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
109 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
110 $data_path, and $header_path variables; fixed documentation bugs and
113 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
114 exipick to access non-standard spools, including the "frozen" queue
117 NM/01 Bugzilla 965: Support mysql stored procedures.
118 Patch from Alain Williams
120 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
122 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
123 Patch from Andreas Metzler
125 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
126 Patch from Kirill Miazine
128 NM/05 Bugzilla 671: Added umask to procmail example.
130 JJ/03 installed exipick 20100323.0, fixing doc bug
132 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
133 directory. Notification and patch from Dan Rosenberg.
135 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
137 TK/02 Improve log output when DKIM signing operation fails.
139 MH/01 Treat the transport option dkim_domain as a colon separated
140 list, not as a single string, and sign the message with each element,
141 omitting multiple occurences of the same signer.
143 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
144 Bugzilla 985, 986. Patch by Simon Arlott
146 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
147 Patch by Simon Arlott
149 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
150 MBX locking. Notification from Dan Rosenberg.
156 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
158 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
160 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
162 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
164 NM/04 Bugzilla 915: Flag broken perl installation during build.
170 TK/01 Added patch by Johannes Berg that expands the main option
171 "spamd_address" if it starts with a dollar sign.
173 TK/02 Write list of recipients to X-Envelope-Sender header when building
174 the mbox-format spool file for content scanning (suggested by Jakob
177 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
178 (http://www.dcc-servers.net/) support via dccifd. Activated by
179 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
181 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
182 by Mark Daniel Reidel <mr@df.eu>.
184 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
185 When building exim an external PCRE library is now needed -
186 PCRE is a system library on the majority of modern systems.
187 See entry on PCRE_LIBS in EDITME file.
189 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
190 conversation. Added nologin parameter to request.
191 Patch contributed by Kirill Miazine.
193 TF/01 Do not log submission mode rewrites if they do not change the address.
195 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
197 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
198 log files in place. Contributed by Roberto Lima.
200 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
202 TF/03 Bugzilla 615: When checking the local_parts router precondition
203 after a local_part_suffix or local_part_prefix option, Exim now
204 does not use the address's named list lookup cache, since this
205 contains cached lookups for the whole local part.
207 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
208 Robert Millan. Documentation is in experimental-spec.txt.
210 TF/04 Bugzilla 668: Fix parallel build (make -j).
212 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
214 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
215 Patch provided by Jan Srzednicki.
217 TF/05 Leading white space used to be stripped from $spam_report which
218 wrecked the formatting. Now it is preserved.
220 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
221 that they are available at delivery time.
223 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
225 TF/08 TLS error reporting now respects the incoming_interface and
226 incoming_port log selectors.
228 TF/09 Produce a more useful error message if an SMTP transport's hosts
229 setting expands to an empty string.
231 NM/06 Bugzilla 744: EXPN did not work under TLS.
232 Patch provided by Phil Pennock.
234 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
235 Patch provided by Richard Godbee.
237 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
238 acl_smtp_notquit, added index entry.
240 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
241 Patch provided by Eugene Bujak.
243 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
244 accept(). Patch provided by Maxim Dounin.
246 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
247 Patch provided by Phil Pennock.
249 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
251 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
252 Patch provided by Brad "anomie" Jorsch.
254 NM/14 Bugzilla 622: Added timeout setting to transport filter.
255 Patch provided by Dean Brooks.
257 TK/05 Add native DKIM support (does not depend on external libraries).
259 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
260 Patch provided by Graeme Fowler.
262 NM/16 Bugzilla 851: Documentation example syntax fix.
264 NM/17 Changed NOTICE file to remove references to embedded PCRE.
266 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
269 NM/19 Bugzilla 745: TLS version reporting.
270 Patch provided by Phil Pennock.
272 NM/20 Bugzilla 167: bool: condition support.
273 Patch provided by Phil Pennock.
275 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
276 clients. Patch provided by Phil Pennock.
278 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
279 Patch provided by Brad "anomie" Jorsch.
281 NM/23 Bugzilla 687: Fix misparses in eximstats.
282 Patch provided by Heiko Schlittermann.
284 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
285 Patch provided by Heiko Schlittermann.
287 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
288 plus update to original patch.
290 NM/26 Bugzilla 799: Documentation correction for ratelimit.
292 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
293 Patch provided by David Brownlee.
295 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
297 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
299 NM/30 Bugzilla 888: TLS documentation bugfixes.
301 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
303 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
304 Unlike the original bugzilla I have changed all shell scripts in src tree.
306 NM/33 Bugzilla 898: Transport filter timeout fix.
307 Patch by Todd Rinaldo.
309 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
310 Patch by Serge Demonchaux.
312 NM/35 Bugzilla 39: Base64 decode bug fixes.
313 Patch by Jakob Hirsch.
315 NM/36 Bugzilla 909: Correct connect() call in dcc code.
317 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
319 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
321 NM/39 Bugzilla 911: Fixed MakeLinks build script.
327 TK/01 Add preliminary DKIM support. Currently requires a forked version of
328 ALT-N's libdkim that I have put here:
329 http://duncanthrax.net/exim-experimental/
331 Note to Michael Haardt: I had to rename some vars in sieve.c. They
332 were called 'true' and it seems that C99 defines that as a reserved
333 keyword to be used with 'bool' variable types. That means you could
334 not include C99-style headers which use bools without triggering
335 build errors in sieve.c.
337 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
338 as mailq or other aliases. Changed the --help handling significantly
339 to do whats expected. exim_usage() emits usage/help information.
341 SC/01 Added the -bylocaldomain option to eximstats.
343 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
345 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
347 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
353 PH/01 Another patch from the Sieve maintainer.
355 PH/02 When an IPv6 address is converted to a string for single-key lookup
356 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
357 dots are used instead of colons so that keys in lsearch files need not
358 contain colons. This was done some time before quoting was made available
359 in lsearch files. However, iplsearch files do require colons in IPv6 keys
360 (notated using the quote facility) so as to distinguish them from IPv4
361 keys. This meant that lookups for IP addresses in host lists did not work
362 for iplsearch lookups.
364 This has been fixed by arranging for IPv6 addresses to be expressed with
365 colons if the lookup type is iplsearch. This is not incompatible, because
366 previously such lookups could never work.
368 The situation is now rather anomolous, since one *can* have colons in
369 ordinary lsearch keys. However, making the change in all cases is
370 incompatible and would probably break a number of configurations.
372 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
375 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
376 conversion specification without a maximum field width, thereby enabling
377 a rogue spamd server to cause a buffer overflow. While nobody in their
378 right mind would setup Exim to query an untrusted spamd server, an
379 attacker that gains access to a server running spamd could potentially
380 exploit this vulnerability to run arbitrary code as the Exim user.
382 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
383 $primary_hostname instead of what libspf2 thinks the hosts name is.
385 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
386 a directory entry by the name of the lookup key. Previously, if a
387 symlink pointed to a non-existing file or a file in a directory that
388 Exim lacked permissions to read, a lookup for a key matching that
389 symlink would fail. Now it is enough that a matching directory entry
390 exists, symlink or not. (Bugzilla 503.)
392 PH/03 The body_linecount and body_zerocount variables are now exported in the
395 PH/04 Added the $dnslist_matched variable.
397 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
398 This means they are set thereafter only if the connection becomes
401 PH/06 Added the client_condition to authenticators so that some can be skipped
402 by clients under certain conditions.
404 PH/07 The error message for a badly-placed control=no_multiline_responses left
405 "_responses" off the end of the name.
407 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
409 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
410 (without spaces) instead of just copying the configuration text.
412 PH/10 Added the /noupdate option to the ratelimit ACL condition.
414 PH/11 Added $max_received_linelength.
416 PH/12 Added +ignore_defer and +include_defer to host lists.
418 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
419 way in which PCRE > 7.0 is built.
421 PH/14 Implemented queue_only_load_latch.
423 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
424 MAIL command. The effect was to mangle the value on 64-bit systems.
426 PH/16 Another patch from the Sieve maintainer.
428 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
430 PH/18 If a system quota error occurred while trying to create the file for
431 a maildir delivery, the message "Mailbox is full" was not appended to the
432 bounce if the delivery eventually timed out. Change 4.67/27 below applied
433 only to a quota excession during the actual writing of the file.
435 PH/19 It seems that peer DN values may contain newlines (and other non-printing
436 characters?) which causes problems in log lines. The DN values are now
437 passed through string_printing() before being added to log lines.
439 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
440 and InterBase are left for another time.)
442 PH/21 Added message_body_newlines option.
444 PH/22 Guard against possible overflow in moan_check_errorcopy().
446 PH/23 POSIX allows open() to be a macro; guard against that.
448 PH/24 If the recipient of an error message contained an @ in the local part
449 (suitably quoted, of course), incorrect values were put in $domain and
450 $local_part during the evaluation of errors_copy.
456 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
457 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
460 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
461 issue a MAIL command.
463 PH/02 In an ACL statement such as
465 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
467 if a client was not listed at all, or was listed with a value other than
468 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
469 the condition was not true (as it should be), so access was not denied.
470 The bug was that the ! inversion was incorrectly passed on to the second
471 item. This has been fixed.
473 PH/03 Added additional dnslists conditions == and =& which are different from
474 = and & when the dns lookup returns more than one IP address.
476 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
477 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
479 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
480 FSYNC, which compiles an option called disable_fsync that allows for
481 bypassing fsync(). The documentation is heavily laced with warnings.
483 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
485 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
486 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
487 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
488 including adding "make clean"; (3) Added -fPIC when compiling the test
489 dynamically loaded module, to get rid of a warning.
491 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
492 message fails, move_frozen_messages = true and ignore_bounce_errors_after
493 = 0s. The bug is otherwise harmless.
495 PH/07 There was a bug in the dovecot authenticator such that the value of
496 $auth1 could be overwritten, and so not correctly preserved, after a
497 successful authentication. This usually meant that the value preserved by
498 the server_setid option was incorrect.
500 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
502 PH/09 Installed PCRE release 7.0.
504 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
505 run for batched SMTP input. It is now run at the start of every message
506 in the batch. While fixing this I discovered that the process information
507 (output by running exiwhat) was not always getting set for -bs and -bS
508 input. This is fixed, and it now also says "batched" for BSMTP.
510 PH/11 Added control=no_pipelining.
512 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
513 patch, slightly modified), and move the expansion of helo_data till after
514 the connection is made in the smtp transport (so it can use these
517 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
519 PH/14 Added log_selector = +pid.
521 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
523 PH/16 Add ${if forany and ${if forall.
525 PH/17 Added dsn_from option to vary the From: line in DSNs.
527 PH/18 Flush SMTP output before performing a callout, unless control =
528 no_callout_flush is set.
530 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
531 was true (the default) a successful delivery failed to delete the retry
532 item, thus causing premature timeout of the address. The bug is now
535 PH/20 Added hosts_avoid_pipelining to the smtp transport.
537 PH/21 Long custom messages for fakedefer and fakereject are now split up
538 into multiline reponses in the same way that messages for "deny" and
539 other ACL rejections are.
541 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
542 with slight modification.
544 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
545 draft, changing the syntax and factoring some duplicate code".
547 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
548 for deliveries of the second and subsequent messages over the same SMTP
551 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
552 ${reduce, with only minor "tidies".
554 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
556 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
557 expansion side effects.
559 PH/27 When a message times out after an over-quota error from an Exim-imposed
560 quota, the bounce message says "mailbox is full". This message was not
561 being given when it was a system quota that was exceeded. It now should
564 MH/03 Made $recipients available in local_scan(). local_scan() already has
565 better access to the recipient list through recipients_list[], but
566 $recipients can be useful in postmaster-provided expansion strings.
568 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
569 in the case of a MAIL command with additional options following the
570 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
571 were accidentally chopped off.
573 PH/29 SMTP synchronization checks are implemented when a command is read -
574 there is a check that no more input is waiting when there shouldn't be
575 any. However, for some commands, a delay in an ACL can mean that it is
576 some time before the response is written. In this time, more input might
577 arrive, invalidly. So now there are extra checks after an ACL has run for
578 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
579 pipelining has not been advertised.
581 PH/30 MH's patch to allow iscntrl() characters to be list separators.
583 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
584 returned in the SMTP response when smtp_return_error_details was false.
587 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
588 instead of the C I/O that was originally supplied, because problems were
591 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
592 Exim which did not show up earlier: it was assuming that a call to
593 SSL_CTX_set_info_callback() might give an error value. In fact, there is
594 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
595 was a macro that became an assignment, so it seemed to work. This has
596 changed to a proper function call with a void return, hence the compile
597 error. Exim's code has been fixed.
599 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
602 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
604 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
606 PH/37 If a message is not accepted after it has had an id assigned (e.g.
607 because it turns out to be too big or there is a timeout) there is no
608 "Completed" line in the log. When some messages of this type were
609 selected by exigrep, they were listed as "not completed". Others were
610 picked up by some special patterns. I have improved the selection
611 criteria to be more general.
613 PH/38 The host_find_failed option in the manualroute router can now be set
614 to "ignore", to completely ignore a host whose IP address cannot be
615 found. If all hosts are ignored, the behaviour is controlled by the new
616 host_all_ignored option.
618 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
619 homing or because of multiple MX records with /mx) generated more than
620 one IP address, and the following item turned out to be the local host,
621 all the secondary addresses of the first item were incorrectly removed
622 from the list, along with the local host and any following hosts (which
623 is what is supposed to happen).
625 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
626 whoever called Exim into the -H file. In the case of the daemon it was
627 behaving confusingly. When first started, it used values for whoever
628 started the daemon, but after a SIGHUP it used the Exim user (because it
629 calls itself on a restart). I have changed the code so that it now always
632 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
633 message are rejected with the same error (e.g. no authentication or bad
634 sender address), and a DATA command is nevertheless sent (as can happen
635 with PIPELINING or a stupid MUA), the error message that was given to the
636 RCPT commands is included in the rejection of the DATA command. This is
637 intended to be helpful for MUAs that show only the final error to their
640 PH/42 Another patch from the Sieve maintainer.
642 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
643 Eximstats - Fixed some broken HTML links and added missing column headers
645 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
646 columns for Rejects, Temp Rejects, Ham, and Spam rows.
648 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
650 PH/43 Yet another patch from the Sieve maintainer.
652 PH/44 I found a way to check for a TCP/IP connection going away before sending
653 the response to the final '.' that terminates a message, but only in the
654 case where the client has not sent further data following the '.'
655 (unfortunately, this is allowed). However, in many cases there won't be
656 any further data because there won't be any more messages to send. A call
657 to select() can be used: if it shows that the input is "ready", there is
658 either input waiting, or the socket has been closed. An attempt to read
659 the next input character can distinguish the two cases. Previously, Exim
660 would have sent an OK response which the client would never have see.
661 This could lead to message repetition. This fix should cure that, at
662 least in a lot of common cases.
664 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
665 advertised in response to EHLO.
671 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
672 fixed by 4.65/MH/01 (is this a record?) are fixed:
674 (i) An empty string was always treated as zero by the numeric comparison
675 operators. This behaviour has been restored.
677 (ii) It is documented that the numeric comparison operators always treat
678 their arguments as decimal numbers. This was broken in that numbers
679 starting with 0 were being interpreted as octal.
681 While fixing these problems I realized that there was another issue that
682 hadn't been noticed. Values of message_size_limit (both the global option
683 and the transport option) were treated as octal if they started with 0.
684 The documentation was vague. These values are now always treated as
685 decimal, and I will make that clear in the documentation.
691 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
692 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
695 MH/01 Don't check that the operands of numeric comparison operators are
696 integers when their expansion is in "skipping" mode (fixes bug
697 introduced by 4.64-PH/07).
699 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
700 child addresses, Exim now panics and dies. Previously, because the count
701 is held in a short int, deliveries were likely to be lost. As such a
702 large number of recipients for a single message is ridiculous
703 (performance will be very, very poor), I have chosen to impose a limit
704 rather than extend the field.
710 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
711 leftover -K file (the existence of which was triggered by #402).
712 While we were at it, introduced process PID as part of the -K
713 filename. This should rule out race conditions when creating
716 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
717 processing considerably. Previous code took too long for large mails,
718 triggering a timeout which in turn triggers #401.
720 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
721 in the DK code in transports.c. sendfile() is not really portable,
722 hence the _LINUX specificness.
724 TF/01 In the add_headers option to the mail command in an Exim filter,
725 there was a bug that Exim would claim a syntax error in any
726 header after the first one which had an odd number of characters
729 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
730 callout verification, Exim cached a "reject" for the entire domain. This
731 is correct for most verifications, but it is not correct for a recipient
732 verification with use_sender or use_postmaster set, because in that case
733 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
734 case of MAIL FROM:<> rejection from other early rejections (e.g.
735 rejection of HELO). When verifying a recipient using a non-null MAIL
736 address, the cache is ignored if it shows MAIL FROM:<> rejection.
737 Whatever the result of the callout, the value of the domain cache is
738 left unchanged (for any other kind of callout, getting as far as trying
739 RCPT means that the domain itself is ok).
741 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
744 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
745 manifest itself as EPIPE rather than ECONNECT. When tidying away a
746 session, the daemon ignores ECONNECT errors and logs others; it now
747 ignores EPIPE as well.
749 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
750 (quoted-printable decoding).
752 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
753 later the small subsequent patch to fix an introduced bug.
755 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
757 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
759 PH/08 An error is now given if message_size_limit is specified negative.
761 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
762 to be given (somewhat) arbitrary names.
764 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
767 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
768 miscellaneous code fixes
770 PH/10 Added the log_reject_target ACL modifier to specify where to log
773 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
774 hostname. This is wrong, because it relates to the incoming message (and
775 probably the interface on which it is arriving) and not to the outgoing
776 callout (which could be using a different interface). This has been
777 changed to use the value of the helo_data option from the smtp transport
778 instead - this is what is used when a message is actually being sent. If
779 there is no remote transport (possible with a router that sets up host
780 addresses), $smtp_active_hostname is used.
782 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
783 tweaks were necessary in order to get it to work (see also 21 below):
784 (a) The code assumed that strncpy() returns a negative number on buffer
785 overflow, which isn't the case. Replaced with Exim's string_format()
787 (b) There were several signed/unsigned issues. I just did the minimum
788 hacking in of casts. There is scope for a larger refactoring.
789 (c) The code used strcasecmp() which is not a standard C function.
790 Replaced with Exim's strcmpic() function.
791 (d) The code set only $1; it now sets $auth1 as well.
792 (e) A simple test gave the error "authentication client didn't specify
793 service in request". It would seem that Dovecot has changed its
794 interface. Fortunately there's a specification; I followed it and
795 changed what the client sends and it appears to be working now.
797 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
800 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
801 address A is aliased to B and C, where B exists and C does not. Without
802 -v the output is "A verified" because verification stops after a
803 successful redirection if more than one address is generated. However,
804 with -v the child addresses are also verified. Exim was outputting "A
805 failed to verify" and then showing the successful verification for C,
806 with its parentage. It now outputs "B failed to verify", showing B's
807 parentage before showing the successful verification of C.
809 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
810 look up a TXT record in a specific list after matching in a combined
813 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
814 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
815 they consult the DNS. I had assumed they would set it the way they
816 wanted; and indeed my experiments on Linux seem to show that in some
817 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
818 To be on the safe side, however, I have now made the interface to
819 host_find_byname() similar to host_find_bydns(), with an argument
820 containing the DNS resolver options. The host_find_byname() function now
821 sets these options at its start, just as host_find_bydns() does. The smtp
822 transport options dns_qualify_single and dns_search_parents are passed to
823 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
824 of host_find_byname() use the default settings of RES_DEFNAMES
825 (qualify_single) but not RES_DNSRCH (search_parents).
827 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
828 spool_read_header() do less string testing, by means of a preliminary
829 switch on the second character of optional "-foo" lines. (This is
830 overdue, caused by the large number of possibilities that now exist.
831 Originally there were few.) While I was there, I also converted the
832 str(n)cmp tests so they don't re-test the leading "-" and the first
833 character, in the hope this might squeeze out yet more improvement.
835 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
836 flag allowing group syntax was set by the header_syntax check but not
837 turned off, possible causing trouble later; (2) The flag was not being
838 set at all for the header_verify test, causing "group"-style headers to
839 be rejected. I have now set it in this case, and also caused header_
840 verify to ignore an empty address taken from a group. While doing this, I
841 came across some other cases where the code for allowing group syntax
842 while scanning a header line wasn't quite right (mostly, not resetting
843 the flag correctly in the right place). These bugs could have caused
844 trouble for malformed header lines. I hope it is now all correct.
846 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
847 with the "reply" argument non-NULL. The code, however (which originally
848 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
849 but it didn't always do it. This confused somebody who was copying the
850 code for some other use. I have removed all the tests.
852 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
853 feature that was used to support insecure browsers during the U.S. crypto
854 embargo. It requires special client support, and Exim is probably the
855 only MTA that supported it -- and would never use it because real RSA is
856 always available. This code has been removed, because it had the bad
857 effect of slowing Exim down by computing (never used) parameters for the
858 RSA_EXPORT functionality.
860 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
861 authenticator to fail if there's a tab character in the incoming data
862 (there should never be unless someone is messing about, as it's supposed
863 to be base64-encoded). Also added, on Timo's advice, the "secured" option
864 if the connection is using TLS or if the remote IP is the same as the
865 local IP, and the "valid-client-cert option" if a client certificate has
868 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
869 authenticators. This can be used for authorization after authentication
870 succeeds. (In the case of plaintext, it servers for both authentication
873 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
874 if any retry times were supplied.
876 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
877 connection, where there is no client IP address to check. In this
878 situation, the verify now always succeeds.
880 PH/25 Applied John Jetmore's -Mset patch.
882 PH/26 Added -bem to be like -Mset, but loading a message from a file.
884 PH/27 In a string expansion for a processed (not raw) header when multiple
885 headers of the same name were present, leading whitespace was being
886 removed from all of them, but trailing whitespace was being removed only
887 from the last one. Now trailing whitespace is removed from each header
888 before concatenation. Completely empty headers in a concatenation (as
891 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
892 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
894 PH/29 [Removed. This was a change that I later backed out, and forgot to
895 correct the ChangeLog entry (that I had efficiently created) before
896 committing the later change.]
898 PH/30 Exim was sometimes attempting to deliver messages that had suffered
899 address errors (4xx response to RCPT) over the same connection as other
900 messages routed to the same hosts. Such deliveries are always "forced",
901 so retry times are not inspected. This resulted in far too many retries
902 for the affected addresses. The effect occurred only when there were more
903 hosts than the hosts_max_try setting in the smtp transport when it had
904 the 4xx errors. Those hosts that it had tried were not added to the list
905 of hosts for which the message was waiting, so if all were tried, there
906 was no problem. Two fixes have been applied:
908 (i) If there are any address or message errors in an SMTP delivery, none
909 of the hosts (tried or untried) are now added to the list of hosts
910 for which the message is waiting, so the message should not be a
911 candidate for sending over the same connection that was used for a
912 successful delivery of some other message. This seems entirely
913 reasonable: after all the message is NOT "waiting for some host".
914 This is so "obvious" that I'm not sure why it wasn't done
915 previously. Hope I haven't missed anything, but it can't do any
916 harm, as the worst effect is to miss an optimization.
918 (ii) If, despite (i), such a delivery is accidentally attempted, the
919 routing retry time is respected, so at least it doesn't keep
920 hammering the server.
922 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
923 in ${readsocket because some servers need this prod.
925 PH/32 Added some extra debug output when updating a wait-xxx database.
927 PH/33 The hint "could be header name not terminated by colon", which has been
928 given for certain expansion errors for a long time, was not being given
929 for the ${if def:h_colon_omitted{... case.
931 PH/34 The spec says: "With one important exception, whenever a domain list is
932 being scanned, $domain contains the subject domain." There was at least
933 one case where this was not true.
935 PH/35 The error "getsockname() failed: connection reset by peer" was being
936 written to the panic log as well as the main log, but it isn't really
937 panic-worthy as it just means the connection died rather early on. I have
938 removed the panic log writing for the ECONNRESET error when getsockname()
941 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
942 runs only) independently of the message's sender address. This meant
943 that, if the 4xx error was in fact related to the sender, a different
944 message to the same recipient with a different sender could confuse
945 things. In particualar, this can happen when sending to a greylisting
946 server, but other circumstances could also provoke similar problems.
947 I have changed the default so that the retry time for these errors is now
948 based a combination of the sender and recipient addresses. This change
949 can be overridden by setting address_retry_include_sender=false in the
952 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
953 remote server are returned as part of bounce messages. This was not
954 happening for LMTP over a pipe (the lmtp transport), but now it is the
955 same for both kinds of LMTP.
957 PH/38 Despite being documented as not happening, Exim was rewriting addresses
958 in header lines that were in fact CNAMEs. This is no longer the case.
960 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
961 and queue runs started by the daemon processed all messages. This has
962 been fixed so that -R and -S can now usefully be given with -q<time>.
964 PH/40 Import PCRE release 6.7 (fixes some bugs).
966 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
968 PH/42 Give an error if -q is specified more than once.
970 PH/43 Renamed the variables $interface_address and $interface_port as
971 $received_ip_address and $received_port, to make it clear that these
972 values apply to message reception, and not to the outgoing interface when
973 a message is delivered. (The old names remain recognized, of course.)
975 PH/44 There was no timeout on the connect() call when using a Unix domain
976 socket in the ${readsocket expansion. There now is.
978 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
979 be meaningful with "accept".
981 SC/01 Eximstats V1.43
982 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
984 SC/02 Eximstats V1.44
985 Use a glob alias rather than an array ref in the generated
986 parser. This improves both readability and performance.
988 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
989 Collect SpamAssassin and rejection statistics.
990 Don't display local sender or destination tables unless
991 there is data to show.
992 Added average volumes into the top table text output.
994 SC/04 Eximstats V1.46
995 Collect data on the number of addresses (recipients)
996 as well as the number of messages.
998 SC/05 Eximstats V1.47
999 Added 'Message too big' to the list of mail rejection
1000 reasons (thanks to Marco Gaiarin).
1002 SC/06 Eximstats V1.48
1003 Mainlog lines which have GMT offsets and are too short to
1004 have a flag are now skipped.
1006 SC/07 Eximstats V1.49 (Alain Williams)
1007 Added the -emptyok flag.
1009 SC/08 Eximstats V1.50
1010 Fixes for obtaining the IP address from reject messages.
1012 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1013 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1014 whitesspace changes from 4.64-PH/27
1016 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1019 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1020 are found, allow negative numbers in numeric criteria)
1022 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1024 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1025 to match changes made in 4.64-PH/43
1027 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1029 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1030 "domain" is actually an IP address, and force a failure. This locks out
1031 those revolvers/nameservers that support "A-for-A" lookups, in
1032 contravention of the specifications.
1034 PH/48 When a host name was looked up from an IP address, and the subsequent
1035 forward lookup of the name timed out, the host name was left in
1036 $sender_host_name, contrary to the specification.
1038 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1039 restricted to single-key lookups, Exim was not diagnosing an error if
1040 * or *@ was used with a query-style lookup.
1042 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1044 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1045 long ago, but noone interested enough thought of it. Let's just say that
1046 the "1.1" means that there are some new functions that weren't there at
1047 some point in the past.
1049 PH/51 Error processing for expansion failure of helo_data from an smtp
1050 transport during callout processing was broken.
1052 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1053 tested/used via the -bh/-bhc/-bs options.
1055 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1056 bug, fixed in subsequent PCRE releases).
1058 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1059 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1061 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1067 SC/01 Use a glob alias rather than an array ref in eximstats generated
1068 parser. This improves both readability and performance.
1070 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1071 Don't display local sender or destination tables in eximstats unless
1072 there is data to show.
1073 Added average volumes into the eximstats top table text output.
1075 SC/03 Collect data on the number of addresses (recipients) as well
1076 as the number of messages in eximstats.
1078 TF/01 Correct an error in the documentation for the redirect router. Exim
1079 does (usually) call initgroups() when daemonizing.
1081 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1082 with consistent privilege compared to when running as a daemon.
1084 TF/03 Note in the spec that $authenticated_id is not set for local
1085 submissions from trusted users.
1087 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1088 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1090 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1091 by adding some example configuration directives to the default
1092 configuration file. A little bit of work is required to uncomment the
1093 directives and define how usernames and passwords are checked, but
1094 there is now a framework to start from.
1096 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1097 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1098 without this. I don't know how relevant this is to other LDAP libraries.
1100 PH/02 Add the verb name to the "unknown ACL verb" error.
1102 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1104 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1106 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1107 directory not expanded when it should be if an expanded home directory
1108 was set for the address (which is overridden by the transport).
1110 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1113 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1114 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1115 because it is too late at that time, and has no effect.
1117 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1118 security issue with \' (bugzilla #107). I could not use the
1119 PQescapeStringConn() function, because it needs a PGconn value as one of
1122 PH/08 When testing addresses using -bt, indicate those final addresses that
1123 are duplicates that would not cause an additional delivery. At least one
1124 person was confused, thinking that -bt output corresponded to deliveries.
1125 (Suppressing duplicates isn't a good idea as you lose the information
1126 about possibly different redirections that led to the duplicates.)
1128 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1129 systems where poll() doesn't work, in particular OS X.
1131 PH/10 Added more information to debugging output for retry time not reached.
1133 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1134 operations in malware.c.
1136 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1139 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1140 syslog_duplication set false, log lines that would normally be written
1141 both the the main log and to the reject log were not written to syslog at
1144 PH/14 In the default configuration, change the use of "message" in ACL warn
1145 statements to "add_header".
1147 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1148 not followed by a command (e.g. "seen endif").
1150 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1151 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1154 PH/17 Added extra conditions to the default value of delay_warning_condition
1158 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1159 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1160 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1163 The Auto-Submitted: and various List- headers are standardised, whereas I
1164 don't think Precedence: ever was.
1166 PH/18 Refactored debugging code in route_finduser() to show more information,
1167 in particular, the error code if getpwnam() issues one.
1169 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1170 This is apparently needed in addition to the PH/07 change above to avoid
1171 any possible encoding problems.
1173 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1174 but not after initializing Perl.
1176 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1177 output them only if debugging. By default they are written stderr,
1178 apparently, which is not desirable.
1180 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1183 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1186 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1188 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1189 authenticated or an ident call has been made. Suppress the default
1190 values for $authenticated_id and $authenticated_sender (but permit -oMai
1191 and -oMas) when testing with -bh.
1193 PH/24 Re-jigged the order of the tests in the default configuration so that the
1194 tests for valid domains and recipients precede the DNS black list and CSA
1195 tests, on the grounds that those ones are more expensive.
1197 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1198 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1199 This bug exists in every version of Exim that I still have, right back to
1202 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1203 However, an attempt to turn on case-sensitivity in a regex key by
1204 including (?-i) didn't work because the subject string was already
1205 lowercased, and the effects were non-intuitive. It turns out that a
1206 one-line patch can be used to allow (?-i) to work as expected.
1212 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1213 other effects) broke the use of negated acl sub-conditions.
1215 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1218 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1219 "Deny" causes Exim to reject the incoming connection with a 554 error.
1220 Unfortunately, if there is a major crisis, such as a disk failure,
1221 tcp-wrappers gives "deny", whereas what one would like would be some
1222 kind of temporary error. A kludge has been added to help with this.
1223 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1224 554 error is used if errno is still zero or contains ENOENT (which occurs
1225 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1228 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1230 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1231 errors. Otherwise a message that provokes a temporary error (when other
1232 messages do not) can cause a whole host to time out.
1234 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1235 addresses were routed directly to files or pipes from a redirect router.
1236 File deliveries just didn't batch; pipe deliveries might have suffered
1239 PH/06 A failure to get a lock for a hints database would erroneously always say
1240 "Failed to get write lock", even when it was really a read lock.
1242 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1243 of 0600. This has been changed to use the value of the lockfile_mode
1244 option (which defaults to 0600).
1246 PH/08 Applied small patch from the Sieve maintainer.
1248 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1249 folder from quota calculations, a direct delivery into this folder messed
1250 up the contents of the maildirsize file. This was because the regex was
1251 used only to exclude .Trash (or whatever) when the size of the mailbox
1252 was calculated. There was no check that a delivery was happening into an
1253 excluded directory. This bug has been fixed by ignoring all quota
1254 processing for deliveries into excluded directories.
1256 PH/10 Added the maildirfolder_create_regex option to appendfile.
1262 PH/01 The code for finding all the local interface addresses on a FreeBSD
1263 system running IPv6 was broken. This may well have applied to all BSD
1264 systems, as well as to others that have similar system calls. The broken
1265 code found IPv4 interfaces correctly, but gave incorrect values for the
1266 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1267 that it would not match correctly against @[] and not recognize the IPv6
1270 PH/02 The ipliteral router was not recognizing addresses of the form user@
1271 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1273 PH/03 Added disable_ipv6.
1275 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1276 decoded form, because it is most often used to construct To: headers
1277 lines in autoreplies, and the decoded form may well be syntactically
1278 invalid. However, $reply_address has leading white space removed, and all
1279 newlines turned into spaces so that the autoreply transport does not
1282 PH/05 If group was specified without a user on a router, and no group or user
1283 was specified on a transport, the group from the router was ignored.
1285 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1286 for visible compile-time settings that can be used to change these
1287 numbers, for those that want even more. Backwards compatibility with old
1288 spool files has been maintained. However, going back to a previous Exim
1289 release will lost any variables that are in spool files.
1291 PH/07 Two small changes when running in the test harness: increase delay when
1292 passing a TCP/IP connection to a new process, in case the original
1293 process has to generate a bounce, and remove special handling of
1294 127.0.0.2 (sic), which is no longer necessary.
1296 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1297 be the same on different OS.
1299 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1302 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1303 whether --show-vars was specified or not
1305 JJ/02 exipick: Added support for new ACL variable spool format introduced
1308 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1309 syntactically invalid From: or Reply-to: line, and a filter used this to
1310 generate an autoreply, and therefore failed to obtain an address for the
1311 autoreply, Exim could try to deliver to a non-existent relative file
1312 name, causing unrelated and misleading errors. What now happens is that
1313 it logs this as a hard delivery error, but does not attempt to create a
1316 PH/11 The exinext utility has a -C option for testing purposes, but although
1317 the given file was scanned by exinext itself; it wasn't being passed on
1318 when Exim was called.
1320 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1321 an end-of-file indication when reading a command response.
1323 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1324 compiled. In many other places in Exim, IPv6 addresses are always
1325 recognized, so I have changed this. It also means that IPv4 domain
1326 literals of the form [IPV4:n.n.n.n] are now always recognized.
1328 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1329 used if the router is not running as root, for example, when verifying at
1330 ACL time, or when using -bh. The debugging output from this situation was
1331 non-existent - all you got was a failure to exec. I have made two
1334 (a) Failures to set uid/gid, the current directory, or a process leader
1335 in a subprocess such as that created by queryprogram now generate
1336 suitable debugging ouput when -d is set.
1338 (b) The queryprogram router detects when it is not running as root,
1339 outputs suitable debugging information if -d is set, and then runs
1340 the subprocess without attempting to change uid/gid.
1342 PH/15 Minor change to Makefile for building test_host (undocumented testing
1345 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1346 additional section of a DNS packet that returns MX or SRV records.
1347 Instead, it always explicitly searches for A/AAAA records. This avoids
1348 major problems that occur when a DNS server includes only records of one
1349 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1350 fixed another bug: if SRV records were looked up and the corresponding
1351 address records were *not* found in the additional section, the port
1352 values from the SRV records were lost.
1354 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1355 using the correct key (the original address) when searching the retry
1356 rules in order to find which one to use for generating the retry hint.
1358 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1359 adding the default one. Similarly, if it contains a Reply-To: header, the
1360 errors_reply_to option, if set, is not used.
1362 PH/19 When calculating a retry time, Exim used to measure the "time since
1363 failure" by looking at the "first failed" field in the retry record. Now
1364 it does not use this if it is later than than the arrival time of the
1365 message. Instead it uses the arrival time. This makes for better
1366 behaviour in cases where some deliveries succeed, thus re-setting the
1367 "first failed" field. An example is a quota failure for a huge message
1368 when small messages continue to be delivered. Without this change, the
1369 "time since failure" will always be short, possible causing more frequent
1370 delivery attempts for the huge message than are intended.
1371 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1373 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1374 $1, $2, $3) because the numerical variables can be reset during some
1375 expansion items (e.g. "match"), thereby losing the authentication data.
1377 PH/21 Make -bV show the size of off_t variables so that the test suite can
1378 decide whether to run tests for quotas > 2G.
1380 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1381 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1382 filecount value is greater than 2G or if a quota value is greater than 2G
1383 on a system where the size of off_t is not greater than 4, a panic error
1386 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1387 never match. The debug and -bh output now contains an explicit error
1388 message indicating a malformed IPv4 address or mask.
1390 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1391 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1392 PH/23 above applies.
1394 PH/25 Do not write to syslog when running in the test harness. The only
1395 occasion when this arises is a failure to open the main or panic logs
1396 (for which there is an explicit test).
1398 PH/26 Added the /no_tell option to "control=freeze".
1400 PH/27 If a host name lookup failed very early in a connection, for example, if
1401 the IP address matched host_lookup and the reverse lookup yielded a name
1402 that did not have a forward lookup, an error message of the form "no IP
1403 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1404 could be logged. Now it outputs the IP address instead of "NULL".
1406 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1407 allows the sender and the authenticated sender to be set when
1408 submitting a message from within Exim. Since child_open_exim() is
1409 documented for local_scan(), the new function should be too.
1411 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1412 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1413 results in an empty string is now treated as unset.
1415 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1417 PH/31 Added sender_verify_fail logging option.
1419 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1420 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1421 tidied the source and removed it altogether.
1423 PH/33 When a queue run was abandoned because the load average was too high, a
1424 log line was always written; now it is written only if the queue_run log
1425 selector is set. In addition, the log line for abandonment now contains
1426 information about the queue run such as the pid. This is always present
1427 in "start" and "stop" lines but was omitted from the "abandon" line.
1429 PH/34 Omit spaces between a header name and the colon in the error message that
1430 is given when verify = headers_syntax fails (if there are lots of them,
1431 the message gets confusing).
1433 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1434 names, as there are now some PTR records that contain slashes. This check
1435 is only to protect against broken name servers that fall over on strange
1436 characters, so the fact that it applies to all lookups doesn't matter.
1438 PH/36 Now that the new test suite is complete, we can remove some of the
1439 special code in Exim that was needed for the old test suite. For example,
1440 sorting DNS records because real resolvers return them in an arbitrary
1441 order. The new test suite's fake resolver always returns records in the
1444 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1445 bounces) except when queue_only is set, to avoid logging races between
1446 the different processes.
1448 PH/38 Panic-die if .include specifies a non-absolute path.
1450 PH/39 A tweak to the "H" retry rule from its user.
1452 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1453 a label. They prevented compilation on older perls.
1455 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1456 a warning to be raised on newish perls.
1458 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1459 on queue. Changes to match documented behaviour of showing count of
1460 messages matching specified criteria.
1462 PH/40 Changed the default ident timeout from 30s to 5s.
1464 PH/41 Added support for the use of login_cap features, on those BSD systems
1465 that have them, for controlling the resources used by pipe deliveries.
1467 PH/42 The content-scanning code uses fopen() to create files in which to put
1468 message data. Previously it was not paying any attention to the mode of
1469 the files. Exim runs with umask(0) because the rest of the code creates
1470 files with open(), and sets the required mode explicitly. Thus, these
1471 files were ending up world-writeable. This was not a big issue, because,
1472 being within the spool directory, they were not world-accessible. I have
1473 created a function called modefopen, which takes an additional mode
1474 argument. It sets umask(777), creates the file, chmods it to the required
1475 mode, then resets the umask. All the relevant calls to fopen() in the
1476 content scanning code have been changed to use this function.
1478 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1479 to 24 hours. This avoids potential overflow problems when processing G
1480 and H retry rules. I suspect nobody ever tinkers with this value.
1482 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1484 PH/45 When the plaintext authenticator is running as a client, the server's
1485 challenges are checked to ensure they are valid base64 strings. By
1486 default, the authentication attempt is cancelled if an invalid string is
1487 received. Setting client_ignore_invalid_base64 true ignores these errors.
1488 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1489 they are received. Thus, the responses can be made to depend on the
1490 challenges. If an invalid string is ignored, an empty string is placed in
1493 PH/46 Messages that are created by the autoreply transport now contains a
1494 References: header, in accordance with RFCs 2822 and 3834.
1496 PH/47 Added authenticated_sender_force to the smtp transport.
1498 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1500 PH/49 Installed latest patch from the Sieve maintainer.
1502 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1503 was also set, the appendfile transport was unnecessarily scanning a
1504 directory of message files (e.g. for maildir delivery) to find the count
1505 of files (along with the size), even though it did not need this
1506 information. It now does the scan only if it needs to find either the
1507 size of the count of files.
1509 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1511 PH/52 Two bugs concerned with error handling when the smtp transport is
1514 (i) Exim was not creating retry information for temporary errors given
1515 for individual recipients after the DATA command when the smtp transport
1516 was used in LMTP mode. This meant that they could be retried too
1517 frequently, and not timed out correctly.
1519 (ii) Exim was setting the flag that allows error details to be returned
1520 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1521 recipients that were returned after the DATA command.
1523 PH/53 This is related to PH/52, but is more general: for any failing address,
1524 when detailed error information was permitted to be returned to the
1525 sender, but the error was temporary, then after the final timeout, only
1526 "retry timeout exceeded" was returned. Now it returns the full error as
1527 well as "retry timeout exceeded".
1529 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1530 do this, and (what is worse) MTAs that accept it.
1532 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1533 will now be deprecated.
1535 PH/56 New os.c-cygwin from the Cygwin maintainer.
1537 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1538 formats (previously only available in exim formats via -bpr, -bpru,
1539 and -bpra. Now also available in native and exiqgrep formats)
1541 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1542 with very large, slow to parse queues
1544 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1546 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1548 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1549 responses to authentication challenges, though it was showing the
1550 challenges; (ii) I've removed the CR characters from the debug output for
1553 PH/58 Allow for the insertion of a newline as well as a space when a string
1554 is turned into more than one encoded-word during RFC 2047 encoding. The
1555 Sieve code now uses this.
1557 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1558 data_4xx, lost_connection, tls_required.
1560 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1561 message was being sent as an SMTP response.
1563 PH/61 Add -l and -k options to exicyclog.
1565 PH/62 When verifying, if an address was redirected to one new address, so that
1566 verification continued, and the new address failed or deferred after
1567 having set something in $address_data, the value of $address_data was not
1568 passed back to the ACL. This was different to the case when no
1569 redirection occurred. The value is now passed back in both cases.
1571 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1572 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1573 use login_cap.h, so on its own it isn't the distinguishing feature. The
1574 new name refers directly to the setclassresources() function.
1576 PH/65 Added configuration files for NetBSD3.
1578 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1580 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1581 is preferred over IPv4.
1583 PH/68 The bounce_return_message and bounce_return_body options were not being
1584 honoured for bounces generated during the reception of non-SMTP messages.
1585 In particular, this applied to messages rejected by the ACL. This bug has
1586 been fixed. However, if bounce_return_message is true and bounce_return_
1587 body is false, the headers that are returned for a non-SMTP message
1588 include only those that have been read before the error was detected.
1589 (In the case of an ACL rejection, they have all been read.)
1591 PH/69 The HTML version of the specification is now built in a directory called
1592 spec_html instead of spec.html, because the latter looks like a path with
1593 a MIME-type, and this confuses some software.
1595 PH/70 Catch two compiler warnings in sieve.c.
1597 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1598 function verify_get_ident() calls ip_connect() to connect a socket, but
1599 if the "connect()" function timed out, ip_connect() used to close the
1600 socket. However, verify_get_ident() also closes the socket later, and in
1601 between Exim writes to the log, which may get opened at this point. When
1602 the socket was closed in ip_connect(), the log could get the same file
1603 descriptor number as the socket. This naturally causes chaos. The fix is
1604 not to close the socket in ip_connect(); the socket should be closed by
1605 the function that creates it. There was only one place in the code where
1606 this was missing, in the iplookup router, which I don't think anybody now
1607 uses, but I've fixed it anyway.
1609 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1610 well as to direct DNS lookups. Otherwise the handling of names in host
1611 lists is inconsistent and therefore confusing.
1617 PH/01 Two changes to the default runtime configuration:
1619 (1) Move the checks for relay_from_hosts and authenticated clients from
1620 after to before the (commented out) DNS black list checks.
1622 (2) Add control=submission to the relay_from_hosts and authenticated
1623 clients checks, on the grounds that messages accepted by these
1624 statements are most likely to be submissions.
1626 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1628 (1) Generate an error if the third argument for the ${prvs expansion is
1631 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1634 (3) Reset the variables that are obtained from the first argument of
1635 ${prvscheck and used in the second argument before leaving the code,
1636 because their memory is reclaimed, so using them afterwards may do
1639 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1640 one (it's much easier than Tom thought :-).
1642 (5) Because of (4), we can now allow for the use of $prvscheck_result
1643 inside the third argument.
1645 PH/03 For some reason, the default setting of PATH when running a command from
1646 a pipe transport was just "/usr/bin". I have changed it to
1649 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1650 anything to be listed in the output from -bV.
1652 PH/05 When a filter generated an autoreply, the entire To: header line was
1653 quoted in the delivery log line, like this:
1655 => >A.N.Other <ano@some.domain> <original@ddress> ...
1657 This has been changed so that it extracts the operative address. There
1658 may be more than one such address. If so, they are comma-separated, like
1661 => >ano@some.domain,ona@other.domain <original@ddress> ...
1663 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1664 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1665 not being looked up in the rDNS to get a host name, Exim was showing the
1666 IP address twice in Received: lines, even though the IP addresses were
1667 identical. For example:
1669 Received: from [1.2.3.4] (helo=[1.2.3.4])
1671 However, if the real host name was known, it was omitting the HELO data
1672 if it matched the actual IP address. This has been tidied up so that it
1673 doesn't show the same IP address twice.
1675 PH/07 When both +timestamp and +memory debugging was on, the value given by
1676 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1677 called by the debug printing, thereby overwriting the timestamp buffer.
1678 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1680 PH/08 When the original message was included in an autoreply transport, it
1681 always said "this is a copy of the message, including all the headers",
1682 even if body_only or headers_only was set. It now gives an appropriate
1685 PH/09 Applied a patch from the Sieve maintainer which:
1687 o fixes some comments
1688 o adds the (disabled) notify extension core
1689 o adds some debug output for the result of if/elsif tests
1690 o points to the current vacation draft in the documentation
1691 and documents the missing references header update
1695 o fixes a bug in processing the envelope test (when testing
1696 multiple envelope elements, the last element determinted the
1699 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1700 Electronic Mail") by including:
1702 Auto-submitted: auto-generated
1704 in the messages that it generates (bounce messages and others, such as
1705 warnings). In the case of bounce messages for non-SMTP mesages, there was
1706 also a typo: it was using "Auto_submitted" (underscore instead of
1707 hyphen). Since every message generated by Exim is necessarily in response
1708 to another message, thes have all been changed to:
1710 Auto-Submitted: auto-replied
1712 in accordance with these statements in the RFC:
1714 The auto-replied keyword:
1716 - SHOULD be used on messages sent in direct response to another
1717 message by an automatic process,
1719 - MUST NOT be used on manually-generated messages,
1721 - MAY be used on Delivery Status Notifications (DSNs) and Message
1722 Disposition Notifications (MDNs),
1724 - MUST NOT be used on messages generated by automatic or periodic
1725 processes, except for messages which are automatic responses to
1728 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1729 to the default Received: header definition.
1731 PH/12 Added log selector acl_warn_skipped (default on).
1733 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1734 variables because (a) they are in the wrong storage pool and (b) even if
1735 they were copied, it wouldn't work properly because of the caching.
1737 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1738 checking when decoding. Apparently there are clients that generate
1739 overlong encoded strings. Why am I not surprised?
1741 PH/15 If the first argument of "${if match_address" was not empty, but did not
1742 contain an "@" character, Exim crashed. Now it writes a panic log message
1743 and treats the condition as false.
1745 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1747 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1748 extension "envelope-auth". The code is finished and in agreement with
1749 other implementations, but there is no documentation so far and in fact,
1750 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1751 not changing the active code.
1753 Print executed "if" and "elsif" statements when debugging is used. This
1754 helps a great deal to understand what a filter does.
1756 Document more things not specified clearly in RFC3028. I had all this
1757 sorted out, when out of a sudden new issues came to my mind. Oops."
1759 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1762 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1763 canonical form (as documented). However, after a host name lookup from
1764 the IP address, check_host() was doing a simple string comparison with
1765 addresses acquired from the DNS when checking that the found name did
1766 have the original IP as one of its addresses. Since any found IPv6
1767 addresses are likely to be in abbreviated form, the comparison could
1768 fail. Luckily, there already exists a function for doing the comparison
1769 by converting both addresses to binary, so now that is used instead of
1770 the text comparison.
1772 PH/20 There was another similar case to PH/19, when a complete host name was
1773 given in a host list; looking up its IP address could give an abbreviated
1774 form, whereas the current host's name might or might not be abbreviated.
1775 The same fix has been applied.
1781 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1782 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1785 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1786 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1788 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1790 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1791 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1792 Also, for "Auto-Submitted", treat anything other than "no" as
1793 non-personal, in accordance with RFC 3834. (Previously it treated
1794 anything starting "auto-" as non-personal.)
1796 TF/01 The control=submission/name=... option had a problem with syntax
1797 errors if the name included a slash character. The /name= option
1798 now slurps the rest of the string, so it can include any characters
1799 but it must come last in the list of options (after /sender_retain
1802 PH/05 Some modifications to the interface to the fake nameserver for the new
1810 TK/01 Added the "success_on_redirect" address verification option. See
1811 NewStuff for rationale and an example.
1813 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1815 PH/02 Patch to exigrep to allow it to work on syslog lines.
1817 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1818 fread() to skip over the body file's header line, because in Cygwin the
1819 header line is locked and is inaccessible.
1821 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1822 co-exist for some time) to make it clear that it is the Exim ID that is
1823 referenced, not the Message-ID: header line.
1825 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1826 string_format() function, because snprintf() does not exist on all
1829 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1830 ${stat: expansion item.
1832 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1833 protocol synchronization error", to keep the pedants happy.
1835 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1836 well as for IRIX systems, when gcc is being used. See the host.c source
1839 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1841 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1844 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1845 transport and to the smtp transport in LMTP mode.
1847 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1849 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1850 run a filter in a subprocess. This could lead to confusion in subsequent
1851 lookups in the parent process. There should also be a search_tidyup() at
1852 the end of the subprocess.
1854 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1855 only if the host matched helo_try_verify_hosts, which caused the
1856 verification to occur when the EHLO/HELO command was issued. The ACL just
1857 tested the remembered result. Now, if a previous verification attempt has
1858 not happened, "verify = helo" does it there and then.
1860 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1862 TK/03 Fix log output including CR from clamd.
1864 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1865 exist provoked a memory error which could cause a segfault.
1867 PH/15 Installed PCRE 6.2
1869 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1871 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1872 of the problem. Specifically, suggested +O2 rather than +O1 for the
1875 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1877 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1878 record, the port was not correctly logged when the outgoing_port log
1879 selector was set (it logged the transort's default port).
1881 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1882 fallback_hosts, and "hosts" in the smtp transport.
1884 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1885 host errors such as "Connection refused".
1887 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1888 authentication with radiusclient 0.4.9:
1890 - Error returned from rc_read_config was caught wrongly
1891 - Username/password not passed on to radius server due to wrong length.
1893 The presumption is that some radiusclient API changes for 4.51/PH/17
1894 were not taken care of correctly. The code is still untested by me (my
1895 Linux distribution still has 0.3.2 of radiusclient), but it was
1896 contributed by a Radius user.
1898 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1899 expanding the "port" option of the smtp transport.
1901 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1902 while reading a MIME header. Thanks to Tom Hughes for a patch.
1904 PH/24 Include config.h inside local_scan.h so that configuration settings are
1907 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1908 that in an ACL for RCPT (for example), you can examine exactly what was
1911 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1912 commands, but it was not correctly comparing the address with the actual
1913 client host address. Thus, it would show the EHLO address in Received:
1914 header lines when this was not necessary.
1916 PH/27 Added the % operator to ${eval:}.
1918 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1919 it should be ignoring failures (because with -C, for example, it has lost
1920 privilege). It wasn't ignoring creation failures other than "already
1923 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1926 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1927 because an input line was too long, either on its own, or by virtue of
1928 too many continuations, the temporary file was not being removed, and the
1929 return code was incorrect.
1931 PH/31 Missing "BOOL" in function definition in filtertest.c.
1933 PH/32 Applied Sieve patches from the maintainer.
1935 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1937 PH/33 Added "verify = not_blind".
1939 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1940 Local/Makefile (with some defaults set). These are used in built scripts
1941 such as exicyclog, but they have never been used in the exim_install
1942 script (though there are many overriding facilities there). I have
1943 arranged that the exim_install script now takes note of these two
1946 PH/35 Installed configuration files for Dragonfly.
1948 PH/36 When a locally submitted message by a trusted user did not contain a
1949 From: header, and the sender address was obtained from -f or from an SMTP
1950 MAIL command, and the trusted user did not use -F to supply a sender
1951 name, $originator_name was incorrectly used when constructing a From:
1952 header. Furthermore, $originator_name was used for submission mode
1953 messages from external hosts without From: headers in a similar way,
1954 which is clearly wrong.
1956 PH/37 Added control=suppress_local_fixups.
1958 PH/38 When log_selector = +received_sender was set, and the addition of the
1959 sender made the log line's construction buffer exactly full, or one byte
1960 less than full, an overflow happened when the terminating "\n" was
1963 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1964 when the result of a list match is failure because a DNS lookup failed.
1966 PH/40 RM_COMMAND is now used in the building process.
1968 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1969 the "build-* directories that it finds.
1971 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1972 address in a domain literal was a prefix of an interface address.
1974 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1975 when verifying a sender address, unless rewrite_headers is false.
1977 PH/44 Wrote a long comment about why errors_to addresses are verified as
1978 recipients, not senders.
1980 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1981 the ratelimit ACL was added.
1983 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1985 PH/46 Added extra information about PostgreSQL errors to the error string.
1987 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1988 avoiding the need to install special zones in a real server. This is
1989 backwards compatible; if it can't find the fake resolver, it drops back.
1990 Thus, both old and new test suites can be run.
1992 TF/02 Added util/ratelimit.pl
1994 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1995 clock is set back in time.
1997 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1998 Candler <B.Candler@pobox.com>.
2000 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2001 OK for addresses that are the result of redirections.
2003 PH/48 A number of further additions for the benefit of the new test suite,
2004 including a fake gethostbyname() that interfaces to the fake DNS resolver
2007 TF/06 The fix for widen_domains has also been applied to qualify_single and
2008 search_parents which are the other dnslookup options that can cause
2011 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2014 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2016 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2017 service" if it encountered an MX record with an empty target hostname.
2018 The message is now "an MX or SRV record indicated no SMTP service".
2020 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2021 if the DNS of the sending site is misconfigured. This is quite a
2022 common situation. This change restores the behaviour of treating a
2023 helo verification defer as a failure.
2025 PH/51 If self=fail was set on a router, the bounce message did not include the
2026 actual error message.
2032 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2034 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2035 command itself ended in error, the underlying message about the transport
2036 filter timeout was being overwritten with the pipe command error. Now the
2037 underlying error message should be appended to the second error message.
2039 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2041 PH/02 Reduce the amount of output that "make" produces by default. Full output
2042 can still be requested.
2044 PH/03 The warning log line about a condition test deferring for a "warn" verb
2045 was being output only once per connection, rather than after each
2046 occurrence (because it was using the same function as for successful
2047 "warn" verbs). This seems wrong, so I have changed it.
2049 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2050 it should not have, which might have caused a crash in the right
2051 circumstances, but probably never did.
2053 PH/04 Installed a modified version of Tony Finch's patch to make submission
2054 mode fix the return path as well as the Sender: header line, and to
2055 add a /name= option so that you can make the user's friendly name appear
2058 TF/03 Added the control = fakedefer ACL modifier.
2060 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2061 Mark Lowes for thorough testing.
2063 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2065 TK/03 Merged latest SRS patch from Miles Wilton.
2067 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2068 with the definition in sysexits.h (which is #included earlier).
2069 Fortunately, Exim does not actually use EX_OK. The code used to try to
2070 preserve the sysexits.h value, by assumimg that macro definitions were
2071 scanned for macro replacements. I have been disabused of this notion,
2072 so now the code just undefines EX_OK before #including unistd.h.
2074 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2075 in the smtp transport. When a block could not be written in a single
2076 write() function, the timeout was being re-applied to each part-write.
2077 This seems wrong - if the receiver was accepting one byte at a time it
2078 would take for ever. The timeout is now adjusted when this happens. It
2079 doesn't have to be particularly precise.
2081 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2082 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2084 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2085 without a domain if the check to <postmaster@domain> fails.
2087 SC/01 Eximstats: added -xls and the ability to specify output files
2088 (patch written by Frank Heydlauf).
2090 SC/02 Eximstats: use FileHandles for outputing results.
2092 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2094 SC/04 Eximstats: fixed display of large numbers with -nvr option
2096 SC/05 Eximstats: fixed merging of reports with empty tables.
2098 SC/06 Eximstats: added the -include_original_destination flag
2100 SC/07 Eximstats: removed tabs and trailing whitespace.
2102 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2104 TK/06 MBOX spool code: Add real "From " MBOX separator line
2105 so the .eml file is really in mbox format (even though
2106 most programs do not really care). Patch from Alex Miller.
2108 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2109 The latter is generated from $received_to and is only set if the
2110 message has one envelope recipient. SA can use these headers,
2111 obviously out-of-the-box. Patch from Alex Miller.
2113 PH/08 The ${def test on a variable was returning false if the variable's
2114 value was "0", contrary to what the specification has always said!
2115 The result should be true unless the variable is empty.
2117 PH/09 The syntax error of a character other than { following "${if
2118 def:variable_name" (after optional whitespace) was not being diagnosed.
2119 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2120 accidental colon was present, for example, could give incorrect results.
2122 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2123 result is used (not including appendfile, where other changes are about
2126 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2127 This involved changing a lot of size variables from int to off_t. It
2128 should work with maildirs and everything.
2130 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2131 spamd dying while we are connected to it.
2133 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2136 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2137 function for BDB 4.3.
2139 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2141 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2142 the sources for PCRE have been re-arranged and more files are now
2145 PH/15 The code I had for printing potentially long long variables in PH/11
2146 above was not the best (it lost precision). The length of off_t variables
2147 is now inspected at build time, and an appropriate printing format (%ld
2148 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2149 to be "long long int" or "long int". This is needed for the internal
2150 formatting function string_vformat().
2152 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2153 the configuration file to be ":syslog", then the script "guesses" where
2154 the logs files are, rather than using the compiled in default. In our
2155 case the guess is not the same as the compiled default, so the script
2156 suddenly stopped working when I started to use syslog. The patch checks
2157 to see if log_file_path is "". If so, it attempts to read it from exim
2158 with no configuration file to get the compiled in version, before it
2159 falls back to the previous guessing code."
2161 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2162 implementing BATV in an Exim configuration. See NewStuff for the gory
2165 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2166 Makefile that are specific to HP-UX.
2168 PH/18 If the "use_postmaster" option was set for a recipient callout together
2169 with the "random" option, the postmaster address was used as the MAIL
2170 FROM address for the random test, but not for the subsequent recipient
2171 test. It is now used for both.
2173 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2174 patch removes a few documentation additions to RFC 3028, because the
2175 latest draft now contains them. It adds the new en;ascii-case comparator
2176 and a new error check for 8bit text in MIME parts. Comparator and
2177 require names are now matched exactly. I enabled the subaddress
2178 extension, but it is not well tested yet (read: it works for me)."
2180 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2181 rework some of the code of TK/09 above to avoid the hardwired use of
2182 "%lld" and "long long". Replaced the call to snprintf() with a call to
2185 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2186 records point to non-existent hosts", "retry timeout exceeded", and
2187 "retry time not reached for any host after a long failure period".
2189 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2190 experimental DomainKeys support:
2192 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2193 (2) On an error such as an illegally used "control", the wrong name for
2194 the control was given.
2196 These problems did NOT occur unless DomainKeys support was compiled.
2198 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2200 PH/24 Added ${if match_ip condition.
2202 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2203 databases so that it will be absolutely obvious if a crash occurs in the
2204 DB library. This is a regular occurrence (often caused by mis-matched
2207 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2208 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2209 picked up on a user's system that detects such things. There doesn't seem
2210 to be a gcc warning option for this - only an attribute that has to be
2211 put on the function's prototype. It seems that in Fedora Core 4 they have
2212 set this on a number of new functions. No doubt there will be more in due
2215 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2216 specify a transport. However, if an address that was verified by such a
2217 router was the subject of a callout, Exim crashed because it tried to
2218 read the rcpt_include_affixes from the non-existent transport. Now it
2219 just assumes that the setting of that option is false. This bug was
2220 introduced by 4.51/PH/31.
2222 PH/28 Changed -d+all to exclude +memory, because that information is very
2223 rarely of interest, but it makes the output a lot bigger. People tend to
2224 do -d+all out of habit.
2226 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2227 code in os-type was giving problems when libc.so lives in lib64, like on
2230 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2231 aren't the modern standard, and it seems that some systems' include files
2232 don't always have them. Exim was already checking for some of the newer
2233 ones like T_AAAA, and defining it itself. I've added checks for all the
2234 record types that Exim uses.
2236 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2237 not automatically generating a new one, as it is supposed to. This
2238 prevented TLS from working. If the file did exist, but contained invalid
2239 data, a new version was generated, as expected. It was only the case of a
2240 non-existent file that was broken.
2242 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2243 with a change in libdomainkeys > 0.64.
2245 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2246 from DNS. If the selector record carries the flag, it now has
2247 precedence over the domain-wide flag.
2249 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2251 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2252 the use of a port name that isn't defined in /etc/services) occurred, the
2253 message was deferred as in a normal delivery, and thus remained on the
2254 spool, instead of being failed because of the mua_wrapper setting. This
2255 is now fixed, and I tidied up some of the mua_wrapper messages at the
2258 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2259 the messages in a hash of arrays rather than using individual hashes.
2260 This is a bit cleaner and results in dramatic memory savings, albeit
2261 at a slight CPU cost.
2263 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2264 as requested by Marc Sherman.
2266 SC/10 Eximstats: added histograms for user specified patterns as requested
2269 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2271 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2272 fopen() in the content-scanning modules that did not already have it.
2278 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2279 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2281 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2283 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2285 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2286 to test Sieve filters that use "vacation".
2288 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2289 that changes the way the GnuTLS parameters are stored in the cache file.
2290 The new format can be generated externally. For backward compatibility,
2291 if the data in the cache doesn't make sense, Exim assumes it has read an
2292 old-format file, and it generates new data and writes a new file. This
2293 means that you can't go back to an older release without removing the
2296 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2297 work if there are any delivery delays because "one_time" forces the
2298 parent to be marked "delivered", so its unseen clone is never tried
2299 again. For this reason, Exim now forbids the simultaneous setting of
2302 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2303 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2304 entry if you want to know the details. The fix, however, neglected to
2305 consider the case when local delivery batching is involved. The test for
2306 "previously delivered" was not happening when checking to see if an
2307 address could be batched with a previous (undelivered) one; under
2308 certain circumstances this could lead to multiple deliveries to the same
2311 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2312 in its include files, and this causes problems building Exim.
2314 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2315 header_syntax) but Exim was just ignoring anything given after a slash.
2316 In particular, this caused confusion with an attempt to use "verify =
2317 reverse_host_lookup/defer_ok". An error is now given when options are
2318 supplied for verify items that do not have them. (Maybe reverse_host_
2319 lookup should have a defer_ok option, but that's a different point.)
2321 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2322 defined by RFC 821) to 2048, because there were problems with some AUTH
2323 commands, and RFC 1869 says the size should be increased for extended
2324 SMTP commands that take arguments.
2326 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2329 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2330 "unknown" error; now it says that the functionality isn't in the binary.
2332 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2333 an address' error message when a string expansion fails (syntax or
2334 whatever). Otherwise the password may appear in the log. Following change
2335 PH/42 below, there is no longer a chance of it appearing in a bounce
2338 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2340 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2341 of its addresses was ever tried. (Bugzilla bug #2.)
2343 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2344 the result incorrectly in the debug output. (It correctly added a newline
2345 to what was transported.)
2347 TF/01 Added $received_time.
2349 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2350 commented out examples of how to interface to a virus scanner and to
2351 SpamAssassin. Also added commented examples of av_scanner and
2352 spamd_address settings.
2354 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2355 and controls are allowed in which ACLs. There were a couple of minor
2356 errors. Some of the entries in the conditions table (which is a table of
2357 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2358 negation of where the condition IS allowed.
2360 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2362 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2363 header file does not have a version number, so I've had to invent a new
2364 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2365 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2366 radiusclient), but it was contributed by a Radius user.
2368 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2369 files or directories, for OpenSSL.
2371 PH/19 When an Exim process that is running as root has to create an Exim log
2372 file, it does so in a subprocess that runs as exim:exim so as to get the
2373 ownership right at creation (otherwise, other Exim processes might see
2374 the file with the wrong ownership). There was no test for failure of this
2375 fork() call, which would lead to the process getting stuck as it waited
2376 for a non-existent subprocess. Forks do occasionally fail when resources
2377 run out. I reviewed all the other calls to fork(); they all seem to check
2380 PH/20 When checking for unexpected SMTP input at connect time (before writing
2381 the banner), Exim was not dealing correctly with a non-positive return
2382 from the read() function. If the client had disconnected by this time,
2383 the result was a log entry for a synchronization error with an empty
2384 string after "input=" when read() returned zero. If read() returned -1
2385 (an event I could not check), uninitialized data bytes were printed.
2386 There were reports of junk text (parts of files, etc) appearing after
2389 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2391 PH/22 Added support for macro redefinition, and (re)definition in between
2392 driver and ACL definitions.
2394 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2395 forgetting to use the resulting value; it was using the unexpanded value.
2397 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2398 hadn't been configured. The fix is from Juergen Kreileder, who
2399 understands it better than I do:
2401 "Here's what I see happening with three configured cyrus_sasl
2402 authenticators configured (plain, login, cram-md5):
2404 On startup auth_cyrus_sasl_init() gets called for each of these.
2405 This means three calls to sasl_listmech() without a specified mech_list.
2406 => SASL tests which mechs of all available mechs actually work
2407 => three warnings about OTP not working
2408 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2410 With the patch, sasl_listmech() also gets called three times. But now
2411 SASL's mech_list option is set to the server_mech specified in the the
2412 authenticator. Or in other words, the answer from sasl_listmech()
2413 gets limited to just the mech you're testing for (which is different
2415 => the return list contains just 'plain' or 'login', 'cram-md5' or
2416 nothing depending on the value of ob->server_mech.
2418 I've just tested the patch: Authentication still works fine,
2419 unavailable mechs specified in the exim configuration are still
2420 caught, and the auth.log warnings about OTP are gone."
2422 PH/25 When debugging is enabled, the contents of the command line are added
2423 to the debugging output, even when log_selector=+arguments is not
2426 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2427 answer is "GNU", and only if the return is "GNU/something" is the answer
2430 PH/27 $acl_verify_message is now set immediately after the failure of a
2431 verification in an ACL, and so is available in subsequent modifiers. In
2432 particular, the message can be preserved by coding like this:
2434 warn !verify = sender
2435 set acl_m0 = $acl_verify_message
2437 Previously, $acl_verify_message was set only while expanding "message"
2438 and "log_message" when a very denied access.
2440 PH/28 Modified OS/os.c-Linux with
2442 -#ifndef OS_LOAD_AVERAGE
2443 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2445 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2446 nomenclature these days.)
2448 PH/29 Installed patch from the Sieve maintainer that adds the options
2449 sieve_useraddress and sieve_subaddress to the redirect router.
2451 PH/30 In these circumstances:
2452 . Two addresses routed to the same list of hosts;
2453 . First host does not offer TLS;
2454 . First host accepts first address;
2455 . First host gives temporary error to second address;
2456 . Second host offers TLS and a TLS session is established;
2457 . Second host accepts second address.
2458 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2459 and peerdn, if requested) that were in fact used only for the second
2462 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2463 attention to any local part prefix or suffix that was matched by the
2464 router that accepted the address. It now behaves in the same way as it
2465 does for delivery: the affixes are removed from the local part unless
2466 rcpt_include_affixes is set on the transport.
2468 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2469 timeout during the DATA phase of an incoming message.
2471 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2472 applied a patch sanctioned by the Sieve maintainer.
2474 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2475 the uid or gid is negative. A case of a negative gid caused this to be
2476 noticed. The fix allows for either to be negative.
2478 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2479 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2482 PH/36 The change PH/12 above was broken. Fixed it.
2484 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2485 the grounds that routing the same address twice would always produce the
2486 same answer. This might have been true once, but it is certainly no
2487 longer true now. Routing a child address may depend on the previous
2488 routing that produced that child. Some complicated redirection strategies
2489 went wrong when messages had multiple recipients, and made Exim's
2490 behaviour dependent on the order in which the addresses were given.
2492 I have moved the duplicate checking until after the routing is complete.
2493 Exim scans the addresses that are assigned to local and remote
2494 transports, and removes any duplicates. This means that more work will be
2495 done, as duplicates will always all be routed, but duplicates are
2496 presumably rare, so I don't expect this is of any significance.
2498 For deliveries to pipes, files, and autoreplies, the duplicate checking
2499 still happens during the routing process, since they are not going to be
2502 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2503 It corrects a timeout issue with spamd. This is Ian's comment: "The
2504 background is that sometimes spamd either never reads data from a
2505 connection it has accepted, or it never writes response data. The exiscan
2506 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2507 blindly assumes that writes won't block so it may never time out."
2509 PH/39 Allow G after quota size as well as K and M.
2511 PH/40 The value set for $authenticated_id in an authenticator may not contain
2512 binary zeroes or newlines because the value is written to log lines and
2513 to spool files. There was no check on this. Now the value is run through
2514 the string_printing() function so that such characters are converted to
2515 printable escape sequences.
2517 PH/41 $message_linecount is a new variable that contains the total number of
2518 lines in the message. Compare $body_linecount, which is the count for the
2521 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2522 bounce and delay warning messages, except in certain special cases, which
2525 (a) An SMTP error message from a remote host;
2526 (b) A message specified in a :fail: redirection;
2527 (c) A message specified in a "fail" command in a system filter;
2528 (d) A message specified in a FAIL return from the queryprogram router;
2529 (e) A message specified by the cannot_route_message router option.
2531 In these cases only, Exim does include the error details in bounce and
2532 warning messages. There are also a few cases where bland messages such
2533 as "unrouteable address" or "local delivery error" are given.
2535 PH/43 $value is now also set for the "else" part of a ${run expansion.
2537 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2538 being worked on, but at least Exim now implements the latest version to
2541 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2542 process to complete was treated as a delivery failure, a timeout while
2543 writing the message to the pipe was logged, but erroneously treated as a
2544 successful delivery. Such timeouts include transport filter timeouts. For
2545 consistency with the overall process timeout, these timeouts are now
2546 treated as errors, giving rise to delivery failures by default. However,
2547 there is now a new Boolean option for the pipe transport called
2548 timeout_defer, which, if set TRUE, converts the failures into defers for
2549 both kinds of timeout. A transport filter timeout is now identified in
2552 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2553 systems where "make" and "gmake" are different, calling "gmake" at top
2554 level broke things. I've arranged for the value of $(MAKE) to be passed
2555 from the Makefile to this script so that it can call the same version of
2559 A note about Exim versions 4.44 and 4.50
2560 ----------------------------------------
2562 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2563 changes of various kinds. As a consequence, a big documentation update was
2564 needed. This delayed the release for rather longer than seemed good, especially
2565 in the light of a couple of (minor) security issues. Therefore, the changes
2566 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2567 release. So 4.44 and 4.50 are in effect two different branches that both start
2570 I have left the 4.50 change log unchanged; it contains all the changes since
2571 4.43. The change log for 4.44 is below; many of its items are identical to
2572 those for 4.50. This seems to be the most sensible way to preserve the
2573 historical information.
2579 1. Minor wording change to the doc/README.SIEVE file.
2581 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2582 computation of the current number of files was incorrect.
2584 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2585 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2588 4. Give more explanation in the error message when the command for a transport
2589 filter fails to execute.
2591 5. There are several places where Exim runs a non-Exim command in a
2592 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2593 was being done only for the command run by the queryprogram router. It is
2594 now done for all such subprocesses. The other cases are: ${run, transport
2595 filters, and the commands run by the lmtp and pipe transports.
2597 6. Added CONFIGURE_GROUP build-time option.
2599 7. Some older OS have a limit of 256 on the maximum number of file
2600 descriptors. Exim was using setrlimit() to set 1000 as a large value
2601 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2602 systems. I've change it so that if it can't get 1000, it tries for 256.
2604 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2605 was an oversight, and furthermore, ever since the addition of extra
2606 controls (e.g. 4.43/32), the checks on when to allow different forms of
2607 "control" were broken. There should now be diagnostics for all cases when a
2608 control that does not make sense is encountered.
2610 9. Added the /retain_sender option to "control=submission".
2612 10. $recipients is now available in the predata ACL (oversight).
2614 11. Tidy the search cache before the fork to do a delivery from a message
2615 received from the command line. Otherwise the child will trigger a lookup
2616 failure and thereby defer the delivery if it tries to use (for example) a
2617 cached ldap connection that the parent has called unbind on.
2619 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2620 of $address_data from the recipient verification was clobbered by the
2621 sender verification.
2623 13. The value of address_data from a sender verification is now available in
2624 $sender_address_data in subsequent conditions in the ACL statement.
2626 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2628 15. Added a new option "connect=<time>" to callout options, to set a different
2631 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2632 was its contents. (It was OK if the option was not defined at all.)
2634 17. A "Completed" log line is now written for messages that are removed from
2635 the spool by the -Mrm option.
2637 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2638 information about exactly what failed.
2640 19. Added -dd to debug only the daemon process.
2642 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2643 handles timeouts, both on the server side and network timeouts. Renamed the
2644 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2646 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2647 It is now set to "smtps".
2649 22. $host_address is now set to the target address during the checking of
2650 ignore_target_hosts.
2652 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2653 being passed; this would have caused $sender_host_name to have been used if
2654 matching the list had actually called for a host name (not very likely,
2655 since this list is usually IP addresses). A host name is now passed as
2658 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2659 code that specifies a non-restarting handler (typically sigaction() in
2660 modern systems) in an attempt to fix a rare and obscure crash bug.
2662 25. Narrowed the window for a race in the daemon that could cause it to ignore
2663 SIGCHLD signals. This is not a major problem, because they are used only to
2664 wake it up if nothing else does.
2666 26. A malformed maildirsize file could cause Exim to calculate negative values
2667 for the mailbox size or file count. Odd effects could occur as a result.
2668 The maildirsize information is now recalculated if the size or filecount
2671 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2672 support for a long time. Removed HAVE_SYS_VFS_H.
2674 28. Installed the latest version of exipick from John Jetmore.
2676 29. In an address list, if the pattern was not a regular expression, an empty
2677 subject address (from a bounce message) matched only if the pattern was an
2678 empty string. Non-empty patterns were not even tested. This was the wrong
2679 because it is perfectly reasonable to use an empty address as part of a
2680 database query. An empty address is now tested by patterns that are
2681 lookups. However, all the other forms of pattern expect the subject to
2682 contain a local part and a domain, and therefore, for them, an empty
2683 address still always fails if the pattern is not itself empty.
2685 30. Exim went into a mad DNS loop when attempting to do a callout where the
2686 host was specified on an smtp transport, and looking it up yielded more
2687 than one IP address.
2689 31. Re-factored the code for checking spool and log partition space into a
2690 function that finds that data and another that does the check. The former
2691 is then used to implement four new variables: $spool_space, $log_space,
2692 $spool_inodes, and $log_inodes.
2694 32. The RFC2047 encoding function was originally intended for short strings
2695 such as real names; it was not keeping to the 75-character limit for
2696 encoded words that the RFC imposes. It now respects the limit, and
2697 generates multiple encoded words if necessary. To be on the safe side, I
2698 have increased the buffer size for the ${rfc2047: expansion operator from
2701 33. It is now permitted to omit both strings after an "if" condition; if the
2702 condition is true, the result is "true". As before, when the second string
2703 is omitted, a false condition yields an empty string. This makes it less
2704 cumbersome to write custom ACL and router conditions.
2706 34. Failure to deliver a bounce message always caused it to be frozen, even if
2707 there was an errors_to setting on the router. The errors_to setting is now
2710 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2711 canonical form (fully expanded) before being placed in
2712 $sender_host_address.
2714 36. The table in the code that translates DNS record types into text (T_A to
2715 "A" for instance) was missing entries for NS and CNAME. It is just possible
2716 that this could have caused confusion if both these types were looked up
2717 for the same domain, because the text type is used as part of Exim's
2718 per-process caching. But the chance of anyone hitting this buglet seems
2721 37. The dnsdb lookup has been extended in a number of ways.
2723 (1) There is a new type, "zns", which walks up the domain tree until it
2724 finds some nameserver records. It should be used with care.
2726 (2) There is a new type, "mxh", which is like "mx" except that it returns
2727 just the host names, not the priorities.
2729 (3) It is now possible to give a list of domains (or IP addresses) to be
2730 looked up. The behaviour when one of the lookups defers can be
2731 controlled by a keyword.
2733 (4) It is now possible to specify the separator character for use when
2734 multiple records are returned.
2736 38. The dnslists ACL condition has been extended: it is now possible to supply
2737 a list of IP addresses and/or domains to be looked up in a particular DNS
2740 39. Added log_selector=+queue_time_overall.
2742 40. When running the queue in the test harness, wait just a tad after forking a
2743 delivery process, to get repeatability of debugging output.
2745 41. Include certificate and key file names in error message when GnuTLS fails
2746 to set them up, because the GnuTLS error message doesn't include the name
2747 of the failing file when there is a problem reading it.
2749 42. Allow both -bf and -bF in the same test run.
2751 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2753 44. The "Exiscan patch" is now merged into the mainline Exim source.
2755 45. Sometimes the final signoff response after QUIT could fail to get
2756 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2757 < 0 before doing a fflush(). This bug looks as though it goes back to the
2758 introduction of TLS in release 3.20, but "sometimes" must have been rare
2759 because the tests only now provoked it.
2761 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2762 (this can affect the format of dates).
2764 47. exim_tidydb, when checking for the continued existence of a message for
2765 which it has found a message-specific retry record, was not finding
2766 messages that were in split spool directories. Consequently, it was
2767 deleting retry records that should have stayed in existence.
2769 48. Steve fixed some bugs in eximstats.
2771 49. The SPA authentication driver was not abandoning authentication and moving
2772 on to the next authenticator when an expansion was forced to fail,
2773 contradicting the general specification for all authenticators. Instead it
2774 was generating a temporary error. It now behaves as specified.
2776 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2777 (the order specifies the preference for clients). The order is now AES256,
2778 AES128, 3DES, ARCFOUR128.
2780 51. Small patch to Sieve code - explicitly set From: when generating an
2783 52. Exim crashed if a remote delivery caused a very long error message to be
2784 recorded - for instance if somebody sent an entire SpamAssassin report back
2785 as a large number of 550 error lines. This bug was coincidentally fixed by
2786 increasing the size of one of Exim's internal buffers (big_buffer) that
2787 happened as part of the Exiscan merge. However, to be on the safe side, I
2788 have made the code more robust (and fixed the comments that describe what
2791 53. Now that there can be additional text after "Completed" in log lines (if
2792 the queue_time_overall log selector is set), a one-byte patch to exigrep
2793 was needed to allow it to recognize "Completed" as not the last thing in
2796 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2797 patch that reportedly fixes this has been added. I am not expert enough to
2798 create a test for it. This is what the patch creator wrote:
2800 "I found a little strange behaviour of ldap code when working with
2801 Windows 2003 AD Domain, where users was placed in more than one
2802 Organization Units. When I tried to give exim partial DN, the exit code
2803 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2804 But simultaneously result of request was absolutely normal ldap result,
2805 so I produce this patch..."
2807 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2808 REFERENCE, so I have modified the code to exclude the patch when that macro
2811 55. Some experimental protocols are using DNS PTR records for new purposes. The
2812 keys for these records are domain names, not reversed IP addresses. The
2813 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2814 leaves it alone. Component reversal etc. now happens only for IP addresses.
2817 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2819 57. Double the size of the debug message buffer (to 2048) so that more of very
2820 long debug lines gets shown.
2822 58. The exicyclog utility now does better if the number of log files to keep
2823 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2825 59. Two changes related to the smtp_active_hostname option:
2827 (1) $smtp_active_hostname is now available as a variable.
2828 (2) The default for smtp_banner uses $smtp_active_hostname instead
2829 of $primary_hostname.
2831 60. The host_aton() function is supposed to be passed a string that is known
2832 to be a valid IP address. However, in the case of IPv6 addresses, it was
2833 not checking this. This is a hostage to fortune. Exim now panics and dies
2834 if the condition is not met. A case was found where this could be provoked
2835 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2836 components; fortuitously, this particular loophole had already been fixed
2837 by change 4.50/55 above.
2839 If there are any other similar loopholes, the new check in host_aton()
2840 itself should stop them being exploited. The report I received stated that
2841 data on the command line could provoke the exploit when Exim was running as
2842 exim, but did not say which command line option was involved. All I could
2843 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2844 running as the user.
2847 61. There was a buffer overflow vulnerability in the SPA authentication code
2848 (which came originally from the Samba project). I have added a test to the
2849 spa_base64_to_bits() function which I hope fixes it.
2852 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2853 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2855 63. The daemon start-up calls getloadavg() while still root for those OS that
2856 need the first call to be done as root, but it missed one case: when
2857 deliver_queue_load_max is set with deliver_drop_privilege. This is
2858 necessary for the benefit of the queue runner, because there is no re-exec
2859 when deliver_drop_privilege is set.
2861 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2862 This has been fixed.
2864 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2865 was in use, was not putting the data itself into the right store pool;
2866 consequently, it could be overwritten for a subsequent message in the same
2867 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2870 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2872 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2873 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2874 the function were treating the return as a boolean value, which happened to
2875 work because 0=false and not-0=true, but is not correct code.
2877 68. The host_aton() function was not handling scoped IPv6 addresses (those
2878 with, for example, "%eth0" on the end) correctly.
2880 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2881 negated items (that is, ~something) in unsigned ints. Some compilers
2882 apparently mutter when there is no cast.
2884 70. If an address verification called from an ACL failed, and did not produce a
2885 user-specific message (i.e. there was only a "system" message), nothing was
2886 put in $acl_verify_message. In this situation, it now puts the system
2889 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2890 session; change 4.31/43 added the unwanted input to the log line - except
2891 that it did not do this in the start of session case. It now does.
2893 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2894 This is wrong and can cause the other end to generate a synchronization
2895 error if it is another Exim or anything else that does the synchronization
2896 check. A QUIT command is no longer sent after a timeout.
2898 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2899 during host lookups.
2901 74. The defer_ok option of callout verification was not working if it was used
2902 when verifying addresses in header lines, that is, for this case:
2904 verify = header_sender/callout=defer_ok
2906 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2907 those file descriptors could be used for SMTP connections. If anything
2908 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2909 could be sent to the SMTP client, causing chaos. The daemon now opens
2910 stdin, stdout, and stderr to /dev/null when it puts itself into the
2913 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2914 log by default. The user can override this with suitable Perl magic.
2916 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2917 the log message when discard triggers, was not working for the DATA ACL or
2918 for the non-SMTP ACL.
2920 78. Error message wording change in sieve.c.
2922 79. If smtp_accept_max_per_host was set, the number of connections could be
2923 restricted to fewer than expected, because the daemon was trying to set up
2924 a new connection before checking whether the processes handling previous
2925 connections had finished. The check for completed processes is now done
2926 earlier. On busy systems, this bug wouldn't be noticed because something
2927 else would have woken the daemon, and it would have reaped the completed
2930 80. If a message was submitted locally by a user whose login name contained one
2931 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2932 It caused a spool format error. I have fixed the spool reading code. A
2933 related problem was that the "from" clause in the Received: line became
2934 illegal because of the space(s). It is now covered by ${quote_local_part.
2936 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2939 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2941 83. Patch from Sieve maintainer to fix unterminated string problem in
2942 "vacation" handling.
2944 84. Some minor changes to the Linux configuration files to help with other
2945 OS variants using glibc.
2947 85. One more patch for Sieve to update vacation handling to latest spec.
2950 ----------------------------------------------------
2951 See the note above about the 4.44 and 4.50 releases.
2952 ----------------------------------------------------
2958 1. Change 4.43/35 introduced a bug that caused file counts to be
2959 incorrectly computed when quota_filecount was set in an appendfile
2962 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2963 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2966 3. Give more explanation in the error message when the command for a transport
2967 filter fails to execute.
2969 4. There are several places where Exim runs a non-Exim command in a
2970 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2971 was being done only for the command run by the queryprogram router. It is
2972 now done for all such subprocesses. The other cases are: ${run, transport
2973 filters, and the commands run by the lmtp and pipe transports.
2975 5. Some older OS have a limit of 256 on the maximum number of file
2976 descriptors. Exim was using setrlimit() to set 1000 as a large value
2977 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2978 systems. I've change it so that if it can't get 1000, it tries for 256.
2980 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2981 was an oversight, and furthermore, ever since the addition of extra
2982 controls (e.g. 4.43/32), the checks on when to allow different forms of
2983 "control" were broken. There should now be diagnostics for all cases when a
2984 control that does not make sense is encountered.
2986 7. $recipients is now available in the predata ACL (oversight).
2988 8. Tidy the search cache before the fork to do a delivery from a message
2989 received from the command line. Otherwise the child will trigger a lookup
2990 failure and thereby defer the delivery if it tries to use (for example) a
2991 cached ldap connection that the parent has called unbind on.
2993 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2994 of $address_data from the recipient verification was clobbered by the
2995 sender verification.
2997 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2998 was its contents. (It was OK if the option was not defined at all.)
3000 11. A "Completed" log line is now written for messages that are removed from
3001 the spool by the -Mrm option.
3003 12. $host_address is now set to the target address during the checking of
3004 ignore_target_hosts.
3006 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3007 being passed; this would have caused $sender_host_name to have been used if
3008 matching the list had actually called for a host name (not very likely,
3009 since this list is usually IP addresses). A host name is now passed as
3012 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3013 code that specifies a non-restarting handler (typically sigaction() in
3014 modern systems) in an attempt to fix a rare and obscure crash bug.
3016 15. Narrowed the window for a race in the daemon that could cause it to ignore
3017 SIGCHLD signals. This is not a major problem, because they are used only to
3018 wake it up if nothing else does.
3020 16. A malformed maildirsize file could cause Exim to calculate negative values
3021 for the mailbox size or file count. Odd effects could occur as a result.
3022 The maildirsize information is now recalculated if the size or filecount
3025 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3026 support for a long time. Removed HAVE_SYS_VFS_H.
3028 18. Updated exipick to current release from John Jetmore.
3030 19. Allow an empty sender to be matched against a lookup in an address list.
3031 Previously the only cases considered were a regular expression, or an
3034 20. Exim went into a mad DNS lookup loop when doing a callout where the
3035 host was specified on the transport, if the DNS lookup yielded more than
3038 21. The RFC2047 encoding function was originally intended for short strings
3039 such as real names; it was not keeping to the 75-character limit for
3040 encoded words that the RFC imposes. It now respects the limit, and
3041 generates multiple encoded words if necessary. To be on the safe side, I
3042 have increased the buffer size for the ${rfc2047: expansion operator from
3045 22. Failure to deliver a bounce message always caused it to be frozen, even if
3046 there was an errors_to setting on the router. The errors_to setting is now
3049 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3050 canonical form (fully expanded) before being placed in
3051 $sender_host_address.
3053 24. Updated eximstats to version 1.33
3055 25. Include certificate and key file names in error message when GnuTLS fails
3056 to set them up, because the GnuTLS error message doesn't include the name
3057 of the failing file when there is a problem reading it.
3059 26. Expand error message when OpenSSL has problems setting up cert/key files.
3062 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3063 (this can affect the format of dates).
3065 28. exim_tidydb, when checking for the continued existence of a message for
3066 which it has found a message-specific retry record, was not finding
3067 messages that were in split spool directories. Consequently, it was
3068 deleting retry records that should have stayed in existence.
3070 29. eximstats updated to version 1.35
3071 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3072 1.35 - bugfix such that pie charts by volume are generated correctly
3074 30. The SPA authentication driver was not abandoning authentication and moving
3075 on to the next authenticator when an expansion was forced to fail,
3076 contradicting the general specification for all authenticators. Instead it
3077 was generating a temporary error. It now behaves as specified.
3079 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3080 (the order specifies the preference for clients). The order is now AES256,
3081 AES128, 3DES, ARCFOUR128.
3083 31. Small patch to Sieve code - explicitly set From: when generating an
3086 32. Exim crashed if a remote delivery caused a very long error message to be
3087 recorded - for instance if somebody sent an entire SpamAssassin report back
3088 as a large number of 550 error lines. This bug was coincidentally fixed by
3089 increasing the size of one of Exim's internal buffers (big_buffer) that
3090 happened as part of the Exiscan merge. However, to be on the safe side, I
3091 have made the code more robust (and fixed the comments that describe what
3094 33. Some experimental protocols are using DNS PTR records for new purposes. The
3095 keys for these records are domain names, not reversed IP addresses. The
3096 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3097 leaves it alone. Component reversal etc. now happens only for IP addresses.
3100 34. The host_aton() function is supposed to be passed a string that is known
3101 to be a valid IP address. However, in the case of IPv6 addresses, it was
3102 not checking this. This is a hostage to fortune. Exim now panics and dies
3103 if the condition is not met. A case was found where this could be provoked
3104 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3105 components; fortuitously, this particular loophole had already been fixed
3106 by change 4.50/55 or 4.44/33 above.
3108 If there are any other similar loopholes, the new check in host_aton()
3109 itself should stop them being exploited. The report I received stated that
3110 data on the command line could provoke the exploit when Exim was running as
3111 exim, but did not say which command line option was involved. All I could
3112 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3113 running as the user.
3116 35. There was a buffer overflow vulnerability in the SPA authentication code
3117 (which came originally from the Samba project). I have added a test to the
3118 spa_base64_to_bits() function which I hope fixes it.
3121 36. The daemon start-up calls getloadavg() while still root for those OS that
3122 need the first call to be done as root, but it missed one case: when
3123 deliver_queue_load_max is set with deliver_drop_privilege. This is
3124 necessary for the benefit of the queue runner, because there is no re-exec
3125 when deliver_drop_privilege is set.
3127 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3128 was in use, was not putting the data itself into the right store pool;
3129 consequently, it could be overwritten for a subsequent message in the same
3130 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3133 38. Sometimes the final signoff response after QUIT could fail to get
3134 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3135 < 0 before doing a fflush(). This bug looks as though it goes back to the
3136 introduction of TLS in release 3.20, but "sometimes" must have been rare
3137 because the tests only now provoked it.
3143 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3144 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3145 Now it returns an integer. A number of calls were still expecting a T/F
3146 return. Fortuitously, in all cases, the tests worked in OK situations,
3147 which is the norm. However, things would have gone wrong on any write
3148 failures on the smtp file descriptor. This function is used when sending
3149 messages over SMTP and also when doing verify callouts.
3151 2. When Exim is called to do synchronous delivery of a locally submitted
3152 message (the -odf or -odi options), it no longer closes stderr before doing
3155 3. Implemented the mua_wrapper option.
3157 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3159 5. Implemented the functions header_remove(), header_testname(),
3160 header_add_at_position(), and receive_remove_recipient(), and exported them
3163 6. If an ACL "warn" statement specified the addition of headers, Exim already
3164 inserted X-ACL-Warn: at the start if there was no header name. However, it
3165 was not making this test for the second and subsequent header lines if
3166 there were newlines in the string. This meant that an invalid header could
3167 be inserted if Exim was badly configured.
3169 7. Allow an ACL "warn" statement to add header lines at the start or after all
3170 the Received: headers, as well as at the end.
3172 8. Added the rcpt_4xx retry error code.
3174 9. Added postmaster_mailfrom=xxx to callout verification option.
3176 10. Added mailfrom=xxxx to the callout verification option, for verify=
3179 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3180 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3182 12. Inserted some casts to stop certain compilers warning when using pointer
3183 differences as field lengths or precisions in printf-type calls (mostly
3184 affecting debugging statements).
3186 13. Added optional readline() support for -be (dynamically loaded).
3188 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3189 same clock tick as a message's arrival, so that its received time was the
3190 same as the "first fail" time on the retry record, and that message
3191 remained on the queue past the ultimate address timeout, every queue runner
3192 would try a delivery (because it was past the ultimate address timeout) but
3193 after another failure, the ultimate address timeout, which should have then
3194 bounced the address, did not kick in. This was a "< instead of <=" error;
3195 in most cases the first failure would have been in the next clock tick
3196 after the received time, and all would be well.
3198 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3199 being recognized when the domain list was tested by the match_domain
3200 condition in an expansion string.
3202 16. Added the ${str2b64: operator.
3204 17. Exim was always calling setrlimit() to set a large limit for the number of
3205 processes, without checking whether the existing limit was already
3206 adequate. (It did check for the limit on file descriptors.) Furthermore,
3207 errors from getrlimit() and setrlimit() were being ignored. Now they are
3208 logged to the main and panic logs, but Exim does carry on, to try to do its
3209 job under whatever limits there are.
3211 18. Imported PCRE 5.0.
3213 19. Trivial typo in log message " temporarily refused connection" (the leading
3216 20. If the log selector return_path_on_delivery was set and an address was
3217 redirected to /dev/null, the delivery process crashed because it assumed
3218 that a return path would always be set for a "successful" delivery. In this
3219 case, the whole delivery is bypassed as an optimization, and therefore no
3222 21. Internal re-arrangement: the function for sending a challenge and reading
3223 a response while authentication was assuming a zero-terminated challenge
3224 string. It's now changed to take a pointer and a length, to allow for
3225 binary data in such strings.
3227 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3229 23. Exim was not respecting finduser_retries when seeking the login of the
3230 uid under which it was called; it was always trying 10 times. (The default
3231 setting of finduser_retries is zero.) Also, it was sleeping after the final
3232 failure, which is pointless.
3234 24. Implemented tls_on_connect_ports.
3236 25. Implemented acl_smtp_predata.
3238 26. If the domain in control=submission is set empty, Exim assumes that the
3239 authenticated id is a complete email address when it generates From: or
3240 Sender: header lines.
3242 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3243 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3244 chown and chgrp in /bin and hostname in /usr/bin.
3246 28. Exim was keeping the "process log" file open after each use, just as it
3247 does for the main log. This opens the possibility of it remaining open for
3248 long periods when the USR1 signal hits a daemon. Occasional processlog
3249 errors were reported, that could have been caused by this. Anyway, it seems
3250 much more sensible not to leave this file open at all, so that is what now
3253 29. The long-running daemon process does not normally write to the log once it
3254 has entered its main loop, and it closes the log before doing so. This is
3255 so that log files can straightforwardly be renamed and moved. However,
3256 there are a couple of unusual error situations where the daemon does write
3257 log entries, and I had neglected to close the log afterwards.
3259 30. The text of an SMTP error response that was received during a remote
3260 delivery was being truncated at 512 bytes. This is too short for some of
3261 the long messages that one sometimes sees. I've increased the limit to
3264 31. It is now possible to make retry rules that apply only when a message has a
3265 specific sender, in particular, an empty sender.
3267 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3268 it possible to be selective about when SMTP synchronization is enforced.
3270 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3272 32. Implemented hosts_connection_nolog.
3274 33. Added an ACL for QUIT.
3276 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3279 35. Added mailbox_size and mailbox_filecount to appendfile.
3281 36. Added control = no_multiline_responses to ACLs.
3283 37. There was a bug in the logic of the code that waits for the clock to tick
3284 in the case where the clock went backwards by a substantial amount such
3285 that the microsecond fraction of "now" was more than the microsecond
3286 fraction of "then" (but the whole seconds number was less).
3288 38. Added support for the libradius Radius client library this is found on
3289 FreeBSD (previously only the radiusclient library was supported).
3295 1. When certain lookups returned multiple values in the form name=value, the
3296 quoting of the values was not always being done properly. Specifically:
3297 (a) If the value started with a double quote, but contained no whitespace,
3299 (b) If the value contained whitespace other than a space character (i.e.
3300 tabs or newlines or carriage returns) it was not quoted.
3301 This fix has been applied to the mysql and pgsql lookups by writing a
3302 separate quoting function and calling it from the lookup code. The fix
3303 should probably also be applied to nisplus, ibase and oracle lookups, but
3304 since I cannot test any of those, I have not disturbed their existing code.
3306 2. A hit in the callout cache for a specific address caused a log line with no
3307 reason for rejecting RCPT. Now it says "Previous (cached) callout
3308 verification failure".
3310 3. There was an off-by-one bug in the queryprogram router. An over-long
3311 return line was truncated at 256 instead of 255 characters, thereby
3312 overflowing its buffer with the terminating zero. As well as fixing this, I
3313 have increased the buffer size to 1024 (and made a note to document this).
3315 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3316 when Exim is waiting for an SMTP response from a remote server, Exim
3317 restarts its select() call on the socket, thereby resetting its timeout.
3318 This is not a problem when such interrupts are rare. Somebody set up a cron
3319 job to run exiwhat every 2 minutes, which is less than the normal select()
3320 timeout (5 or 10 minutes). This meant that the select() timeout never
3321 kicked in because it was always reset. I have fixed this by comparing the
3322 time when an interrupt arrives with the time at the start of the first call
3323 to select(). If more time than the timeout has elapsed, the interrupt is
3324 treated as a timeout.
3326 5. Some internal re-factoring in preparation for the addition of Sieve
3327 extensions (by MH). In particular, the "personal" test is moved to a
3328 separate function, and given an option for scanning Cc: and Bcc: (which is
3329 not set for Exim filters).
3331 6. When Exim created an email address using the login of the caller as the
3332 local part (e.g. when creating a From: or Sender: header line), it was not
3333 quoting the local part when it contained special characters such as @.
3335 7. Installed new OpenBSD configuration files.
3337 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3338 try to make them clearer.
3340 9. Callout options, other than the timeout value, were being ignored when
3341 verifying sender addresses in header lines. For example, when using
3343 verify = header_sender/callout=no_cache
3345 the cache was (incorrectly) being used.
3347 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3348 only the Cygwin environment.
3350 11. When return_path_on_delivery was set as a log selector, if different remote
3351 addresses in the same message used different return paths and parallel
3352 remote delivery occurred, the wrong values would sometimes be logged.
3353 (Whenever a remote delivery process finished, the return path value from
3354 the most recently started remote delivery process was logged.)
3356 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3357 lines when AUTH and/or TLS are in use. This is the "received protocol"
3358 field. Exim used to use "asmtp" for authenticated SMTP, without any
3359 indication (in the protocol name) for TLS use. Now it follows the RFC and
3360 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3361 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3362 names appear in log lines as well as in Received: header lines.
3364 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3365 extensions, and comparison tests, and to fix some bugs.
3367 14. Changes to the "personal" filter test:
3369 (1) The test was buggy in that it was just doing the equivalent of
3370 "contains" tests on header lines. For example, if a user's address was
3371 anne@some.where, the "personal" test would incorrectly be true for
3373 To: susanne@some.where
3375 This test is now done by extracting each address from the header in turn,
3376 and checking the entire address. Other tests that are part of "personal"
3377 are now done using regular expressions (for example, to check local parts
3378 of addresses in From: header lines).
3380 (2) The list of non-personal local parts in From: addresses has been
3381 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3382 taken from the Sieve specification recommendations.
3384 (3) If the message contains any header line starting with "List-" it is
3385 treated as non-personal.
3387 (4) The test for "circular" in the Subject: header line has been removed
3388 because it now seems ill-conceived.
3390 15. Minor typos in src/EDITME comments corrected.
3392 16. Installed latest exipick from John Jetmore.
3394 17. If headers_add on a router specified a text string that was too long for
3395 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3396 of string_sprintf() is now avoided.
3398 18. $message_body_size was not set (it was always zero) when running the DATA
3399 ACL and the local_scan() function.
3401 19. For the "mail" command in an Exim filter, no default was being set for
3402 the once_repeat time, causing a random time value to be used if "once" was
3403 specified. (If the value happened to be <= 0, no repeat happened.) The
3404 default is now 0s, meaning "never repeat". The "vacation" command was OK
3405 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3406 (I found it when inspecting the code).
3408 20. There is now an overall timeout for performing a callout verification. It
3409 defaults to 4 times the callout timeout, which applies to individual SMTP
3410 commands during the callout. The overall timeout applies when there is more
3411 than one host that can be tried. The timeout is checked before trying the
3412 next host. This prevents very long delays if there are a large number of
3413 hosts and all are timing out (e.g. when the network connections are timing
3414 out). The value of the overall timeout can be changed by specifying an
3415 additional sub-option for "callout", called "maxwait". For example:
3417 verify = sender/callout=5s,maxwait=20s
3419 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3420 to the end before writing, but this should make it even safer).
3422 22. Exim was forgetting that it had advertised PIPELINING for the second and
3423 subsequent messages on an SMTP connection. It was also not resetting its
3424 memory on STARTTLS and an internal HELO.
3426 23. When Exim logs an SMTP synchronization error within a session, it now
3427 records whether PIPELINING has been advertised or not.
3429 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3430 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3431 rather than long int.
3433 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3435 26. Added the never_mail option to autoreply.
3441 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3442 crash if the getsockname() call failed; this can happen if a connection is
3443 closed very soon after it is established. The problem was simply in the
3444 order in which certain operations were done, causing Exim to try to write
3445 to the SMTP stream before it had set up the file descriptor. The bug has
3446 been fixed by making things happen in the correct order.
3452 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3453 before the connection was closed, thus losing the rejection response.
3455 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3456 some early Solaris releases, but causes trouble in current releases where
3457 socklen_t is defined.
3459 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3462 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3465 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3466 that releasing the top of it at the end releases what was used for sub-
3467 expansions (unless the block got too big). However, discard this block if
3468 the first thing is a variable or header, so that we can use its block when
3469 it is dynamic (useful for very large $message_headers, for example).
3471 6. Lookups now cache *every* query, not just the most recent. A new, separate
3472 store pool is used for this. It can be recovered when all lookup caches are
3473 flushed. Lookups now release memory at the end of their result strings.
3474 This has involved some general refactoring of the lookup sources.
3476 7. Some code has been added to the store_xxx() functions to reduce the amount
3477 of flapping under certain conditions.
3479 8. log_incoming_interface used to affect only the <= reception log lines. Now
3480 it causes the local interface and port to be added to several more SMTP log
3481 lines, for example "SMTP connection from", and rejection lines.
3483 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3485 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3487 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3488 could be overwritten at the end of the current message (or the start of a
3489 new message if it was set in a HELO ACL). The value is now preserved for
3490 the duration of the SMTP connection.
3492 12. If a transport had a headers_rewrite setting, and a matching header line
3493 contained an unqualified address, that address was qualified, even if it
3494 did not match any rewriting rules. The underlying bug was that the values
3495 of the flags that permit the existence of unqualified sender and recipient
3496 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3497 non-local messages, and by -bnq for local messages) were not being
3498 preserved with the message after it was received.
3500 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3501 "next input=" as part of the text comprising the host identity instead of
3502 the correct text. The code was using the same buffer for two different
3503 strings. However, depending on which order the printing function evaluated
3504 its arguments, the bug did not always show up. Under Linux, for example, my
3505 test suite worked just fine.
3507 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3508 doesn't work with some older versions of Perl. It has been changed to "my",
3509 which in any case is probably the better facility to use.
3511 15. A really picky compiler found some instances of statements for creating
3512 error messages that either had too many or two few arguments for the format
3515 16. The size of the buffer for calls to the DNS resolver has been increased
3516 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3517 for addresses that have a lot of PTR records. This alleviates a problem; it
3518 does not fully solve it.
3520 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3521 buffer now truncates the list and logs the incident, which is the same
3522 action as happens when Exim is looking up a host name and its aliases.
3523 Previously in this situation something unpredictable would happen;
3524 sometimes it was "internal error: store_reset failed".
3526 18. If a server dropped the connection unexpectedly when an Exim client was
3527 using GnuTLS and trying to read a response, the client delivery process
3528 crashed while trying to generate an error log message.
3530 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3531 string, for example:
3533 warn message = H1: something\nH2: something
3535 the text was added as a single header line from Exim's point of view
3536 though it ended up OK in the delivered message. However, searching for the
3537 second and subsequent header lines using $h_h2: did not work. This has been
3538 fixed. Similarly, if a system filter added multiple headers in this way,
3539 the routers could not see them.
3541 20. Expanded the error message when iplsearch is called with an invalid key to
3542 suggest using net-iplsearch in a host list.
3544 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3545 ACLs are no longer actually imposed (and a message to that effect is
3548 22. If a "gecos" field in a passwd entry contained escaped characters, in
3549 particular, if it contained a \" sequence, Exim got it wrong when building
3550 a From: or a Sender: header from that name. A second bug also caused
3551 incorrect handling when an unquoted " was present following a character
3552 that needed quoting.
3554 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3555 was not being matched caselessly.
3557 24. Arranged for all hyphens in the exim.8 source to be escaped with
3560 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3561 properly. Recipient callout cache records were still being keyed to include
3562 the sender, even when use_sender was set false. This led to far more
3563 callouts that were necessary. The sender is no longer included in the key
3564 when use_sender is false.
3566 26. Added "control = submission" modifier to ACLs.
3568 27. Added the ${base62d: operator to decode base 62 numbers.
3570 28. dnsdb lookups can now access SRV records.
3572 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3573 the configuration file.
3575 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3576 (-v) mode. This makes the output for a verbose queue run more intelligible.
3578 31. Added a use_postmaster feature to recipient callouts.
3580 32. Added the $body_zerocount variable, containing the number of binary zero
3581 bytes in the message body.
3583 33. The time of last modification of the "new" subdirectory is now used as the
3584 "mailbox time last read" when there is a quota error for a maildir
3587 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3589 35. Added +ignore_unknown as a special item in host lists.
3591 36. Code for decoding IPv6 addresses in host lists is now included, even if
3592 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3593 address was recognized as an IP address, but was then not correctly decoded
3594 into binary, causing unexpected and incorrect effects when compared with
3601 1. Very minor rewording of debugging text in manualroute to say "list of
3602 hosts" instead of "hostlist".
3604 2. If verify=header_syntax was set, and a header line with an unqualified
3605 address (no domain) and a large number of spaces between the end of the
3606 name and the colon was received, the reception process suffered a buffer
3607 overflow, and (when I tested it) crashed. This was caused by some obsolete
3608 code that should have been removed. The fix is to remove it!
3610 3. When running in the test harness, delay a bit after writing a bounce
3611 message to get a bit more predictability in the log output.
3613 4. Added a call to search_tidyup() just before forking a reception process. In
3614 theory, someone could use a lookup in the expansion of smtp_accept_max_
3615 per_host which, without the tidyup, could leave open a database connection.
3617 5. Added the variables $recipient_data and $sender_data which get set from a
3618 lookup success in an ACL "recipients" or "senders" condition, or a router
3619 "senders" option, similar to $domain_data and $local_part_data.
3621 6. Moved the writing of debug_print from before to after the "senders" test
3624 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3625 problems for message scanning, either using a data ACL, or using
3626 local_scan() because the Received: header was not generated till after they
3627 were called (in order to set the time as the time of reception completion).
3628 I have revised the way this works. The header is now generated after the
3629 body is received, but before the ACL or local_scan() are called. After they
3630 are run, the timestamp in the header is updated.
3636 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3637 before starting a queue runner without re-exec. This happened only when
3638 deliver_drop_privilege was set or when the Exim user was set to root. The
3639 effect of the bug was that timeouts during subsequent deliveries caused
3640 crashes instead of being properly handled. The handler is now left at its
3641 default (and expected) setting.
3643 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3644 message, again when deliver_drop_privilege is set or Exim is run as root.
3645 The bug described in (1) was not present in this case, but the tidying up
3646 of the other signals was missing. I have made the two cases consistent.
3648 3. The ignore_target_hosts setting on a manualroute router was being ignored
3649 for hosts that were looked up using the /MX notation.
3651 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3654 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3655 operated on the sender address. After changing the $sender_address to <>
3656 for the sender address verify, Exim was re-instated it as the original
3657 (before rewriting) address, but remembering that it had rewritten it, so it
3658 wasn't rewriting it again. This bug also had the effect of breaking the
3659 sender address verification caching when the sender address was rewritten.
3661 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3662 This has been changed so that if the ip literal address matches
3663 ignore_target_hosts, the router declines.
3665 7. Added expansion conditions match_domain, match_address, and match_local_
3666 part (NOT match_host).
3668 8. The placeholder for the Received: header didn't have a length field set.
3670 9. Added code to Exim itself and to exim_lock to test for a specific race
3671 condition that could lead to file corruption when using MBX delivery. The
3672 issue is with the lockfile that is created in /tmp. If this file is removed
3673 after a process has opened it but before that process has acquired a lock,
3674 there is the potential for a second process to recreate the file and also
3675 acquire a lock. This could lead to two Exim processes writing to the file
3676 at the same time. The added code performs the same test as UW imapd; it
3677 checks after acquiring the lock that its file descriptor still refers to
3678 the same named file.
3680 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3681 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3684 11. Added the smtp_active_hostname option. If used, this will typically be made
3685 to depend on the incoming interface address. Because $interface_address is
3686 not set up until the daemon has forked a reception process, error responses
3687 that can happen earlier (such as "too many connections") no longer contain
3690 12. If an expansion in a condition on a "warn" statement fails because a lookup
3691 defers, the "warn" statement is abandoned, and the next ACL statement is
3692 processed. Previously this caused the whole ACL to be aborted.
3694 13. Added the iplsearch lookup type.
3696 14. Added ident_timeout as a log selector.
3698 15. Added tls_certificate_verified as a log selector.
3700 16. Added a global option tls_require_ciphers (compare the smtp transport
3701 option of the same name). This controls incoming TLS connections.
3703 17. I finally figured out how to make tls_require_ciphers do a similar thing
3704 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3705 before starting the TLS session.
3707 18. Tabs are now shown as \t in -bP output.
3709 19. If the log selector return_path_on_delivery was set, Exim crashed when
3710 bouncing a message because it had too many Received: header lines.
3712 20. If two routers both had headers_remove settings, and the first one included
3713 a superfluous trailing colon, the final name in the first list and the
3714 first name in the second list were incorrectly joined into one item (with a
3715 colon in the middle).
3721 1. Added -C and -D options to the exinext utility, mainly to make it easier
3722 to include in the automated testing, but these could be helpful when
3723 multiple configurations are in use.
3725 2. The exinext utility was not formatting the output nicely when there was
3726 an alternate port involved in the retry record key, nor when there was a
3727 message id as well (for retries that were specific to a specific message
3728 and a specific host). It was also confused by IPv6 addresses, because of
3729 the additional colons they contain. I have fixed the IPv4 problem, and
3730 patched it up to do a reasonable job for IPv6.
3732 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3733 delivery, the log line now contains "pipelined" if PIPELINING was used.
3735 4. An SMTP transport process used to panic and die if the bind() call to set
3736 an explicit outgoing interface failed. This has been changed; it is now
3737 treated in the same way as a connect() failure.
3739 5. A reference to $sender_host_name in the part of a conditional expansion
3740 that was being skipped was still causing a DNS lookup. This no longer
3743 6. The def: expansion condition was not recognizing references to header lines
3744 that used bh_ and bheader_.
3746 7. Added the _cache feature to named lists.
3748 8. The code for checking quota_filecount in the appendfile transport was
3749 allowing one more file than it should have been.
3751 9. For compatibility with Sendmail, the command line option
3759 and sets the incoming protocol and host name (for trusted callers). The
3760 host name and its colon can be omitted when only the protocol is to be set.
3761 Note the Exim already has two private options, -pd and -ps, that refer to
3762 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3763 "s", but I don't think that's a major issue.
3765 10. A number of refactoring changes to the code, none of which should affect
3768 (a) The number of logging options was getting close to filling up the
3769 32-bit word that was used as a bit map. I have split them into two classes:
3770 those that are passed in the argument to log_write(), and those that are
3771 only ever tested independently outside of that function. These are now in
3772 separate 32-bit words, so there is plenty of room for expansion again.
3773 There is no change in the user interface or the logging behaviour.
3775 (b) When building, for example, log lines, the code previously used a
3776 macro that called string_cat() twice, in order to add two strings. This is
3777 not really sufficiently general. Furthermore, there was one instance where
3778 it was actually wrong because one of the argument was used twice, and in
3779 one call a function was used. (As it happened, calling the function twice
3780 did not affect the overall behaviour.) The macro has been replaced by a
3781 function that can join an arbitrary number of extra strings onto a growing
3784 (c) The code for expansion conditions now uses a table and a binary chop
3785 instead of a serial search (which was left over from when there were very
3786 few conditions). Also, it now recognizes conditions like "pam" even when
3787 the relevant support is not compiled in: a suitably worded error message is
3788 given if an attempt is made to use such a condition.
3790 11. Added ${time_interval:xxxxx}.
3792 12. A bug was causing one of the ddress fields not to be passed back correctly
3793 from remote delivery subprocesses. The field in question was not being
3794 subsequently used, so this caused to problems in practice.
3796 13. Added new log selectors queue_time and deliver_time.
3798 14. Might have fixed a bug in maildirsizefile handling that threw up
3799 "unexpected character" debug warnings, and recalculated the data
3800 unnecessarily. In any case, I expanded the warning message to give more
3803 15. Added the message "Restricted characters in address" to the statements in
3804 the default ACL that block characters like @ and % in local parts.
3806 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3807 Three changes have been made:
3809 (a) There was a serious bug; a negative response to MAIL caused the whole
3810 recipient domain to be cached as invalid, thereby blocking all messages
3811 to all local parts at the same domain, from all senders. This bug has
3812 been fixed. The domain is no longer cached after a negative response to
3813 MAIL if the sender used is not empty.
3815 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3818 (c) A new callout option, "use_sender" has been added for people who want
3819 the modified behaviour.
3825 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3828 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3829 indeed breaks things for older releases.
3831 3. Added additional logging to the case where there is a problem reading data
3832 from a filter that is running in a subprocess using a pipe, in order to
3833 try to track down a specific problem.
3835 4. Testing facility fudge: when running in the test harness and attempting
3836 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3837 getting "No route to host". Convert this to a timeout.
3839 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3842 6. Some OS don't have socklen_t but use size_t instead. This affects the
3843 fifth argument of getsockopt() amongst other things. This is now
3844 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3845 can be set for individual OS. I have set it for SunOS5, OSF1, and
3846 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3847 some earlier ones do not.
3849 7. Change 4.30/15 was not doing the test caselessly.
3851 8. The standard form for an IPv6 address literal was being rejected by address
3852 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3853 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3854 this, as well as the form without the "IPv6" on the front (but only when
3855 address literals are enabled, of course).
3857 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3859 10. Exim crashed if a message with an empty sender address specified by -f
3860 encountered a router with an errors_to setting. This could be provoked only
3861 by a command such as
3865 where an empty string was supplied; "<>" did not hit this bug.
3867 11. Installed PCRE release 4.5.
3869 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3870 remained set. It is now erased.
3872 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3873 times from message ids (which are base 36 rather than the normal 62).
3875 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3876 were being counted as actual protocol errors, and logged if the log
3877 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3878 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3879 and DATA following a set of rejected RCPTs do not count as protocol errors.
3880 In other words, Exim assumes they were pipelined, though this may not
3881 actually be the case. Of course, in all cases the client gets an
3882 appropriate error code.
3884 15. If a lookup fails in an ACL condition, a message about the failure may
3885 be available; it is used if testing the ACL cannot continue, because most
3886 such messages specify what the cause of the deferral is. However, some
3887 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3888 that caused an old message to be retained and used if a later statement
3889 caused a defer, replacing the real cause of the deferral.
3891 16. If an IP address had so many PTR records that the DNS lookup buffer
3892 was not large enough to hold them, Exim could crash while trying to process
3893 the truncated data. It now detects and logs this case.
3895 17. Further to 4.21/58, another change has been made: if (and only if) the
3896 first line of a message (the first header line) ends with CRLF, a bare LF
3897 in a subsequent header line has a space inserted after it, so as not to
3898 terminate the header.
3900 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3901 unnecessarily, used atoi() instead of strtol(), and didn't check the
3902 termination when getting file sizes from file names by regex.
3904 19. Completely re-implemented the support for maildirsize files, in the light
3905 of a number of problems with the previous contributed implementation
3906 (4.30/29). In particular:
3908 . If the quota is zero, the maildirsize file is maintained, but no quota is
3911 . If the maildir directory does not exist, it is created before any attempt
3912 to write a maildirsize file.
3914 . The quota value in the file is just a cache; if the quota is changed in
3915 the transport, the new value overrides.
3917 . A regular expression is available for excluding directories from the
3920 20. The autoreply transport checks the characters in options that define the
3921 message's headers; it allows continued headers, but it was checking with
3922 isspace() after an embedded newline instead of explicitly looking for a
3925 21. If all the "regular" hosts to which an address was routed had passed their
3926 expiry times, and had not reached their retry times, the address was
3927 bounced, even if fallback hosts were defined. Now Exim should go on to try
3930 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3931 equivalent code in the SMTP transport. Some hosts send humungous responses
3932 to HELO/EHLO, more than 1024 it seems.
3934 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3935 gives compiler warnings in some environments. I've now done it "properly",
3938 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3939 (because of problems with the built-in one) was declared to return uschar *
3940 instead of char *, causing compiler failure.
3942 25. Fixed a file descriptor leak when processing alias/forward files.
3944 26. Fixed a minor format string issue in dbfn.c.
3946 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3948 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3949 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3952 29. When a sender address is verified, it is cached, to save repeating the test
3953 when there is more than one recipient in a message. However, when the
3954 verification involves a callout, it is possible for different callout
3955 options to be set for different recipients. It is too complicated to keep
3956 track of this in the cache, so now Exim always runs a verification when a
3957 callout is required, relying on the callout cache for the optimization.
3958 The overhead is duplication of the address routing, but this should not be
3961 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3962 to be verified with callout=postmaster, and the main callout worked but the
3963 postmaster check failed, the verification correctly failed. However, if a
3964 subsequent RCPT command asked for sender verification *without* the
3965 postmaster check, incorrect caching caused this verification also to fail,
3968 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3969 it was not caching the DNS options (qualify_single, search_parents) that
3970 were used when the lookup failed. A subsequent lookup with different
3971 options therefore always gave the same answer, though there were cases
3972 where it should not have. (Example: a "domains = !$mx_any" option on a
3973 dnslookup router: the "domains" option is always processed without any
3974 widening, but the router might have qualify_single set.) Now Exim uses the
3975 cached value only when the same options are set.
3977 32. Added John Jetmore's "exipick" utility to the distribution.
3979 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3980 than a timeout (e.g. a certificate is required, and is not provided), an
3981 Exim server now closes the connection immediately. Previously it waited for
3982 the client to close - but if the client is SSL, it seems that they each
3983 wait for each other, leading to a delay before one of them times out.
3985 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3986 maintained 0.8.x compatibility because I don't think many are using it, and
3987 it is clearly obsolete.
3989 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3992 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3993 client certificate was expired. A simple patch fixes this, though I don't
3994 understand the full logic of why the verify callback is called multiple
3997 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3998 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3999 which causes problems with some clients (such as the Certicom SSL Plus
4000 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4001 disables the coutermeasure allowing Eudora to connect."
4003 38. Exim was not checking that a write() to a log file succeeded. This could
4004 lead to Bad Things if a log got too big, in particular if it hit a file
4005 size limit. Exim now panics and dies if it cannot write to a log file, just
4006 as it does if it cannot open a log file.
4008 39. Modified OS/Makefile-Linux so that it now contains
4010 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4012 The two -D definitions ensure that Exim is compiled with large file
4013 support, which makes it possible to handle log files that are bigger than
4016 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4017 instance) a domain was checked against a named list that involved a lookup,
4018 causing $domain_data to be set, then another domain was checked against the
4019 same list, then the first domain was re-checked, the value of $domain_data
4020 after the final check could be wrong. In particular, if the second check
4021 failed, it could be set empty. This bug probably also applied to
4024 41. The strip_trailing_dot option was not being applied to the address given
4025 with the -f command-line option.
4027 42. The code for reading a message's header from the spool was incrementing
4028 $received_count, but never initializing it. This meant that the value was
4029 incorrect (doubled) while delivering a message in the same process in which
4030 it was received. In the most common configuration of Exim, this never
4031 happens - a fresh exec is done - but it can happen when
4032 deliver_drop_privilege is set.
4034 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4035 it now includes up to 150 characters of the unexpected data in the log
4038 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4039 and building data strings. The size of both of these buffers was 10 000
4040 bytes - far larger than anybody would *ever* want, thought I. Needless to
4041 say, somebody hit the limit. I have increased the maximum line length to
4042 20 000 and the maximum data length of concatenated lines to 100 000. I have
4043 also fixed two bugs, because there was no checking on these buffers. Tsk,
4044 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4045 buffer is too small.
4047 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4048 lsearch lookups. Now it does.
4050 46. When parsing a route_list item in a manualroute router, a fixed-length
4051 buffer was used for the list of hosts. I made this 1024 bytes long,
4052 thinking that nobody would ever have a list of hosts that long. Wrong.
4053 Somebody had a whole pile of complicated expansion conditions, and the
4054 string was silently truncated, leading to an expansion error. It turns out
4055 that it is easier to change to an unlimited length (owing to other changes
4056 that have happened since this code was originally written) than to build
4057 structure for giving a limitation error. The length of the item that
4058 expands into the list of hosts is now unlimited.
4060 47. The lsearch lookup could not handle data where the length of text line was
4061 more than 4095 characters. Such lines were truncated, leading to shortened
4062 data being returned. It should now handle lines of any length.
4064 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4065 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4068 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4069 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4070 no longer happens in comment lines. A list of replacements is now placed
4071 at the head of all of the source files, except those whose only change is
4072 to replace PERL_COMMAND in the very first #! line.
4074 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4075 messages on the queue, with a bottom-up merge sort, using code contributed
4076 by Michael Haardt. This should make operations like -bp somewhat faster on
4077 large queues. It won't affect queue runners, except when queue_run_in_order
4080 51. Installed eximstats 1.31 in the distribution.
4082 52. Added support for SRV lookups to the dnslookup router.
4084 53. If an ACL referred to $message_body or $message_body_end, the value was not
4085 reset for any messages that followed in the same SMTP session.
4087 54. The store-handling optimization for building very long strings was not
4088 differentiating between the different store pools. I don't think this
4089 actually made any difference in practice, but I've tidied it.
4091 55. While running the routers to verify a sender address, $sender_address
4092 was still set to the sender address. This is wrong, because when routing to
4093 send a bounce to the sender, it would be empty. Therefore, I have changed
4094 it so that, while verifying a sender address, $sender_address is set to <>.
4095 (There is no change to what happens when verifying a recipient address.)
4097 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4098 target A or AAAA records (if not already returned) without resetting the
4099 qualify_single or search_parents options of the DNS resolver. These are
4100 inappropriate in this case because the targets of MX and SRV records must
4101 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4102 target that, when qualified, matched something in the local domain. These
4103 two options are now turned off when doing these lookups.
4105 57. It seems that at least some releases of Reiserfs (which does not have the
4106 concept of a fixed number of inodes) returns zero and not -1 for the
4107 number of available inodes. This interacted badly with check_spool_inodes,
4108 which assumed that -1 was the "no such thing" setting. What I have done is
4109 to check that the total number of inodes is greater than zero before doing
4110 the test of how many are available.
4112 58. When a "warn" ACL statement has a log_message modifier, the message is
4113 remembered, and not repeated. This is to avoid a lot of repetition when a
4114 message has many recipients that cause the same warning to be written.
4115 Howewer, Exim was preserving the list of already written lines for an
4116 entire SMTP session, which doesn't seem right. The memory is now reset if a
4117 new message is started.
4119 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4120 debugging output unless log_rewrite was also set.
4122 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4123 of (int)(handle) when we know that handle contains (void *)(-1).
4125 61. The Exim daemon panic-logs an error return when it closes the incoming
4126 connection. However "connection reset by peer" seems to be common, and
4127 isn't really an error worthy of noting specially, so that particular error
4130 62. When Exim is trying to find all the local interfaces, it used to panic and
4131 die if the ioctl to get the interface flags failed. However, it seems that
4132 on at least one OS (Solaris 9) it is possible to have an interface that is
4133 included in the list of interfaces, but for which you get a failure error
4134 for this call. This happens when the interface is not "plumbed" into a
4135 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4136 failure of the "get flags" call assumes that the interface is down.
4138 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4139 makes life easier for people who are doing arithmetic on fields extracted
4140 from dates, where you often get leading zeros that should not be
4141 interpreted as octal.
4143 64. Added qualify_domain to the redirect router, to override the global
4146 65. If a pathologically long header line contained very many addresses (the
4147 report of this problem mentioned 10 000) and each of them was rewritten,
4148 Exim could use up a very large amount of memory. (It kept on making new
4149 copies of the header line as it rewrote, and never released the old ones.)
4150 At the expense of a bit more processing, the header rewriting function has
4151 been changed so that it no longer eats memory in this way.
4153 66. The generation of the Received: header has been moved from the time that a
4154 message starts to be received, to the time that it finishes. The timestamp
4155 in the Received: header should now be very close to that of the <= log
4156 line. There are two side-effects of this change:
4158 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4159 logged header lines no longer include the local Received: line, because
4160 it has not yet been created. The same applies to a copy of the message
4161 that is returned to a non-SMTP sender when a message is rejected.
4163 (b) When a filter file is tested using -bf, no additional Received: header
4164 is added to the test message. After some thought, I decided that this
4167 This change does not affect the value of $received_for. It is still set
4168 after address rewriting, but before local_scan() is called.
4170 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4172 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4173 gave an unhelpful panic error message, and a defer error. I have managed to
4174 change this behaviour so that it now rejects any supplied certificate,
4175 which seems right, as the list of acceptable certificates is empty.
4177 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4178 gave an unhelpful defer error. I have not managed to make this reject any
4179 supplied certificates, but the error message it gives is "no certificate
4180 supplied", which is not helpful.
4182 70. exigrep's output now also includes lines that are not associated with any
4183 message, but which match the given pattern. Implemented by a patch from
4184 Martin Sluka, which also tidied up the Perl a bit.
4186 71. Recipient callout verification, like sender verification, was using <> in
4187 the MAIL FROM command. This isn't really the right thing, since the actual
4188 sender may affect whether the remote host accepts the recipient or not. I
4189 have changed it to use the actual sender in the callout; this means that
4190 the cache record is now keyed on a recipient/sender pair, not just the
4191 recipient address. There doesn't seem to be a real danger of callout loops,
4192 since a callout by the remote host to check the sender would use <>.
4193 [SEE ABOVE: changed after hitting problems.]
4195 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4196 temporary errors. However, in the case of such a code being given after
4197 the end of a data transmission (i.e. after ".") Exim was failing to write
4198 a retry record for the message. (Yes, there was some broken host that was
4199 actually sending 8xx at this point.)
4201 73. An unknown lookup type in a host list could cause Exim to panic-die when
4202 the list was checked. (An example that provoked this was putting <; in the
4203 middle of a list instead of at the start.) If this happened during a DATA
4204 ACL check, a -D file could be left lying around. This kind of configuration
4205 error no longer causes Exim to die; instead it causes a defer errror. The
4206 incident is still logged to the main and panic logs.
4208 74. Buglet left over from Exim 3 conversion. The message "too many messages
4209 in one connection" was written to the rejectlog but not the mainlog, except
4210 when address rewriting (yes!) was being logged.
4212 75. Added write_rejectlog option.
4214 76. When a system filter was run not as root (that is, when system_filter_user
4215 was set), the values of the $n variables were not being returned to the
4216 main process; thus, they were not subsequently available in the $sn
4219 77. Added +return_path_on_delivery log selector.
4221 78. A connection timeout was being treated differently from recipients deferred
4222 when testing hosts_max_try with a message that was older than the host's
4223 retry timeout. (The host should not be counted, thus allowing all hosts to
4224 be tried at least once before bouncing.) This may have been the cause of an
4225 occasionally reported bug whereby a message would remain on the queue
4226 longer than the retry timeout, but would be bounced if a delivery was
4227 forced. I say "may" because I never totally pinned down the problem;
4228 setting up timeout/retry tests is difficult. See also the next item.
4230 79. The ultimate address timeout was not being applied to errors that involved
4231 a combination of host plus message (for example, a timeout on a MAIL
4232 command). When an address resolved to a number of possible hosts, and they
4233 were not all tried for each delivery (e.g. because of hosts_max_try), a
4234 message could remain on the queue longer than the retry timeout.
4236 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4239 81. Fixed an obscure SMTP outgoing bug which required at least the following
4240 conditions: (a) there was another message waiting for the same server;
4241 (b) the server returned 5xx to all RCPT commands in the first message so
4242 that the message was not completed; (c) the server dropped the connection
4243 or gave a negative response to the RSET that Exim sends to abort the
4244 transaction. The observed case was a dropped connection after DATA that had
4245 been sent in pipelining mode. That is, the server had advertised PIPELINING
4246 but was not implementing it correctly. The effect of the bug was incorrect
4247 behaviour, such as trying another host, and this could lead to a crash.
4253 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4254 and daemon.c were passed as pointers to ints; they should have been
4255 pointers to socklen_t variables (which are typically unsigned ints).
4257 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4260 3. Fixed a really odd bug that affected only the testing scheme; patching a
4261 certain fixed string in the binary changed the value of another string that
4262 happened to be identical to the end of the original first string.
4264 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4265 name", it returns that address as the IP address. On some operating
4266 systems (e.g. Solaris), it also passes back the IP address string as the
4267 "host name". However, on others (e.g. Linux), it passes back an empty
4268 string. Exim wasn't checking for this, and was changing the host name to an
4269 empty string, assuming it had been canonicized.
4271 5. Although rare, it is permitted to have more than one PTR record for a given
4272 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4273 all the names associated with an address, because they do in Solaris.
4274 However, it seems that they do not in Linux for data that comes from the
4275 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4276 I found this out when I moved to a new Linux workstation and tried to run
4277 the Exim test suite.
4279 To get round this problem I have changed the code so that it now does its
4280 own call to the DNS to look up PTR records when searching for a host name.
4281 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4282 addresses that are only in /etc/hosts are still found.
4284 This behaviour is, however, controlled by an option called host_lookup_
4285 order, which defaults to "bydns:byaddr". If people want to use the other
4286 order, or indeed, just use one or the other means of lookup, they can
4287 specify it in this variable.
4289 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4290 some operating systems, this comes back from gethostbyaddr() as an empty
4291 string, and this is what Exim used to test for. However, it seems that in
4292 other systems, "." is yielded. Exim now tests for this case too.
4294 7. The values of check_spool_space and check_log_space are now held internally
4295 as a number of kilobytes instead of an absolute number of bytes. If a
4296 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4297 kilobyte. This means that much larger values can be stored.
4299 8. Exim monitor: an attempt to get the action menu when not actually pointing
4300 at a message produces an empty menu entitled "No message selected". This
4301 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4302 no entries in it ("Shell widget menu has zero width and/or height"). So I
4303 have added a single, blank menu entry in this case.
4305 9. Added ${quote_local_part.
4307 10. MIME decoding is now applied to the contents of Subject: header lines when
4310 11. Now that a reference to $sender_host_address automatically causes a reverse
4311 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4312 host lookup before query-style lookups in lists that might use this
4313 variable. This has therefore been abolished, and the "net-" prefix is no
4314 longer necessary for query-style lookups.
4316 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4317 have been a typo for LFLAGS, so it has been changed.
4319 13. The install script calls Exim with "-C /dev/null" in order to find the
4320 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4321 to be output. Howeve, since Exim outputs its version number before the
4322 error, it didn't break the script. It just looked ugly. I fixed this by
4323 always allowing "-C /dev/null" if the caller is root.
4325 14. Ignore overlarge ACL variable number when reading spool file - insurance
4326 against a later release with more variables having written the file.
4328 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4329 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4330 this, as well as the form without the "IPv6" on the front.
4332 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4333 OS/Makefile-Darwin file.
4335 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4336 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4338 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4339 variable. However, some OS use ints for inodes. I've added cast to long int
4340 to get rid of the compiler warning.
4342 19. I had forgotten to lock out "/../" in configuration file names when
4343 ALT_CONFIG_PREFIX was set.
4345 20. Routers used for verification do not need to specify transports. However,
4346 if such a router generated a host list, and callout was configured, Exim
4347 crashed, because it could not find a port number from the (non-existent)
4348 transport. It now assumes port 25 in this circumstance.
4350 21. Added the -t option to exigrep.
4352 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4353 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4354 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4355 them is set, LOOKUP_LSEARCH is forced.
4357 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4359 24. Added sender and host information to the "rejected by local_scan()" log
4360 line; previously there was no indication of these.
4362 25. Added .include_if_exists.
4364 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4365 Linux. It turns out that not all file systems support this. Apparently some
4366 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4367 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4368 not supported on the file descriptor, is now ignored when Exim is trying to
4369 sync a directory. This applies only to Linux.
4371 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4373 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4374 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4377 29. Added support for maildirsize files from supplied patch (modified a bit).
4379 30. The use of :fail: followed by an empty string could lead Exim to respond to
4380 sender verification failures with (e.g.):
4382 550 Verification failed for <xxx>
4383 550 Sender verify failed
4385 where the first response line was missing the '-' that indicates it is not
4386 the final line of the response.
4388 31. The loop for finding the name of the user that called Exim had a hardwired
4389 limit of 10; it now uses the value of finduser_retries, which is used for
4390 all other user lookups.
4392 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4395 33. Exim was neglecting to zero errno before one call of strtol() when
4396 expanding a string and expecting an integer value. On some systems this
4397 resulted in spurious "integer overflow" errors. Also, it was casting the
4398 result into an int without checking.
4400 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4401 did not work. The code looks as if it has *never* worked, though it appears
4402 to have been documented since at least releast 1.62. I have made it work.
4404 35. The "timeout_DNS" error in retry rules, also documented since at least
4405 1.62, also never worked. As it isn't clear exactly what this means, and
4406 clearly it isn't a major issue, I have abolished the feature by treating it
4407 as "timeout", and writing a warning to the main and panic logs.
4409 36. The display of retry rules for -brt wasn't always showing the error code
4412 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4413 timeout_connect_A, timeout_connect_MX.
4415 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4416 to the empty sender.
4418 39. The daemon was not analysing the content of -oX till after it had closed
4419 stderr and disconnected from the controlling terminal. This meant that any
4420 syntax errors were only noted on the panic log, and the return code from
4421 the command was 0. By re-arranging the code a little, I've made the
4422 decoding happen first, so such errors now appear on stderr, and the return
4423 code is 1. However, the actual setting up of the sockets still happens in
4424 the disconnected process, so errors there are still only recorded on the
4427 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4428 connections (as happens on some IP stacks) was logged at start up time as
4429 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4430 it from "IPv6 and IPv4", which means that two separate sockets are being
4433 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4434 says whether AF_INET or AF_INET6 was passed as an argument.
4436 42. Exiwhat output was messed up when time zones were included in log
4439 43. Exiwhat now gives more information about the daemon's listening ports,
4440 and whether -tls-on-connect was used.
4442 44. The "port" option of the smtp transport is now expanded.
4444 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4445 silently ignored. Now an error message is written to the main and panic
4448 46. There's a new ACL modifier called "logwrite" which writes to a log file
4449 as soon as it is encountered.
4451 47. Added $local_user_uid and $local_user_gid at routing time.
4453 48. Exim crashed when trying to verify a sender address that was being
4456 49. Exim was recognizing only a space character after ".include". It now also
4457 recognizes a tab character.
4459 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4460 extracting the relevant information from the specification. The man page no
4461 longer contains scrambled data for the -d option, and I've added a section
4462 at the front about calling Exim under different names.
4464 51. Added "extra_headers" argument to the "mail" command in filter files.
4466 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4469 53. Installed eximstats 1.29.
4471 54. Added transport_filter_timeout as a generic transport option.
4473 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4474 Cc: header lines. This was required by RFC 822, but it not required by RFC
4477 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4478 incoming messages that did not have them. Now it does so only if the
4479 message originates locally, that is, if there is no associated remote host
4480 address. When Resent- header lines are present, this applies to the Resent-
4481 lines rather than the non-Resent- lines.
4483 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4484 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4486 58. Messages for configuration errors now include the name of the main
4487 configuration file - useful now that there may be more than one file in a
4488 list (.included file names were always shown).
4490 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4491 for those rare installations that do not start the daemon as root or run it
4492 setuid root. I've cut out the call to initgroups() if the daemon is not
4495 60. The Exim user and group can now be bound into the binary as text strings
4496 that are looked up at the start of Exim's processing.
4498 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4500 62. Added $mailstore_basename variable.
4502 63. Installed patch to sieve.c from Michael Haardt.
4504 64. When Exim failed to open the panic log after failing to open the main log,
4505 the original message it was trying to log was written to stderr and debug
4506 output, but if they were not available (the usual case in production), it
4507 was lost. Now it is written to syslog before the two lines that record the
4508 failures to open the logs.
4510 65. Users' Exim filters run in subprocesses under the user's uid. It is
4511 possible for a "deliver" command or an alias in a "personal" command to
4512 provoke an address rewrite. If logging of address rewriting is configured,
4513 this fails because the process is not running as root or exim. There may be
4514 a better way of dealing with this, but for the moment (because 4.30 needs
4515 to be released), I have disabled address rewrite logging when running a
4516 filter in a non-root, non-exim process.
4522 1. The buildconfig auxiliary program wasn't quoting the value set for
4523 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4524 not defined. This bug was masked in 4.22 by the effect that was fixed in
4527 2. Some messages that were rejected after a message id was allocated were
4528 shown as "incomplete" by exigrep. It no longer does this for messages that
4529 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4531 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4532 have allow_domain_literals set, the ID did not get logged in the <= line.
4533 Domain literals are now always recognized in Message-ID: header lines.
4535 4. The first argument for a ${extract expansion item is the key name or field
4536 number. Leading and trailing spaces in this item were not being ignored,
4537 causing some misleading effects.
4539 5. When deliver_drop_privilege was set, single queue runner processes started
4540 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4541 same command in the process it spins off) were not dropping privilege.
4543 6. When the daemon running as "exim" started a queue runner, it always
4544 re-executed Exim in the spun-off process. This is a waste of effort when
4545 deliver_drop_privilege is set. The new process now just calls the
4546 queue-runner function directly.
4552 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4555 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4556 ignored. Though the use of -oP was forcing the writing of a pid file, it
4557 was always written to the default place.
4559 3. If the message "no IP address found for host xxxx" is generated during
4560 incoming verification, it is now followed by identification of the incoming
4561 connection (so you can more easily find what provoked it).
4563 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4565 5. Added some features to "harden" Exim a bit more against certain attacks:
4567 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4568 be put in Local/Makefile. This is like the never_users runtime option,
4569 but it cannot be overridden. The default setting is "root".
4571 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4572 prefix string with which any file named in a -C command line option
4575 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4576 is retained for -C and -D only if the caller of Exim is root. Without
4577 it, the exim user may also use -C and -D and retain privilege.
4579 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4580 command line option is disabled.
4582 6. Macro names set by the -D option must start with an upper case letter, just
4583 like macro names defined in the configuration file.
4585 7. Added "dereference=" facility to LDAP.
4587 8. Two instances of the typo "uknown" in the source files are fixed.
4589 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4590 the Configure-Makefile script screwed up while processing it.
4592 10. Incorporated PCRE 4.4.
4594 11. The SMTP synchronization check was not operating right at the start of an
4595 SMTP session. For example, it could not catch a HELO sent before the client
4596 waited for the greeting. There is now a check for outstanding input at the
4597 point when the greeting is written. Because of the duplex, asynchronous
4598 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4599 way, but not yet received, when the check is performed.
4601 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4602 on TCP/IP sockets, because this apparently causes some broken clients to
4605 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4606 unchanged) from the Cygwin maintainer.
4608 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4609 was supported instead of testing for mbx. Effectively a typo.
4611 15. The spa authenticator server code was not checking that the input it
4612 received was valid base64.
4614 16. The debug output line for the "set" modifier in ACLs was not showing the
4615 name of the variable that was being set.
4617 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4619 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4620 lookup on its own. Something else was needed to trigger the lookup. For
4621 example, a match in host_lookup or the need for a host name in a host list.
4622 Now, if $sender_host_name is referenced and the host name has not yet been
4623 looked up, a lookup is performed. If the lookup fails, the variable remains
4624 empty, and $host_lookup_failed is set to "1".
4626 19. Added "eqi" as a case-independent comparison operator.
4628 20. The saslauthd authentication condition could segfault if neither service
4629 nor realm was specified.
4631 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4632 error message that was logged was misleading, and incoming SMTP
4633 connections were dropped. The message is now more accurate, and temporary
4634 errors are given to SMTP connections.
4636 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4637 (except RSET) with a 421 error, until QUIT is received. However, it was
4638 failing to send a response to QUIT.
4640 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4641 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4642 ACL is now run after the helo_try_verify_hosts code.
4644 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4645 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4646 case-independent, so other case variants are also recognized). Apparently
4647 some systems use these upper case variants.
4649 25. If more than two messages were waiting for the same host, and a transport
4650 filter was specified for the transport, Exim sent two messages over the
4651 same TCP/IP connection, and then failed with "socket operation on non-
4652 socket" when it tried to send the third.
4654 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4656 27. The extern definition of crypt16() in expand.c was not being excluded when
4657 the OS had its own crypt16() function.
4659 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4660 as a preferred synonym for return_size_limit, both as an option and as an
4663 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4665 30. Changed the default configuration ACL to relax the local part checking rule
4666 for addresses that are not in any local domains. For these addresses,
4667 slashes and pipe symbols are allowed within local parts, but the sequence
4668 /../ is explicitly forbidden.
4670 31. SPA server authentication was not clearing the challenge buffer before
4673 32. log_message in a "warn" ACL statement was writing to the reject log as
4674 well as to the main log, which contradicts the documentation and doesn't
4675 seem right (because no rejection is happening). So I have stopped it.
4677 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4678 However, I am unable to do any testing of this.
4680 34. Fixed an infelicity in the appendfile transport. When checking directories
4681 for a mailbox, to see if any needed to be created, it was accidentally
4682 using path names with one or more superfluous leading slashes; tracing
4683 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4685 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4686 contents are added to the log line that is written for every discarded
4687 recipient. (Previously a log_message setting was ignored.)
4689 36. The ${quote: operator now quotes the string if it is empty.
4691 37. The install script runs exim in order to find its version number. If for
4692 some reason other than non-existence or emptiness, which it checks, it
4693 could not run './exim', it was installing it with an empty version number,
4694 i.e. as "exim-". This error state is now caught, and the installation is
4697 38. An argument was missing from the function that creates an error message
4698 when Exim fails to connect to the socket for saslauthd authentication.
4699 This could cause Exim to crash, or give a corrupted message.
4701 39. Added isip, isip4, and isip6 to ${if conditions.
4703 40. The ACL variables $acl_xx are now saved with the message, and can be
4704 accessed later in routers, transports, and filters.
4706 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4707 strings in the file are not string-expanded.
4709 42. If a MAIL command specified a SIZE value that was too large to fit into an
4710 int variable, the check against message_size_limit failed. Such values are
4711 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4712 day this will have to be increased, but I don't think I want to be around
4713 when emails are that large.
4720 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4721 iconv() is not standard in FreeBSD.
4723 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4724 IPv6 enabled. The observed symptom was a segmentation fault on return from
4725 the function os_common_find_running_interfaces() in src/os.c.
4727 3. In the check_special_case() function in daemon.c I had used "errno" as an
4728 argument name, which causes warnings on some systems. This was basically a
4729 typo, since it was named "eno" in the comments!
4731 4. The code that waits for the clock to tick (at a resolution of some fraction
4732 of a second) so as to ensure message-id uniqueness was always waiting for
4733 at least one whole tick, when it could have waited for less. [This is
4734 almost certainly not relevant at current processor speeds, where it is
4735 unlikely to ever wait at all. But we try to future-proof.]
4737 5. The function that sleeps for a time interval that includes fractions of a
4738 second contained a race. It did not block SIGALRM between setting the
4739 timer, and suspending (a couple of lines later). If the interval was short
4740 and the sigsuspend() was delayed until after it had expired, the suspension
4741 never ended. On busy systems this could lead to processes getting stuck for
4744 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4745 process, before it forks any delivery processes. The open lookup caching
4746 mechanism meant that the open file or database connection was passed into
4747 the delivery process. The problem was that delivery processes always tidy
4748 up cached lookup data. This could cause a problem for the next delivery
4749 process started by the queue runner, because the external queue runner
4750 process does not know about the closure. So the next delivery process
4751 still has data in the lookup cache. In the case of a file lookup, there was
4752 no problem because closing a file descriptor in a subprocess doesn't affect
4753 the parent. However, if the lookup was caching a connection to a database,
4754 the connection was closed, and the second delivery process was likely to
4755 see errors such as "PGSQL: query failed: server closed the connection
4756 unexpectedly". The problem has been fixed by closing all cached lookups
4757 in a queue runner before running a delivery process.
4759 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4760 seem to have the "const" qualifier which it has on other OS. I've
4763 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4764 *of the same type* (client or server) with the same public name that an
4765 error should be diagnosed.
4767 9. When Exim looked up a host name for an IP address, but failed to find the
4768 original IP address when looking up the host name (a safety check), it
4769 output the message "<ip address> does not match any IP for NULL", which was
4770 confusing, to say the least. The bug was that the host name should have
4771 appeared instead of "NULL".
4773 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4774 user that is built into the binary, and the -C or -D options is used, root
4775 privilege is dropped before the configuration file is read. In addition,
4776 logging is switched to stderr instead of the normal log files. If the
4777 configuration then re-defines the Exim user, the unprivileged environment
4778 is probably not what is expected, so Exim logs a panic warning message (but
4781 However, if deliver_drop_privilege is set, the unprivileged state may well
4782 be exactly what is intended, so the warning has been cut out in that case,
4783 and Exim is allowed to try to write to its normal log files.
4789 1. smtp_return_error_details was not giving details for temporary sender
4790 or receiver verification errors.
4792 2. Diagnose a configuration error if two authenticators have the same public
4795 3. Exim used not to create the message log file for a message until the first
4796 delivery attempt. This could be confusing when incoming messages were held
4797 for policy or load reasons. The message log file is now created at the time
4798 the message is received, and an initial "Received" line is written to it.
4800 4. The automatically generated man page for command line options had a minor
4801 bug that caused no ill effects; however, a more serious problem was that
4802 the procedure for building the man page automatically didn't always
4803 operate. Consequently, release 4.20 contains an out-of-date version. This
4804 shouldn't happen again.
4806 5. When building Exim with embedded Perl support, the script that builds the
4807 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4808 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4810 6. The freeze_tell option was not being used for messages that were frozen on
4811 arrival, either by an ACL or by local_scan().
4813 7. Added the smtp_incomplete_transaction log selector.
4815 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4816 was accepting AUTH without a new EHLO.
4818 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4819 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4822 10. Logging of TCP/IP connections (when configured) now happens in the main
4823 daemon process instead of the child process, so that the TCP/IP connection
4824 count is more accurate (but it can never be perfect).
4826 11. The use of "drop" in a nested ACL was not being handled correctly in the
4827 outer ACL. Now, if condition failure induced by the nested "drop" causes
4828 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4829 or "require"), the connection is dropped.
4831 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4832 that yield "discard" can now be used with an "accept" or a "discard" verb,
4833 but an error is generated for any others (because I can't see a useful way
4834 to define what should happen).
4836 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4837 are now processed in the same way as lines in the Exim configuration file.
4838 In particular, continuation lines are supported.
4840 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4842 15. Added -ti meaning -t -i.
4844 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4845 domains, and warn by logging if others are found.
4847 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4848 returned from GIFCONF when Exim is trying to find the list of interfaces on
4849 a host. The code in os.c has been modified to copy each ifreq to an aligned
4850 structure in all cases.
4852 Also, in some cases, the returned ifreq's were being copied to a 'struct
4853 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4854 means the last couple of bytes of an IPv6 address could be chopped if the
4855 ifreq contained only a normal sockaddr (14 bytes storage).
4857 18. Named domain lists were not supported in the hosts_treat_as_local option.
4858 An entry such as +xxxx was not recognized, and was treated as a literal
4861 19. Ensure that header lines added by a DATA ACL are included in the reject log
4862 if the ACL subsequently rejects the message.
4864 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4865 MD5 (which is deprecated).
4867 21. When testing a filter file using -bf, Exim was writing a message when it
4868 took the sender from a "From " line in the message, but it was not doing so
4869 when it took $return_path from a Return-Path: header line. It now does.
4871 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4872 with a valid header line field name (a series of printing characters
4873 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4875 23. Changed "disc" in the source to "disk" to conform to the documentation and
4876 the book and for uniformity.
4878 24. Ignore Sendmail's -Ooption=value command line item.
4880 25. When execve() failed while trying to run a command in a pipe transport,
4881 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4882 could be confused with a return value of 69 from the command itself. This
4883 has been changed to 127, the value the shell returns if it is asked to run
4884 a non-existent command. The wording for the related log line suggests a
4885 non-existent command as the problem.
4887 26. If received_header_text expands to an empty string, do not add a Received:
4888 header line to the message. (Well, it adds a token one on the spool, but
4889 marks it "old" so that it doesn't get used or transmitted.)
4891 27. Installed eximstats 1.28 (addition of -nt option).
4893 28. There was no check for failure on the call to getsockname() in the daemon
4894 code. This can fail if there is a shortage of resources on the system, with
4895 ENOMEM, for example. A temporary error is now given on failure.
4897 29. Contrary to the C standard, it seems that in some environments, the
4898 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4899 program. Exim now does this explicitly; it affects the formatting of
4900 timestamps using strftime().
4902 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4903 complaints. I've now initialized all the variables, to avoid this.
4905 32. Header lines added by a system filter were not being "seen" during
4906 transport-time rewrites.
4908 33. The info_callback() function passed to OpenSSL is set up with type void
4909 (*)(SSL *, int, int), as described somewhere. However, when calling the
4910 function (actually a macro) that sets it up, the type void(*)() is
4911 expected. I've put in a cast to prevent warnings from picky compilers.
4913 34. If a DNS black list lookup found a CNAME record, but there were no A
4914 records associated with the domain it pointed at, Exim crashed.
4916 35. If a DNS black list lookup returned more than one A record, Exim ignored
4917 all but the first. It now scans all returned addresses if a particular IP
4918 value is being sought. In this situation, the contents of the
4919 $dnslist_value variable are a list of all the addresses, separated by a
4922 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4923 to accept a host name and all its aliases if the forward lookup for any of
4924 them yielded the IP address of the incoming connection. Now it accepts only
4925 those names whose forward lookup yields the correct IP address. Any other
4926 names are discarded. This closes a loophole whereby a rogue DNS
4927 administrator could create reverse DNS records to break through a
4928 wildcarded host restriction in an ACL.
4930 37. If a user filter or a system filter that ran in a subprocess used any of
4931 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4932 the wrong values were passed to the pipe command ($thisaddress had the
4933 value of $0, $0 had the value of $1, etc). This bug was introduced by
4934 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4936 38. Improved the line breaking for long SMTP error messages from ACLs.
4937 Previously, if there was no break point between 40 and 75 characters, Exim
4938 left the rest of the message alone. Two changes have been made: (a) I've
4939 reduced the minimum length to 35 characters; (b) if it can't find a break
4940 point between 35 and 75 characters, it looks ahead and uses the first one
4941 that it finds. This may give the occasional overlong line, but at least the
4942 remaining text gets split now.
4944 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4945 file descriptors might be low, and that setting 1000 would always raise it.
4946 It turns out that in some environments, the limit is already over 1000 and
4947 that lowering it causes trouble. So now Exim takes care not to decrease it.
4949 40. When delivering a message, the value of $return_path is set to $sender_
4950 address at the start of routing (routers may change the value). By an
4951 oversight, this default was not being set up when an address was tested by
4952 -bt or -bv, which affected the outcome if any router or filter referred to
4955 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4956 log only when "message" or "log_message" are set. However, if one of the
4957 conditions was an address verification, or a call to a nested ACL, the
4958 messages generated by the underlying test were being passed through. This
4959 no longer happens. The underlying message is available in $acl_verify_
4960 message for both "message" and "log_message" expansions, so it can be
4961 passed through if needed.
4963 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4964 new expansion $bh_ to give the encoded byte string without charset
4965 translation. Translation happens only if iconv() is available; HAVE_ICONV
4966 indicates this at build time. HEADERS_CHARSET gives the charset to
4967 translate to; headers_charset can change it in the configuration, and
4968 "headers charset" can change it in an individual filter file.
4970 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4971 that creates RFC 2047 encoded "words" labels them as that charset instead
4972 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4973 expansion operator; (ii) when Exim creates a From: line for a local
4974 message; (iii) when a header line is rewritten to include a "phrase" part.
4976 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4977 buggy, causing it to skip the first lines of messages whose message ID
4978 ended in 'D'. This would not have bitten before Exim release 4.14, because
4979 message IDs were unlikely to end in 'D' before then. The effect was to have
4980 incorrect size information for certain domains.
4982 45. #include "config.h" was missing at the start of the crypt16.c module. This
4983 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4986 46. If there was a timeout during a "random" callout check, Exim treated it as
4987 a failure of the random address, and carried on sending RSET and the real
4988 address. If the delay was just some slowness somewhere, the response to the
4989 original RCPT would be taken as a response to RSET and so on, causing
4990 mayhem of various kinds.
4992 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4993 when I implemented it. It didn't allow for the fact that some option values
4994 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4995 the right test for positive values.
4997 48. Domain names in DNS records are case-independent. Exim always looks them up
4998 in lower case. Some resolvers return domain names in exactly the case they
4999 appear in the zone file, that is, they may contain uppercase letters. Not
5000 all resolvers do this - some return always lower case. Exim was treating a
5001 change of case by a resolver as a change of domain, similar to a widening
5002 of a domain abbreviation. This triggered its re-routing code and so it was
5003 trying to route what was effectively the same domain again. This normally
5004 caused routing to fail (because the router wouldn't handle the domain
5005 twice). Now Exim checks for this case specially, and just changes the
5006 casing of the domain that it ultimately uses when it transmits the message
5009 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5012 50. If a filter generated a file delivery with a non-absolute name (possible if
5013 no home directory exists for the router), the forbid_file option was not
5016 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5017 the existing equality matching.
5019 52. Exim was using ints instead of ino_t variables in some places where it was
5020 dealing with inode numbers.
5022 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5023 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5024 called TMPDIR, and if it finds it is different, it changes its value.
5026 54. The smtp_printf() function is now made available to local_scan() so
5027 additional output lines can be written before returning. There is also an
5028 smtp_fflush() function to enable the detection of a dropped connection.
5029 The variables smtp_input and smtp_batched_input are exported to
5032 55. Changed the default runtime configuration: the message "Unknown user"
5033 has been removed from the ACL, and instead placed on the localuser router,
5034 using the cannot_route_message feature. This means that any verification
5035 failures that generate their own messages won't get overridden. Similarly,
5036 the "Unrouteable address" message that was in the ACL for unverifiable
5037 relay addresses has also been removed.
5039 56. Added hosts_avoid_esmtp to the smtp transport.
5041 57. The exicyclog script was not checking for the esoteric option
5042 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5043 will work only if exicyclog is run under the appropriate euid.
5045 58. Following a discussion on the list, the rules by which Exim recognises line
5046 endings on incoming messages have been changed. The -dropcr and drop_cr
5047 options are now no-ops, retained only for backwards compatibility. The
5048 following line terminators are recognized: LF CRLF CR. However, special
5049 processing applies to CR:
5051 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5052 nor a local message in the state where . is a terminator.
5054 (ii) If a bare CR is encountered in a header line, an extra space is added
5055 after the line terminator so as not to end the header. The reasoning
5056 behind this is that bare CRs in header lines are most likely either
5057 to be mistakes, or people trying to play silly games.
5059 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5060 was being incorrectly given as 18 bytes larger than it should have been.
5061 This is a VOB (very old bug).
5063 60. This may never have affected anything current, but just in case it has:
5064 When the local host is found other than at the start of a list of hosts,
5065 the local host, those with the same MX, and any that follow, are discarded.
5066 When the list in question was part of a longer list of hosts, the following
5067 hosts (not currently being processed) were also being discarded. This no
5068 longer happens. I'm not sure if this situation could ever has previously
5071 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5074 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5075 header. This is something that is recommended in a new Internet Draft, and
5076 is something that is documented as being done by Sendmail. There are two
5077 possible values. For messages generated by the autoreply transport, Exim
5080 Auto-Submitted: auto-replied
5082 whereas for all other generated messages (e.g. bounces) it adds
5084 Auto-Submitted: auto-generated
5086 63. The "personal" condition in filters now includes a test for the
5087 Auto-Submitted: header. If it contains the string "auto-" the message it
5088 not considered personal.
5090 64. Added rcpt_include_affixes as a generic transport option.
5092 65. Added queue_only_override (default true).
5094 66. Added the syslog_duplication option.
5096 67. If what should have been the first header line of a message consisted of
5097 a space followed by a colon, Exim was mis-interpreting it as a header line.
5098 It isn't of course - it is syntactically invalid and should therefore be
5099 treated as the start of the message body. The misbehaviour could have
5100 caused a number of strange effects, including loss of data in subsequent
5101 header lines, and spool format errors.
5103 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5104 client host had authenticated. This control can now be exercised by an ACL
5105 for more flexibility.
5107 69. By default, callouts do not happen when testing with -bh. There is now a
5108 variant, -bhc, which does actually run the callout code, including
5109 consulting and updating the callout cache.
5111 70. Added support for saslauthd authentication, courtesy of Alexander
5114 71. If statvfs() failed on the spool or log directories while checking their
5115 size for availability, Exim confusingly gave the error "space shortage".
5116 Furthermore, in debugging mode it crashed with a floating point exception.
5117 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5118 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5119 serious problem, Exim now writes to the main and panic logs when this
5120 happens, with details of the failure. It then refuses to accept the
5121 incoming message, giving the message "spool directory problem" or "log
5122 directory problem" with a 421 code for SMTP messages.
5124 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5125 0, 1, and 2 exist, because some OS complain for execs without them (see
5126 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5127 descriptors. However, the code omitted to check that the open succeeded,
5128 causing mysterious errors if for some reason the permissions on /dev/null
5129 got screwed. Now Exim writes a message to the main and panic logs, and
5130 bombs out if it can't open /dev/null.
5132 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5133 interact so that it is all more flexible. It is supposed to remain
5134 backwards compatible. Also added extra_local_interfaces.
5136 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5137 to bomb out with an assertion failure - to the client this appears as a
5138 connection drop. This problem occurs in the part of the code that was taken
5139 from the Samba project. Fortunately, the assertion is in a very simple
5140 function, so I have fixed this by reproducing the function inline in the
5141 one place where it is called, and arranging for authentication to fail
5142 instead of killing the process with assert().
5144 75. The SPA client code was not working when the server requested OEM rather
5145 than Unicode encoding.
5147 76. Added code to make require_files with a specific uid setting more usable in
5148 the case where statting the file as root fails - usually a non-root-mounted
5149 NFS file system. When this happens and the failure is EACCES, Exim now
5150 forks a subprocess and does the per-uid checking as the relevant uid.
5152 77. Added process_log_path.
5154 78. If log_file_path was not explicitly set, a setting of check_log_space or
5155 check_log_inodes was ignored.
5157 79. If a space check for the spool or log partitions fails, the incident is now
5158 logged. Of course, in the latter case the data may get lost...
5160 80. Added the %p formatting code to string_format() so that it can be used to
5161 print addresses in debug_print(). Adjusted all the address printing in the
5162 debugging in store.c to use %p rather than %d.
5164 81. There was a concern that a line of code in smtp_in.c could overflow a
5165 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5166 initially expressed, the concern was not well-founded, because trailing
5167 spaces are removed early. However, if the trailing spaces were followed by
5168 a NULL, they did not get removed, so the overflow was possible. Two fixes
5171 (a) I re-wrote the offending code in a cleaner fashion.
5172 (b) If an incoming SMTP command contains a NULL character, it is rejected
5175 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5176 runs initgroups(), so that if the Exim user is in any additional groups,
5177 they will be used during message reception.
5183 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob