1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
74 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
77 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
78 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
79 configuration file which is writeable by the Exim user or group.
81 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
82 of configuration files to cover files specified with the -C option if
83 they are going to be used with root privileges, not just the default
86 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
87 option (effectively making it always true).
89 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
90 files to be used while preserving root privileges.
92 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
93 that rogue child processes cannot use them.
95 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
96 run-time user, instead of root.
98 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
99 Exim run-time user without dropping privileges.
101 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
102 result string, instead of calling string_vformat() twice with the same
105 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
106 for other users. Others should always drop root privileges if they use
107 -C on the command line, even for a whitelisted configure file.
109 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
111 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
117 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
118 $data_path, and $header_path variables; fixed documentation bugs and
121 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
122 exipick to access non-standard spools, including the "frozen" queue
125 NM/01 Bugzilla 965: Support mysql stored procedures.
126 Patch from Alain Williams
128 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
130 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
131 Patch from Andreas Metzler
133 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
134 Patch from Kirill Miazine
136 NM/05 Bugzilla 671: Added umask to procmail example.
138 JJ/03 installed exipick 20100323.0, fixing doc bug
140 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
141 directory. Notification and patch from Dan Rosenberg.
143 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
145 TK/02 Improve log output when DKIM signing operation fails.
147 MH/01 Treat the transport option dkim_domain as a colon separated
148 list, not as a single string, and sign the message with each element,
149 omitting multiple occurences of the same signer.
151 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
152 Bugzilla 985, 986. Patch by Simon Arlott
154 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
155 Patch by Simon Arlott
157 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
158 MBX locking. Notification from Dan Rosenberg.
164 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
166 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
168 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
170 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
172 NM/04 Bugzilla 915: Flag broken perl installation during build.
178 TK/01 Added patch by Johannes Berg that expands the main option
179 "spamd_address" if it starts with a dollar sign.
181 TK/02 Write list of recipients to X-Envelope-Sender header when building
182 the mbox-format spool file for content scanning (suggested by Jakob
185 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
186 (http://www.dcc-servers.net/) support via dccifd. Activated by
187 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
189 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
190 by Mark Daniel Reidel <mr@df.eu>.
192 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
193 When building exim an external PCRE library is now needed -
194 PCRE is a system library on the majority of modern systems.
195 See entry on PCRE_LIBS in EDITME file.
197 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
198 conversation. Added nologin parameter to request.
199 Patch contributed by Kirill Miazine.
201 TF/01 Do not log submission mode rewrites if they do not change the address.
203 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
205 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
206 log files in place. Contributed by Roberto Lima.
208 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
210 TF/03 Bugzilla 615: When checking the local_parts router precondition
211 after a local_part_suffix or local_part_prefix option, Exim now
212 does not use the address's named list lookup cache, since this
213 contains cached lookups for the whole local part.
215 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
216 Robert Millan. Documentation is in experimental-spec.txt.
218 TF/04 Bugzilla 668: Fix parallel build (make -j).
220 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
222 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
223 Patch provided by Jan Srzednicki.
225 TF/05 Leading white space used to be stripped from $spam_report which
226 wrecked the formatting. Now it is preserved.
228 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
229 that they are available at delivery time.
231 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
233 TF/08 TLS error reporting now respects the incoming_interface and
234 incoming_port log selectors.
236 TF/09 Produce a more useful error message if an SMTP transport's hosts
237 setting expands to an empty string.
239 NM/06 Bugzilla 744: EXPN did not work under TLS.
240 Patch provided by Phil Pennock.
242 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
243 Patch provided by Richard Godbee.
245 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
246 acl_smtp_notquit, added index entry.
248 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
249 Patch provided by Eugene Bujak.
251 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
252 accept(). Patch provided by Maxim Dounin.
254 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
255 Patch provided by Phil Pennock.
257 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
259 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
260 Patch provided by Brad "anomie" Jorsch.
262 NM/14 Bugzilla 622: Added timeout setting to transport filter.
263 Patch provided by Dean Brooks.
265 TK/05 Add native DKIM support (does not depend on external libraries).
267 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
268 Patch provided by Graeme Fowler.
270 NM/16 Bugzilla 851: Documentation example syntax fix.
272 NM/17 Changed NOTICE file to remove references to embedded PCRE.
274 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
277 NM/19 Bugzilla 745: TLS version reporting.
278 Patch provided by Phil Pennock.
280 NM/20 Bugzilla 167: bool: condition support.
281 Patch provided by Phil Pennock.
283 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
284 clients. Patch provided by Phil Pennock.
286 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
287 Patch provided by Brad "anomie" Jorsch.
289 NM/23 Bugzilla 687: Fix misparses in eximstats.
290 Patch provided by Heiko Schlittermann.
292 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
293 Patch provided by Heiko Schlittermann.
295 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
296 plus update to original patch.
298 NM/26 Bugzilla 799: Documentation correction for ratelimit.
300 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
301 Patch provided by David Brownlee.
303 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
305 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
307 NM/30 Bugzilla 888: TLS documentation bugfixes.
309 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
311 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
312 Unlike the original bugzilla I have changed all shell scripts in src tree.
314 NM/33 Bugzilla 898: Transport filter timeout fix.
315 Patch by Todd Rinaldo.
317 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
318 Patch by Serge Demonchaux.
320 NM/35 Bugzilla 39: Base64 decode bug fixes.
321 Patch by Jakob Hirsch.
323 NM/36 Bugzilla 909: Correct connect() call in dcc code.
325 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
327 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
329 NM/39 Bugzilla 911: Fixed MakeLinks build script.
335 TK/01 Add preliminary DKIM support. Currently requires a forked version of
336 ALT-N's libdkim that I have put here:
337 http://duncanthrax.net/exim-experimental/
339 Note to Michael Haardt: I had to rename some vars in sieve.c. They
340 were called 'true' and it seems that C99 defines that as a reserved
341 keyword to be used with 'bool' variable types. That means you could
342 not include C99-style headers which use bools without triggering
343 build errors in sieve.c.
345 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
346 as mailq or other aliases. Changed the --help handling significantly
347 to do whats expected. exim_usage() emits usage/help information.
349 SC/01 Added the -bylocaldomain option to eximstats.
351 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
353 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
355 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
361 PH/01 Another patch from the Sieve maintainer.
363 PH/02 When an IPv6 address is converted to a string for single-key lookup
364 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
365 dots are used instead of colons so that keys in lsearch files need not
366 contain colons. This was done some time before quoting was made available
367 in lsearch files. However, iplsearch files do require colons in IPv6 keys
368 (notated using the quote facility) so as to distinguish them from IPv4
369 keys. This meant that lookups for IP addresses in host lists did not work
370 for iplsearch lookups.
372 This has been fixed by arranging for IPv6 addresses to be expressed with
373 colons if the lookup type is iplsearch. This is not incompatible, because
374 previously such lookups could never work.
376 The situation is now rather anomolous, since one *can* have colons in
377 ordinary lsearch keys. However, making the change in all cases is
378 incompatible and would probably break a number of configurations.
380 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
383 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
384 conversion specification without a maximum field width, thereby enabling
385 a rogue spamd server to cause a buffer overflow. While nobody in their
386 right mind would setup Exim to query an untrusted spamd server, an
387 attacker that gains access to a server running spamd could potentially
388 exploit this vulnerability to run arbitrary code as the Exim user.
390 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
391 $primary_hostname instead of what libspf2 thinks the hosts name is.
393 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
394 a directory entry by the name of the lookup key. Previously, if a
395 symlink pointed to a non-existing file or a file in a directory that
396 Exim lacked permissions to read, a lookup for a key matching that
397 symlink would fail. Now it is enough that a matching directory entry
398 exists, symlink or not. (Bugzilla 503.)
400 PH/03 The body_linecount and body_zerocount variables are now exported in the
403 PH/04 Added the $dnslist_matched variable.
405 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
406 This means they are set thereafter only if the connection becomes
409 PH/06 Added the client_condition to authenticators so that some can be skipped
410 by clients under certain conditions.
412 PH/07 The error message for a badly-placed control=no_multiline_responses left
413 "_responses" off the end of the name.
415 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
417 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
418 (without spaces) instead of just copying the configuration text.
420 PH/10 Added the /noupdate option to the ratelimit ACL condition.
422 PH/11 Added $max_received_linelength.
424 PH/12 Added +ignore_defer and +include_defer to host lists.
426 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
427 way in which PCRE > 7.0 is built.
429 PH/14 Implemented queue_only_load_latch.
431 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
432 MAIL command. The effect was to mangle the value on 64-bit systems.
434 PH/16 Another patch from the Sieve maintainer.
436 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
438 PH/18 If a system quota error occurred while trying to create the file for
439 a maildir delivery, the message "Mailbox is full" was not appended to the
440 bounce if the delivery eventually timed out. Change 4.67/27 below applied
441 only to a quota excession during the actual writing of the file.
443 PH/19 It seems that peer DN values may contain newlines (and other non-printing
444 characters?) which causes problems in log lines. The DN values are now
445 passed through string_printing() before being added to log lines.
447 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
448 and InterBase are left for another time.)
450 PH/21 Added message_body_newlines option.
452 PH/22 Guard against possible overflow in moan_check_errorcopy().
454 PH/23 POSIX allows open() to be a macro; guard against that.
456 PH/24 If the recipient of an error message contained an @ in the local part
457 (suitably quoted, of course), incorrect values were put in $domain and
458 $local_part during the evaluation of errors_copy.
464 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
465 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
468 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
469 issue a MAIL command.
471 PH/02 In an ACL statement such as
473 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
475 if a client was not listed at all, or was listed with a value other than
476 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
477 the condition was not true (as it should be), so access was not denied.
478 The bug was that the ! inversion was incorrectly passed on to the second
479 item. This has been fixed.
481 PH/03 Added additional dnslists conditions == and =& which are different from
482 = and & when the dns lookup returns more than one IP address.
484 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
485 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
487 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
488 FSYNC, which compiles an option called disable_fsync that allows for
489 bypassing fsync(). The documentation is heavily laced with warnings.
491 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
493 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
494 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
495 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
496 including adding "make clean"; (3) Added -fPIC when compiling the test
497 dynamically loaded module, to get rid of a warning.
499 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
500 message fails, move_frozen_messages = true and ignore_bounce_errors_after
501 = 0s. The bug is otherwise harmless.
503 PH/07 There was a bug in the dovecot authenticator such that the value of
504 $auth1 could be overwritten, and so not correctly preserved, after a
505 successful authentication. This usually meant that the value preserved by
506 the server_setid option was incorrect.
508 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
510 PH/09 Installed PCRE release 7.0.
512 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
513 run for batched SMTP input. It is now run at the start of every message
514 in the batch. While fixing this I discovered that the process information
515 (output by running exiwhat) was not always getting set for -bs and -bS
516 input. This is fixed, and it now also says "batched" for BSMTP.
518 PH/11 Added control=no_pipelining.
520 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
521 patch, slightly modified), and move the expansion of helo_data till after
522 the connection is made in the smtp transport (so it can use these
525 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
527 PH/14 Added log_selector = +pid.
529 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
531 PH/16 Add ${if forany and ${if forall.
533 PH/17 Added dsn_from option to vary the From: line in DSNs.
535 PH/18 Flush SMTP output before performing a callout, unless control =
536 no_callout_flush is set.
538 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
539 was true (the default) a successful delivery failed to delete the retry
540 item, thus causing premature timeout of the address. The bug is now
543 PH/20 Added hosts_avoid_pipelining to the smtp transport.
545 PH/21 Long custom messages for fakedefer and fakereject are now split up
546 into multiline reponses in the same way that messages for "deny" and
547 other ACL rejections are.
549 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
550 with slight modification.
552 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
553 draft, changing the syntax and factoring some duplicate code".
555 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
556 for deliveries of the second and subsequent messages over the same SMTP
559 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
560 ${reduce, with only minor "tidies".
562 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
564 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
565 expansion side effects.
567 PH/27 When a message times out after an over-quota error from an Exim-imposed
568 quota, the bounce message says "mailbox is full". This message was not
569 being given when it was a system quota that was exceeded. It now should
572 MH/03 Made $recipients available in local_scan(). local_scan() already has
573 better access to the recipient list through recipients_list[], but
574 $recipients can be useful in postmaster-provided expansion strings.
576 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
577 in the case of a MAIL command with additional options following the
578 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
579 were accidentally chopped off.
581 PH/29 SMTP synchronization checks are implemented when a command is read -
582 there is a check that no more input is waiting when there shouldn't be
583 any. However, for some commands, a delay in an ACL can mean that it is
584 some time before the response is written. In this time, more input might
585 arrive, invalidly. So now there are extra checks after an ACL has run for
586 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
587 pipelining has not been advertised.
589 PH/30 MH's patch to allow iscntrl() characters to be list separators.
591 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
592 returned in the SMTP response when smtp_return_error_details was false.
595 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
596 instead of the C I/O that was originally supplied, because problems were
599 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
600 Exim which did not show up earlier: it was assuming that a call to
601 SSL_CTX_set_info_callback() might give an error value. In fact, there is
602 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
603 was a macro that became an assignment, so it seemed to work. This has
604 changed to a proper function call with a void return, hence the compile
605 error. Exim's code has been fixed.
607 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
610 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
612 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
614 PH/37 If a message is not accepted after it has had an id assigned (e.g.
615 because it turns out to be too big or there is a timeout) there is no
616 "Completed" line in the log. When some messages of this type were
617 selected by exigrep, they were listed as "not completed". Others were
618 picked up by some special patterns. I have improved the selection
619 criteria to be more general.
621 PH/38 The host_find_failed option in the manualroute router can now be set
622 to "ignore", to completely ignore a host whose IP address cannot be
623 found. If all hosts are ignored, the behaviour is controlled by the new
624 host_all_ignored option.
626 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
627 homing or because of multiple MX records with /mx) generated more than
628 one IP address, and the following item turned out to be the local host,
629 all the secondary addresses of the first item were incorrectly removed
630 from the list, along with the local host and any following hosts (which
631 is what is supposed to happen).
633 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
634 whoever called Exim into the -H file. In the case of the daemon it was
635 behaving confusingly. When first started, it used values for whoever
636 started the daemon, but after a SIGHUP it used the Exim user (because it
637 calls itself on a restart). I have changed the code so that it now always
640 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
641 message are rejected with the same error (e.g. no authentication or bad
642 sender address), and a DATA command is nevertheless sent (as can happen
643 with PIPELINING or a stupid MUA), the error message that was given to the
644 RCPT commands is included in the rejection of the DATA command. This is
645 intended to be helpful for MUAs that show only the final error to their
648 PH/42 Another patch from the Sieve maintainer.
650 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
651 Eximstats - Fixed some broken HTML links and added missing column headers
653 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
654 columns for Rejects, Temp Rejects, Ham, and Spam rows.
656 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
658 PH/43 Yet another patch from the Sieve maintainer.
660 PH/44 I found a way to check for a TCP/IP connection going away before sending
661 the response to the final '.' that terminates a message, but only in the
662 case where the client has not sent further data following the '.'
663 (unfortunately, this is allowed). However, in many cases there won't be
664 any further data because there won't be any more messages to send. A call
665 to select() can be used: if it shows that the input is "ready", there is
666 either input waiting, or the socket has been closed. An attempt to read
667 the next input character can distinguish the two cases. Previously, Exim
668 would have sent an OK response which the client would never have see.
669 This could lead to message repetition. This fix should cure that, at
670 least in a lot of common cases.
672 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
673 advertised in response to EHLO.
679 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
680 fixed by 4.65/MH/01 (is this a record?) are fixed:
682 (i) An empty string was always treated as zero by the numeric comparison
683 operators. This behaviour has been restored.
685 (ii) It is documented that the numeric comparison operators always treat
686 their arguments as decimal numbers. This was broken in that numbers
687 starting with 0 were being interpreted as octal.
689 While fixing these problems I realized that there was another issue that
690 hadn't been noticed. Values of message_size_limit (both the global option
691 and the transport option) were treated as octal if they started with 0.
692 The documentation was vague. These values are now always treated as
693 decimal, and I will make that clear in the documentation.
699 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
700 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
703 MH/01 Don't check that the operands of numeric comparison operators are
704 integers when their expansion is in "skipping" mode (fixes bug
705 introduced by 4.64-PH/07).
707 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
708 child addresses, Exim now panics and dies. Previously, because the count
709 is held in a short int, deliveries were likely to be lost. As such a
710 large number of recipients for a single message is ridiculous
711 (performance will be very, very poor), I have chosen to impose a limit
712 rather than extend the field.
718 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
719 leftover -K file (the existence of which was triggered by #402).
720 While we were at it, introduced process PID as part of the -K
721 filename. This should rule out race conditions when creating
724 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
725 processing considerably. Previous code took too long for large mails,
726 triggering a timeout which in turn triggers #401.
728 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
729 in the DK code in transports.c. sendfile() is not really portable,
730 hence the _LINUX specificness.
732 TF/01 In the add_headers option to the mail command in an Exim filter,
733 there was a bug that Exim would claim a syntax error in any
734 header after the first one which had an odd number of characters
737 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
738 callout verification, Exim cached a "reject" for the entire domain. This
739 is correct for most verifications, but it is not correct for a recipient
740 verification with use_sender or use_postmaster set, because in that case
741 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
742 case of MAIL FROM:<> rejection from other early rejections (e.g.
743 rejection of HELO). When verifying a recipient using a non-null MAIL
744 address, the cache is ignored if it shows MAIL FROM:<> rejection.
745 Whatever the result of the callout, the value of the domain cache is
746 left unchanged (for any other kind of callout, getting as far as trying
747 RCPT means that the domain itself is ok).
749 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
752 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
753 manifest itself as EPIPE rather than ECONNECT. When tidying away a
754 session, the daemon ignores ECONNECT errors and logs others; it now
755 ignores EPIPE as well.
757 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
758 (quoted-printable decoding).
760 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
761 later the small subsequent patch to fix an introduced bug.
763 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
765 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
767 PH/08 An error is now given if message_size_limit is specified negative.
769 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
770 to be given (somewhat) arbitrary names.
772 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
775 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
776 miscellaneous code fixes
778 PH/10 Added the log_reject_target ACL modifier to specify where to log
781 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
782 hostname. This is wrong, because it relates to the incoming message (and
783 probably the interface on which it is arriving) and not to the outgoing
784 callout (which could be using a different interface). This has been
785 changed to use the value of the helo_data option from the smtp transport
786 instead - this is what is used when a message is actually being sent. If
787 there is no remote transport (possible with a router that sets up host
788 addresses), $smtp_active_hostname is used.
790 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
791 tweaks were necessary in order to get it to work (see also 21 below):
792 (a) The code assumed that strncpy() returns a negative number on buffer
793 overflow, which isn't the case. Replaced with Exim's string_format()
795 (b) There were several signed/unsigned issues. I just did the minimum
796 hacking in of casts. There is scope for a larger refactoring.
797 (c) The code used strcasecmp() which is not a standard C function.
798 Replaced with Exim's strcmpic() function.
799 (d) The code set only $1; it now sets $auth1 as well.
800 (e) A simple test gave the error "authentication client didn't specify
801 service in request". It would seem that Dovecot has changed its
802 interface. Fortunately there's a specification; I followed it and
803 changed what the client sends and it appears to be working now.
805 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
808 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
809 address A is aliased to B and C, where B exists and C does not. Without
810 -v the output is "A verified" because verification stops after a
811 successful redirection if more than one address is generated. However,
812 with -v the child addresses are also verified. Exim was outputting "A
813 failed to verify" and then showing the successful verification for C,
814 with its parentage. It now outputs "B failed to verify", showing B's
815 parentage before showing the successful verification of C.
817 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
818 look up a TXT record in a specific list after matching in a combined
821 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
822 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
823 they consult the DNS. I had assumed they would set it the way they
824 wanted; and indeed my experiments on Linux seem to show that in some
825 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
826 To be on the safe side, however, I have now made the interface to
827 host_find_byname() similar to host_find_bydns(), with an argument
828 containing the DNS resolver options. The host_find_byname() function now
829 sets these options at its start, just as host_find_bydns() does. The smtp
830 transport options dns_qualify_single and dns_search_parents are passed to
831 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
832 of host_find_byname() use the default settings of RES_DEFNAMES
833 (qualify_single) but not RES_DNSRCH (search_parents).
835 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
836 spool_read_header() do less string testing, by means of a preliminary
837 switch on the second character of optional "-foo" lines. (This is
838 overdue, caused by the large number of possibilities that now exist.
839 Originally there were few.) While I was there, I also converted the
840 str(n)cmp tests so they don't re-test the leading "-" and the first
841 character, in the hope this might squeeze out yet more improvement.
843 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
844 flag allowing group syntax was set by the header_syntax check but not
845 turned off, possible causing trouble later; (2) The flag was not being
846 set at all for the header_verify test, causing "group"-style headers to
847 be rejected. I have now set it in this case, and also caused header_
848 verify to ignore an empty address taken from a group. While doing this, I
849 came across some other cases where the code for allowing group syntax
850 while scanning a header line wasn't quite right (mostly, not resetting
851 the flag correctly in the right place). These bugs could have caused
852 trouble for malformed header lines. I hope it is now all correct.
854 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
855 with the "reply" argument non-NULL. The code, however (which originally
856 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
857 but it didn't always do it. This confused somebody who was copying the
858 code for some other use. I have removed all the tests.
860 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
861 feature that was used to support insecure browsers during the U.S. crypto
862 embargo. It requires special client support, and Exim is probably the
863 only MTA that supported it -- and would never use it because real RSA is
864 always available. This code has been removed, because it had the bad
865 effect of slowing Exim down by computing (never used) parameters for the
866 RSA_EXPORT functionality.
868 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
869 authenticator to fail if there's a tab character in the incoming data
870 (there should never be unless someone is messing about, as it's supposed
871 to be base64-encoded). Also added, on Timo's advice, the "secured" option
872 if the connection is using TLS or if the remote IP is the same as the
873 local IP, and the "valid-client-cert option" if a client certificate has
876 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
877 authenticators. This can be used for authorization after authentication
878 succeeds. (In the case of plaintext, it servers for both authentication
881 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
882 if any retry times were supplied.
884 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
885 connection, where there is no client IP address to check. In this
886 situation, the verify now always succeeds.
888 PH/25 Applied John Jetmore's -Mset patch.
890 PH/26 Added -bem to be like -Mset, but loading a message from a file.
892 PH/27 In a string expansion for a processed (not raw) header when multiple
893 headers of the same name were present, leading whitespace was being
894 removed from all of them, but trailing whitespace was being removed only
895 from the last one. Now trailing whitespace is removed from each header
896 before concatenation. Completely empty headers in a concatenation (as
899 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
900 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
902 PH/29 [Removed. This was a change that I later backed out, and forgot to
903 correct the ChangeLog entry (that I had efficiently created) before
904 committing the later change.]
906 PH/30 Exim was sometimes attempting to deliver messages that had suffered
907 address errors (4xx response to RCPT) over the same connection as other
908 messages routed to the same hosts. Such deliveries are always "forced",
909 so retry times are not inspected. This resulted in far too many retries
910 for the affected addresses. The effect occurred only when there were more
911 hosts than the hosts_max_try setting in the smtp transport when it had
912 the 4xx errors. Those hosts that it had tried were not added to the list
913 of hosts for which the message was waiting, so if all were tried, there
914 was no problem. Two fixes have been applied:
916 (i) If there are any address or message errors in an SMTP delivery, none
917 of the hosts (tried or untried) are now added to the list of hosts
918 for which the message is waiting, so the message should not be a
919 candidate for sending over the same connection that was used for a
920 successful delivery of some other message. This seems entirely
921 reasonable: after all the message is NOT "waiting for some host".
922 This is so "obvious" that I'm not sure why it wasn't done
923 previously. Hope I haven't missed anything, but it can't do any
924 harm, as the worst effect is to miss an optimization.
926 (ii) If, despite (i), such a delivery is accidentally attempted, the
927 routing retry time is respected, so at least it doesn't keep
928 hammering the server.
930 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
931 in ${readsocket because some servers need this prod.
933 PH/32 Added some extra debug output when updating a wait-xxx database.
935 PH/33 The hint "could be header name not terminated by colon", which has been
936 given for certain expansion errors for a long time, was not being given
937 for the ${if def:h_colon_omitted{... case.
939 PH/34 The spec says: "With one important exception, whenever a domain list is
940 being scanned, $domain contains the subject domain." There was at least
941 one case where this was not true.
943 PH/35 The error "getsockname() failed: connection reset by peer" was being
944 written to the panic log as well as the main log, but it isn't really
945 panic-worthy as it just means the connection died rather early on. I have
946 removed the panic log writing for the ECONNRESET error when getsockname()
949 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
950 runs only) independently of the message's sender address. This meant
951 that, if the 4xx error was in fact related to the sender, a different
952 message to the same recipient with a different sender could confuse
953 things. In particualar, this can happen when sending to a greylisting
954 server, but other circumstances could also provoke similar problems.
955 I have changed the default so that the retry time for these errors is now
956 based a combination of the sender and recipient addresses. This change
957 can be overridden by setting address_retry_include_sender=false in the
960 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
961 remote server are returned as part of bounce messages. This was not
962 happening for LMTP over a pipe (the lmtp transport), but now it is the
963 same for both kinds of LMTP.
965 PH/38 Despite being documented as not happening, Exim was rewriting addresses
966 in header lines that were in fact CNAMEs. This is no longer the case.
968 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
969 and queue runs started by the daemon processed all messages. This has
970 been fixed so that -R and -S can now usefully be given with -q<time>.
972 PH/40 Import PCRE release 6.7 (fixes some bugs).
974 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
976 PH/42 Give an error if -q is specified more than once.
978 PH/43 Renamed the variables $interface_address and $interface_port as
979 $received_ip_address and $received_port, to make it clear that these
980 values apply to message reception, and not to the outgoing interface when
981 a message is delivered. (The old names remain recognized, of course.)
983 PH/44 There was no timeout on the connect() call when using a Unix domain
984 socket in the ${readsocket expansion. There now is.
986 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
987 be meaningful with "accept".
989 SC/01 Eximstats V1.43
990 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
992 SC/02 Eximstats V1.44
993 Use a glob alias rather than an array ref in the generated
994 parser. This improves both readability and performance.
996 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
997 Collect SpamAssassin and rejection statistics.
998 Don't display local sender or destination tables unless
999 there is data to show.
1000 Added average volumes into the top table text output.
1002 SC/04 Eximstats V1.46
1003 Collect data on the number of addresses (recipients)
1004 as well as the number of messages.
1006 SC/05 Eximstats V1.47
1007 Added 'Message too big' to the list of mail rejection
1008 reasons (thanks to Marco Gaiarin).
1010 SC/06 Eximstats V1.48
1011 Mainlog lines which have GMT offsets and are too short to
1012 have a flag are now skipped.
1014 SC/07 Eximstats V1.49 (Alain Williams)
1015 Added the -emptyok flag.
1017 SC/08 Eximstats V1.50
1018 Fixes for obtaining the IP address from reject messages.
1020 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1021 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1022 whitesspace changes from 4.64-PH/27
1024 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1027 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1028 are found, allow negative numbers in numeric criteria)
1030 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1032 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1033 to match changes made in 4.64-PH/43
1035 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1037 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1038 "domain" is actually an IP address, and force a failure. This locks out
1039 those revolvers/nameservers that support "A-for-A" lookups, in
1040 contravention of the specifications.
1042 PH/48 When a host name was looked up from an IP address, and the subsequent
1043 forward lookup of the name timed out, the host name was left in
1044 $sender_host_name, contrary to the specification.
1046 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1047 restricted to single-key lookups, Exim was not diagnosing an error if
1048 * or *@ was used with a query-style lookup.
1050 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1052 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1053 long ago, but noone interested enough thought of it. Let's just say that
1054 the "1.1" means that there are some new functions that weren't there at
1055 some point in the past.
1057 PH/51 Error processing for expansion failure of helo_data from an smtp
1058 transport during callout processing was broken.
1060 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1061 tested/used via the -bh/-bhc/-bs options.
1063 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1064 bug, fixed in subsequent PCRE releases).
1066 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1067 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1069 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1075 SC/01 Use a glob alias rather than an array ref in eximstats generated
1076 parser. This improves both readability and performance.
1078 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1079 Don't display local sender or destination tables in eximstats unless
1080 there is data to show.
1081 Added average volumes into the eximstats top table text output.
1083 SC/03 Collect data on the number of addresses (recipients) as well
1084 as the number of messages in eximstats.
1086 TF/01 Correct an error in the documentation for the redirect router. Exim
1087 does (usually) call initgroups() when daemonizing.
1089 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1090 with consistent privilege compared to when running as a daemon.
1092 TF/03 Note in the spec that $authenticated_id is not set for local
1093 submissions from trusted users.
1095 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1096 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1098 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1099 by adding some example configuration directives to the default
1100 configuration file. A little bit of work is required to uncomment the
1101 directives and define how usernames and passwords are checked, but
1102 there is now a framework to start from.
1104 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1105 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1106 without this. I don't know how relevant this is to other LDAP libraries.
1108 PH/02 Add the verb name to the "unknown ACL verb" error.
1110 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1112 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1114 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1115 directory not expanded when it should be if an expanded home directory
1116 was set for the address (which is overridden by the transport).
1118 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1121 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1122 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1123 because it is too late at that time, and has no effect.
1125 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1126 security issue with \' (bugzilla #107). I could not use the
1127 PQescapeStringConn() function, because it needs a PGconn value as one of
1130 PH/08 When testing addresses using -bt, indicate those final addresses that
1131 are duplicates that would not cause an additional delivery. At least one
1132 person was confused, thinking that -bt output corresponded to deliveries.
1133 (Suppressing duplicates isn't a good idea as you lose the information
1134 about possibly different redirections that led to the duplicates.)
1136 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1137 systems where poll() doesn't work, in particular OS X.
1139 PH/10 Added more information to debugging output for retry time not reached.
1141 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1142 operations in malware.c.
1144 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1147 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1148 syslog_duplication set false, log lines that would normally be written
1149 both the the main log and to the reject log were not written to syslog at
1152 PH/14 In the default configuration, change the use of "message" in ACL warn
1153 statements to "add_header".
1155 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1156 not followed by a command (e.g. "seen endif").
1158 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1159 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1162 PH/17 Added extra conditions to the default value of delay_warning_condition
1166 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1167 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1168 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1171 The Auto-Submitted: and various List- headers are standardised, whereas I
1172 don't think Precedence: ever was.
1174 PH/18 Refactored debugging code in route_finduser() to show more information,
1175 in particular, the error code if getpwnam() issues one.
1177 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1178 This is apparently needed in addition to the PH/07 change above to avoid
1179 any possible encoding problems.
1181 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1182 but not after initializing Perl.
1184 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1185 output them only if debugging. By default they are written stderr,
1186 apparently, which is not desirable.
1188 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1191 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1194 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1196 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1197 authenticated or an ident call has been made. Suppress the default
1198 values for $authenticated_id and $authenticated_sender (but permit -oMai
1199 and -oMas) when testing with -bh.
1201 PH/24 Re-jigged the order of the tests in the default configuration so that the
1202 tests for valid domains and recipients precede the DNS black list and CSA
1203 tests, on the grounds that those ones are more expensive.
1205 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1206 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1207 This bug exists in every version of Exim that I still have, right back to
1210 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1211 However, an attempt to turn on case-sensitivity in a regex key by
1212 including (?-i) didn't work because the subject string was already
1213 lowercased, and the effects were non-intuitive. It turns out that a
1214 one-line patch can be used to allow (?-i) to work as expected.
1220 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1221 other effects) broke the use of negated acl sub-conditions.
1223 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1226 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1227 "Deny" causes Exim to reject the incoming connection with a 554 error.
1228 Unfortunately, if there is a major crisis, such as a disk failure,
1229 tcp-wrappers gives "deny", whereas what one would like would be some
1230 kind of temporary error. A kludge has been added to help with this.
1231 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1232 554 error is used if errno is still zero or contains ENOENT (which occurs
1233 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1236 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1238 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1239 errors. Otherwise a message that provokes a temporary error (when other
1240 messages do not) can cause a whole host to time out.
1242 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1243 addresses were routed directly to files or pipes from a redirect router.
1244 File deliveries just didn't batch; pipe deliveries might have suffered
1247 PH/06 A failure to get a lock for a hints database would erroneously always say
1248 "Failed to get write lock", even when it was really a read lock.
1250 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1251 of 0600. This has been changed to use the value of the lockfile_mode
1252 option (which defaults to 0600).
1254 PH/08 Applied small patch from the Sieve maintainer.
1256 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1257 folder from quota calculations, a direct delivery into this folder messed
1258 up the contents of the maildirsize file. This was because the regex was
1259 used only to exclude .Trash (or whatever) when the size of the mailbox
1260 was calculated. There was no check that a delivery was happening into an
1261 excluded directory. This bug has been fixed by ignoring all quota
1262 processing for deliveries into excluded directories.
1264 PH/10 Added the maildirfolder_create_regex option to appendfile.
1270 PH/01 The code for finding all the local interface addresses on a FreeBSD
1271 system running IPv6 was broken. This may well have applied to all BSD
1272 systems, as well as to others that have similar system calls. The broken
1273 code found IPv4 interfaces correctly, but gave incorrect values for the
1274 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1275 that it would not match correctly against @[] and not recognize the IPv6
1278 PH/02 The ipliteral router was not recognizing addresses of the form user@
1279 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1281 PH/03 Added disable_ipv6.
1283 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1284 decoded form, because it is most often used to construct To: headers
1285 lines in autoreplies, and the decoded form may well be syntactically
1286 invalid. However, $reply_address has leading white space removed, and all
1287 newlines turned into spaces so that the autoreply transport does not
1290 PH/05 If group was specified without a user on a router, and no group or user
1291 was specified on a transport, the group from the router was ignored.
1293 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1294 for visible compile-time settings that can be used to change these
1295 numbers, for those that want even more. Backwards compatibility with old
1296 spool files has been maintained. However, going back to a previous Exim
1297 release will lost any variables that are in spool files.
1299 PH/07 Two small changes when running in the test harness: increase delay when
1300 passing a TCP/IP connection to a new process, in case the original
1301 process has to generate a bounce, and remove special handling of
1302 127.0.0.2 (sic), which is no longer necessary.
1304 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1305 be the same on different OS.
1307 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1310 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1311 whether --show-vars was specified or not
1313 JJ/02 exipick: Added support for new ACL variable spool format introduced
1316 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1317 syntactically invalid From: or Reply-to: line, and a filter used this to
1318 generate an autoreply, and therefore failed to obtain an address for the
1319 autoreply, Exim could try to deliver to a non-existent relative file
1320 name, causing unrelated and misleading errors. What now happens is that
1321 it logs this as a hard delivery error, but does not attempt to create a
1324 PH/11 The exinext utility has a -C option for testing purposes, but although
1325 the given file was scanned by exinext itself; it wasn't being passed on
1326 when Exim was called.
1328 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1329 an end-of-file indication when reading a command response.
1331 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1332 compiled. In many other places in Exim, IPv6 addresses are always
1333 recognized, so I have changed this. It also means that IPv4 domain
1334 literals of the form [IPV4:n.n.n.n] are now always recognized.
1336 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1337 used if the router is not running as root, for example, when verifying at
1338 ACL time, or when using -bh. The debugging output from this situation was
1339 non-existent - all you got was a failure to exec. I have made two
1342 (a) Failures to set uid/gid, the current directory, or a process leader
1343 in a subprocess such as that created by queryprogram now generate
1344 suitable debugging ouput when -d is set.
1346 (b) The queryprogram router detects when it is not running as root,
1347 outputs suitable debugging information if -d is set, and then runs
1348 the subprocess without attempting to change uid/gid.
1350 PH/15 Minor change to Makefile for building test_host (undocumented testing
1353 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1354 additional section of a DNS packet that returns MX or SRV records.
1355 Instead, it always explicitly searches for A/AAAA records. This avoids
1356 major problems that occur when a DNS server includes only records of one
1357 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1358 fixed another bug: if SRV records were looked up and the corresponding
1359 address records were *not* found in the additional section, the port
1360 values from the SRV records were lost.
1362 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1363 using the correct key (the original address) when searching the retry
1364 rules in order to find which one to use for generating the retry hint.
1366 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1367 adding the default one. Similarly, if it contains a Reply-To: header, the
1368 errors_reply_to option, if set, is not used.
1370 PH/19 When calculating a retry time, Exim used to measure the "time since
1371 failure" by looking at the "first failed" field in the retry record. Now
1372 it does not use this if it is later than than the arrival time of the
1373 message. Instead it uses the arrival time. This makes for better
1374 behaviour in cases where some deliveries succeed, thus re-setting the
1375 "first failed" field. An example is a quota failure for a huge message
1376 when small messages continue to be delivered. Without this change, the
1377 "time since failure" will always be short, possible causing more frequent
1378 delivery attempts for the huge message than are intended.
1379 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1381 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1382 $1, $2, $3) because the numerical variables can be reset during some
1383 expansion items (e.g. "match"), thereby losing the authentication data.
1385 PH/21 Make -bV show the size of off_t variables so that the test suite can
1386 decide whether to run tests for quotas > 2G.
1388 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1389 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1390 filecount value is greater than 2G or if a quota value is greater than 2G
1391 on a system where the size of off_t is not greater than 4, a panic error
1394 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1395 never match. The debug and -bh output now contains an explicit error
1396 message indicating a malformed IPv4 address or mask.
1398 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1399 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1400 PH/23 above applies.
1402 PH/25 Do not write to syslog when running in the test harness. The only
1403 occasion when this arises is a failure to open the main or panic logs
1404 (for which there is an explicit test).
1406 PH/26 Added the /no_tell option to "control=freeze".
1408 PH/27 If a host name lookup failed very early in a connection, for example, if
1409 the IP address matched host_lookup and the reverse lookup yielded a name
1410 that did not have a forward lookup, an error message of the form "no IP
1411 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1412 could be logged. Now it outputs the IP address instead of "NULL".
1414 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1415 allows the sender and the authenticated sender to be set when
1416 submitting a message from within Exim. Since child_open_exim() is
1417 documented for local_scan(), the new function should be too.
1419 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1420 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1421 results in an empty string is now treated as unset.
1423 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1425 PH/31 Added sender_verify_fail logging option.
1427 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1428 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1429 tidied the source and removed it altogether.
1431 PH/33 When a queue run was abandoned because the load average was too high, a
1432 log line was always written; now it is written only if the queue_run log
1433 selector is set. In addition, the log line for abandonment now contains
1434 information about the queue run such as the pid. This is always present
1435 in "start" and "stop" lines but was omitted from the "abandon" line.
1437 PH/34 Omit spaces between a header name and the colon in the error message that
1438 is given when verify = headers_syntax fails (if there are lots of them,
1439 the message gets confusing).
1441 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1442 names, as there are now some PTR records that contain slashes. This check
1443 is only to protect against broken name servers that fall over on strange
1444 characters, so the fact that it applies to all lookups doesn't matter.
1446 PH/36 Now that the new test suite is complete, we can remove some of the
1447 special code in Exim that was needed for the old test suite. For example,
1448 sorting DNS records because real resolvers return them in an arbitrary
1449 order. The new test suite's fake resolver always returns records in the
1452 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1453 bounces) except when queue_only is set, to avoid logging races between
1454 the different processes.
1456 PH/38 Panic-die if .include specifies a non-absolute path.
1458 PH/39 A tweak to the "H" retry rule from its user.
1460 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1461 a label. They prevented compilation on older perls.
1463 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1464 a warning to be raised on newish perls.
1466 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1467 on queue. Changes to match documented behaviour of showing count of
1468 messages matching specified criteria.
1470 PH/40 Changed the default ident timeout from 30s to 5s.
1472 PH/41 Added support for the use of login_cap features, on those BSD systems
1473 that have them, for controlling the resources used by pipe deliveries.
1475 PH/42 The content-scanning code uses fopen() to create files in which to put
1476 message data. Previously it was not paying any attention to the mode of
1477 the files. Exim runs with umask(0) because the rest of the code creates
1478 files with open(), and sets the required mode explicitly. Thus, these
1479 files were ending up world-writeable. This was not a big issue, because,
1480 being within the spool directory, they were not world-accessible. I have
1481 created a function called modefopen, which takes an additional mode
1482 argument. It sets umask(777), creates the file, chmods it to the required
1483 mode, then resets the umask. All the relevant calls to fopen() in the
1484 content scanning code have been changed to use this function.
1486 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1487 to 24 hours. This avoids potential overflow problems when processing G
1488 and H retry rules. I suspect nobody ever tinkers with this value.
1490 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1492 PH/45 When the plaintext authenticator is running as a client, the server's
1493 challenges are checked to ensure they are valid base64 strings. By
1494 default, the authentication attempt is cancelled if an invalid string is
1495 received. Setting client_ignore_invalid_base64 true ignores these errors.
1496 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1497 they are received. Thus, the responses can be made to depend on the
1498 challenges. If an invalid string is ignored, an empty string is placed in
1501 PH/46 Messages that are created by the autoreply transport now contains a
1502 References: header, in accordance with RFCs 2822 and 3834.
1504 PH/47 Added authenticated_sender_force to the smtp transport.
1506 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1508 PH/49 Installed latest patch from the Sieve maintainer.
1510 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1511 was also set, the appendfile transport was unnecessarily scanning a
1512 directory of message files (e.g. for maildir delivery) to find the count
1513 of files (along with the size), even though it did not need this
1514 information. It now does the scan only if it needs to find either the
1515 size of the count of files.
1517 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1519 PH/52 Two bugs concerned with error handling when the smtp transport is
1522 (i) Exim was not creating retry information for temporary errors given
1523 for individual recipients after the DATA command when the smtp transport
1524 was used in LMTP mode. This meant that they could be retried too
1525 frequently, and not timed out correctly.
1527 (ii) Exim was setting the flag that allows error details to be returned
1528 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1529 recipients that were returned after the DATA command.
1531 PH/53 This is related to PH/52, but is more general: for any failing address,
1532 when detailed error information was permitted to be returned to the
1533 sender, but the error was temporary, then after the final timeout, only
1534 "retry timeout exceeded" was returned. Now it returns the full error as
1535 well as "retry timeout exceeded".
1537 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1538 do this, and (what is worse) MTAs that accept it.
1540 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1541 will now be deprecated.
1543 PH/56 New os.c-cygwin from the Cygwin maintainer.
1545 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1546 formats (previously only available in exim formats via -bpr, -bpru,
1547 and -bpra. Now also available in native and exiqgrep formats)
1549 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1550 with very large, slow to parse queues
1552 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1554 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1556 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1557 responses to authentication challenges, though it was showing the
1558 challenges; (ii) I've removed the CR characters from the debug output for
1561 PH/58 Allow for the insertion of a newline as well as a space when a string
1562 is turned into more than one encoded-word during RFC 2047 encoding. The
1563 Sieve code now uses this.
1565 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1566 data_4xx, lost_connection, tls_required.
1568 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1569 message was being sent as an SMTP response.
1571 PH/61 Add -l and -k options to exicyclog.
1573 PH/62 When verifying, if an address was redirected to one new address, so that
1574 verification continued, and the new address failed or deferred after
1575 having set something in $address_data, the value of $address_data was not
1576 passed back to the ACL. This was different to the case when no
1577 redirection occurred. The value is now passed back in both cases.
1579 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1580 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1581 use login_cap.h, so on its own it isn't the distinguishing feature. The
1582 new name refers directly to the setclassresources() function.
1584 PH/65 Added configuration files for NetBSD3.
1586 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1588 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1589 is preferred over IPv4.
1591 PH/68 The bounce_return_message and bounce_return_body options were not being
1592 honoured for bounces generated during the reception of non-SMTP messages.
1593 In particular, this applied to messages rejected by the ACL. This bug has
1594 been fixed. However, if bounce_return_message is true and bounce_return_
1595 body is false, the headers that are returned for a non-SMTP message
1596 include only those that have been read before the error was detected.
1597 (In the case of an ACL rejection, they have all been read.)
1599 PH/69 The HTML version of the specification is now built in a directory called
1600 spec_html instead of spec.html, because the latter looks like a path with
1601 a MIME-type, and this confuses some software.
1603 PH/70 Catch two compiler warnings in sieve.c.
1605 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1606 function verify_get_ident() calls ip_connect() to connect a socket, but
1607 if the "connect()" function timed out, ip_connect() used to close the
1608 socket. However, verify_get_ident() also closes the socket later, and in
1609 between Exim writes to the log, which may get opened at this point. When
1610 the socket was closed in ip_connect(), the log could get the same file
1611 descriptor number as the socket. This naturally causes chaos. The fix is
1612 not to close the socket in ip_connect(); the socket should be closed by
1613 the function that creates it. There was only one place in the code where
1614 this was missing, in the iplookup router, which I don't think anybody now
1615 uses, but I've fixed it anyway.
1617 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1618 well as to direct DNS lookups. Otherwise the handling of names in host
1619 lists is inconsistent and therefore confusing.
1625 PH/01 Two changes to the default runtime configuration:
1627 (1) Move the checks for relay_from_hosts and authenticated clients from
1628 after to before the (commented out) DNS black list checks.
1630 (2) Add control=submission to the relay_from_hosts and authenticated
1631 clients checks, on the grounds that messages accepted by these
1632 statements are most likely to be submissions.
1634 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1636 (1) Generate an error if the third argument for the ${prvs expansion is
1639 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1642 (3) Reset the variables that are obtained from the first argument of
1643 ${prvscheck and used in the second argument before leaving the code,
1644 because their memory is reclaimed, so using them afterwards may do
1647 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1648 one (it's much easier than Tom thought :-).
1650 (5) Because of (4), we can now allow for the use of $prvscheck_result
1651 inside the third argument.
1653 PH/03 For some reason, the default setting of PATH when running a command from
1654 a pipe transport was just "/usr/bin". I have changed it to
1657 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1658 anything to be listed in the output from -bV.
1660 PH/05 When a filter generated an autoreply, the entire To: header line was
1661 quoted in the delivery log line, like this:
1663 => >A.N.Other <ano@some.domain> <original@ddress> ...
1665 This has been changed so that it extracts the operative address. There
1666 may be more than one such address. If so, they are comma-separated, like
1669 => >ano@some.domain,ona@other.domain <original@ddress> ...
1671 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1672 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1673 not being looked up in the rDNS to get a host name, Exim was showing the
1674 IP address twice in Received: lines, even though the IP addresses were
1675 identical. For example:
1677 Received: from [1.2.3.4] (helo=[1.2.3.4])
1679 However, if the real host name was known, it was omitting the HELO data
1680 if it matched the actual IP address. This has been tidied up so that it
1681 doesn't show the same IP address twice.
1683 PH/07 When both +timestamp and +memory debugging was on, the value given by
1684 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1685 called by the debug printing, thereby overwriting the timestamp buffer.
1686 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1688 PH/08 When the original message was included in an autoreply transport, it
1689 always said "this is a copy of the message, including all the headers",
1690 even if body_only or headers_only was set. It now gives an appropriate
1693 PH/09 Applied a patch from the Sieve maintainer which:
1695 o fixes some comments
1696 o adds the (disabled) notify extension core
1697 o adds some debug output for the result of if/elsif tests
1698 o points to the current vacation draft in the documentation
1699 and documents the missing references header update
1703 o fixes a bug in processing the envelope test (when testing
1704 multiple envelope elements, the last element determinted the
1707 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1708 Electronic Mail") by including:
1710 Auto-submitted: auto-generated
1712 in the messages that it generates (bounce messages and others, such as
1713 warnings). In the case of bounce messages for non-SMTP mesages, there was
1714 also a typo: it was using "Auto_submitted" (underscore instead of
1715 hyphen). Since every message generated by Exim is necessarily in response
1716 to another message, thes have all been changed to:
1718 Auto-Submitted: auto-replied
1720 in accordance with these statements in the RFC:
1722 The auto-replied keyword:
1724 - SHOULD be used on messages sent in direct response to another
1725 message by an automatic process,
1727 - MUST NOT be used on manually-generated messages,
1729 - MAY be used on Delivery Status Notifications (DSNs) and Message
1730 Disposition Notifications (MDNs),
1732 - MUST NOT be used on messages generated by automatic or periodic
1733 processes, except for messages which are automatic responses to
1736 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1737 to the default Received: header definition.
1739 PH/12 Added log selector acl_warn_skipped (default on).
1741 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1742 variables because (a) they are in the wrong storage pool and (b) even if
1743 they were copied, it wouldn't work properly because of the caching.
1745 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1746 checking when decoding. Apparently there are clients that generate
1747 overlong encoded strings. Why am I not surprised?
1749 PH/15 If the first argument of "${if match_address" was not empty, but did not
1750 contain an "@" character, Exim crashed. Now it writes a panic log message
1751 and treats the condition as false.
1753 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1755 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1756 extension "envelope-auth". The code is finished and in agreement with
1757 other implementations, but there is no documentation so far and in fact,
1758 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1759 not changing the active code.
1761 Print executed "if" and "elsif" statements when debugging is used. This
1762 helps a great deal to understand what a filter does.
1764 Document more things not specified clearly in RFC3028. I had all this
1765 sorted out, when out of a sudden new issues came to my mind. Oops."
1767 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1770 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1771 canonical form (as documented). However, after a host name lookup from
1772 the IP address, check_host() was doing a simple string comparison with
1773 addresses acquired from the DNS when checking that the found name did
1774 have the original IP as one of its addresses. Since any found IPv6
1775 addresses are likely to be in abbreviated form, the comparison could
1776 fail. Luckily, there already exists a function for doing the comparison
1777 by converting both addresses to binary, so now that is used instead of
1778 the text comparison.
1780 PH/20 There was another similar case to PH/19, when a complete host name was
1781 given in a host list; looking up its IP address could give an abbreviated
1782 form, whereas the current host's name might or might not be abbreviated.
1783 The same fix has been applied.
1789 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1790 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1793 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1794 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1796 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1798 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1799 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1800 Also, for "Auto-Submitted", treat anything other than "no" as
1801 non-personal, in accordance with RFC 3834. (Previously it treated
1802 anything starting "auto-" as non-personal.)
1804 TF/01 The control=submission/name=... option had a problem with syntax
1805 errors if the name included a slash character. The /name= option
1806 now slurps the rest of the string, so it can include any characters
1807 but it must come last in the list of options (after /sender_retain
1810 PH/05 Some modifications to the interface to the fake nameserver for the new
1818 TK/01 Added the "success_on_redirect" address verification option. See
1819 NewStuff for rationale and an example.
1821 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1823 PH/02 Patch to exigrep to allow it to work on syslog lines.
1825 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1826 fread() to skip over the body file's header line, because in Cygwin the
1827 header line is locked and is inaccessible.
1829 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1830 co-exist for some time) to make it clear that it is the Exim ID that is
1831 referenced, not the Message-ID: header line.
1833 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1834 string_format() function, because snprintf() does not exist on all
1837 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1838 ${stat: expansion item.
1840 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1841 protocol synchronization error", to keep the pedants happy.
1843 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1844 well as for IRIX systems, when gcc is being used. See the host.c source
1847 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1849 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1852 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1853 transport and to the smtp transport in LMTP mode.
1855 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1857 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1858 run a filter in a subprocess. This could lead to confusion in subsequent
1859 lookups in the parent process. There should also be a search_tidyup() at
1860 the end of the subprocess.
1862 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1863 only if the host matched helo_try_verify_hosts, which caused the
1864 verification to occur when the EHLO/HELO command was issued. The ACL just
1865 tested the remembered result. Now, if a previous verification attempt has
1866 not happened, "verify = helo" does it there and then.
1868 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1870 TK/03 Fix log output including CR from clamd.
1872 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1873 exist provoked a memory error which could cause a segfault.
1875 PH/15 Installed PCRE 6.2
1877 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1879 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1880 of the problem. Specifically, suggested +O2 rather than +O1 for the
1883 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1885 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1886 record, the port was not correctly logged when the outgoing_port log
1887 selector was set (it logged the transort's default port).
1889 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1890 fallback_hosts, and "hosts" in the smtp transport.
1892 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1893 host errors such as "Connection refused".
1895 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1896 authentication with radiusclient 0.4.9:
1898 - Error returned from rc_read_config was caught wrongly
1899 - Username/password not passed on to radius server due to wrong length.
1901 The presumption is that some radiusclient API changes for 4.51/PH/17
1902 were not taken care of correctly. The code is still untested by me (my
1903 Linux distribution still has 0.3.2 of radiusclient), but it was
1904 contributed by a Radius user.
1906 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1907 expanding the "port" option of the smtp transport.
1909 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1910 while reading a MIME header. Thanks to Tom Hughes for a patch.
1912 PH/24 Include config.h inside local_scan.h so that configuration settings are
1915 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1916 that in an ACL for RCPT (for example), you can examine exactly what was
1919 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1920 commands, but it was not correctly comparing the address with the actual
1921 client host address. Thus, it would show the EHLO address in Received:
1922 header lines when this was not necessary.
1924 PH/27 Added the % operator to ${eval:}.
1926 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1927 it should be ignoring failures (because with -C, for example, it has lost
1928 privilege). It wasn't ignoring creation failures other than "already
1931 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1934 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1935 because an input line was too long, either on its own, or by virtue of
1936 too many continuations, the temporary file was not being removed, and the
1937 return code was incorrect.
1939 PH/31 Missing "BOOL" in function definition in filtertest.c.
1941 PH/32 Applied Sieve patches from the maintainer.
1943 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1945 PH/33 Added "verify = not_blind".
1947 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1948 Local/Makefile (with some defaults set). These are used in built scripts
1949 such as exicyclog, but they have never been used in the exim_install
1950 script (though there are many overriding facilities there). I have
1951 arranged that the exim_install script now takes note of these two
1954 PH/35 Installed configuration files for Dragonfly.
1956 PH/36 When a locally submitted message by a trusted user did not contain a
1957 From: header, and the sender address was obtained from -f or from an SMTP
1958 MAIL command, and the trusted user did not use -F to supply a sender
1959 name, $originator_name was incorrectly used when constructing a From:
1960 header. Furthermore, $originator_name was used for submission mode
1961 messages from external hosts without From: headers in a similar way,
1962 which is clearly wrong.
1964 PH/37 Added control=suppress_local_fixups.
1966 PH/38 When log_selector = +received_sender was set, and the addition of the
1967 sender made the log line's construction buffer exactly full, or one byte
1968 less than full, an overflow happened when the terminating "\n" was
1971 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1972 when the result of a list match is failure because a DNS lookup failed.
1974 PH/40 RM_COMMAND is now used in the building process.
1976 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1977 the "build-* directories that it finds.
1979 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1980 address in a domain literal was a prefix of an interface address.
1982 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1983 when verifying a sender address, unless rewrite_headers is false.
1985 PH/44 Wrote a long comment about why errors_to addresses are verified as
1986 recipients, not senders.
1988 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1989 the ratelimit ACL was added.
1991 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1993 PH/46 Added extra information about PostgreSQL errors to the error string.
1995 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1996 avoiding the need to install special zones in a real server. This is
1997 backwards compatible; if it can't find the fake resolver, it drops back.
1998 Thus, both old and new test suites can be run.
2000 TF/02 Added util/ratelimit.pl
2002 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2003 clock is set back in time.
2005 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2006 Candler <B.Candler@pobox.com>.
2008 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2009 OK for addresses that are the result of redirections.
2011 PH/48 A number of further additions for the benefit of the new test suite,
2012 including a fake gethostbyname() that interfaces to the fake DNS resolver
2015 TF/06 The fix for widen_domains has also been applied to qualify_single and
2016 search_parents which are the other dnslookup options that can cause
2019 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2022 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2024 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2025 service" if it encountered an MX record with an empty target hostname.
2026 The message is now "an MX or SRV record indicated no SMTP service".
2028 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2029 if the DNS of the sending site is misconfigured. This is quite a
2030 common situation. This change restores the behaviour of treating a
2031 helo verification defer as a failure.
2033 PH/51 If self=fail was set on a router, the bounce message did not include the
2034 actual error message.
2040 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2042 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2043 command itself ended in error, the underlying message about the transport
2044 filter timeout was being overwritten with the pipe command error. Now the
2045 underlying error message should be appended to the second error message.
2047 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2049 PH/02 Reduce the amount of output that "make" produces by default. Full output
2050 can still be requested.
2052 PH/03 The warning log line about a condition test deferring for a "warn" verb
2053 was being output only once per connection, rather than after each
2054 occurrence (because it was using the same function as for successful
2055 "warn" verbs). This seems wrong, so I have changed it.
2057 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2058 it should not have, which might have caused a crash in the right
2059 circumstances, but probably never did.
2061 PH/04 Installed a modified version of Tony Finch's patch to make submission
2062 mode fix the return path as well as the Sender: header line, and to
2063 add a /name= option so that you can make the user's friendly name appear
2066 TF/03 Added the control = fakedefer ACL modifier.
2068 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2069 Mark Lowes for thorough testing.
2071 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2073 TK/03 Merged latest SRS patch from Miles Wilton.
2075 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2076 with the definition in sysexits.h (which is #included earlier).
2077 Fortunately, Exim does not actually use EX_OK. The code used to try to
2078 preserve the sysexits.h value, by assumimg that macro definitions were
2079 scanned for macro replacements. I have been disabused of this notion,
2080 so now the code just undefines EX_OK before #including unistd.h.
2082 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2083 in the smtp transport. When a block could not be written in a single
2084 write() function, the timeout was being re-applied to each part-write.
2085 This seems wrong - if the receiver was accepting one byte at a time it
2086 would take for ever. The timeout is now adjusted when this happens. It
2087 doesn't have to be particularly precise.
2089 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2090 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2092 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2093 without a domain if the check to <postmaster@domain> fails.
2095 SC/01 Eximstats: added -xls and the ability to specify output files
2096 (patch written by Frank Heydlauf).
2098 SC/02 Eximstats: use FileHandles for outputing results.
2100 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2102 SC/04 Eximstats: fixed display of large numbers with -nvr option
2104 SC/05 Eximstats: fixed merging of reports with empty tables.
2106 SC/06 Eximstats: added the -include_original_destination flag
2108 SC/07 Eximstats: removed tabs and trailing whitespace.
2110 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2112 TK/06 MBOX spool code: Add real "From " MBOX separator line
2113 so the .eml file is really in mbox format (even though
2114 most programs do not really care). Patch from Alex Miller.
2116 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2117 The latter is generated from $received_to and is only set if the
2118 message has one envelope recipient. SA can use these headers,
2119 obviously out-of-the-box. Patch from Alex Miller.
2121 PH/08 The ${def test on a variable was returning false if the variable's
2122 value was "0", contrary to what the specification has always said!
2123 The result should be true unless the variable is empty.
2125 PH/09 The syntax error of a character other than { following "${if
2126 def:variable_name" (after optional whitespace) was not being diagnosed.
2127 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2128 accidental colon was present, for example, could give incorrect results.
2130 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2131 result is used (not including appendfile, where other changes are about
2134 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2135 This involved changing a lot of size variables from int to off_t. It
2136 should work with maildirs and everything.
2138 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2139 spamd dying while we are connected to it.
2141 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2144 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2145 function for BDB 4.3.
2147 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2149 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2150 the sources for PCRE have been re-arranged and more files are now
2153 PH/15 The code I had for printing potentially long long variables in PH/11
2154 above was not the best (it lost precision). The length of off_t variables
2155 is now inspected at build time, and an appropriate printing format (%ld
2156 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2157 to be "long long int" or "long int". This is needed for the internal
2158 formatting function string_vformat().
2160 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2161 the configuration file to be ":syslog", then the script "guesses" where
2162 the logs files are, rather than using the compiled in default. In our
2163 case the guess is not the same as the compiled default, so the script
2164 suddenly stopped working when I started to use syslog. The patch checks
2165 to see if log_file_path is "". If so, it attempts to read it from exim
2166 with no configuration file to get the compiled in version, before it
2167 falls back to the previous guessing code."
2169 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2170 implementing BATV in an Exim configuration. See NewStuff for the gory
2173 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2174 Makefile that are specific to HP-UX.
2176 PH/18 If the "use_postmaster" option was set for a recipient callout together
2177 with the "random" option, the postmaster address was used as the MAIL
2178 FROM address for the random test, but not for the subsequent recipient
2179 test. It is now used for both.
2181 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2182 patch removes a few documentation additions to RFC 3028, because the
2183 latest draft now contains them. It adds the new en;ascii-case comparator
2184 and a new error check for 8bit text in MIME parts. Comparator and
2185 require names are now matched exactly. I enabled the subaddress
2186 extension, but it is not well tested yet (read: it works for me)."
2188 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2189 rework some of the code of TK/09 above to avoid the hardwired use of
2190 "%lld" and "long long". Replaced the call to snprintf() with a call to
2193 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2194 records point to non-existent hosts", "retry timeout exceeded", and
2195 "retry time not reached for any host after a long failure period".
2197 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2198 experimental DomainKeys support:
2200 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2201 (2) On an error such as an illegally used "control", the wrong name for
2202 the control was given.
2204 These problems did NOT occur unless DomainKeys support was compiled.
2206 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2208 PH/24 Added ${if match_ip condition.
2210 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2211 databases so that it will be absolutely obvious if a crash occurs in the
2212 DB library. This is a regular occurrence (often caused by mis-matched
2215 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2216 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2217 picked up on a user's system that detects such things. There doesn't seem
2218 to be a gcc warning option for this - only an attribute that has to be
2219 put on the function's prototype. It seems that in Fedora Core 4 they have
2220 set this on a number of new functions. No doubt there will be more in due
2223 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2224 specify a transport. However, if an address that was verified by such a
2225 router was the subject of a callout, Exim crashed because it tried to
2226 read the rcpt_include_affixes from the non-existent transport. Now it
2227 just assumes that the setting of that option is false. This bug was
2228 introduced by 4.51/PH/31.
2230 PH/28 Changed -d+all to exclude +memory, because that information is very
2231 rarely of interest, but it makes the output a lot bigger. People tend to
2232 do -d+all out of habit.
2234 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2235 code in os-type was giving problems when libc.so lives in lib64, like on
2238 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2239 aren't the modern standard, and it seems that some systems' include files
2240 don't always have them. Exim was already checking for some of the newer
2241 ones like T_AAAA, and defining it itself. I've added checks for all the
2242 record types that Exim uses.
2244 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2245 not automatically generating a new one, as it is supposed to. This
2246 prevented TLS from working. If the file did exist, but contained invalid
2247 data, a new version was generated, as expected. It was only the case of a
2248 non-existent file that was broken.
2250 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2251 with a change in libdomainkeys > 0.64.
2253 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2254 from DNS. If the selector record carries the flag, it now has
2255 precedence over the domain-wide flag.
2257 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2259 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2260 the use of a port name that isn't defined in /etc/services) occurred, the
2261 message was deferred as in a normal delivery, and thus remained on the
2262 spool, instead of being failed because of the mua_wrapper setting. This
2263 is now fixed, and I tidied up some of the mua_wrapper messages at the
2266 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2267 the messages in a hash of arrays rather than using individual hashes.
2268 This is a bit cleaner and results in dramatic memory savings, albeit
2269 at a slight CPU cost.
2271 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2272 as requested by Marc Sherman.
2274 SC/10 Eximstats: added histograms for user specified patterns as requested
2277 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2279 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2280 fopen() in the content-scanning modules that did not already have it.
2286 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2287 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2289 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2291 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2293 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2294 to test Sieve filters that use "vacation".
2296 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2297 that changes the way the GnuTLS parameters are stored in the cache file.
2298 The new format can be generated externally. For backward compatibility,
2299 if the data in the cache doesn't make sense, Exim assumes it has read an
2300 old-format file, and it generates new data and writes a new file. This
2301 means that you can't go back to an older release without removing the
2304 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2305 work if there are any delivery delays because "one_time" forces the
2306 parent to be marked "delivered", so its unseen clone is never tried
2307 again. For this reason, Exim now forbids the simultaneous setting of
2310 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2311 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2312 entry if you want to know the details. The fix, however, neglected to
2313 consider the case when local delivery batching is involved. The test for
2314 "previously delivered" was not happening when checking to see if an
2315 address could be batched with a previous (undelivered) one; under
2316 certain circumstances this could lead to multiple deliveries to the same
2319 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2320 in its include files, and this causes problems building Exim.
2322 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2323 header_syntax) but Exim was just ignoring anything given after a slash.
2324 In particular, this caused confusion with an attempt to use "verify =
2325 reverse_host_lookup/defer_ok". An error is now given when options are
2326 supplied for verify items that do not have them. (Maybe reverse_host_
2327 lookup should have a defer_ok option, but that's a different point.)
2329 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2330 defined by RFC 821) to 2048, because there were problems with some AUTH
2331 commands, and RFC 1869 says the size should be increased for extended
2332 SMTP commands that take arguments.
2334 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2337 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2338 "unknown" error; now it says that the functionality isn't in the binary.
2340 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2341 an address' error message when a string expansion fails (syntax or
2342 whatever). Otherwise the password may appear in the log. Following change
2343 PH/42 below, there is no longer a chance of it appearing in a bounce
2346 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2348 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2349 of its addresses was ever tried. (Bugzilla bug #2.)
2351 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2352 the result incorrectly in the debug output. (It correctly added a newline
2353 to what was transported.)
2355 TF/01 Added $received_time.
2357 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2358 commented out examples of how to interface to a virus scanner and to
2359 SpamAssassin. Also added commented examples of av_scanner and
2360 spamd_address settings.
2362 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2363 and controls are allowed in which ACLs. There were a couple of minor
2364 errors. Some of the entries in the conditions table (which is a table of
2365 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2366 negation of where the condition IS allowed.
2368 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2370 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2371 header file does not have a version number, so I've had to invent a new
2372 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2373 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2374 radiusclient), but it was contributed by a Radius user.
2376 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2377 files or directories, for OpenSSL.
2379 PH/19 When an Exim process that is running as root has to create an Exim log
2380 file, it does so in a subprocess that runs as exim:exim so as to get the
2381 ownership right at creation (otherwise, other Exim processes might see
2382 the file with the wrong ownership). There was no test for failure of this
2383 fork() call, which would lead to the process getting stuck as it waited
2384 for a non-existent subprocess. Forks do occasionally fail when resources
2385 run out. I reviewed all the other calls to fork(); they all seem to check
2388 PH/20 When checking for unexpected SMTP input at connect time (before writing
2389 the banner), Exim was not dealing correctly with a non-positive return
2390 from the read() function. If the client had disconnected by this time,
2391 the result was a log entry for a synchronization error with an empty
2392 string after "input=" when read() returned zero. If read() returned -1
2393 (an event I could not check), uninitialized data bytes were printed.
2394 There were reports of junk text (parts of files, etc) appearing after
2397 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2399 PH/22 Added support for macro redefinition, and (re)definition in between
2400 driver and ACL definitions.
2402 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2403 forgetting to use the resulting value; it was using the unexpanded value.
2405 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2406 hadn't been configured. The fix is from Juergen Kreileder, who
2407 understands it better than I do:
2409 "Here's what I see happening with three configured cyrus_sasl
2410 authenticators configured (plain, login, cram-md5):
2412 On startup auth_cyrus_sasl_init() gets called for each of these.
2413 This means three calls to sasl_listmech() without a specified mech_list.
2414 => SASL tests which mechs of all available mechs actually work
2415 => three warnings about OTP not working
2416 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2418 With the patch, sasl_listmech() also gets called three times. But now
2419 SASL's mech_list option is set to the server_mech specified in the the
2420 authenticator. Or in other words, the answer from sasl_listmech()
2421 gets limited to just the mech you're testing for (which is different
2423 => the return list contains just 'plain' or 'login', 'cram-md5' or
2424 nothing depending on the value of ob->server_mech.
2426 I've just tested the patch: Authentication still works fine,
2427 unavailable mechs specified in the exim configuration are still
2428 caught, and the auth.log warnings about OTP are gone."
2430 PH/25 When debugging is enabled, the contents of the command line are added
2431 to the debugging output, even when log_selector=+arguments is not
2434 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2435 answer is "GNU", and only if the return is "GNU/something" is the answer
2438 PH/27 $acl_verify_message is now set immediately after the failure of a
2439 verification in an ACL, and so is available in subsequent modifiers. In
2440 particular, the message can be preserved by coding like this:
2442 warn !verify = sender
2443 set acl_m0 = $acl_verify_message
2445 Previously, $acl_verify_message was set only while expanding "message"
2446 and "log_message" when a very denied access.
2448 PH/28 Modified OS/os.c-Linux with
2450 -#ifndef OS_LOAD_AVERAGE
2451 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2453 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2454 nomenclature these days.)
2456 PH/29 Installed patch from the Sieve maintainer that adds the options
2457 sieve_useraddress and sieve_subaddress to the redirect router.
2459 PH/30 In these circumstances:
2460 . Two addresses routed to the same list of hosts;
2461 . First host does not offer TLS;
2462 . First host accepts first address;
2463 . First host gives temporary error to second address;
2464 . Second host offers TLS and a TLS session is established;
2465 . Second host accepts second address.
2466 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2467 and peerdn, if requested) that were in fact used only for the second
2470 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2471 attention to any local part prefix or suffix that was matched by the
2472 router that accepted the address. It now behaves in the same way as it
2473 does for delivery: the affixes are removed from the local part unless
2474 rcpt_include_affixes is set on the transport.
2476 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2477 timeout during the DATA phase of an incoming message.
2479 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2480 applied a patch sanctioned by the Sieve maintainer.
2482 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2483 the uid or gid is negative. A case of a negative gid caused this to be
2484 noticed. The fix allows for either to be negative.
2486 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2487 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2490 PH/36 The change PH/12 above was broken. Fixed it.
2492 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2493 the grounds that routing the same address twice would always produce the
2494 same answer. This might have been true once, but it is certainly no
2495 longer true now. Routing a child address may depend on the previous
2496 routing that produced that child. Some complicated redirection strategies
2497 went wrong when messages had multiple recipients, and made Exim's
2498 behaviour dependent on the order in which the addresses were given.
2500 I have moved the duplicate checking until after the routing is complete.
2501 Exim scans the addresses that are assigned to local and remote
2502 transports, and removes any duplicates. This means that more work will be
2503 done, as duplicates will always all be routed, but duplicates are
2504 presumably rare, so I don't expect this is of any significance.
2506 For deliveries to pipes, files, and autoreplies, the duplicate checking
2507 still happens during the routing process, since they are not going to be
2510 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2511 It corrects a timeout issue with spamd. This is Ian's comment: "The
2512 background is that sometimes spamd either never reads data from a
2513 connection it has accepted, or it never writes response data. The exiscan
2514 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2515 blindly assumes that writes won't block so it may never time out."
2517 PH/39 Allow G after quota size as well as K and M.
2519 PH/40 The value set for $authenticated_id in an authenticator may not contain
2520 binary zeroes or newlines because the value is written to log lines and
2521 to spool files. There was no check on this. Now the value is run through
2522 the string_printing() function so that such characters are converted to
2523 printable escape sequences.
2525 PH/41 $message_linecount is a new variable that contains the total number of
2526 lines in the message. Compare $body_linecount, which is the count for the
2529 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2530 bounce and delay warning messages, except in certain special cases, which
2533 (a) An SMTP error message from a remote host;
2534 (b) A message specified in a :fail: redirection;
2535 (c) A message specified in a "fail" command in a system filter;
2536 (d) A message specified in a FAIL return from the queryprogram router;
2537 (e) A message specified by the cannot_route_message router option.
2539 In these cases only, Exim does include the error details in bounce and
2540 warning messages. There are also a few cases where bland messages such
2541 as "unrouteable address" or "local delivery error" are given.
2543 PH/43 $value is now also set for the "else" part of a ${run expansion.
2545 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2546 being worked on, but at least Exim now implements the latest version to
2549 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2550 process to complete was treated as a delivery failure, a timeout while
2551 writing the message to the pipe was logged, but erroneously treated as a
2552 successful delivery. Such timeouts include transport filter timeouts. For
2553 consistency with the overall process timeout, these timeouts are now
2554 treated as errors, giving rise to delivery failures by default. However,
2555 there is now a new Boolean option for the pipe transport called
2556 timeout_defer, which, if set TRUE, converts the failures into defers for
2557 both kinds of timeout. A transport filter timeout is now identified in
2560 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2561 systems where "make" and "gmake" are different, calling "gmake" at top
2562 level broke things. I've arranged for the value of $(MAKE) to be passed
2563 from the Makefile to this script so that it can call the same version of
2567 A note about Exim versions 4.44 and 4.50
2568 ----------------------------------------
2570 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2571 changes of various kinds. As a consequence, a big documentation update was
2572 needed. This delayed the release for rather longer than seemed good, especially
2573 in the light of a couple of (minor) security issues. Therefore, the changes
2574 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2575 release. So 4.44 and 4.50 are in effect two different branches that both start
2578 I have left the 4.50 change log unchanged; it contains all the changes since
2579 4.43. The change log for 4.44 is below; many of its items are identical to
2580 those for 4.50. This seems to be the most sensible way to preserve the
2581 historical information.
2587 1. Minor wording change to the doc/README.SIEVE file.
2589 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2590 computation of the current number of files was incorrect.
2592 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2593 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2596 4. Give more explanation in the error message when the command for a transport
2597 filter fails to execute.
2599 5. There are several places where Exim runs a non-Exim command in a
2600 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2601 was being done only for the command run by the queryprogram router. It is
2602 now done for all such subprocesses. The other cases are: ${run, transport
2603 filters, and the commands run by the lmtp and pipe transports.
2605 6. Added CONFIGURE_GROUP build-time option.
2607 7. Some older OS have a limit of 256 on the maximum number of file
2608 descriptors. Exim was using setrlimit() to set 1000 as a large value
2609 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2610 systems. I've change it so that if it can't get 1000, it tries for 256.
2612 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2613 was an oversight, and furthermore, ever since the addition of extra
2614 controls (e.g. 4.43/32), the checks on when to allow different forms of
2615 "control" were broken. There should now be diagnostics for all cases when a
2616 control that does not make sense is encountered.
2618 9. Added the /retain_sender option to "control=submission".
2620 10. $recipients is now available in the predata ACL (oversight).
2622 11. Tidy the search cache before the fork to do a delivery from a message
2623 received from the command line. Otherwise the child will trigger a lookup
2624 failure and thereby defer the delivery if it tries to use (for example) a
2625 cached ldap connection that the parent has called unbind on.
2627 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2628 of $address_data from the recipient verification was clobbered by the
2629 sender verification.
2631 13. The value of address_data from a sender verification is now available in
2632 $sender_address_data in subsequent conditions in the ACL statement.
2634 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2636 15. Added a new option "connect=<time>" to callout options, to set a different
2639 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2640 was its contents. (It was OK if the option was not defined at all.)
2642 17. A "Completed" log line is now written for messages that are removed from
2643 the spool by the -Mrm option.
2645 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2646 information about exactly what failed.
2648 19. Added -dd to debug only the daemon process.
2650 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2651 handles timeouts, both on the server side and network timeouts. Renamed the
2652 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2654 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2655 It is now set to "smtps".
2657 22. $host_address is now set to the target address during the checking of
2658 ignore_target_hosts.
2660 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2661 being passed; this would have caused $sender_host_name to have been used if
2662 matching the list had actually called for a host name (not very likely,
2663 since this list is usually IP addresses). A host name is now passed as
2666 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2667 code that specifies a non-restarting handler (typically sigaction() in
2668 modern systems) in an attempt to fix a rare and obscure crash bug.
2670 25. Narrowed the window for a race in the daemon that could cause it to ignore
2671 SIGCHLD signals. This is not a major problem, because they are used only to
2672 wake it up if nothing else does.
2674 26. A malformed maildirsize file could cause Exim to calculate negative values
2675 for the mailbox size or file count. Odd effects could occur as a result.
2676 The maildirsize information is now recalculated if the size or filecount
2679 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2680 support for a long time. Removed HAVE_SYS_VFS_H.
2682 28. Installed the latest version of exipick from John Jetmore.
2684 29. In an address list, if the pattern was not a regular expression, an empty
2685 subject address (from a bounce message) matched only if the pattern was an
2686 empty string. Non-empty patterns were not even tested. This was the wrong
2687 because it is perfectly reasonable to use an empty address as part of a
2688 database query. An empty address is now tested by patterns that are
2689 lookups. However, all the other forms of pattern expect the subject to
2690 contain a local part and a domain, and therefore, for them, an empty
2691 address still always fails if the pattern is not itself empty.
2693 30. Exim went into a mad DNS loop when attempting to do a callout where the
2694 host was specified on an smtp transport, and looking it up yielded more
2695 than one IP address.
2697 31. Re-factored the code for checking spool and log partition space into a
2698 function that finds that data and another that does the check. The former
2699 is then used to implement four new variables: $spool_space, $log_space,
2700 $spool_inodes, and $log_inodes.
2702 32. The RFC2047 encoding function was originally intended for short strings
2703 such as real names; it was not keeping to the 75-character limit for
2704 encoded words that the RFC imposes. It now respects the limit, and
2705 generates multiple encoded words if necessary. To be on the safe side, I
2706 have increased the buffer size for the ${rfc2047: expansion operator from
2709 33. It is now permitted to omit both strings after an "if" condition; if the
2710 condition is true, the result is "true". As before, when the second string
2711 is omitted, a false condition yields an empty string. This makes it less
2712 cumbersome to write custom ACL and router conditions.
2714 34. Failure to deliver a bounce message always caused it to be frozen, even if
2715 there was an errors_to setting on the router. The errors_to setting is now
2718 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2719 canonical form (fully expanded) before being placed in
2720 $sender_host_address.
2722 36. The table in the code that translates DNS record types into text (T_A to
2723 "A" for instance) was missing entries for NS and CNAME. It is just possible
2724 that this could have caused confusion if both these types were looked up
2725 for the same domain, because the text type is used as part of Exim's
2726 per-process caching. But the chance of anyone hitting this buglet seems
2729 37. The dnsdb lookup has been extended in a number of ways.
2731 (1) There is a new type, "zns", which walks up the domain tree until it
2732 finds some nameserver records. It should be used with care.
2734 (2) There is a new type, "mxh", which is like "mx" except that it returns
2735 just the host names, not the priorities.
2737 (3) It is now possible to give a list of domains (or IP addresses) to be
2738 looked up. The behaviour when one of the lookups defers can be
2739 controlled by a keyword.
2741 (4) It is now possible to specify the separator character for use when
2742 multiple records are returned.
2744 38. The dnslists ACL condition has been extended: it is now possible to supply
2745 a list of IP addresses and/or domains to be looked up in a particular DNS
2748 39. Added log_selector=+queue_time_overall.
2750 40. When running the queue in the test harness, wait just a tad after forking a
2751 delivery process, to get repeatability of debugging output.
2753 41. Include certificate and key file names in error message when GnuTLS fails
2754 to set them up, because the GnuTLS error message doesn't include the name
2755 of the failing file when there is a problem reading it.
2757 42. Allow both -bf and -bF in the same test run.
2759 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2761 44. The "Exiscan patch" is now merged into the mainline Exim source.
2763 45. Sometimes the final signoff response after QUIT could fail to get
2764 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2765 < 0 before doing a fflush(). This bug looks as though it goes back to the
2766 introduction of TLS in release 3.20, but "sometimes" must have been rare
2767 because the tests only now provoked it.
2769 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2770 (this can affect the format of dates).
2772 47. exim_tidydb, when checking for the continued existence of a message for
2773 which it has found a message-specific retry record, was not finding
2774 messages that were in split spool directories. Consequently, it was
2775 deleting retry records that should have stayed in existence.
2777 48. Steve fixed some bugs in eximstats.
2779 49. The SPA authentication driver was not abandoning authentication and moving
2780 on to the next authenticator when an expansion was forced to fail,
2781 contradicting the general specification for all authenticators. Instead it
2782 was generating a temporary error. It now behaves as specified.
2784 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2785 (the order specifies the preference for clients). The order is now AES256,
2786 AES128, 3DES, ARCFOUR128.
2788 51. Small patch to Sieve code - explicitly set From: when generating an
2791 52. Exim crashed if a remote delivery caused a very long error message to be
2792 recorded - for instance if somebody sent an entire SpamAssassin report back
2793 as a large number of 550 error lines. This bug was coincidentally fixed by
2794 increasing the size of one of Exim's internal buffers (big_buffer) that
2795 happened as part of the Exiscan merge. However, to be on the safe side, I
2796 have made the code more robust (and fixed the comments that describe what
2799 53. Now that there can be additional text after "Completed" in log lines (if
2800 the queue_time_overall log selector is set), a one-byte patch to exigrep
2801 was needed to allow it to recognize "Completed" as not the last thing in
2804 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2805 patch that reportedly fixes this has been added. I am not expert enough to
2806 create a test for it. This is what the patch creator wrote:
2808 "I found a little strange behaviour of ldap code when working with
2809 Windows 2003 AD Domain, where users was placed in more than one
2810 Organization Units. When I tried to give exim partial DN, the exit code
2811 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2812 But simultaneously result of request was absolutely normal ldap result,
2813 so I produce this patch..."
2815 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2816 REFERENCE, so I have modified the code to exclude the patch when that macro
2819 55. Some experimental protocols are using DNS PTR records for new purposes. The
2820 keys for these records are domain names, not reversed IP addresses. The
2821 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2822 leaves it alone. Component reversal etc. now happens only for IP addresses.
2825 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2827 57. Double the size of the debug message buffer (to 2048) so that more of very
2828 long debug lines gets shown.
2830 58. The exicyclog utility now does better if the number of log files to keep
2831 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2833 59. Two changes related to the smtp_active_hostname option:
2835 (1) $smtp_active_hostname is now available as a variable.
2836 (2) The default for smtp_banner uses $smtp_active_hostname instead
2837 of $primary_hostname.
2839 60. The host_aton() function is supposed to be passed a string that is known
2840 to be a valid IP address. However, in the case of IPv6 addresses, it was
2841 not checking this. This is a hostage to fortune. Exim now panics and dies
2842 if the condition is not met. A case was found where this could be provoked
2843 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2844 components; fortuitously, this particular loophole had already been fixed
2845 by change 4.50/55 above.
2847 If there are any other similar loopholes, the new check in host_aton()
2848 itself should stop them being exploited. The report I received stated that
2849 data on the command line could provoke the exploit when Exim was running as
2850 exim, but did not say which command line option was involved. All I could
2851 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2852 running as the user.
2855 61. There was a buffer overflow vulnerability in the SPA authentication code
2856 (which came originally from the Samba project). I have added a test to the
2857 spa_base64_to_bits() function which I hope fixes it.
2860 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2861 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2863 63. The daemon start-up calls getloadavg() while still root for those OS that
2864 need the first call to be done as root, but it missed one case: when
2865 deliver_queue_load_max is set with deliver_drop_privilege. This is
2866 necessary for the benefit of the queue runner, because there is no re-exec
2867 when deliver_drop_privilege is set.
2869 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2870 This has been fixed.
2872 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2873 was in use, was not putting the data itself into the right store pool;
2874 consequently, it could be overwritten for a subsequent message in the same
2875 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2878 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2880 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2881 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2882 the function were treating the return as a boolean value, which happened to
2883 work because 0=false and not-0=true, but is not correct code.
2885 68. The host_aton() function was not handling scoped IPv6 addresses (those
2886 with, for example, "%eth0" on the end) correctly.
2888 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2889 negated items (that is, ~something) in unsigned ints. Some compilers
2890 apparently mutter when there is no cast.
2892 70. If an address verification called from an ACL failed, and did not produce a
2893 user-specific message (i.e. there was only a "system" message), nothing was
2894 put in $acl_verify_message. In this situation, it now puts the system
2897 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2898 session; change 4.31/43 added the unwanted input to the log line - except
2899 that it did not do this in the start of session case. It now does.
2901 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2902 This is wrong and can cause the other end to generate a synchronization
2903 error if it is another Exim or anything else that does the synchronization
2904 check. A QUIT command is no longer sent after a timeout.
2906 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2907 during host lookups.
2909 74. The defer_ok option of callout verification was not working if it was used
2910 when verifying addresses in header lines, that is, for this case:
2912 verify = header_sender/callout=defer_ok
2914 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2915 those file descriptors could be used for SMTP connections. If anything
2916 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2917 could be sent to the SMTP client, causing chaos. The daemon now opens
2918 stdin, stdout, and stderr to /dev/null when it puts itself into the
2921 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2922 log by default. The user can override this with suitable Perl magic.
2924 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2925 the log message when discard triggers, was not working for the DATA ACL or
2926 for the non-SMTP ACL.
2928 78. Error message wording change in sieve.c.
2930 79. If smtp_accept_max_per_host was set, the number of connections could be
2931 restricted to fewer than expected, because the daemon was trying to set up
2932 a new connection before checking whether the processes handling previous
2933 connections had finished. The check for completed processes is now done
2934 earlier. On busy systems, this bug wouldn't be noticed because something
2935 else would have woken the daemon, and it would have reaped the completed
2938 80. If a message was submitted locally by a user whose login name contained one
2939 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2940 It caused a spool format error. I have fixed the spool reading code. A
2941 related problem was that the "from" clause in the Received: line became
2942 illegal because of the space(s). It is now covered by ${quote_local_part.
2944 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2947 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2949 83. Patch from Sieve maintainer to fix unterminated string problem in
2950 "vacation" handling.
2952 84. Some minor changes to the Linux configuration files to help with other
2953 OS variants using glibc.
2955 85. One more patch for Sieve to update vacation handling to latest spec.
2958 ----------------------------------------------------
2959 See the note above about the 4.44 and 4.50 releases.
2960 ----------------------------------------------------
2966 1. Change 4.43/35 introduced a bug that caused file counts to be
2967 incorrectly computed when quota_filecount was set in an appendfile
2970 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2971 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2974 3. Give more explanation in the error message when the command for a transport
2975 filter fails to execute.
2977 4. There are several places where Exim runs a non-Exim command in a
2978 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2979 was being done only for the command run by the queryprogram router. It is
2980 now done for all such subprocesses. The other cases are: ${run, transport
2981 filters, and the commands run by the lmtp and pipe transports.
2983 5. Some older OS have a limit of 256 on the maximum number of file
2984 descriptors. Exim was using setrlimit() to set 1000 as a large value
2985 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2986 systems. I've change it so that if it can't get 1000, it tries for 256.
2988 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2989 was an oversight, and furthermore, ever since the addition of extra
2990 controls (e.g. 4.43/32), the checks on when to allow different forms of
2991 "control" were broken. There should now be diagnostics for all cases when a
2992 control that does not make sense is encountered.
2994 7. $recipients is now available in the predata ACL (oversight).
2996 8. Tidy the search cache before the fork to do a delivery from a message
2997 received from the command line. Otherwise the child will trigger a lookup
2998 failure and thereby defer the delivery if it tries to use (for example) a
2999 cached ldap connection that the parent has called unbind on.
3001 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3002 of $address_data from the recipient verification was clobbered by the
3003 sender verification.
3005 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3006 was its contents. (It was OK if the option was not defined at all.)
3008 11. A "Completed" log line is now written for messages that are removed from
3009 the spool by the -Mrm option.
3011 12. $host_address is now set to the target address during the checking of
3012 ignore_target_hosts.
3014 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3015 being passed; this would have caused $sender_host_name to have been used if
3016 matching the list had actually called for a host name (not very likely,
3017 since this list is usually IP addresses). A host name is now passed as
3020 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3021 code that specifies a non-restarting handler (typically sigaction() in
3022 modern systems) in an attempt to fix a rare and obscure crash bug.
3024 15. Narrowed the window for a race in the daemon that could cause it to ignore
3025 SIGCHLD signals. This is not a major problem, because they are used only to
3026 wake it up if nothing else does.
3028 16. A malformed maildirsize file could cause Exim to calculate negative values
3029 for the mailbox size or file count. Odd effects could occur as a result.
3030 The maildirsize information is now recalculated if the size or filecount
3033 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3034 support for a long time. Removed HAVE_SYS_VFS_H.
3036 18. Updated exipick to current release from John Jetmore.
3038 19. Allow an empty sender to be matched against a lookup in an address list.
3039 Previously the only cases considered were a regular expression, or an
3042 20. Exim went into a mad DNS lookup loop when doing a callout where the
3043 host was specified on the transport, if the DNS lookup yielded more than
3046 21. The RFC2047 encoding function was originally intended for short strings
3047 such as real names; it was not keeping to the 75-character limit for
3048 encoded words that the RFC imposes. It now respects the limit, and
3049 generates multiple encoded words if necessary. To be on the safe side, I
3050 have increased the buffer size for the ${rfc2047: expansion operator from
3053 22. Failure to deliver a bounce message always caused it to be frozen, even if
3054 there was an errors_to setting on the router. The errors_to setting is now
3057 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3058 canonical form (fully expanded) before being placed in
3059 $sender_host_address.
3061 24. Updated eximstats to version 1.33
3063 25. Include certificate and key file names in error message when GnuTLS fails
3064 to set them up, because the GnuTLS error message doesn't include the name
3065 of the failing file when there is a problem reading it.
3067 26. Expand error message when OpenSSL has problems setting up cert/key files.
3070 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3071 (this can affect the format of dates).
3073 28. exim_tidydb, when checking for the continued existence of a message for
3074 which it has found a message-specific retry record, was not finding
3075 messages that were in split spool directories. Consequently, it was
3076 deleting retry records that should have stayed in existence.
3078 29. eximstats updated to version 1.35
3079 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3080 1.35 - bugfix such that pie charts by volume are generated correctly
3082 30. The SPA authentication driver was not abandoning authentication and moving
3083 on to the next authenticator when an expansion was forced to fail,
3084 contradicting the general specification for all authenticators. Instead it
3085 was generating a temporary error. It now behaves as specified.
3087 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3088 (the order specifies the preference for clients). The order is now AES256,
3089 AES128, 3DES, ARCFOUR128.
3091 31. Small patch to Sieve code - explicitly set From: when generating an
3094 32. Exim crashed if a remote delivery caused a very long error message to be
3095 recorded - for instance if somebody sent an entire SpamAssassin report back
3096 as a large number of 550 error lines. This bug was coincidentally fixed by
3097 increasing the size of one of Exim's internal buffers (big_buffer) that
3098 happened as part of the Exiscan merge. However, to be on the safe side, I
3099 have made the code more robust (and fixed the comments that describe what
3102 33. Some experimental protocols are using DNS PTR records for new purposes. The
3103 keys for these records are domain names, not reversed IP addresses. The
3104 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3105 leaves it alone. Component reversal etc. now happens only for IP addresses.
3108 34. The host_aton() function is supposed to be passed a string that is known
3109 to be a valid IP address. However, in the case of IPv6 addresses, it was
3110 not checking this. This is a hostage to fortune. Exim now panics and dies
3111 if the condition is not met. A case was found where this could be provoked
3112 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3113 components; fortuitously, this particular loophole had already been fixed
3114 by change 4.50/55 or 4.44/33 above.
3116 If there are any other similar loopholes, the new check in host_aton()
3117 itself should stop them being exploited. The report I received stated that
3118 data on the command line could provoke the exploit when Exim was running as
3119 exim, but did not say which command line option was involved. All I could
3120 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3121 running as the user.
3124 35. There was a buffer overflow vulnerability in the SPA authentication code
3125 (which came originally from the Samba project). I have added a test to the
3126 spa_base64_to_bits() function which I hope fixes it.
3129 36. The daemon start-up calls getloadavg() while still root for those OS that
3130 need the first call to be done as root, but it missed one case: when
3131 deliver_queue_load_max is set with deliver_drop_privilege. This is
3132 necessary for the benefit of the queue runner, because there is no re-exec
3133 when deliver_drop_privilege is set.
3135 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3136 was in use, was not putting the data itself into the right store pool;
3137 consequently, it could be overwritten for a subsequent message in the same
3138 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3141 38. Sometimes the final signoff response after QUIT could fail to get
3142 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3143 < 0 before doing a fflush(). This bug looks as though it goes back to the
3144 introduction of TLS in release 3.20, but "sometimes" must have been rare
3145 because the tests only now provoked it.
3151 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3152 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3153 Now it returns an integer. A number of calls were still expecting a T/F
3154 return. Fortuitously, in all cases, the tests worked in OK situations,
3155 which is the norm. However, things would have gone wrong on any write
3156 failures on the smtp file descriptor. This function is used when sending
3157 messages over SMTP and also when doing verify callouts.
3159 2. When Exim is called to do synchronous delivery of a locally submitted
3160 message (the -odf or -odi options), it no longer closes stderr before doing
3163 3. Implemented the mua_wrapper option.
3165 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3167 5. Implemented the functions header_remove(), header_testname(),
3168 header_add_at_position(), and receive_remove_recipient(), and exported them
3171 6. If an ACL "warn" statement specified the addition of headers, Exim already
3172 inserted X-ACL-Warn: at the start if there was no header name. However, it
3173 was not making this test for the second and subsequent header lines if
3174 there were newlines in the string. This meant that an invalid header could
3175 be inserted if Exim was badly configured.
3177 7. Allow an ACL "warn" statement to add header lines at the start or after all
3178 the Received: headers, as well as at the end.
3180 8. Added the rcpt_4xx retry error code.
3182 9. Added postmaster_mailfrom=xxx to callout verification option.
3184 10. Added mailfrom=xxxx to the callout verification option, for verify=
3187 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3188 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3190 12. Inserted some casts to stop certain compilers warning when using pointer
3191 differences as field lengths or precisions in printf-type calls (mostly
3192 affecting debugging statements).
3194 13. Added optional readline() support for -be (dynamically loaded).
3196 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3197 same clock tick as a message's arrival, so that its received time was the
3198 same as the "first fail" time on the retry record, and that message
3199 remained on the queue past the ultimate address timeout, every queue runner
3200 would try a delivery (because it was past the ultimate address timeout) but
3201 after another failure, the ultimate address timeout, which should have then
3202 bounced the address, did not kick in. This was a "< instead of <=" error;
3203 in most cases the first failure would have been in the next clock tick
3204 after the received time, and all would be well.
3206 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3207 being recognized when the domain list was tested by the match_domain
3208 condition in an expansion string.
3210 16. Added the ${str2b64: operator.
3212 17. Exim was always calling setrlimit() to set a large limit for the number of
3213 processes, without checking whether the existing limit was already
3214 adequate. (It did check for the limit on file descriptors.) Furthermore,
3215 errors from getrlimit() and setrlimit() were being ignored. Now they are
3216 logged to the main and panic logs, but Exim does carry on, to try to do its
3217 job under whatever limits there are.
3219 18. Imported PCRE 5.0.
3221 19. Trivial typo in log message " temporarily refused connection" (the leading
3224 20. If the log selector return_path_on_delivery was set and an address was
3225 redirected to /dev/null, the delivery process crashed because it assumed
3226 that a return path would always be set for a "successful" delivery. In this
3227 case, the whole delivery is bypassed as an optimization, and therefore no
3230 21. Internal re-arrangement: the function for sending a challenge and reading
3231 a response while authentication was assuming a zero-terminated challenge
3232 string. It's now changed to take a pointer and a length, to allow for
3233 binary data in such strings.
3235 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3237 23. Exim was not respecting finduser_retries when seeking the login of the
3238 uid under which it was called; it was always trying 10 times. (The default
3239 setting of finduser_retries is zero.) Also, it was sleeping after the final
3240 failure, which is pointless.
3242 24. Implemented tls_on_connect_ports.
3244 25. Implemented acl_smtp_predata.
3246 26. If the domain in control=submission is set empty, Exim assumes that the
3247 authenticated id is a complete email address when it generates From: or
3248 Sender: header lines.
3250 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3251 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3252 chown and chgrp in /bin and hostname in /usr/bin.
3254 28. Exim was keeping the "process log" file open after each use, just as it
3255 does for the main log. This opens the possibility of it remaining open for
3256 long periods when the USR1 signal hits a daemon. Occasional processlog
3257 errors were reported, that could have been caused by this. Anyway, it seems
3258 much more sensible not to leave this file open at all, so that is what now
3261 29. The long-running daemon process does not normally write to the log once it
3262 has entered its main loop, and it closes the log before doing so. This is
3263 so that log files can straightforwardly be renamed and moved. However,
3264 there are a couple of unusual error situations where the daemon does write
3265 log entries, and I had neglected to close the log afterwards.
3267 30. The text of an SMTP error response that was received during a remote
3268 delivery was being truncated at 512 bytes. This is too short for some of
3269 the long messages that one sometimes sees. I've increased the limit to
3272 31. It is now possible to make retry rules that apply only when a message has a
3273 specific sender, in particular, an empty sender.
3275 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3276 it possible to be selective about when SMTP synchronization is enforced.
3278 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3280 32. Implemented hosts_connection_nolog.
3282 33. Added an ACL for QUIT.
3284 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3287 35. Added mailbox_size and mailbox_filecount to appendfile.
3289 36. Added control = no_multiline_responses to ACLs.
3291 37. There was a bug in the logic of the code that waits for the clock to tick
3292 in the case where the clock went backwards by a substantial amount such
3293 that the microsecond fraction of "now" was more than the microsecond
3294 fraction of "then" (but the whole seconds number was less).
3296 38. Added support for the libradius Radius client library this is found on
3297 FreeBSD (previously only the radiusclient library was supported).
3303 1. When certain lookups returned multiple values in the form name=value, the
3304 quoting of the values was not always being done properly. Specifically:
3305 (a) If the value started with a double quote, but contained no whitespace,
3307 (b) If the value contained whitespace other than a space character (i.e.
3308 tabs or newlines or carriage returns) it was not quoted.
3309 This fix has been applied to the mysql and pgsql lookups by writing a
3310 separate quoting function and calling it from the lookup code. The fix
3311 should probably also be applied to nisplus, ibase and oracle lookups, but
3312 since I cannot test any of those, I have not disturbed their existing code.
3314 2. A hit in the callout cache for a specific address caused a log line with no
3315 reason for rejecting RCPT. Now it says "Previous (cached) callout
3316 verification failure".
3318 3. There was an off-by-one bug in the queryprogram router. An over-long
3319 return line was truncated at 256 instead of 255 characters, thereby
3320 overflowing its buffer with the terminating zero. As well as fixing this, I
3321 have increased the buffer size to 1024 (and made a note to document this).
3323 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3324 when Exim is waiting for an SMTP response from a remote server, Exim
3325 restarts its select() call on the socket, thereby resetting its timeout.
3326 This is not a problem when such interrupts are rare. Somebody set up a cron
3327 job to run exiwhat every 2 minutes, which is less than the normal select()
3328 timeout (5 or 10 minutes). This meant that the select() timeout never
3329 kicked in because it was always reset. I have fixed this by comparing the
3330 time when an interrupt arrives with the time at the start of the first call
3331 to select(). If more time than the timeout has elapsed, the interrupt is
3332 treated as a timeout.
3334 5. Some internal re-factoring in preparation for the addition of Sieve
3335 extensions (by MH). In particular, the "personal" test is moved to a
3336 separate function, and given an option for scanning Cc: and Bcc: (which is
3337 not set for Exim filters).
3339 6. When Exim created an email address using the login of the caller as the
3340 local part (e.g. when creating a From: or Sender: header line), it was not
3341 quoting the local part when it contained special characters such as @.
3343 7. Installed new OpenBSD configuration files.
3345 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3346 try to make them clearer.
3348 9. Callout options, other than the timeout value, were being ignored when
3349 verifying sender addresses in header lines. For example, when using
3351 verify = header_sender/callout=no_cache
3353 the cache was (incorrectly) being used.
3355 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3356 only the Cygwin environment.
3358 11. When return_path_on_delivery was set as a log selector, if different remote
3359 addresses in the same message used different return paths and parallel
3360 remote delivery occurred, the wrong values would sometimes be logged.
3361 (Whenever a remote delivery process finished, the return path value from
3362 the most recently started remote delivery process was logged.)
3364 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3365 lines when AUTH and/or TLS are in use. This is the "received protocol"
3366 field. Exim used to use "asmtp" for authenticated SMTP, without any
3367 indication (in the protocol name) for TLS use. Now it follows the RFC and
3368 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3369 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3370 names appear in log lines as well as in Received: header lines.
3372 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3373 extensions, and comparison tests, and to fix some bugs.
3375 14. Changes to the "personal" filter test:
3377 (1) The test was buggy in that it was just doing the equivalent of
3378 "contains" tests on header lines. For example, if a user's address was
3379 anne@some.where, the "personal" test would incorrectly be true for
3381 To: susanne@some.where
3383 This test is now done by extracting each address from the header in turn,
3384 and checking the entire address. Other tests that are part of "personal"
3385 are now done using regular expressions (for example, to check local parts
3386 of addresses in From: header lines).
3388 (2) The list of non-personal local parts in From: addresses has been
3389 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3390 taken from the Sieve specification recommendations.
3392 (3) If the message contains any header line starting with "List-" it is
3393 treated as non-personal.
3395 (4) The test for "circular" in the Subject: header line has been removed
3396 because it now seems ill-conceived.
3398 15. Minor typos in src/EDITME comments corrected.
3400 16. Installed latest exipick from John Jetmore.
3402 17. If headers_add on a router specified a text string that was too long for
3403 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3404 of string_sprintf() is now avoided.
3406 18. $message_body_size was not set (it was always zero) when running the DATA
3407 ACL and the local_scan() function.
3409 19. For the "mail" command in an Exim filter, no default was being set for
3410 the once_repeat time, causing a random time value to be used if "once" was
3411 specified. (If the value happened to be <= 0, no repeat happened.) The
3412 default is now 0s, meaning "never repeat". The "vacation" command was OK
3413 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3414 (I found it when inspecting the code).
3416 20. There is now an overall timeout for performing a callout verification. It
3417 defaults to 4 times the callout timeout, which applies to individual SMTP
3418 commands during the callout. The overall timeout applies when there is more
3419 than one host that can be tried. The timeout is checked before trying the
3420 next host. This prevents very long delays if there are a large number of
3421 hosts and all are timing out (e.g. when the network connections are timing
3422 out). The value of the overall timeout can be changed by specifying an
3423 additional sub-option for "callout", called "maxwait". For example:
3425 verify = sender/callout=5s,maxwait=20s
3427 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3428 to the end before writing, but this should make it even safer).
3430 22. Exim was forgetting that it had advertised PIPELINING for the second and
3431 subsequent messages on an SMTP connection. It was also not resetting its
3432 memory on STARTTLS and an internal HELO.
3434 23. When Exim logs an SMTP synchronization error within a session, it now
3435 records whether PIPELINING has been advertised or not.
3437 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3438 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3439 rather than long int.
3441 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3443 26. Added the never_mail option to autoreply.
3449 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3450 crash if the getsockname() call failed; this can happen if a connection is
3451 closed very soon after it is established. The problem was simply in the
3452 order in which certain operations were done, causing Exim to try to write
3453 to the SMTP stream before it had set up the file descriptor. The bug has
3454 been fixed by making things happen in the correct order.
3460 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3461 before the connection was closed, thus losing the rejection response.
3463 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3464 some early Solaris releases, but causes trouble in current releases where
3465 socklen_t is defined.
3467 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3470 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3473 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3474 that releasing the top of it at the end releases what was used for sub-
3475 expansions (unless the block got too big). However, discard this block if
3476 the first thing is a variable or header, so that we can use its block when
3477 it is dynamic (useful for very large $message_headers, for example).
3479 6. Lookups now cache *every* query, not just the most recent. A new, separate
3480 store pool is used for this. It can be recovered when all lookup caches are
3481 flushed. Lookups now release memory at the end of their result strings.
3482 This has involved some general refactoring of the lookup sources.
3484 7. Some code has been added to the store_xxx() functions to reduce the amount
3485 of flapping under certain conditions.
3487 8. log_incoming_interface used to affect only the <= reception log lines. Now
3488 it causes the local interface and port to be added to several more SMTP log
3489 lines, for example "SMTP connection from", and rejection lines.
3491 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3493 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3495 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3496 could be overwritten at the end of the current message (or the start of a
3497 new message if it was set in a HELO ACL). The value is now preserved for
3498 the duration of the SMTP connection.
3500 12. If a transport had a headers_rewrite setting, and a matching header line
3501 contained an unqualified address, that address was qualified, even if it
3502 did not match any rewriting rules. The underlying bug was that the values
3503 of the flags that permit the existence of unqualified sender and recipient
3504 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3505 non-local messages, and by -bnq for local messages) were not being
3506 preserved with the message after it was received.
3508 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3509 "next input=" as part of the text comprising the host identity instead of
3510 the correct text. The code was using the same buffer for two different
3511 strings. However, depending on which order the printing function evaluated
3512 its arguments, the bug did not always show up. Under Linux, for example, my
3513 test suite worked just fine.
3515 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3516 doesn't work with some older versions of Perl. It has been changed to "my",
3517 which in any case is probably the better facility to use.
3519 15. A really picky compiler found some instances of statements for creating
3520 error messages that either had too many or two few arguments for the format
3523 16. The size of the buffer for calls to the DNS resolver has been increased
3524 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3525 for addresses that have a lot of PTR records. This alleviates a problem; it
3526 does not fully solve it.
3528 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3529 buffer now truncates the list and logs the incident, which is the same
3530 action as happens when Exim is looking up a host name and its aliases.
3531 Previously in this situation something unpredictable would happen;
3532 sometimes it was "internal error: store_reset failed".
3534 18. If a server dropped the connection unexpectedly when an Exim client was
3535 using GnuTLS and trying to read a response, the client delivery process
3536 crashed while trying to generate an error log message.
3538 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3539 string, for example:
3541 warn message = H1: something\nH2: something
3543 the text was added as a single header line from Exim's point of view
3544 though it ended up OK in the delivered message. However, searching for the
3545 second and subsequent header lines using $h_h2: did not work. This has been
3546 fixed. Similarly, if a system filter added multiple headers in this way,
3547 the routers could not see them.
3549 20. Expanded the error message when iplsearch is called with an invalid key to
3550 suggest using net-iplsearch in a host list.
3552 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3553 ACLs are no longer actually imposed (and a message to that effect is
3556 22. If a "gecos" field in a passwd entry contained escaped characters, in
3557 particular, if it contained a \" sequence, Exim got it wrong when building
3558 a From: or a Sender: header from that name. A second bug also caused
3559 incorrect handling when an unquoted " was present following a character
3560 that needed quoting.
3562 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3563 was not being matched caselessly.
3565 24. Arranged for all hyphens in the exim.8 source to be escaped with
3568 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3569 properly. Recipient callout cache records were still being keyed to include
3570 the sender, even when use_sender was set false. This led to far more
3571 callouts that were necessary. The sender is no longer included in the key
3572 when use_sender is false.
3574 26. Added "control = submission" modifier to ACLs.
3576 27. Added the ${base62d: operator to decode base 62 numbers.
3578 28. dnsdb lookups can now access SRV records.
3580 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3581 the configuration file.
3583 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3584 (-v) mode. This makes the output for a verbose queue run more intelligible.
3586 31. Added a use_postmaster feature to recipient callouts.
3588 32. Added the $body_zerocount variable, containing the number of binary zero
3589 bytes in the message body.
3591 33. The time of last modification of the "new" subdirectory is now used as the
3592 "mailbox time last read" when there is a quota error for a maildir
3595 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3597 35. Added +ignore_unknown as a special item in host lists.
3599 36. Code for decoding IPv6 addresses in host lists is now included, even if
3600 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3601 address was recognized as an IP address, but was then not correctly decoded
3602 into binary, causing unexpected and incorrect effects when compared with
3609 1. Very minor rewording of debugging text in manualroute to say "list of
3610 hosts" instead of "hostlist".
3612 2. If verify=header_syntax was set, and a header line with an unqualified
3613 address (no domain) and a large number of spaces between the end of the
3614 name and the colon was received, the reception process suffered a buffer
3615 overflow, and (when I tested it) crashed. This was caused by some obsolete
3616 code that should have been removed. The fix is to remove it!
3618 3. When running in the test harness, delay a bit after writing a bounce
3619 message to get a bit more predictability in the log output.
3621 4. Added a call to search_tidyup() just before forking a reception process. In
3622 theory, someone could use a lookup in the expansion of smtp_accept_max_
3623 per_host which, without the tidyup, could leave open a database connection.
3625 5. Added the variables $recipient_data and $sender_data which get set from a
3626 lookup success in an ACL "recipients" or "senders" condition, or a router
3627 "senders" option, similar to $domain_data and $local_part_data.
3629 6. Moved the writing of debug_print from before to after the "senders" test
3632 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3633 problems for message scanning, either using a data ACL, or using
3634 local_scan() because the Received: header was not generated till after they
3635 were called (in order to set the time as the time of reception completion).
3636 I have revised the way this works. The header is now generated after the
3637 body is received, but before the ACL or local_scan() are called. After they
3638 are run, the timestamp in the header is updated.
3644 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3645 before starting a queue runner without re-exec. This happened only when
3646 deliver_drop_privilege was set or when the Exim user was set to root. The
3647 effect of the bug was that timeouts during subsequent deliveries caused
3648 crashes instead of being properly handled. The handler is now left at its
3649 default (and expected) setting.
3651 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3652 message, again when deliver_drop_privilege is set or Exim is run as root.
3653 The bug described in (1) was not present in this case, but the tidying up
3654 of the other signals was missing. I have made the two cases consistent.
3656 3. The ignore_target_hosts setting on a manualroute router was being ignored
3657 for hosts that were looked up using the /MX notation.
3659 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3662 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3663 operated on the sender address. After changing the $sender_address to <>
3664 for the sender address verify, Exim was re-instated it as the original
3665 (before rewriting) address, but remembering that it had rewritten it, so it
3666 wasn't rewriting it again. This bug also had the effect of breaking the
3667 sender address verification caching when the sender address was rewritten.
3669 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3670 This has been changed so that if the ip literal address matches
3671 ignore_target_hosts, the router declines.
3673 7. Added expansion conditions match_domain, match_address, and match_local_
3674 part (NOT match_host).
3676 8. The placeholder for the Received: header didn't have a length field set.
3678 9. Added code to Exim itself and to exim_lock to test for a specific race
3679 condition that could lead to file corruption when using MBX delivery. The
3680 issue is with the lockfile that is created in /tmp. If this file is removed
3681 after a process has opened it but before that process has acquired a lock,
3682 there is the potential for a second process to recreate the file and also
3683 acquire a lock. This could lead to two Exim processes writing to the file
3684 at the same time. The added code performs the same test as UW imapd; it
3685 checks after acquiring the lock that its file descriptor still refers to
3686 the same named file.
3688 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3689 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3692 11. Added the smtp_active_hostname option. If used, this will typically be made
3693 to depend on the incoming interface address. Because $interface_address is
3694 not set up until the daemon has forked a reception process, error responses
3695 that can happen earlier (such as "too many connections") no longer contain
3698 12. If an expansion in a condition on a "warn" statement fails because a lookup
3699 defers, the "warn" statement is abandoned, and the next ACL statement is
3700 processed. Previously this caused the whole ACL to be aborted.
3702 13. Added the iplsearch lookup type.
3704 14. Added ident_timeout as a log selector.
3706 15. Added tls_certificate_verified as a log selector.
3708 16. Added a global option tls_require_ciphers (compare the smtp transport
3709 option of the same name). This controls incoming TLS connections.
3711 17. I finally figured out how to make tls_require_ciphers do a similar thing
3712 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3713 before starting the TLS session.
3715 18. Tabs are now shown as \t in -bP output.
3717 19. If the log selector return_path_on_delivery was set, Exim crashed when
3718 bouncing a message because it had too many Received: header lines.
3720 20. If two routers both had headers_remove settings, and the first one included
3721 a superfluous trailing colon, the final name in the first list and the
3722 first name in the second list were incorrectly joined into one item (with a
3723 colon in the middle).
3729 1. Added -C and -D options to the exinext utility, mainly to make it easier
3730 to include in the automated testing, but these could be helpful when
3731 multiple configurations are in use.
3733 2. The exinext utility was not formatting the output nicely when there was
3734 an alternate port involved in the retry record key, nor when there was a
3735 message id as well (for retries that were specific to a specific message
3736 and a specific host). It was also confused by IPv6 addresses, because of
3737 the additional colons they contain. I have fixed the IPv4 problem, and
3738 patched it up to do a reasonable job for IPv6.
3740 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3741 delivery, the log line now contains "pipelined" if PIPELINING was used.
3743 4. An SMTP transport process used to panic and die if the bind() call to set
3744 an explicit outgoing interface failed. This has been changed; it is now
3745 treated in the same way as a connect() failure.
3747 5. A reference to $sender_host_name in the part of a conditional expansion
3748 that was being skipped was still causing a DNS lookup. This no longer
3751 6. The def: expansion condition was not recognizing references to header lines
3752 that used bh_ and bheader_.
3754 7. Added the _cache feature to named lists.
3756 8. The code for checking quota_filecount in the appendfile transport was
3757 allowing one more file than it should have been.
3759 9. For compatibility with Sendmail, the command line option
3767 and sets the incoming protocol and host name (for trusted callers). The
3768 host name and its colon can be omitted when only the protocol is to be set.
3769 Note the Exim already has two private options, -pd and -ps, that refer to
3770 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3771 "s", but I don't think that's a major issue.
3773 10. A number of refactoring changes to the code, none of which should affect
3776 (a) The number of logging options was getting close to filling up the
3777 32-bit word that was used as a bit map. I have split them into two classes:
3778 those that are passed in the argument to log_write(), and those that are
3779 only ever tested independently outside of that function. These are now in
3780 separate 32-bit words, so there is plenty of room for expansion again.
3781 There is no change in the user interface or the logging behaviour.
3783 (b) When building, for example, log lines, the code previously used a
3784 macro that called string_cat() twice, in order to add two strings. This is
3785 not really sufficiently general. Furthermore, there was one instance where
3786 it was actually wrong because one of the argument was used twice, and in
3787 one call a function was used. (As it happened, calling the function twice
3788 did not affect the overall behaviour.) The macro has been replaced by a
3789 function that can join an arbitrary number of extra strings onto a growing
3792 (c) The code for expansion conditions now uses a table and a binary chop
3793 instead of a serial search (which was left over from when there were very
3794 few conditions). Also, it now recognizes conditions like "pam" even when
3795 the relevant support is not compiled in: a suitably worded error message is
3796 given if an attempt is made to use such a condition.
3798 11. Added ${time_interval:xxxxx}.
3800 12. A bug was causing one of the ddress fields not to be passed back correctly
3801 from remote delivery subprocesses. The field in question was not being
3802 subsequently used, so this caused to problems in practice.
3804 13. Added new log selectors queue_time and deliver_time.
3806 14. Might have fixed a bug in maildirsizefile handling that threw up
3807 "unexpected character" debug warnings, and recalculated the data
3808 unnecessarily. In any case, I expanded the warning message to give more
3811 15. Added the message "Restricted characters in address" to the statements in
3812 the default ACL that block characters like @ and % in local parts.
3814 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3815 Three changes have been made:
3817 (a) There was a serious bug; a negative response to MAIL caused the whole
3818 recipient domain to be cached as invalid, thereby blocking all messages
3819 to all local parts at the same domain, from all senders. This bug has
3820 been fixed. The domain is no longer cached after a negative response to
3821 MAIL if the sender used is not empty.
3823 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3826 (c) A new callout option, "use_sender" has been added for people who want
3827 the modified behaviour.
3833 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3836 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3837 indeed breaks things for older releases.
3839 3. Added additional logging to the case where there is a problem reading data
3840 from a filter that is running in a subprocess using a pipe, in order to
3841 try to track down a specific problem.
3843 4. Testing facility fudge: when running in the test harness and attempting
3844 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3845 getting "No route to host". Convert this to a timeout.
3847 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3850 6. Some OS don't have socklen_t but use size_t instead. This affects the
3851 fifth argument of getsockopt() amongst other things. This is now
3852 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3853 can be set for individual OS. I have set it for SunOS5, OSF1, and
3854 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3855 some earlier ones do not.
3857 7. Change 4.30/15 was not doing the test caselessly.
3859 8. The standard form for an IPv6 address literal was being rejected by address
3860 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3861 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3862 this, as well as the form without the "IPv6" on the front (but only when
3863 address literals are enabled, of course).
3865 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3867 10. Exim crashed if a message with an empty sender address specified by -f
3868 encountered a router with an errors_to setting. This could be provoked only
3869 by a command such as
3873 where an empty string was supplied; "<>" did not hit this bug.
3875 11. Installed PCRE release 4.5.
3877 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3878 remained set. It is now erased.
3880 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3881 times from message ids (which are base 36 rather than the normal 62).
3883 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3884 were being counted as actual protocol errors, and logged if the log
3885 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3886 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3887 and DATA following a set of rejected RCPTs do not count as protocol errors.
3888 In other words, Exim assumes they were pipelined, though this may not
3889 actually be the case. Of course, in all cases the client gets an
3890 appropriate error code.
3892 15. If a lookup fails in an ACL condition, a message about the failure may
3893 be available; it is used if testing the ACL cannot continue, because most
3894 such messages specify what the cause of the deferral is. However, some
3895 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3896 that caused an old message to be retained and used if a later statement
3897 caused a defer, replacing the real cause of the deferral.
3899 16. If an IP address had so many PTR records that the DNS lookup buffer
3900 was not large enough to hold them, Exim could crash while trying to process
3901 the truncated data. It now detects and logs this case.
3903 17. Further to 4.21/58, another change has been made: if (and only if) the
3904 first line of a message (the first header line) ends with CRLF, a bare LF
3905 in a subsequent header line has a space inserted after it, so as not to
3906 terminate the header.
3908 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3909 unnecessarily, used atoi() instead of strtol(), and didn't check the
3910 termination when getting file sizes from file names by regex.
3912 19. Completely re-implemented the support for maildirsize files, in the light
3913 of a number of problems with the previous contributed implementation
3914 (4.30/29). In particular:
3916 . If the quota is zero, the maildirsize file is maintained, but no quota is
3919 . If the maildir directory does not exist, it is created before any attempt
3920 to write a maildirsize file.
3922 . The quota value in the file is just a cache; if the quota is changed in
3923 the transport, the new value overrides.
3925 . A regular expression is available for excluding directories from the
3928 20. The autoreply transport checks the characters in options that define the
3929 message's headers; it allows continued headers, but it was checking with
3930 isspace() after an embedded newline instead of explicitly looking for a
3933 21. If all the "regular" hosts to which an address was routed had passed their
3934 expiry times, and had not reached their retry times, the address was
3935 bounced, even if fallback hosts were defined. Now Exim should go on to try
3938 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3939 equivalent code in the SMTP transport. Some hosts send humungous responses
3940 to HELO/EHLO, more than 1024 it seems.
3942 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3943 gives compiler warnings in some environments. I've now done it "properly",
3946 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3947 (because of problems with the built-in one) was declared to return uschar *
3948 instead of char *, causing compiler failure.
3950 25. Fixed a file descriptor leak when processing alias/forward files.
3952 26. Fixed a minor format string issue in dbfn.c.
3954 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3956 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3957 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3960 29. When a sender address is verified, it is cached, to save repeating the test
3961 when there is more than one recipient in a message. However, when the
3962 verification involves a callout, it is possible for different callout
3963 options to be set for different recipients. It is too complicated to keep
3964 track of this in the cache, so now Exim always runs a verification when a
3965 callout is required, relying on the callout cache for the optimization.
3966 The overhead is duplication of the address routing, but this should not be
3969 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3970 to be verified with callout=postmaster, and the main callout worked but the
3971 postmaster check failed, the verification correctly failed. However, if a
3972 subsequent RCPT command asked for sender verification *without* the
3973 postmaster check, incorrect caching caused this verification also to fail,
3976 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3977 it was not caching the DNS options (qualify_single, search_parents) that
3978 were used when the lookup failed. A subsequent lookup with different
3979 options therefore always gave the same answer, though there were cases
3980 where it should not have. (Example: a "domains = !$mx_any" option on a
3981 dnslookup router: the "domains" option is always processed without any
3982 widening, but the router might have qualify_single set.) Now Exim uses the
3983 cached value only when the same options are set.
3985 32. Added John Jetmore's "exipick" utility to the distribution.
3987 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3988 than a timeout (e.g. a certificate is required, and is not provided), an
3989 Exim server now closes the connection immediately. Previously it waited for
3990 the client to close - but if the client is SSL, it seems that they each
3991 wait for each other, leading to a delay before one of them times out.
3993 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3994 maintained 0.8.x compatibility because I don't think many are using it, and
3995 it is clearly obsolete.
3997 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4000 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4001 client certificate was expired. A simple patch fixes this, though I don't
4002 understand the full logic of why the verify callback is called multiple
4005 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4006 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4007 which causes problems with some clients (such as the Certicom SSL Plus
4008 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4009 disables the coutermeasure allowing Eudora to connect."
4011 38. Exim was not checking that a write() to a log file succeeded. This could
4012 lead to Bad Things if a log got too big, in particular if it hit a file
4013 size limit. Exim now panics and dies if it cannot write to a log file, just
4014 as it does if it cannot open a log file.
4016 39. Modified OS/Makefile-Linux so that it now contains
4018 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4020 The two -D definitions ensure that Exim is compiled with large file
4021 support, which makes it possible to handle log files that are bigger than
4024 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4025 instance) a domain was checked against a named list that involved a lookup,
4026 causing $domain_data to be set, then another domain was checked against the
4027 same list, then the first domain was re-checked, the value of $domain_data
4028 after the final check could be wrong. In particular, if the second check
4029 failed, it could be set empty. This bug probably also applied to
4032 41. The strip_trailing_dot option was not being applied to the address given
4033 with the -f command-line option.
4035 42. The code for reading a message's header from the spool was incrementing
4036 $received_count, but never initializing it. This meant that the value was
4037 incorrect (doubled) while delivering a message in the same process in which
4038 it was received. In the most common configuration of Exim, this never
4039 happens - a fresh exec is done - but it can happen when
4040 deliver_drop_privilege is set.
4042 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4043 it now includes up to 150 characters of the unexpected data in the log
4046 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4047 and building data strings. The size of both of these buffers was 10 000
4048 bytes - far larger than anybody would *ever* want, thought I. Needless to
4049 say, somebody hit the limit. I have increased the maximum line length to
4050 20 000 and the maximum data length of concatenated lines to 100 000. I have
4051 also fixed two bugs, because there was no checking on these buffers. Tsk,
4052 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4053 buffer is too small.
4055 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4056 lsearch lookups. Now it does.
4058 46. When parsing a route_list item in a manualroute router, a fixed-length
4059 buffer was used for the list of hosts. I made this 1024 bytes long,
4060 thinking that nobody would ever have a list of hosts that long. Wrong.
4061 Somebody had a whole pile of complicated expansion conditions, and the
4062 string was silently truncated, leading to an expansion error. It turns out
4063 that it is easier to change to an unlimited length (owing to other changes
4064 that have happened since this code was originally written) than to build
4065 structure for giving a limitation error. The length of the item that
4066 expands into the list of hosts is now unlimited.
4068 47. The lsearch lookup could not handle data where the length of text line was
4069 more than 4095 characters. Such lines were truncated, leading to shortened
4070 data being returned. It should now handle lines of any length.
4072 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4073 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4076 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4077 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4078 no longer happens in comment lines. A list of replacements is now placed
4079 at the head of all of the source files, except those whose only change is
4080 to replace PERL_COMMAND in the very first #! line.
4082 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4083 messages on the queue, with a bottom-up merge sort, using code contributed
4084 by Michael Haardt. This should make operations like -bp somewhat faster on
4085 large queues. It won't affect queue runners, except when queue_run_in_order
4088 51. Installed eximstats 1.31 in the distribution.
4090 52. Added support for SRV lookups to the dnslookup router.
4092 53. If an ACL referred to $message_body or $message_body_end, the value was not
4093 reset for any messages that followed in the same SMTP session.
4095 54. The store-handling optimization for building very long strings was not
4096 differentiating between the different store pools. I don't think this
4097 actually made any difference in practice, but I've tidied it.
4099 55. While running the routers to verify a sender address, $sender_address
4100 was still set to the sender address. This is wrong, because when routing to
4101 send a bounce to the sender, it would be empty. Therefore, I have changed
4102 it so that, while verifying a sender address, $sender_address is set to <>.
4103 (There is no change to what happens when verifying a recipient address.)
4105 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4106 target A or AAAA records (if not already returned) without resetting the
4107 qualify_single or search_parents options of the DNS resolver. These are
4108 inappropriate in this case because the targets of MX and SRV records must
4109 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4110 target that, when qualified, matched something in the local domain. These
4111 two options are now turned off when doing these lookups.
4113 57. It seems that at least some releases of Reiserfs (which does not have the
4114 concept of a fixed number of inodes) returns zero and not -1 for the
4115 number of available inodes. This interacted badly with check_spool_inodes,
4116 which assumed that -1 was the "no such thing" setting. What I have done is
4117 to check that the total number of inodes is greater than zero before doing
4118 the test of how many are available.
4120 58. When a "warn" ACL statement has a log_message modifier, the message is
4121 remembered, and not repeated. This is to avoid a lot of repetition when a
4122 message has many recipients that cause the same warning to be written.
4123 Howewer, Exim was preserving the list of already written lines for an
4124 entire SMTP session, which doesn't seem right. The memory is now reset if a
4125 new message is started.
4127 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4128 debugging output unless log_rewrite was also set.
4130 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4131 of (int)(handle) when we know that handle contains (void *)(-1).
4133 61. The Exim daemon panic-logs an error return when it closes the incoming
4134 connection. However "connection reset by peer" seems to be common, and
4135 isn't really an error worthy of noting specially, so that particular error
4138 62. When Exim is trying to find all the local interfaces, it used to panic and
4139 die if the ioctl to get the interface flags failed. However, it seems that
4140 on at least one OS (Solaris 9) it is possible to have an interface that is
4141 included in the list of interfaces, but for which you get a failure error
4142 for this call. This happens when the interface is not "plumbed" into a
4143 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4144 failure of the "get flags" call assumes that the interface is down.
4146 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4147 makes life easier for people who are doing arithmetic on fields extracted
4148 from dates, where you often get leading zeros that should not be
4149 interpreted as octal.
4151 64. Added qualify_domain to the redirect router, to override the global
4154 65. If a pathologically long header line contained very many addresses (the
4155 report of this problem mentioned 10 000) and each of them was rewritten,
4156 Exim could use up a very large amount of memory. (It kept on making new
4157 copies of the header line as it rewrote, and never released the old ones.)
4158 At the expense of a bit more processing, the header rewriting function has
4159 been changed so that it no longer eats memory in this way.
4161 66. The generation of the Received: header has been moved from the time that a
4162 message starts to be received, to the time that it finishes. The timestamp
4163 in the Received: header should now be very close to that of the <= log
4164 line. There are two side-effects of this change:
4166 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4167 logged header lines no longer include the local Received: line, because
4168 it has not yet been created. The same applies to a copy of the message
4169 that is returned to a non-SMTP sender when a message is rejected.
4171 (b) When a filter file is tested using -bf, no additional Received: header
4172 is added to the test message. After some thought, I decided that this
4175 This change does not affect the value of $received_for. It is still set
4176 after address rewriting, but before local_scan() is called.
4178 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4180 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4181 gave an unhelpful panic error message, and a defer error. I have managed to
4182 change this behaviour so that it now rejects any supplied certificate,
4183 which seems right, as the list of acceptable certificates is empty.
4185 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4186 gave an unhelpful defer error. I have not managed to make this reject any
4187 supplied certificates, but the error message it gives is "no certificate
4188 supplied", which is not helpful.
4190 70. exigrep's output now also includes lines that are not associated with any
4191 message, but which match the given pattern. Implemented by a patch from
4192 Martin Sluka, which also tidied up the Perl a bit.
4194 71. Recipient callout verification, like sender verification, was using <> in
4195 the MAIL FROM command. This isn't really the right thing, since the actual
4196 sender may affect whether the remote host accepts the recipient or not. I
4197 have changed it to use the actual sender in the callout; this means that
4198 the cache record is now keyed on a recipient/sender pair, not just the
4199 recipient address. There doesn't seem to be a real danger of callout loops,
4200 since a callout by the remote host to check the sender would use <>.
4201 [SEE ABOVE: changed after hitting problems.]
4203 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4204 temporary errors. However, in the case of such a code being given after
4205 the end of a data transmission (i.e. after ".") Exim was failing to write
4206 a retry record for the message. (Yes, there was some broken host that was
4207 actually sending 8xx at this point.)
4209 73. An unknown lookup type in a host list could cause Exim to panic-die when
4210 the list was checked. (An example that provoked this was putting <; in the
4211 middle of a list instead of at the start.) If this happened during a DATA
4212 ACL check, a -D file could be left lying around. This kind of configuration
4213 error no longer causes Exim to die; instead it causes a defer errror. The
4214 incident is still logged to the main and panic logs.
4216 74. Buglet left over from Exim 3 conversion. The message "too many messages
4217 in one connection" was written to the rejectlog but not the mainlog, except
4218 when address rewriting (yes!) was being logged.
4220 75. Added write_rejectlog option.
4222 76. When a system filter was run not as root (that is, when system_filter_user
4223 was set), the values of the $n variables were not being returned to the
4224 main process; thus, they were not subsequently available in the $sn
4227 77. Added +return_path_on_delivery log selector.
4229 78. A connection timeout was being treated differently from recipients deferred
4230 when testing hosts_max_try with a message that was older than the host's
4231 retry timeout. (The host should not be counted, thus allowing all hosts to
4232 be tried at least once before bouncing.) This may have been the cause of an
4233 occasionally reported bug whereby a message would remain on the queue
4234 longer than the retry timeout, but would be bounced if a delivery was
4235 forced. I say "may" because I never totally pinned down the problem;
4236 setting up timeout/retry tests is difficult. See also the next item.
4238 79. The ultimate address timeout was not being applied to errors that involved
4239 a combination of host plus message (for example, a timeout on a MAIL
4240 command). When an address resolved to a number of possible hosts, and they
4241 were not all tried for each delivery (e.g. because of hosts_max_try), a
4242 message could remain on the queue longer than the retry timeout.
4244 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4247 81. Fixed an obscure SMTP outgoing bug which required at least the following
4248 conditions: (a) there was another message waiting for the same server;
4249 (b) the server returned 5xx to all RCPT commands in the first message so
4250 that the message was not completed; (c) the server dropped the connection
4251 or gave a negative response to the RSET that Exim sends to abort the
4252 transaction. The observed case was a dropped connection after DATA that had
4253 been sent in pipelining mode. That is, the server had advertised PIPELINING
4254 but was not implementing it correctly. The effect of the bug was incorrect
4255 behaviour, such as trying another host, and this could lead to a crash.
4261 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4262 and daemon.c were passed as pointers to ints; they should have been
4263 pointers to socklen_t variables (which are typically unsigned ints).
4265 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4268 3. Fixed a really odd bug that affected only the testing scheme; patching a
4269 certain fixed string in the binary changed the value of another string that
4270 happened to be identical to the end of the original first string.
4272 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4273 name", it returns that address as the IP address. On some operating
4274 systems (e.g. Solaris), it also passes back the IP address string as the
4275 "host name". However, on others (e.g. Linux), it passes back an empty
4276 string. Exim wasn't checking for this, and was changing the host name to an
4277 empty string, assuming it had been canonicized.
4279 5. Although rare, it is permitted to have more than one PTR record for a given
4280 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4281 all the names associated with an address, because they do in Solaris.
4282 However, it seems that they do not in Linux for data that comes from the
4283 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4284 I found this out when I moved to a new Linux workstation and tried to run
4285 the Exim test suite.
4287 To get round this problem I have changed the code so that it now does its
4288 own call to the DNS to look up PTR records when searching for a host name.
4289 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4290 addresses that are only in /etc/hosts are still found.
4292 This behaviour is, however, controlled by an option called host_lookup_
4293 order, which defaults to "bydns:byaddr". If people want to use the other
4294 order, or indeed, just use one or the other means of lookup, they can
4295 specify it in this variable.
4297 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4298 some operating systems, this comes back from gethostbyaddr() as an empty
4299 string, and this is what Exim used to test for. However, it seems that in
4300 other systems, "." is yielded. Exim now tests for this case too.
4302 7. The values of check_spool_space and check_log_space are now held internally
4303 as a number of kilobytes instead of an absolute number of bytes. If a
4304 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4305 kilobyte. This means that much larger values can be stored.
4307 8. Exim monitor: an attempt to get the action menu when not actually pointing
4308 at a message produces an empty menu entitled "No message selected". This
4309 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4310 no entries in it ("Shell widget menu has zero width and/or height"). So I
4311 have added a single, blank menu entry in this case.
4313 9. Added ${quote_local_part.
4315 10. MIME decoding is now applied to the contents of Subject: header lines when
4318 11. Now that a reference to $sender_host_address automatically causes a reverse
4319 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4320 host lookup before query-style lookups in lists that might use this
4321 variable. This has therefore been abolished, and the "net-" prefix is no
4322 longer necessary for query-style lookups.
4324 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4325 have been a typo for LFLAGS, so it has been changed.
4327 13. The install script calls Exim with "-C /dev/null" in order to find the
4328 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4329 to be output. Howeve, since Exim outputs its version number before the
4330 error, it didn't break the script. It just looked ugly. I fixed this by
4331 always allowing "-C /dev/null" if the caller is root.
4333 14. Ignore overlarge ACL variable number when reading spool file - insurance
4334 against a later release with more variables having written the file.
4336 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4337 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4338 this, as well as the form without the "IPv6" on the front.
4340 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4341 OS/Makefile-Darwin file.
4343 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4344 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4346 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4347 variable. However, some OS use ints for inodes. I've added cast to long int
4348 to get rid of the compiler warning.
4350 19. I had forgotten to lock out "/../" in configuration file names when
4351 ALT_CONFIG_PREFIX was set.
4353 20. Routers used for verification do not need to specify transports. However,
4354 if such a router generated a host list, and callout was configured, Exim
4355 crashed, because it could not find a port number from the (non-existent)
4356 transport. It now assumes port 25 in this circumstance.
4358 21. Added the -t option to exigrep.
4360 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4361 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4362 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4363 them is set, LOOKUP_LSEARCH is forced.
4365 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4367 24. Added sender and host information to the "rejected by local_scan()" log
4368 line; previously there was no indication of these.
4370 25. Added .include_if_exists.
4372 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4373 Linux. It turns out that not all file systems support this. Apparently some
4374 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4375 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4376 not supported on the file descriptor, is now ignored when Exim is trying to
4377 sync a directory. This applies only to Linux.
4379 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4381 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4382 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4385 29. Added support for maildirsize files from supplied patch (modified a bit).
4387 30. The use of :fail: followed by an empty string could lead Exim to respond to
4388 sender verification failures with (e.g.):
4390 550 Verification failed for <xxx>
4391 550 Sender verify failed
4393 where the first response line was missing the '-' that indicates it is not
4394 the final line of the response.
4396 31. The loop for finding the name of the user that called Exim had a hardwired
4397 limit of 10; it now uses the value of finduser_retries, which is used for
4398 all other user lookups.
4400 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4403 33. Exim was neglecting to zero errno before one call of strtol() when
4404 expanding a string and expecting an integer value. On some systems this
4405 resulted in spurious "integer overflow" errors. Also, it was casting the
4406 result into an int without checking.
4408 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4409 did not work. The code looks as if it has *never* worked, though it appears
4410 to have been documented since at least releast 1.62. I have made it work.
4412 35. The "timeout_DNS" error in retry rules, also documented since at least
4413 1.62, also never worked. As it isn't clear exactly what this means, and
4414 clearly it isn't a major issue, I have abolished the feature by treating it
4415 as "timeout", and writing a warning to the main and panic logs.
4417 36. The display of retry rules for -brt wasn't always showing the error code
4420 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4421 timeout_connect_A, timeout_connect_MX.
4423 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4424 to the empty sender.
4426 39. The daemon was not analysing the content of -oX till after it had closed
4427 stderr and disconnected from the controlling terminal. This meant that any
4428 syntax errors were only noted on the panic log, and the return code from
4429 the command was 0. By re-arranging the code a little, I've made the
4430 decoding happen first, so such errors now appear on stderr, and the return
4431 code is 1. However, the actual setting up of the sockets still happens in
4432 the disconnected process, so errors there are still only recorded on the
4435 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4436 connections (as happens on some IP stacks) was logged at start up time as
4437 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4438 it from "IPv6 and IPv4", which means that two separate sockets are being
4441 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4442 says whether AF_INET or AF_INET6 was passed as an argument.
4444 42. Exiwhat output was messed up when time zones were included in log
4447 43. Exiwhat now gives more information about the daemon's listening ports,
4448 and whether -tls-on-connect was used.
4450 44. The "port" option of the smtp transport is now expanded.
4452 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4453 silently ignored. Now an error message is written to the main and panic
4456 46. There's a new ACL modifier called "logwrite" which writes to a log file
4457 as soon as it is encountered.
4459 47. Added $local_user_uid and $local_user_gid at routing time.
4461 48. Exim crashed when trying to verify a sender address that was being
4464 49. Exim was recognizing only a space character after ".include". It now also
4465 recognizes a tab character.
4467 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4468 extracting the relevant information from the specification. The man page no
4469 longer contains scrambled data for the -d option, and I've added a section
4470 at the front about calling Exim under different names.
4472 51. Added "extra_headers" argument to the "mail" command in filter files.
4474 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4477 53. Installed eximstats 1.29.
4479 54. Added transport_filter_timeout as a generic transport option.
4481 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4482 Cc: header lines. This was required by RFC 822, but it not required by RFC
4485 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4486 incoming messages that did not have them. Now it does so only if the
4487 message originates locally, that is, if there is no associated remote host
4488 address. When Resent- header lines are present, this applies to the Resent-
4489 lines rather than the non-Resent- lines.
4491 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4492 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4494 58. Messages for configuration errors now include the name of the main
4495 configuration file - useful now that there may be more than one file in a
4496 list (.included file names were always shown).
4498 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4499 for those rare installations that do not start the daemon as root or run it
4500 setuid root. I've cut out the call to initgroups() if the daemon is not
4503 60. The Exim user and group can now be bound into the binary as text strings
4504 that are looked up at the start of Exim's processing.
4506 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4508 62. Added $mailstore_basename variable.
4510 63. Installed patch to sieve.c from Michael Haardt.
4512 64. When Exim failed to open the panic log after failing to open the main log,
4513 the original message it was trying to log was written to stderr and debug
4514 output, but if they were not available (the usual case in production), it
4515 was lost. Now it is written to syslog before the two lines that record the
4516 failures to open the logs.
4518 65. Users' Exim filters run in subprocesses under the user's uid. It is
4519 possible for a "deliver" command or an alias in a "personal" command to
4520 provoke an address rewrite. If logging of address rewriting is configured,
4521 this fails because the process is not running as root or exim. There may be
4522 a better way of dealing with this, but for the moment (because 4.30 needs
4523 to be released), I have disabled address rewrite logging when running a
4524 filter in a non-root, non-exim process.
4530 1. The buildconfig auxiliary program wasn't quoting the value set for
4531 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4532 not defined. This bug was masked in 4.22 by the effect that was fixed in
4535 2. Some messages that were rejected after a message id was allocated were
4536 shown as "incomplete" by exigrep. It no longer does this for messages that
4537 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4539 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4540 have allow_domain_literals set, the ID did not get logged in the <= line.
4541 Domain literals are now always recognized in Message-ID: header lines.
4543 4. The first argument for a ${extract expansion item is the key name or field
4544 number. Leading and trailing spaces in this item were not being ignored,
4545 causing some misleading effects.
4547 5. When deliver_drop_privilege was set, single queue runner processes started
4548 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4549 same command in the process it spins off) were not dropping privilege.
4551 6. When the daemon running as "exim" started a queue runner, it always
4552 re-executed Exim in the spun-off process. This is a waste of effort when
4553 deliver_drop_privilege is set. The new process now just calls the
4554 queue-runner function directly.
4560 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4563 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4564 ignored. Though the use of -oP was forcing the writing of a pid file, it
4565 was always written to the default place.
4567 3. If the message "no IP address found for host xxxx" is generated during
4568 incoming verification, it is now followed by identification of the incoming
4569 connection (so you can more easily find what provoked it).
4571 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4573 5. Added some features to "harden" Exim a bit more against certain attacks:
4575 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4576 be put in Local/Makefile. This is like the never_users runtime option,
4577 but it cannot be overridden. The default setting is "root".
4579 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4580 prefix string with which any file named in a -C command line option
4583 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4584 is retained for -C and -D only if the caller of Exim is root. Without
4585 it, the exim user may also use -C and -D and retain privilege.
4587 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4588 command line option is disabled.
4590 6. Macro names set by the -D option must start with an upper case letter, just
4591 like macro names defined in the configuration file.
4593 7. Added "dereference=" facility to LDAP.
4595 8. Two instances of the typo "uknown" in the source files are fixed.
4597 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4598 the Configure-Makefile script screwed up while processing it.
4600 10. Incorporated PCRE 4.4.
4602 11. The SMTP synchronization check was not operating right at the start of an
4603 SMTP session. For example, it could not catch a HELO sent before the client
4604 waited for the greeting. There is now a check for outstanding input at the
4605 point when the greeting is written. Because of the duplex, asynchronous
4606 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4607 way, but not yet received, when the check is performed.
4609 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4610 on TCP/IP sockets, because this apparently causes some broken clients to
4613 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4614 unchanged) from the Cygwin maintainer.
4616 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4617 was supported instead of testing for mbx. Effectively a typo.
4619 15. The spa authenticator server code was not checking that the input it
4620 received was valid base64.
4622 16. The debug output line for the "set" modifier in ACLs was not showing the
4623 name of the variable that was being set.
4625 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4627 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4628 lookup on its own. Something else was needed to trigger the lookup. For
4629 example, a match in host_lookup or the need for a host name in a host list.
4630 Now, if $sender_host_name is referenced and the host name has not yet been
4631 looked up, a lookup is performed. If the lookup fails, the variable remains
4632 empty, and $host_lookup_failed is set to "1".
4634 19. Added "eqi" as a case-independent comparison operator.
4636 20. The saslauthd authentication condition could segfault if neither service
4637 nor realm was specified.
4639 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4640 error message that was logged was misleading, and incoming SMTP
4641 connections were dropped. The message is now more accurate, and temporary
4642 errors are given to SMTP connections.
4644 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4645 (except RSET) with a 421 error, until QUIT is received. However, it was
4646 failing to send a response to QUIT.
4648 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4649 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4650 ACL is now run after the helo_try_verify_hosts code.
4652 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4653 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4654 case-independent, so other case variants are also recognized). Apparently
4655 some systems use these upper case variants.
4657 25. If more than two messages were waiting for the same host, and a transport
4658 filter was specified for the transport, Exim sent two messages over the
4659 same TCP/IP connection, and then failed with "socket operation on non-
4660 socket" when it tried to send the third.
4662 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4664 27. The extern definition of crypt16() in expand.c was not being excluded when
4665 the OS had its own crypt16() function.
4667 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4668 as a preferred synonym for return_size_limit, both as an option and as an
4671 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4673 30. Changed the default configuration ACL to relax the local part checking rule
4674 for addresses that are not in any local domains. For these addresses,
4675 slashes and pipe symbols are allowed within local parts, but the sequence
4676 /../ is explicitly forbidden.
4678 31. SPA server authentication was not clearing the challenge buffer before
4681 32. log_message in a "warn" ACL statement was writing to the reject log as
4682 well as to the main log, which contradicts the documentation and doesn't
4683 seem right (because no rejection is happening). So I have stopped it.
4685 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4686 However, I am unable to do any testing of this.
4688 34. Fixed an infelicity in the appendfile transport. When checking directories
4689 for a mailbox, to see if any needed to be created, it was accidentally
4690 using path names with one or more superfluous leading slashes; tracing
4691 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4693 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4694 contents are added to the log line that is written for every discarded
4695 recipient. (Previously a log_message setting was ignored.)
4697 36. The ${quote: operator now quotes the string if it is empty.
4699 37. The install script runs exim in order to find its version number. If for
4700 some reason other than non-existence or emptiness, which it checks, it
4701 could not run './exim', it was installing it with an empty version number,
4702 i.e. as "exim-". This error state is now caught, and the installation is
4705 38. An argument was missing from the function that creates an error message
4706 when Exim fails to connect to the socket for saslauthd authentication.
4707 This could cause Exim to crash, or give a corrupted message.
4709 39. Added isip, isip4, and isip6 to ${if conditions.
4711 40. The ACL variables $acl_xx are now saved with the message, and can be
4712 accessed later in routers, transports, and filters.
4714 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4715 strings in the file are not string-expanded.
4717 42. If a MAIL command specified a SIZE value that was too large to fit into an
4718 int variable, the check against message_size_limit failed. Such values are
4719 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4720 day this will have to be increased, but I don't think I want to be around
4721 when emails are that large.
4728 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4729 iconv() is not standard in FreeBSD.
4731 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4732 IPv6 enabled. The observed symptom was a segmentation fault on return from
4733 the function os_common_find_running_interfaces() in src/os.c.
4735 3. In the check_special_case() function in daemon.c I had used "errno" as an
4736 argument name, which causes warnings on some systems. This was basically a
4737 typo, since it was named "eno" in the comments!
4739 4. The code that waits for the clock to tick (at a resolution of some fraction
4740 of a second) so as to ensure message-id uniqueness was always waiting for
4741 at least one whole tick, when it could have waited for less. [This is
4742 almost certainly not relevant at current processor speeds, where it is
4743 unlikely to ever wait at all. But we try to future-proof.]
4745 5. The function that sleeps for a time interval that includes fractions of a
4746 second contained a race. It did not block SIGALRM between setting the
4747 timer, and suspending (a couple of lines later). If the interval was short
4748 and the sigsuspend() was delayed until after it had expired, the suspension
4749 never ended. On busy systems this could lead to processes getting stuck for
4752 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4753 process, before it forks any delivery processes. The open lookup caching
4754 mechanism meant that the open file or database connection was passed into
4755 the delivery process. The problem was that delivery processes always tidy
4756 up cached lookup data. This could cause a problem for the next delivery
4757 process started by the queue runner, because the external queue runner
4758 process does not know about the closure. So the next delivery process
4759 still has data in the lookup cache. In the case of a file lookup, there was
4760 no problem because closing a file descriptor in a subprocess doesn't affect
4761 the parent. However, if the lookup was caching a connection to a database,
4762 the connection was closed, and the second delivery process was likely to
4763 see errors such as "PGSQL: query failed: server closed the connection
4764 unexpectedly". The problem has been fixed by closing all cached lookups
4765 in a queue runner before running a delivery process.
4767 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4768 seem to have the "const" qualifier which it has on other OS. I've
4771 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4772 *of the same type* (client or server) with the same public name that an
4773 error should be diagnosed.
4775 9. When Exim looked up a host name for an IP address, but failed to find the
4776 original IP address when looking up the host name (a safety check), it
4777 output the message "<ip address> does not match any IP for NULL", which was
4778 confusing, to say the least. The bug was that the host name should have
4779 appeared instead of "NULL".
4781 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4782 user that is built into the binary, and the -C or -D options is used, root
4783 privilege is dropped before the configuration file is read. In addition,
4784 logging is switched to stderr instead of the normal log files. If the
4785 configuration then re-defines the Exim user, the unprivileged environment
4786 is probably not what is expected, so Exim logs a panic warning message (but
4789 However, if deliver_drop_privilege is set, the unprivileged state may well
4790 be exactly what is intended, so the warning has been cut out in that case,
4791 and Exim is allowed to try to write to its normal log files.
4797 1. smtp_return_error_details was not giving details for temporary sender
4798 or receiver verification errors.
4800 2. Diagnose a configuration error if two authenticators have the same public
4803 3. Exim used not to create the message log file for a message until the first
4804 delivery attempt. This could be confusing when incoming messages were held
4805 for policy or load reasons. The message log file is now created at the time
4806 the message is received, and an initial "Received" line is written to it.
4808 4. The automatically generated man page for command line options had a minor
4809 bug that caused no ill effects; however, a more serious problem was that
4810 the procedure for building the man page automatically didn't always
4811 operate. Consequently, release 4.20 contains an out-of-date version. This
4812 shouldn't happen again.
4814 5. When building Exim with embedded Perl support, the script that builds the
4815 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4816 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4818 6. The freeze_tell option was not being used for messages that were frozen on
4819 arrival, either by an ACL or by local_scan().
4821 7. Added the smtp_incomplete_transaction log selector.
4823 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4824 was accepting AUTH without a new EHLO.
4826 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4827 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4830 10. Logging of TCP/IP connections (when configured) now happens in the main
4831 daemon process instead of the child process, so that the TCP/IP connection
4832 count is more accurate (but it can never be perfect).
4834 11. The use of "drop" in a nested ACL was not being handled correctly in the
4835 outer ACL. Now, if condition failure induced by the nested "drop" causes
4836 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4837 or "require"), the connection is dropped.
4839 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4840 that yield "discard" can now be used with an "accept" or a "discard" verb,
4841 but an error is generated for any others (because I can't see a useful way
4842 to define what should happen).
4844 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4845 are now processed in the same way as lines in the Exim configuration file.
4846 In particular, continuation lines are supported.
4848 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4850 15. Added -ti meaning -t -i.
4852 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4853 domains, and warn by logging if others are found.
4855 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4856 returned from GIFCONF when Exim is trying to find the list of interfaces on
4857 a host. The code in os.c has been modified to copy each ifreq to an aligned
4858 structure in all cases.
4860 Also, in some cases, the returned ifreq's were being copied to a 'struct
4861 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4862 means the last couple of bytes of an IPv6 address could be chopped if the
4863 ifreq contained only a normal sockaddr (14 bytes storage).
4865 18. Named domain lists were not supported in the hosts_treat_as_local option.
4866 An entry such as +xxxx was not recognized, and was treated as a literal
4869 19. Ensure that header lines added by a DATA ACL are included in the reject log
4870 if the ACL subsequently rejects the message.
4872 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4873 MD5 (which is deprecated).
4875 21. When testing a filter file using -bf, Exim was writing a message when it
4876 took the sender from a "From " line in the message, but it was not doing so
4877 when it took $return_path from a Return-Path: header line. It now does.
4879 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4880 with a valid header line field name (a series of printing characters
4881 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4883 23. Changed "disc" in the source to "disk" to conform to the documentation and
4884 the book and for uniformity.
4886 24. Ignore Sendmail's -Ooption=value command line item.
4888 25. When execve() failed while trying to run a command in a pipe transport,
4889 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4890 could be confused with a return value of 69 from the command itself. This
4891 has been changed to 127, the value the shell returns if it is asked to run
4892 a non-existent command. The wording for the related log line suggests a
4893 non-existent command as the problem.
4895 26. If received_header_text expands to an empty string, do not add a Received:
4896 header line to the message. (Well, it adds a token one on the spool, but
4897 marks it "old" so that it doesn't get used or transmitted.)
4899 27. Installed eximstats 1.28 (addition of -nt option).
4901 28. There was no check for failure on the call to getsockname() in the daemon
4902 code. This can fail if there is a shortage of resources on the system, with
4903 ENOMEM, for example. A temporary error is now given on failure.
4905 29. Contrary to the C standard, it seems that in some environments, the
4906 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4907 program. Exim now does this explicitly; it affects the formatting of
4908 timestamps using strftime().
4910 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4911 complaints. I've now initialized all the variables, to avoid this.
4913 32. Header lines added by a system filter were not being "seen" during
4914 transport-time rewrites.
4916 33. The info_callback() function passed to OpenSSL is set up with type void
4917 (*)(SSL *, int, int), as described somewhere. However, when calling the
4918 function (actually a macro) that sets it up, the type void(*)() is
4919 expected. I've put in a cast to prevent warnings from picky compilers.
4921 34. If a DNS black list lookup found a CNAME record, but there were no A
4922 records associated with the domain it pointed at, Exim crashed.
4924 35. If a DNS black list lookup returned more than one A record, Exim ignored
4925 all but the first. It now scans all returned addresses if a particular IP
4926 value is being sought. In this situation, the contents of the
4927 $dnslist_value variable are a list of all the addresses, separated by a
4930 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4931 to accept a host name and all its aliases if the forward lookup for any of
4932 them yielded the IP address of the incoming connection. Now it accepts only
4933 those names whose forward lookup yields the correct IP address. Any other
4934 names are discarded. This closes a loophole whereby a rogue DNS
4935 administrator could create reverse DNS records to break through a
4936 wildcarded host restriction in an ACL.
4938 37. If a user filter or a system filter that ran in a subprocess used any of
4939 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4940 the wrong values were passed to the pipe command ($thisaddress had the
4941 value of $0, $0 had the value of $1, etc). This bug was introduced by
4942 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4944 38. Improved the line breaking for long SMTP error messages from ACLs.
4945 Previously, if there was no break point between 40 and 75 characters, Exim
4946 left the rest of the message alone. Two changes have been made: (a) I've
4947 reduced the minimum length to 35 characters; (b) if it can't find a break
4948 point between 35 and 75 characters, it looks ahead and uses the first one
4949 that it finds. This may give the occasional overlong line, but at least the
4950 remaining text gets split now.
4952 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4953 file descriptors might be low, and that setting 1000 would always raise it.
4954 It turns out that in some environments, the limit is already over 1000 and
4955 that lowering it causes trouble. So now Exim takes care not to decrease it.
4957 40. When delivering a message, the value of $return_path is set to $sender_
4958 address at the start of routing (routers may change the value). By an
4959 oversight, this default was not being set up when an address was tested by
4960 -bt or -bv, which affected the outcome if any router or filter referred to
4963 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4964 log only when "message" or "log_message" are set. However, if one of the
4965 conditions was an address verification, or a call to a nested ACL, the
4966 messages generated by the underlying test were being passed through. This
4967 no longer happens. The underlying message is available in $acl_verify_
4968 message for both "message" and "log_message" expansions, so it can be
4969 passed through if needed.
4971 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4972 new expansion $bh_ to give the encoded byte string without charset
4973 translation. Translation happens only if iconv() is available; HAVE_ICONV
4974 indicates this at build time. HEADERS_CHARSET gives the charset to
4975 translate to; headers_charset can change it in the configuration, and
4976 "headers charset" can change it in an individual filter file.
4978 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4979 that creates RFC 2047 encoded "words" labels them as that charset instead
4980 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4981 expansion operator; (ii) when Exim creates a From: line for a local
4982 message; (iii) when a header line is rewritten to include a "phrase" part.
4984 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4985 buggy, causing it to skip the first lines of messages whose message ID
4986 ended in 'D'. This would not have bitten before Exim release 4.14, because
4987 message IDs were unlikely to end in 'D' before then. The effect was to have
4988 incorrect size information for certain domains.
4990 45. #include "config.h" was missing at the start of the crypt16.c module. This
4991 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4994 46. If there was a timeout during a "random" callout check, Exim treated it as
4995 a failure of the random address, and carried on sending RSET and the real
4996 address. If the delay was just some slowness somewhere, the response to the
4997 original RCPT would be taken as a response to RSET and so on, causing
4998 mayhem of various kinds.
5000 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5001 when I implemented it. It didn't allow for the fact that some option values
5002 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5003 the right test for positive values.
5005 48. Domain names in DNS records are case-independent. Exim always looks them up
5006 in lower case. Some resolvers return domain names in exactly the case they
5007 appear in the zone file, that is, they may contain uppercase letters. Not
5008 all resolvers do this - some return always lower case. Exim was treating a
5009 change of case by a resolver as a change of domain, similar to a widening
5010 of a domain abbreviation. This triggered its re-routing code and so it was
5011 trying to route what was effectively the same domain again. This normally
5012 caused routing to fail (because the router wouldn't handle the domain
5013 twice). Now Exim checks for this case specially, and just changes the
5014 casing of the domain that it ultimately uses when it transmits the message
5017 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5020 50. If a filter generated a file delivery with a non-absolute name (possible if
5021 no home directory exists for the router), the forbid_file option was not
5024 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5025 the existing equality matching.
5027 52. Exim was using ints instead of ino_t variables in some places where it was
5028 dealing with inode numbers.
5030 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5031 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5032 called TMPDIR, and if it finds it is different, it changes its value.
5034 54. The smtp_printf() function is now made available to local_scan() so
5035 additional output lines can be written before returning. There is also an
5036 smtp_fflush() function to enable the detection of a dropped connection.
5037 The variables smtp_input and smtp_batched_input are exported to
5040 55. Changed the default runtime configuration: the message "Unknown user"
5041 has been removed from the ACL, and instead placed on the localuser router,
5042 using the cannot_route_message feature. This means that any verification
5043 failures that generate their own messages won't get overridden. Similarly,
5044 the "Unrouteable address" message that was in the ACL for unverifiable
5045 relay addresses has also been removed.
5047 56. Added hosts_avoid_esmtp to the smtp transport.
5049 57. The exicyclog script was not checking for the esoteric option
5050 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5051 will work only if exicyclog is run under the appropriate euid.
5053 58. Following a discussion on the list, the rules by which Exim recognises line
5054 endings on incoming messages have been changed. The -dropcr and drop_cr
5055 options are now no-ops, retained only for backwards compatibility. The
5056 following line terminators are recognized: LF CRLF CR. However, special
5057 processing applies to CR:
5059 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5060 nor a local message in the state where . is a terminator.
5062 (ii) If a bare CR is encountered in a header line, an extra space is added
5063 after the line terminator so as not to end the header. The reasoning
5064 behind this is that bare CRs in header lines are most likely either
5065 to be mistakes, or people trying to play silly games.
5067 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5068 was being incorrectly given as 18 bytes larger than it should have been.
5069 This is a VOB (very old bug).
5071 60. This may never have affected anything current, but just in case it has:
5072 When the local host is found other than at the start of a list of hosts,
5073 the local host, those with the same MX, and any that follow, are discarded.
5074 When the list in question was part of a longer list of hosts, the following
5075 hosts (not currently being processed) were also being discarded. This no
5076 longer happens. I'm not sure if this situation could ever has previously
5079 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5082 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5083 header. This is something that is recommended in a new Internet Draft, and
5084 is something that is documented as being done by Sendmail. There are two
5085 possible values. For messages generated by the autoreply transport, Exim
5088 Auto-Submitted: auto-replied
5090 whereas for all other generated messages (e.g. bounces) it adds
5092 Auto-Submitted: auto-generated
5094 63. The "personal" condition in filters now includes a test for the
5095 Auto-Submitted: header. If it contains the string "auto-" the message it
5096 not considered personal.
5098 64. Added rcpt_include_affixes as a generic transport option.
5100 65. Added queue_only_override (default true).
5102 66. Added the syslog_duplication option.
5104 67. If what should have been the first header line of a message consisted of
5105 a space followed by a colon, Exim was mis-interpreting it as a header line.
5106 It isn't of course - it is syntactically invalid and should therefore be
5107 treated as the start of the message body. The misbehaviour could have
5108 caused a number of strange effects, including loss of data in subsequent
5109 header lines, and spool format errors.
5111 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5112 client host had authenticated. This control can now be exercised by an ACL
5113 for more flexibility.
5115 69. By default, callouts do not happen when testing with -bh. There is now a
5116 variant, -bhc, which does actually run the callout code, including
5117 consulting and updating the callout cache.
5119 70. Added support for saslauthd authentication, courtesy of Alexander
5122 71. If statvfs() failed on the spool or log directories while checking their
5123 size for availability, Exim confusingly gave the error "space shortage".
5124 Furthermore, in debugging mode it crashed with a floating point exception.
5125 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5126 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5127 serious problem, Exim now writes to the main and panic logs when this
5128 happens, with details of the failure. It then refuses to accept the
5129 incoming message, giving the message "spool directory problem" or "log
5130 directory problem" with a 421 code for SMTP messages.
5132 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5133 0, 1, and 2 exist, because some OS complain for execs without them (see
5134 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5135 descriptors. However, the code omitted to check that the open succeeded,
5136 causing mysterious errors if for some reason the permissions on /dev/null
5137 got screwed. Now Exim writes a message to the main and panic logs, and
5138 bombs out if it can't open /dev/null.
5140 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5141 interact so that it is all more flexible. It is supposed to remain
5142 backwards compatible. Also added extra_local_interfaces.
5144 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5145 to bomb out with an assertion failure - to the client this appears as a
5146 connection drop. This problem occurs in the part of the code that was taken
5147 from the Samba project. Fortunately, the assertion is in a very simple
5148 function, so I have fixed this by reproducing the function inline in the
5149 one place where it is called, and arranging for authentication to fail
5150 instead of killing the process with assert().
5152 75. The SPA client code was not working when the server requested OEM rather
5153 than Unicode encoding.
5155 76. Added code to make require_files with a specific uid setting more usable in
5156 the case where statting the file as root fails - usually a non-root-mounted
5157 NFS file system. When this happens and the failure is EACCES, Exim now
5158 forks a subprocess and does the per-uid checking as the relevant uid.
5160 77. Added process_log_path.
5162 78. If log_file_path was not explicitly set, a setting of check_log_space or
5163 check_log_inodes was ignored.
5165 79. If a space check for the spool or log partitions fails, the incident is now
5166 logged. Of course, in the latter case the data may get lost...
5168 80. Added the %p formatting code to string_format() so that it can be used to
5169 print addresses in debug_print(). Adjusted all the address printing in the
5170 debugging in store.c to use %p rather than %d.
5172 81. There was a concern that a line of code in smtp_in.c could overflow a
5173 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5174 initially expressed, the concern was not well-founded, because trailing
5175 spaces are removed early. However, if the trailing spaces were followed by
5176 a NULL, they did not get removed, so the overflow was possible. Two fixes
5179 (a) I re-wrote the offending code in a cleaner fashion.
5180 (b) If an incoming SMTP command contains a NULL character, it is rejected
5183 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5184 runs initgroups(), so that if the Exim user is in any additional groups,
5185 they will be used during message reception.
5191 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob