overwritten at the end of the current message (or the start of a
3541 new message if it was set in a HELO ACL). The value is now preserved for
3542 the duration of the SMTP connection.
3544 12. If a transport had a headers_rewrite setting, and a matching header line
3545 contained an unqualified address, that address was qualified, even if it
3546 did not match any rewriting rules. The underlying bug was that the values
3547 of the flags that permit the existence of unqualified sender and recipient
3548 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3549 non-local messages, and by -bnq for local messages) were not being
3550 preserved with the message after it was received.
3552 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3553 "next input=" as part of the text comprising the host identity instead of
3554 the correct text. The code was using the same buffer for two different
3555 strings. However, depending on which order the printing function evaluated
3556 its arguments, the bug did not always show up. Under Linux, for example, my
3557 test suite worked just fine.
3559 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3560 doesn't work with some older versions of Perl. It has been changed to "my",
3561 which in any case is probably the better facility to use.
3563 15. A really picky compiler found some instances of statements for creating
3564 error messages that either had too many or two few arguments for the format
3567 16. The size of the buffer for calls to the DNS resolver has been increased
3568 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3569 for addresses that have a lot of PTR records. This alleviates a problem; it
3570 does not fully solve it.
3572 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3573 buffer now truncates the list and logs the incident, which is the same
3574 action as happens when Exim is looking up a host name and its aliases.
3575 Previously in this situation something unpredictable would happen;
3576 sometimes it was "internal error: store_reset failed".
3578 18. If a server dropped the connection unexpectedly when an Exim client was
3579 using GnuTLS and trying to read a response, the client delivery process
3580 crashed while trying to generate an error log message.
3582 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3583 string, for example:
3585 warn message = H1: something\nH2: something
3587 the text was added as a single header line from Exim's point of view
3588 though it ended up OK in the delivered message. However, searching for the
3589 second and subsequent header lines using $h_h2: did not work. This has been
3590 fixed. Similarly, if a system filter added multiple headers in this way,
3591 the routers could not see them.
3593 20. Expanded the error message when iplsearch is called with an invalid key to
3594 suggest using net-iplsearch in a host list.
3596 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3597 ACLs are no longer actually imposed (and a message to that effect is
3600 22. If a "gecos" field in a passwd entry contained escaped characters, in
3601 particular, if it contained a \" sequence, Exim got it wrong when building
3602 a From: or a Sender: header from that name. A second bug also caused
3603 incorrect handling when an unquoted " was present following a character
3604 that needed quoting.
3606 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3607 was not being matched caselessly.
3609 24. Arranged for all hyphens in the exim.8 source to be escaped with
3612 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3613 properly. Recipient callout cache records were still being keyed to include
3614 the sender, even when use_sender was set false. This led to far more
3615 callouts that were necessary. The sender is no longer included in the key
3616 when use_sender is false.
3618 26. Added "control = submission" modifier to ACLs.
3620 27. Added the ${base62d: operator to decode base 62 numbers.
3622 28. dnsdb lookups can now access SRV records.
3624 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3625 the configuration file.
3627 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3628 (-v) mode. This makes the output for a verbose queue run more intelligible.
3630 31. Added a use_postmaster feature to recipient callouts.
3632 32. Added the $body_zerocount variable, containing the number of binary zero
3633 bytes in the message body.
3635 33. The time of last modification of the "new" subdirectory is now used as the
3636 "mailbox time last read" when there is a quota error for a maildir
3639 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3641 35. Added +ignore_unknown as a special item in host lists.
3643 36. Code for decoding IPv6 addresses in host lists is now included, even if
3644 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3645 address was recognized as an IP address, but was then not correctly decoded
3646 into binary, causing unexpected and incorrect effects when compared with
3653 1. Very minor rewording of debugging text in manualroute to say "list of
3654 hosts" instead of "hostlist".
3656 2. If verify=header_syntax was set, and a header line with an unqualified
3657 address (no domain) and a large number of spaces between the end of the
3658 name and the colon was received, the reception process suffered a buffer
3659 overflow, and (when I tested it) crashed. This was caused by some obsolete
3660 code that should have been removed. The fix is to remove it!
3662 3. When running in the test harness, delay a bit after writing a bounce
3663 message to get a bit more predictability in the log output.
3665 4. Added a call to search_tidyup() just before forking a reception process. In
3666 theory, someone could use a lookup in the expansion of smtp_accept_max_
3667 per_host which, without the tidyup, could leave open a database connection.
3669 5. Added the variables $recipient_data and $sender_data which get set from a
3670 lookup success in an ACL "recipients" or "senders" condition, or a router
3671 "senders" option, similar to $domain_data and $local_part_data.
3673 6. Moved the writing of debug_print from before to after the "senders" test
3676 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3677 problems for message scanning, either using a data ACL, or using
3678 local_scan() because the Received: header was not generated till after they
3679 were called (in order to set the time as the time of reception completion).
3680 I have revised the way this works. The header is now generated after the
3681 body is received, but before the ACL or local_scan() are called. After they
3682 are run, the timestamp in the header is updated.
3688 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3689 before starting a queue runner without re-exec. This happened only when
3690 deliver_drop_privilege was set or when the Exim user was set to root. The
3691 effect of the bug was that timeouts during subsequent deliveries caused
3692 crashes instead of being properly handled. The handler is now left at its
3693 default (and expected) setting.
3695 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3696 message, again when deliver_drop_privilege is set or Exim is run as root.
3697 The bug described in (1) was not present in this case, but the tidying up
3698 of the other signals was missing. I have made the two cases consistent.
3700 3. The ignore_target_hosts setting on a manualroute router was being ignored
3701 for hosts that were looked up using the /MX notation.
3703 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3706 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3707 operated on the sender address. After changing the $sender_address to <>
3708 for the sender address verify, Exim was re-instated it as the original
3709 (before rewriting) address, but remembering that it had rewritten it, so it
3710 wasn't rewriting it again. This bug also had the effect of breaking the
3711 sender address verification caching when the sender address was rewritten.
3713 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3714 This has been changed so that if the ip literal address matches
3715 ignore_target_hosts, the router declines.
3717 7. Added expansion conditions match_domain, match_address, and match_local_
3718 part (NOT match_host).
3720 8. The placeholder for the Received: header didn't have a length field set.
3722 9. Added code to Exim itself and to exim_lock to test for a specific race
3723 condition that could lead to file corruption when using MBX delivery. The
3724 issue is with the lockfile that is created in /tmp. If this file is removed
3725 after a process has opened it but before that process has acquired a lock,
3726 there is the potential for a second process to recreate the file and also
3727 acquire a lock. This could lead to two Exim processes writing to the file
3728 at the same time. The added code performs the same test as UW imapd; it
3729 checks after acquiring the lock that its file descriptor still refers to
3730 the same named file.
3732 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3733 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3736 11. Added the smtp_active_hostname option. If used, this will typically be made
3737 to depend on the incoming interface address. Because $interface_address is
3738 not set up until the daemon has forked a reception process, error responses
3739 that can happen earlier (such as "too many connections") no longer contain
3742 12. If an expansion in a condition on a "warn" statement fails because a lookup
3743 defers, the "warn" statement is abandoned, and the next ACL statement is
3744 processed. Previously this caused the whole ACL to be aborted.
3746 13. Added the iplsearch lookup type.
3748 14. Added ident_timeout as a log selector.
3750 15. Added tls_certificate_verified as a log selector.
3752 16. Added a global option tls_require_ciphers (compare the smtp transport
3753 option of the same name). This controls incoming TLS connections.
3755 17. I finally figured out how to make tls_require_ciphers do a similar thing
3756 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3757 before starting the TLS session.
3759 18. Tabs are now shown as \t in -bP output.
3761 19. If the log selector return_path_on_delivery was set, Exim crashed when
3762 bouncing a message because it had too many Received: header lines.
3764 20. If two routers both had headers_remove settings, and the first one included
3765 a superfluous trailing colon, the final name in the first list and the
3766 first name in the second list were incorrectly joined into one item (with a
3767 colon in the middle).
3773 1. Added -C and -D options to the exinext utility, mainly to make it easier
3774 to include in the automated testing, but these could be helpful when
3775 multiple configurations are in use.
3777 2. The exinext utility was not formatting the output nicely when there was
3778 an alternate port involved in the retry record key, nor when there was a
3779 message id as well (for retries that were specific to a specific message
3780 and a specific host). It was also confused by IPv6 addresses, because of
3781 the additional colons they contain. I have fixed the IPv4 problem, and
3782 patched it up to do a reasonable job for IPv6.
3784 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3785 delivery, the log line now contains "pipelined" if PIPELINING was used.
3787 4. An SMTP transport process used to panic and die if the bind() call to set
3788 an explicit outgoing interface failed. This has been changed; it is now
3789 treated in the same way as a connect() failure.
3791 5. A reference to $sender_host_name in the part of a conditional expansion
3792 that was being skipped was still causing a DNS lookup. This no longer
3795 6. The def: expansion condition was not recognizing references to header lines
3796 that used bh_ and bheader_.
3798 7. Added the _cache feature to named lists.
3800 8. The code for checking quota_filecount in the appendfile transport was
3801 allowing one more file than it should have been.
3803 9. For compatibility with Sendmail, the command line option
3811 and sets the incoming protocol and host name (for trusted callers). The
3812 host name and its colon can be omitted when only the protocol is to be set.
3813 Note the Exim already has two private options, -pd and -ps, that refer to
3814 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3815 "s", but I don't think that's a major issue.
3817 10. A number of refactoring changes to the code, none of which should affect
3820 (a) The number of logging options was getting close to filling up the
3821 32-bit word that was used as a bit map. I have split them into two classes:
3822 those that are passed in the argument to log_write(), and those that are
3823 only ever tested independently outside of that function. These are now in
3824 separate 32-bit words, so there is plenty of room for expansion again.
3825 There is no change in the user interface or the logging behaviour.
3827 (b) When building, for example, log lines, the code previously used a
3828 macro that called string_cat() twice, in order to add two strings. This is
3829 not really sufficiently general. Furthermore, there was one instance where
3830 it was actually wrong because one of the argument was used twice, and in
3831 one call a function was used. (As it happened, calling the function twice
3832 did not affect the overall behaviour.) The macro has been replaced by a
3833 function that can join an arbitrary number of extra strings onto a growing
3836 (c) The code for expansion conditions now uses a table and a binary chop
3837 instead of a serial search (which was left over from when there were very
3838 few conditions). Also, it now recognizes conditions like "pam" even when
3839 the relevant support is not compiled in: a suitably worded error message is
3840 given if an attempt is made to use such a condition.
3842 11. Added ${time_interval:xxxxx}.
3844 12. A bug was causing one of the ddress fields not to be passed back correctly
3845 from remote delivery subprocesses. The field in question was not being
3846 subsequently used, so this caused to problems in practice.
3848 13. Added new log selectors queue_time and deliver_time.
3850 14. Might have fixed a bug in maildirsizefile handling that threw up
3851 "unexpected character" debug warnings, and recalculated the data
3852 unnecessarily. In any case, I expanded the warning message to give more
3855 15. Added the message "Restricted characters in address" to the statements in
3856 the default ACL that block characters like @ and % in local parts.
3858 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3859 Three changes have been made:
3861 (a) There was a serious bug; a negative response to MAIL caused the whole
3862 recipient domain to be cached as invalid, thereby blocking all messages
3863 to all local parts at the same domain, from all senders. This bug has
3864 been fixed. The domain is no longer cached after a negative response to
3865 MAIL if the sender used is not empty.
3867 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3870 (c) A new callout option, "use_sender" has been added for people who want
3871 the modified behaviour.
3877 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3880 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3881 indeed breaks things for older releases.
3883 3. Added additional logging to the case where there is a problem reading data
3884 from a filter that is running in a subprocess using a pipe, in order to
3885 try to track down a specific problem.
3887 4. Testing facility fudge: when running in the test harness and attempting
3888 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3889 getting "No route to host". Convert this to a timeout.
3891 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3894 6. Some OS don't have socklen_t but use size_t instead. This affects the
3895 fifth argument of getsockopt() amongst other things. This is now
3896 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3897 can be set for individual OS. I have set it for SunOS5, OSF1, and
3898 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3899 some earlier ones do not.
3901 7. Change 4.30/15 was not doing the test caselessly.
3903 8. The standard form for an IPv6 address literal was being rejected by address
3904 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3905 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3906 this, as well as the form without the "IPv6" on the front (but only when
3907 address literals are enabled, of course).
3909 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3911 10. Exim crashed if a message with an empty sender address specified by -f
3912 encountered a router with an errors_to setting. This could be provoked only
3913 by a command such as
3917 where an empty string was supplied; "<>" did not hit this bug.
3919 11. Installed PCRE release 4.5.
3921 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3922 remained set. It is now erased.
3924 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3925 times from message ids (which are base 36 rather than the normal 62).
3927 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3928 were being counted as actual protocol errors, and logged if the log
3929 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3930 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3931 and DATA following a set of rejected RCPTs do not count as protocol errors.
3932 In other words, Exim assumes they were pipelined, though this may not
3933 actually be the case. Of course, in all cases the client gets an
3934 appropriate error code.
3936 15. If a lookup fails in an ACL condition, a message about the failure may
3937 be available; it is used if testing the ACL cannot continue, because most
3938 such messages specify what the cause of the deferral is. However, some
3939 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3940 that caused an old message to be retained and used if a later statement
3941 caused a defer, replacing the real cause of the deferral.
3943 16. If an IP address had so many PTR records that the DNS lookup buffer
3944 was not large enough to hold them, Exim could crash while trying to process
3945 the truncated data. It now detects and logs this case.
3947 17. Further to 4.21/58, another change has been made: if (and only if) the
3948 first line of a message (the first header line) ends with CRLF, a bare LF
3949 in a subsequent header line has a space inserted after it, so as not to
3950 terminate the header.
3952 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3953 unnecessarily, used atoi() instead of strtol(), and didn't check the
3954 termination when getting file sizes from file names by regex.
3956 19. Completely re-implemented the support for maildirsize files, in the light
3957 of a number of problems with the previous contributed implementation
3958 (4.30/29). In particular:
3960 . If the quota is zero, the maildirsize file is maintained, but no quota is
3963 . If the maildir directory does not exist, it is created before any attempt
3964 to write a maildirsize file.
3966 . The quota value in the file is just a cache; if the quota is changed in
3967 the transport, the new value overrides.
3969 . A regular expression is available for excluding directories from the
3972 20. The autoreply transport checks the characters in options that define the
3973 message's headers; it allows continued headers, but it was checking with
3974 isspace() after an embedded newline instead of explicitly looking for a
3977 21. If all the "regular" hosts to which an address was routed had passed their
3978 expiry times, and had not reached their retry times, the address was
3979 bounced, even if fallback hosts were defined. Now Exim should go on to try
3982 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3983 equivalent code in the SMTP transport. Some hosts send humungous responses
3984 to HELO/EHLO, more than 1024 it seems.
3986 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3987 gives compiler warnings in some environments. I've now done it "properly",
3990 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3991 (because of problems with the built-in one) was declared to return uschar *
3992 instead of char *, causing compiler failure.
3994 25. Fixed a file descriptor leak when processing alias/forward files.
3996 26. Fixed a minor format string issue in dbfn.c.
3998 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4000 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4001 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4004 29. When a sender address is verified, it is cached, to save repeating the test
4005 when there is more than one recipient in a message. However, when the
4006 verification involves a callout, it is possible for different callout
4007 options to be set for different recipients. It is too complicated to keep
4008 track of this in the cache, so now Exim always runs a verification when a
4009 callout is required, relying on the callout cache for the optimization.
4010 The overhead is duplication of the address routing, but this should not be
4013 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4014 to be verified with callout=postmaster, and the main callout worked but the
4015 postmaster check failed, the verification correctly failed. However, if a
4016 subsequent RCPT command asked for sender verification *without* the
4017 postmaster check, incorrect caching caused this verification also to fail,
4020 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4021 it was not caching the DNS options (qualify_single, search_parents) that
4022 were used when the lookup failed. A subsequent lookup with different
4023 options therefore always gave the same answer, though there were cases
4024 where it should not have. (Example: a "domains = !$mx_any" option on a
4025 dnslookup router: the "domains" option is always processed without any
4026 widening, but the router might have qualify_single set.) Now Exim uses the
4027 cached value only when the same options are set.
4029 32. Added John Jetmore's "exipick" utility to the distribution.
4031 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4032 than a timeout (e.g. a certificate is required, and is not provided), an
4033 Exim server now closes the connection immediately. Previously it waited for
4034 the client to close - but if the client is SSL, it seems that they each
4035 wait for each other, leading to a delay before one of them times out.
4037 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4038 maintained 0.8.x compatibility because I don't think many are using it, and
4039 it is clearly obsolete.
4041 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4044 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4045 client certificate was expired. A simple patch fixes this, though I don't
4046 understand the full logic of why the verify callback is called multiple
4049 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4050 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4051 which causes problems with some clients (such as the Certicom SSL Plus
4052 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4053 disables the coutermeasure allowing Eudora to connect."
4055 38. Exim was not checking that a write() to a log file succeeded. This could
4056 lead to Bad Things if a log got too big, in particular if it hit a file
4057 size limit. Exim now panics and dies if it cannot write to a log file, just
4058 as it does if it cannot open a log file.
4060 39. Modified OS/Makefile-Linux so that it now contains
4062 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4064 The two -D definitions ensure that Exim is compiled with large file
4065 support, which makes it possible to handle log files that are bigger than
4068 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4069 instance) a domain was checked against a named list that involved a lookup,
4070 causing $domain_data to be set, then another domain was checked against the
4071 same list, then the first domain was re-checked, the value of $domain_data
4072 after the final check could be wrong. In particular, if the second check
4073 failed, it could be set empty. This bug probably also applied to
4076 41. The strip_trailing_dot option was not being applied to the address given
4077 with the -f command-line option.
4079 42. The code for reading a message's header from the spool was incrementing
4080 $received_count, but never initializing it. This meant that the value was
4081 incorrect (doubled) while delivering a message in the same process in which
4082 it was received. In the most common configuration of Exim, this never
4083 happens - a fresh exec is done - but it can happen when
4084 deliver_drop_privilege is set.
4086 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4087 it now includes up to 150 characters of the unexpected data in the log
4090 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4091 and building data strings. The size of both of these buffers was 10 000
4092 bytes - far larger than anybody would *ever* want, thought I. Needless to
4093 say, somebody hit the limit. I have increased the maximum line length to
4094 20 000 and the maximum data length of concatenated lines to 100 000. I have
4095 also fixed two bugs, because there was no checking on these buffers. Tsk,
4096 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4097 buffer is too small.
4099 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4100 lsearch lookups. Now it does.
4102 46. When parsing a route_list item in a manualroute router, a fixed-length
4103 buffer was used for the list of hosts. I made this 1024 bytes long,
4104 thinking that nobody would ever have a list of hosts that long. Wrong.
4105 Somebody had a whole pile of complicated expansion conditions, and the
4106 string was silently truncated, leading to an expansion error. It turns out
4107 that it is easier to change to an unlimited length (owing to other changes
4108 that have happened since this code was originally written) than to build
4109 structure for giving a limitation error. The length of the item that
4110 expands into the list of hosts is now unlimited.
4112 47. The lsearch lookup could not handle data where the length of text line was
4113 more than 4095 characters. Such lines were truncated, leading to shortened
4114 data being returned. It should now handle lines of any length.
4116 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4117 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4120 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4121 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4122 no longer happens in comment lines. A list of replacements is now placed
4123 at the head of all of the source files, except those whose only change is
4124 to replace PERL_COMMAND in the very first #! line.
4126 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4127 messages on the queue, with a bottom-up merge sort, using code contributed
4128 by Michael Haardt. This should make operations like -bp somewhat faster on
4129 large queues. It won't affect queue runners, except when queue_run_in_order
4132 51. Installed eximstats 1.31 in the distribution.
4134 52. Added support for SRV lookups to the dnslookup router.
4136 53. If an ACL referred to $message_body or $message_body_end, the value was not
4137 reset for any messages that followed in the same SMTP session.
4139 54. The store-handling optimization for building very long strings was not
4140 differentiating between the different store pools. I don't think this
4141 actually made any difference in practice, but I've tidied it.
4143 55. While running the routers to verify a sender address, $sender_address
4144 was still set to the sender address. This is wrong, because when routing to
4145 send a bounce to the sender, it would be empty. Therefore, I have changed
4146 it so that, while verifying a sender address, $sender_address is set to <>.
4147 (There is no change to what happens when verifying a recipient address.)
4149 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4150 target A or AAAA records (if not already returned) without resetting the
4151 qualify_single or search_parents options of the DNS resolver. These are
4152 inappropriate in this case because the targets of MX and SRV records must
4153 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4154 target that, when qualified, matched something in the local domain. These
4155 two options are now turned off when doing these lookups.
4157 57. It seems that at least some releases of Reiserfs (which does not have the
4158 concept of a fixed number of inodes) returns zero and not -1 for the
4159 number of available inodes. This interacted badly with check_spool_inodes,
4160 which assumed that -1 was the "no such thing" setting. What I have done is
4161 to check that the total number of inodes is greater than zero before doing
4162 the test of how many are available.
4164 58. When a "warn" ACL statement has a log_message modifier, the message is
4165 remembered, and not repeated. This is to avoid a lot of repetition when a
4166 message has many recipients that cause the same warning to be written.
4167 Howewer, Exim was preserving the list of already written lines for an
4168 entire SMTP session, which doesn't seem right. The memory is now reset if a
4169 new message is started.
4171 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4172 debugging output unless log_rewrite was also set.
4174 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4175 of (int)(handle) when we know that handle contains (void *)(-1).
4177 61. The Exim daemon panic-logs an error return when it closes the incoming
4178 connection. However "connection reset by peer" seems to be common, and
4179 isn't really an error worthy of noting specially, so that particular error
4182 62. When Exim is trying to find all the local interfaces, it used to panic and
4183 die if the ioctl to get the interface flags failed. However, it seems that
4184 on at least one OS (Solaris 9) it is possible to have an interface that is
4185 included in the list of interfaces, but for which you get a failure error
4186 for this call. This happens when the interface is not "plumbed" into a
4187 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4188 failure of the "get flags" call assumes that the interface is down.
4190 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4191 makes life easier for people who are doing arithmetic on fields extracted
4192 from dates, where you often get leading zeros that should not be
4193 interpreted as octal.
4195 64. Added qualify_domain to the redirect router, to override the global
4198 65. If a pathologically long header line contained very many addresses (the
4199 report of this problem mentioned 10 000) and each of them was rewritten,
4200 Exim could use up a very large amount of memory. (It kept on making new
4201 copies of the header line as it rewrote, and never released the old ones.)
4202 At the expense of a bit more processing, the header rewriting function has
4203 been changed so that it no longer eats memory in this way.
4205 66. The generation of the Received: header has been moved from the time that a
4206 message starts to be received, to the time that it finishes. The timestamp
4207 in the Received: header should now be very close to that of the <= log
4208 line. There are two side-effects of this change:
4210 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4211 logged header lines no longer include the local Received: line, because
4212 it has not yet been created. The same applies to a copy of the message
4213 that is returned to a non-SMTP sender when a message is rejected.
4215 (b) When a filter file is tested using -bf, no additional Received: header
4216 is added to the test message. After some thought, I decided that this
4219 This change does not affect the value of $received_for. It is still set
4220 after address rewriting, but before local_scan() is called.
4222 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4224 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4225 gave an unhelpful panic error message, and a defer error. I have managed to
4226 change this behaviour so that it now rejects any supplied certificate,
4227 which seems right, as the list of acceptable certificates is empty.
4229 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4230 gave an unhelpful defer error. I have not managed to make this reject any
4231 supplied certificates, but the error message it gives is "no certificate
4232 supplied", which is not helpful.
4234 70. exigrep's output now also includes lines that are not associated with any
4235 message, but which match the given pattern. Implemented by a patch from
4236 Martin Sluka, which also tidied up the Perl a bit.
4238 71. Recipient callout verification, like sender verification, was using <> in
4239 the MAIL FROM command. This isn't really the right thing, since the actual
4240 sender may affect whether the remote host accepts the recipient or not. I
4241 have changed it to use the actual sender in the callout; this means that
4242 the cache record is now keyed on a recipient/sender pair, not just the
4243 recipient address. There doesn't seem to be a real danger of callout loops,
4244 since a callout by the remote host to check the sender would use <>.
4245 [SEE ABOVE: changed after hitting problems.]
4247 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4248 temporary errors. However, in the case of such a code being given after
4249 the end of a data transmission (i.e. after ".") Exim was failing to write
4250 a retry record for the message. (Yes, there was some broken host that was
4251 actually sending 8xx at this point.)
4253 73. An unknown lookup type in a host list could cause Exim to panic-die when
4254 the list was checked. (An example that provoked this was putting <; in the
4255 middle of a list instead of at the start.) If this happened during a DATA
4256 ACL check, a -D file could be left lying around. This kind of configuration
4257 error no longer causes Exim to die; instead it causes a defer errror. The
4258 incident is still logged to the main and panic logs.
4260 74. Buglet left over from Exim 3 conversion. The message "too many messages
4261 in one connection" was written to the rejectlog but not the mainlog, except
4262 when address rewriting (yes!) was being logged.
4264 75. Added write_rejectlog option.
4266 76. When a system filter was run not as root (that is, when system_filter_user
4267 was set), the values of the $n variables were not being returned to the
4268 main process; thus, they were not subsequently available in the $sn
4271 77. Added +return_path_on_delivery log selector.
4273 78. A connection timeout was being treated differently from recipients deferred
4274 when testing hosts_max_try with a message that was older than the host's
4275 retry timeout. (The host should not be counted, thus allowing all hosts to
4276 be tried at least once before bouncing.) This may have been the cause of an
4277 occasionally reported bug whereby a message would remain on the queue
4278 longer than the retry timeout, but would be bounced if a delivery was
4279 forced. I say "may" because I never totally pinned down the problem;
4280 setting up timeout/retry tests is difficult. See also the next item.
4282 79. The ultimate address timeout was not being applied to errors that involved
4283 a combination of host plus message (for example, a timeout on a MAIL
4284 command). When an address resolved to a number of possible hosts, and they
4285 were not all tried for each delivery (e.g. because of hosts_max_try), a
4286 message could remain on the queue longer than the retry timeout.
4288 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4291 81. Fixed an obscure SMTP outgoing bug which required at least the following
4292 conditions: (a) there was another message waiting for the same server;
4293 (b) the server returned 5xx to all RCPT commands in the first message so
4294 that the message was not completed; (c) the server dropped the connection
4295 or gave a negative response to the RSET that Exim sends to abort the
4296 transaction. The observed case was a dropped connection after DATA that had
4297 been sent in pipelining mode. That is, the server had advertised PIPELINING
4298 but was not implementing it correctly. The effect of the bug was incorrect
4299 behaviour, such as trying another host, and this could lead to a crash.
4305 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4306 and daemon.c were passed as pointers to ints; they should have been
4307 pointers to socklen_t variables (which are typically unsigned ints).
4309 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4312 3. Fixed a really odd bug that affected only the testing scheme; patching a
4313 certain fixed string in the binary changed the value of another string that
4314 happened to be identical to the end of the original first string.
4316 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4317 name", it returns that address as the IP address. On some operating
4318 systems (e.g. Solaris), it also passes back the IP address string as the
4319 "host name". However, on others (e.g. Linux), it passes back an empty
4320 string. Exim wasn't checking for this, and was changing the host name to an
4321 empty string, assuming it had been canonicized.
4323 5. Although rare, it is permitted to have more than one PTR record for a given
4324 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4325 all the names associated with an address, because they do in Solaris.
4326 However, it seems that they do not in Linux for data that comes from the
4327 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4328 I found this out when I moved to a new Linux workstation and tried to run
4329 the Exim test suite.
4331 To get round this problem I have changed the code so that it now does its
4332 own call to the DNS to look up PTR records when searching for a host name.
4333 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4334 addresses that are only in /etc/hosts are still found.
4336 This behaviour is, however, controlled by an option called host_lookup_
4337 order, which defaults to "bydns:byaddr". If people want to use the other
4338 order, or indeed, just use one or the other means of lookup, they can
4339 specify it in this variable.
4341 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4342 some operating systems, this comes back from gethostbyaddr() as an empty
4343 string, and this is what Exim used to test for. However, it seems that in
4344 other systems, "." is yielded. Exim now tests for this case too.
4346 7. The values of check_spool_space and check_log_space are now held internally
4347 as a number of kilobytes instead of an absolute number of bytes. If a
4348 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4349 kilobyte. This means that much larger values can be stored.
4351 8. Exim monitor: an attempt to get the action menu when not actually pointing
4352 at a message produces an empty menu entitled "No message selected". This
4353 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4354 no entries in it ("Shell widget menu has zero width and/or height"). So I
4355 have added a single, blank menu entry in this case.
4357 9. Added ${quote_local_part.
4359 10. MIME decoding is now applied to the contents of Subject: header lines when
4362 11. Now that a reference to $sender_host_address automatically causes a reverse
4363 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4364 host lookup before query-style lookups in lists that might use this
4365 variable. This has therefore been abolished, and the "net-" prefix is no
4366 longer necessary for query-style lookups.
4368 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4369 have been a typo for LFLAGS, so it has been changed.
4371 13. The install script calls Exim with "-C /dev/null" in order to find the
4372 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4373 to be output. Howeve, since Exim outputs its version number before the
4374 error, it didn't break the script. It just looked ugly. I fixed this by
4375 always allowing "-C /dev/null" if the caller is root.
4377 14. Ignore overlarge ACL variable number when reading spool file - insurance
4378 against a later release with more variables having written the file.
4380 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4381 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4382 this, as well as the form without the "IPv6" on the front.
4384 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4385 OS/Makefile-Darwin file.
4387 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4388 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4390 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4391 variable. However, some OS use ints for inodes. I've added cast to long int
4392 to get rid of the compiler warning.
4394 19. I had forgotten to lock out "/../" in configuration file names when
4395 ALT_CONFIG_PREFIX was set.
4397 20. Routers used for verification do not need to specify transports. However,
4398 if such a router generated a host list, and callout was configured, Exim
4399 crashed, because it could not find a port number from the (non-existent)
4400 transport. It now assumes port 25 in this circumstance.
4402 21. Added the -t option to exigrep.
4404 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4405 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4406 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4407 them is set, LOOKUP_LSEARCH is forced.
4409 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4411 24. Added sender and host information to the "rejected by local_scan()" log
4412 line; previously there was no indication of these.
4414 25. Added .include_if_exists.
4416 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4417 Linux. It turns out that not all file systems support this. Apparently some
4418 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4419 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4420 not supported on the file descriptor, is now ignored when Exim is trying to
4421 sync a directory. This applies only to Linux.
4423 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4425 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4426 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4429 29. Added support for maildirsize files from supplied patch (modified a bit).
4431 30. The use of :fail: followed by an empty string could lead Exim to respond to
4432 sender verification failures with (e.g.):
4434 550 Verification failed for <xxx>
4435 550 Sender verify failed
4437 where the first response line was missing the '-' that indicates it is not
4438 the final line of the response.
4440 31. The loop for finding the name of the user that called Exim had a hardwired
4441 limit of 10; it now uses the value of finduser_retries, which is used for
4442 all other user lookups.
4444 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4447 33. Exim was neglecting to zero errno before one call of strtol() when
4448 expanding a string and expecting an integer value. On some systems this
4449 resulted in spurious "integer overflow" errors. Also, it was casting the
4450 result into an int without checking.
4452 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4453 did not work. The code looks as if it has *never* worked, though it appears
4454 to have been documented since at least releast 1.62. I have made it work.
4456 35. The "timeout_DNS" error in retry rules, also documented since at least
4457 1.62, also never worked. As it isn't clear exactly what this means, and
4458 clearly it isn't a major issue, I have abolished the feature by treating it
4459 as "timeout", and writing a warning to the main and panic logs.
4461 36. The display of retry rules for -brt wasn't always showing the error code
4464 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4465 timeout_connect_A, timeout_connect_MX.
4467 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4468 to the empty sender.
4470 39. The daemon was not analysing the content of -oX till after it had closed
4471 stderr and disconnected from the controlling terminal. This meant that any
4472 syntax errors were only noted on the panic log, and the return code from
4473 the command was 0. By re-arranging the code a little, I've made the
4474 decoding happen first, so such errors now appear on stderr, and the return
4475 code is 1. However, the actual setting up of the sockets still happens in
4476 the disconnected process, so errors there are still only recorded on the
4479 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4480 connections (as happens on some IP stacks) was logged at start up time as
4481 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4482 it from "IPv6 and IPv4", which means that two separate sockets are being
4485 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4486 says whether AF_INET or AF_INET6 was passed as an argument.
4488 42. Exiwhat output was messed up when time zones were included in log
4491 43. Exiwhat now gives more information about the daemon's listening ports,
4492 and whether -tls-on-connect was used.
4494 44. The "port" option of the smtp transport is now expanded.
4496 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4497 silently ignored. Now an error message is written to the main and panic
4500 46. There's a new ACL modifier called "logwrite" which writes to a log file
4501 as soon as it is encountered.
4503 47. Added $local_user_uid and $local_user_gid at routing time.
4505 48. Exim crashed when trying to verify a sender address that was being
4508 49. Exim was recognizing only a space character after ".include". It now also
4509 recognizes a tab character.
4511 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4512 extracting the relevant information from the specification. The man page no
4513 longer contains scrambled data for the -d option, and I've added a section
4514 at the front about calling Exim under different names.
4516 51. Added "extra_headers" argument to the "mail" command in filter files.
4518 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4521 53. Installed eximstats 1.29.
4523 54. Added transport_filter_timeout as a generic transport option.
4525 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4526 Cc: header lines. This was required by RFC 822, but it not required by RFC
4529 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4530 incoming messages that did not have them. Now it does so only if the
4531 message originates locally, that is, if there is no associated remote host
4532 address. When Resent- header lines are present, this applies to the Resent-
4533 lines rather than the non-Resent- lines.
4535 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4536 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4538 58. Messages for configuration errors now include the name of the main
4539 configuration file - useful now that there may be more than one file in a
4540 list (.included file names were always shown).
4542 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4543 for those rare installations that do not start the daemon as root or run it
4544 setuid root. I've cut out the call to initgroups() if the daemon is not
4547 60. The Exim user and group can now be bound into the binary as text strings
4548 that are looked up at the start of Exim's processing.
4550 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4552 62. Added $mailstore_basename variable.
4554 63. Installed patch to sieve.c from Michael Haardt.
4556 64. When Exim failed to open the panic log after failing to open the main log,
4557 the original message it was trying to log was written to stderr and debug
4558 output, but if they were not available (the usual case in production), it
4559 was lost. Now it is written to syslog before the two lines that record the
4560 failures to open the logs.
4562 65. Users' Exim filters run in subprocesses under the user's uid. It is
4563 possible for a "deliver" command or an alias in a "personal" command to
4564 provoke an address rewrite. If logging of address rewriting is configured,
4565 this fails because the process is not running as root or exim. There may be
4566 a better way of dealing with this, but for the moment (because 4.30 needs
4567 to be released), I have disabled address rewrite logging when running a
4568 filter in a non-root, non-exim process.
4574 1. The buildconfig auxiliary program wasn't quoting the value set for
4575 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4576 not defined. This bug was masked in 4.22 by the effect that was fixed in
4579 2. Some messages that were rejected after a message id was allocated were
4580 shown as "incomplete" by exigrep. It no longer does this for messages that
4581 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4583 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4584 have allow_domain_literals set, the ID did not get logged in the <= line.
4585 Domain literals are now always recognized in Message-ID: header lines.
4587 4. The first argument for a ${extract expansion item is the key name or field
4588 number. Leading and trailing spaces in this item were not being ignored,
4589 causing some misleading effects.
4591 5. When deliver_drop_privilege was set, single queue runner processes started
4592 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4593 same command in the process it spins off) were not dropping privilege.
4595 6. When the daemon running as "exim" started a queue runner, it always
4596 re-executed Exim in the spun-off process. This is a waste of effort when
4597 deliver_drop_privilege is set. The new process now just calls the
4598 queue-runner function directly.
4604 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4607 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4608 ignored. Though the use of -oP was forcing the writing of a pid file, it
4609 was always written to the default place.
4611 3. If the message "no IP address found for host xxxx" is generated during
4612 incoming verification, it is now followed by identification of the incoming
4613 connection (so you can more easily find what provoked it).
4615 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4617 5. Added some features to "harden" Exim a bit more against certain attacks:
4619 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4620 be put in Local/Makefile. This is like the never_users runtime option,
4621 but it cannot be overridden. The default setting is "root".
4623 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4624 prefix string with which any file named in a -C command line option
4627 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4628 is retained for -C and -D only if the caller of Exim is root. Without
4629 it, the exim user may also use -C and -D and retain privilege.
4631 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4632 command line option is disabled.
4634 6. Macro names set by the -D option must start with an upper case letter, just
4635 like macro names defined in the configuration file.
4637 7. Added "dereference=" facility to LDAP.
4639 8. Two instances of the typo "uknown" in the source files are fixed.
4641 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4642 the Configure-Makefile script screwed up while processing it.
4644 10. Incorporated PCRE 4.4.
4646 11. The SMTP synchronization check was not operating right at the start of an
4647 SMTP session. For example, it could not catch a HELO sent before the client
4648 waited for the greeting. There is now a check for outstanding input at the
4649 point when the greeting is written. Because of the duplex, asynchronous
4650 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4651 way, but not yet received, when the check is performed.
4653 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4654 on TCP/IP sockets, because this apparently causes some broken clients to
4657 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4658 unchanged) from the Cygwin maintainer.
4660 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4661 was supported instead of testing for mbx. Effectively a typo.
4663 15. The spa authenticator server code was not checking that the input it
4664 received was valid base64.
4666 16. The debug output line for the "set" modifier in ACLs was not showing the
4667 name of the variable that was being set.
4669 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4671 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4672 lookup on its own. Something else was needed to trigger the lookup. For
4673 example, a match in host_lookup or the need for a host name in a host list.
4674 Now, if $sender_host_name is referenced and the host name has not yet been
4675 looked up, a lookup is performed. If the lookup fails, the variable remains
4676 empty, and $host_lookup_failed is set to "1".
4678 19. Added "eqi" as a case-independent comparison operator.
4680 20. The saslauthd authentication condition could segfault if neither service
4681 nor realm was specified.
4683 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4684 error message that was logged was misleading, and incoming SMTP
4685 connections were dropped. The message is now more accurate, and temporary
4686 errors are given to SMTP connections.
4688 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4689 (except RSET) with a 421 error, until QUIT is received. However, it was
4690 failing to send a response to QUIT.
4692 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4693 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4694 ACL is now run after the helo_try_verify_hosts code.
4696 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4697 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4698 case-independent, so other case variants are also recognized). Apparently
4699 some systems use these upper case variants.
4701 25. If more than two messages were waiting for the same host, and a transport
4702 filter was specified for the transport, Exim sent two messages over the
4703 same TCP/IP connection, and then failed with "socket operation on non-
4704 socket" when it tried to send the third.
4706 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4708 27. The extern definition of crypt16() in expand.c was not being excluded when
4709 the OS had its own crypt16() function.
4711 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4712 as a preferred synonym for return_size_limit, both as an option and as an
4715 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4717 30. Changed the default configuration ACL to relax the local part checking rule
4718 for addresses that are not in any local domains. For these addresses,
4719 slashes and pipe symbols are allowed within local parts, but the sequence
4720 /../ is explicitly forbidden.
4722 31. SPA server authentication was not clearing the challenge buffer before
4725 32. log_message in a "warn" ACL statement was writing to the reject log as
4726 well as to the main log, which contradicts the documentation and doesn't
4727 seem right (because no rejection is happening). So I have stopped it.
4729 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4730 However, I am unable to do any testing of this.
4732 34. Fixed an infelicity in the appendfile transport. When checking directories
4733 for a mailbox, to see if any needed to be created, it was accidentally
4734 using path names with one or more superfluous leading slashes; tracing
4735 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4737 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4738 contents are added to the log line that is written for every discarded
4739 recipient. (Previously a log_message setting was ignored.)
4741 36. The ${quote: operator now quotes the string if it is empty.
4743 37. The install script runs exim in order to find its version number. If for
4744 some reason other than non-existence or emptiness, which it checks, it
4745 could not run './exim', it was installing it with an empty version number,
4746 i.e. as "exim-". This error state is now caught, and the installation is
4749 38. An argument was missing from the function that creates an error message
4750 when Exim fails to connect to the socket for saslauthd authentication.
4751 This could cause Exim to crash, or give a corrupted message.
4753 39. Added isip, isip4, and isip6 to ${if conditions.
4755 40. The ACL variables $acl_xx are now saved with the message, and can be
4756 accessed later in routers, transports, and filters.
4758 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4759 strings in the file are not string-expanded.
4761 42. If a MAIL command specified a SIZE value that was too large to fit into an
4762 int variable, the check against message_size_limit failed. Such values are
4763 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4764 day this will have to be increased, but I don't think I want to be around
4765 when emails are that large.
4772 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4773 iconv() is not standard in FreeBSD.
4775 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4776 IPv6 enabled. The observed symptom was a segmentation fault on return from
4777 the function os_common_find_running_interfaces() in src/os.c.
4779 3. In the check_special_case() function in daemon.c I had used "errno" as an
4780 argument name, which causes warnings on some systems. This was basically a
4781 typo, since it was named "eno" in the comments!
4783 4. The code that waits for the clock to tick (at a resolution of some fraction
4784 of a second) so as to ensure message-id uniqueness was always waiting for
4785 at least one whole tick, when it could have waited for less. [This is
4786 almost certainly not relevant at current processor speeds, where it is
4787 unlikely to ever wait at all. But we try to future-proof.]
4789 5. The function that sleeps for a time interval that includes fractions of a
4790 second contained a race. It did not block SIGALRM between setting the
4791 timer, and suspending (a couple of lines later). If the interval was short
4792 and the sigsuspend() was delayed until after it had expired, the suspension
4793 never ended. On busy systems this could lead to processes getting stuck for
4796 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4797 process, before it forks any delivery processes. The open lookup caching
4798 mechanism meant that the open file or database connection was passed into
4799 the delivery process. The problem was that delivery processes always tidy
4800 up cached lookup data. This could cause a problem for the next delivery
4801 process started by the queue runner, because the external queue runner
4802 process does not know about the closure. So the next delivery process
4803 still has data in the lookup cache. In the case of a file lookup, there was
4804 no problem because closing a file descriptor in a subprocess doesn't affect
4805 the parent. However, if the lookup was caching a connection to a database,
4806 the connection was closed, and the second delivery process was likely to
4807 see errors such as "PGSQL: query failed: server closed the connection
4808 unexpectedly". The problem has been fixed by closing all cached lookups
4809 in a queue runner before running a delivery process.
4811 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4812 seem to have the "const" qualifier which it has on other OS. I've
4815 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4816 *of the same type* (client or server) with the same public name that an
4817 error should be diagnosed.
4819 9. When Exim looked up a host name for an IP address, but failed to find the
4820 original IP address when looking up the host name (a safety check), it
4821 output the message "<ip address> does not match any IP for NULL", which was
4822 confusing, to say the least. The bug was that the host name should have
4823 appeared instead of "NULL".
4825 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4826 user that is built into the binary, and the -C or -D options is used, root
4827 privilege is dropped before the configuration file is read. In addition,
4828 logging is switched to stderr instead of the normal log files. If the
4829 configuration then re-defines the Exim user, the unprivileged environment
4830 is probably not what is expected, so Exim logs a panic warning message (but
4833 However, if deliver_drop_privilege is set, the unprivileged state may well
4834 be exactly what is intended, so the warning has been cut out in that case,
4835 and Exim is allowed to try to write to its normal log files.
4841 1. smtp_return_error_details was not giving details for temporary sender
4842 or receiver verification errors.
4844 2. Diagnose a configuration error if two authenticators have the same public
4847 3. Exim used not to create the message log file for a message until the first
4848 delivery attempt. This could be confusing when incoming messages were held
4849 for policy or load reasons. The message log file is now created at the time
4850 the message is received, and an initial "Received" line is written to it.
4852 4. The automatically generated man page for command line options had a minor
4853 bug that caused no ill effects; however, a more serious problem was that
4854 the procedure for building the man page automatically didn't always
4855 operate. Consequently, release 4.20 contains an out-of-date version. This
4856 shouldn't happen again.
4858 5. When building Exim with embedded Perl support, the script that builds the
4859 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4860 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4862 6. The freeze_tell option was not being used for messages that were frozen on
4863 arrival, either by an ACL or by local_scan().
4865 7. Added the smtp_incomplete_transaction log selector.
4867 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4868 was accepting AUTH without a new EHLO.
4870 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4871 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4874 10. Logging of TCP/IP connections (when configured) now happens in the main
4875 daemon process instead of the child process, so that the TCP/IP connection
4876 count is more accurate (but it can never be perfect).
4878 11. The use of "drop" in a nested ACL was not being handled correctly in the
4879 outer ACL. Now, if condition failure induced by the nested "drop" causes
4880 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4881 or "require"), the connection is dropped.
4883 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4884 that yield "discard" can now be used with an "accept" or a "discard" verb,
4885 but an error is generated for any others (because I can't see a useful way
4886 to define what should happen).
4888 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4889 are now processed in the same way as lines in the Exim configuration file.
4890 In particular, continuation lines are supported.
4892 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4894 15. Added -ti meaning -t -i.
4896 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4897 domains, and warn by logging if others are found.
4899 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4900 returned from GIFCONF when Exim is trying to find the list of interfaces on
4901 a host. The code in os.c has been modified to copy each ifreq to an aligned
4902 structure in all cases.
4904 Also, in some cases, the returned ifreq's were being copied to a 'struct
4905 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4906 means the last couple of bytes of an IPv6 address could be chopped if the
4907 ifreq contained only a normal sockaddr (14 bytes storage).
4909 18. Named domain lists were not supported in the hosts_treat_as_local option.
4910 An entry such as +xxxx was not recognized, and was treated as a literal
4913 19. Ensure that header lines added by a DATA ACL are included in the reject log
4914 if the ACL subsequently rejects the message.
4916 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4917 MD5 (which is deprecated).
4919 21. When testing a filter file using -bf, Exim was writing a message when it
4920 took the sender from a "From " line in the message, but it was not doing so
4921 when it took $return_path from a Return-Path: header line. It now does.
4923 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4924 with a valid header line field name (a series of printing characters
4925 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4927 23. Changed "disc" in the source to "disk" to conform to the documentation and
4928 the book and for uniformity.
4930 24. Ignore Sendmail's -Ooption=value command line item.
4932 25. When execve() failed while trying to run a command in a pipe transport,
4933 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4934 could be confused with a return value of 69 from the command itself. This
4935 has been changed to 127, the value the shell returns if it is asked to run
4936 a non-existent command. The wording for the related log line suggests a
4937 non-existent command as the problem.
4939 26. If received_header_text expands to an empty string, do not add a Received:
4940 header line to the message. (Well, it adds a token one on the spool, but
4941 marks it "old" so that it doesn't get used or transmitted.)
4943 27. Installed eximstats 1.28 (addition of -nt option).
4945 28. There was no check for failure on the call to getsockname() in the daemon
4946 code. This can fail if there is a shortage of resources on the system, with
4947 ENOMEM, for example. A temporary error is now given on failure.
4949 29. Contrary to the C standard, it seems that in some environments, the
4950 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4951 program. Exim now does this explicitly; it affects the formatting of
4952 timestamps using strftime().
4954 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4955 complaints. I've now initialized all the variables, to avoid this.
4957 32. Header lines added by a system filter were not being "seen" during
4958 transport-time rewrites.
4960 33. The info_callback() function passed to OpenSSL is set up with type void
4961 (*)(SSL *, int, int), as described somewhere. However, when calling the
4962 function (actually a macro) that sets it up, the type void(*)() is
4963 expected. I've put in a cast to prevent warnings from picky compilers.
4965 34. If a DNS black list lookup found a CNAME record, but there were no A
4966 records associated with the domain it pointed at, Exim crashed.
4968 35. If a DNS black list lookup returned more than one A record, Exim ignored
4969 all but the first. It now scans all returned addresses if a particular IP
4970 value is being sought. In this situation, the contents of the
4971 $dnslist_value variable are a list of all the addresses, separated by a
4974 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4975 to accept a host name and all its aliases if the forward lookup for any of
4976 them yielded the IP address of the incoming connection. Now it accepts only
4977 those names whose forward lookup yields the correct IP address. Any other
4978 names are discarded. This closes a loophole whereby a rogue DNS
4979 administrator could create reverse DNS records to break through a
4980 wildcarded host restriction in an ACL.
4982 37. If a user filter or a system filter that ran in a subprocess used any of
4983 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4984 the wrong values were passed to the pipe command ($thisaddress had the
4985 value of $0, $0 had the value of $1, etc). This bug was introduced by
4986 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4988 38. Improved the line breaking for long SMTP error messages from ACLs.
4989 Previously, if there was no break point between 40 and 75 characters, Exim
4990 left the rest of the message alone. Two changes have been made: (a) I've
4991 reduced the minimum length to 35 characters; (b) if it can't find a break
4992 point between 35 and 75 characters, it looks ahead and uses the first one
4993 that it finds. This may give the occasional overlong line, but at least the
4994 remaining text gets split now.
4996 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4997 file descriptors might be low, and that setting 1000 would always raise it.
4998 It turns out that in some environments, the limit is already over 1000 and
4999 that lowering it causes trouble. So now Exim takes care not to decrease it.
5001 40. When delivering a message, the value of $return_path is set to $sender_
5002 address at the start of routing (routers may change the value). By an
5003 oversight, this default was not being set up when an address was tested by
5004 -bt or -bv, which affected the outcome if any router or filter referred to
5007 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5008 log only when "message" or "log_message" are set. However, if one of the
5009 conditions was an address verification, or a call to a nested ACL, the
5010 messages generated by the underlying test were being passed through. This
5011 no longer happens. The underlying message is available in $acl_verify_
5012 message for both "message" and "log_message" expansions, so it can be
5013 passed through if needed.
5015 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5016 new expansion $bh_ to give the encoded byte string without charset
5017 translation. Translation happens only if iconv() is available; HAVE_ICONV
5018 indicates this at build time. HEADERS_CHARSET gives the charset to
5019 translate to; headers_charset can change it in the configuration, and
5020 "headers charset" can change it in an individual filter file.
5022 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5023 that creates RFC 2047 encoded "words" labels them as that charset instead
5024 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5025 expansion operator; (ii) when Exim creates a From: line for a local
5026 message; (iii) when a header line is rewritten to include a "phrase" part.
5028 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5029 buggy, causing it to skip the first lines of messages whose message ID
5030 ended in 'D'. This would not have bitten before Exim release 4.14, because
5031 message IDs were unlikely to end in 'D' before then. The effect was to have
5032 incorrect size information for certain domains.
5034 45. #include "config.h" was missing at the start of the crypt16.c module. This
5035 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5038 46. If there was a timeout during a "random" callout check, Exim treated it as
5039 a failure of the random address, and carried on sending RSET and the real
5040 address. If the delay was just some slowness somewhere, the response to the
5041 original RCPT would be taken as a response to RSET and so on, causing
5042 mayhem of various kinds.
5044 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5045 when I implemented it. It didn't allow for the fact that some option values
5046 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5047 the right test for positive values.
5049 48. Domain names in DNS records are case-independent. Exim always looks them up
5050 in lower case. Some resolvers return domain names in exactly the case they
5051 appear in the zone file, that is, they may contain uppercase letters. Not
5052 all resolvers do this - some return always lower case. Exim was treating a
5053 change of case by a resolver as a change of domain, similar to a widening
5054 of a domain abbreviation. This triggered its re-routing code and so it was
5055 trying to route what was effectively the same domain again. This normally
5056 caused routing to fail (because the router wouldn't handle the domain
5057 twice). Now Exim checks for this case specially, and just changes the
5058 casing of the domain that it ultimately uses when it transmits the message
5061 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5064 50. If a filter generated a file delivery with a non-absolute name (possible if
5065 no home directory exists for the router), the forbid_file option was not
5068 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5069 the existing equality matching.
5071 52. Exim was using ints instead of ino_t variables in some places where it was
5072 dealing with inode numbers.
5074 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5075 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5076 called TMPDIR, and if it finds it is different, it changes its value.
5078 54. The smtp_printf() function is now made available to local_scan() so
5079 additional output lines can be written before returning. There is also an
5080 smtp_fflush() function to enable the detection of a dropped connection.
5081 The variables smtp_input and smtp_batched_input are exported to
5084 55. Changed the default runtime configuration: the message "Unknown user"
5085 has been removed from the ACL, and instead placed on the localuser router,
5086 using the cannot_route_message feature. This means that any verification
5087 failures that generate their own messages won't get overridden. Similarly,
5088 the "Unrouteable address" message that was in the ACL for unverifiable
5089 relay addresses has also been removed.
5091 56. Added hosts_avoid_esmtp to the smtp transport.
5093 57. The exicyclog script was not checking for the esoteric option
5094 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5095 will work only if exicyclog is run under the appropriate euid.
5097 58. Following a discussion on the list, the rules by which Exim recognises line
5098 endings on incoming messages have been changed. The -dropcr and drop_cr
5099 options are now no-ops, retained only for backwards compatibility. The
5100 following line terminators are recognized: LF CRLF CR. However, special
5101 processing applies to CR:
5103 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5104 nor a local message in the state where . is a terminator.
5106 (ii) If a bare CR is encountered in a header line, an extra space is added
5107 after the line terminator so as not to end the header. The reasoning
5108 behind this is that bare CRs in header lines are most likely either
5109 to be mistakes, or people trying to play silly games.
5111 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5112 was being incorrectly given as 18 bytes larger than it should have been.
5113 This is a VOB (very old bug).
5115 60. This may never have affected anything current, but just in case it has:
5116 When the local host is found other than at the start of a list of hosts,
5117 the local host, those with the same MX, and any that follow, are discarded.
5118 When the list in question was part of a longer list of hosts, the following
5119 hosts (not currently being processed) were also being discarded. This no
5120 longer happens. I'm not sure if this situation could ever has previously
5123 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5126 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5127 header. This is something that is recommended in a new Internet Draft, and
5128 is something that is documented as being done by Sendmail. There are two
5129 possible values. For messages generated by the autoreply transport, Exim
5132 Auto-Submitted: auto-replied
5134 whereas for all other generated messages (e.g. bounces) it adds
5136 Auto-Submitted: auto-generated
5138 63. The "personal" condition in filters now includes a test for the
5139 Auto-Submitted: header. If it contains the string "auto-" the message it
5140 not considered personal.
5142 64. Added rcpt_include_affixes as a generic transport option.
5144 65. Added queue_only_override (default true).
5146 66. Added the syslog_duplication option.
5148 67. If what should have been the first header line of a message consisted of
5149 a space followed by a colon, Exim was mis-interpreting it as a header line.
5150 It isn't of course - it is syntactically invalid and should therefore be
5151 treated as the start of the message body. The misbehaviour could have
5152 caused a number of strange effects, including loss of data in subsequent
5153 header lines, and spool format errors.
5155 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5156 client host had authenticated. This control can now be exercised by an ACL
5157 for more flexibility.
5159 69. By default, callouts do not happen when testing with -bh. There is now a
5160 variant, -bhc, which does actually run the callout code, including
5161 consulting and updating the callout cache.
5163 70. Added support for saslauthd authentication, courtesy of Alexander
5166 71. If statvfs() failed on the spool or log directories while checking their
5167 size for availability, Exim confusingly gave the error "space shortage".
5168 Furthermore, in debugging mode it crashed with a floating point exception.
5169 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5170 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5171 serious problem, Exim now writes to the main and panic logs when this
5172 happens, with details of the failure. It then refuses to accept the
5173 incoming message, giving the message "spool directory problem" or "log
5174 directory problem" with a 421 code for SMTP messages.
5176 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5177 0, 1, and 2 exist, because some OS complain for execs without them (see
5178 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5179 descriptors. However, the code omitted to check that the open succeeded,
5180 causing mysterious errors if for some reason the permissions on /dev/null
5181 got screwed. Now Exim writes a message to the main and panic logs, and
5182 bombs out if it can't open /dev/null.
5184 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5185 interact so that it is all more flexible. It is supposed to remain
5186 backwards compatible. Also added extra_local_interfaces.
5188 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5189 to bomb out with an assertion failure - to the client this appears as a
5190 connection drop. This problem occurs in the part of the code that was taken
5191 from the Samba project. Fortunately, the assertion is in a very simple
5192 function, so I have fixed this by reproducing the function inline in the
5193 one place where it is called, and arranging for authentication to fail
5194 instead of killing the process with assert().
5196 75. The SPA client code was not working when the server requested OEM rather
5197 than Unicode encoding.
5199 76. Added code to make require_files with a specific uid setting more usable in
5200 the case where statting the file as root fails - usually a non-root-mounted
5201 NFS file system. When this happens and the failure is EACCES, Exim now
5202 forks a subprocess and does the per-uid checking as the relevant uid.
5204 77. Added process_log_path.
5206 78. If log_file_path was not explicitly set, a setting of check_log_space or
5207 check_log_inodes was ignored.
5209 79. If a space check for the spool or log partitions fails, the incident is now
5210 logged. Of course, in the latter case the data may get lost...
5212 80. Added the %p formatting code to string_format() so that it can be used to
5213 print addresses in debug_print(). Adjusted all the address printing in the
5214 debugging in store.c to use %p rather than %d.
5216 81. There was a concern that a line of code in smtp_in.c could overflow a
5217 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5218 initially expressed, the concern was not well-founded, because trailing
5219 spaces are removed early. However, if the trailing spaces were followed by
5220 a NULL, they did not get removed, so the overflow was possible. Two fixes
5223 (a) I re-wrote the offending code in a cleaner fashion.
5224 (b) If an incoming SMTP command contains a NULL character, it is rejected
5227 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5228 runs initgroups(), so that if the Exim user is in any additional groups,
5229 they will be used during message reception.
5235 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob