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).
93 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
94 $data_path, and $header_path variables; fixed documentation bugs and
97 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
98 exipick to access non-standard spools, including the "frozen" queue
101 NM/01 Bugzilla 965: Support mysql stored procedures.
102 Patch from Alain Williams
104 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
106 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
107 Patch from Andreas Metzler
109 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
110 Patch from Kirill Miazine
112 NM/05 Bugzilla 671: Added umask to procmail example.
114 JJ/03 installed exipick 20100323.0, fixing doc bug
116 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
117 directory. Notification and patch from Dan Rosenberg.
119 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
121 TK/02 Improve log output when DKIM signing operation fails.
123 MH/01 Treat the transport option dkim_domain as a colon separated
124 list, not as a single string, and sign the message with each element,
125 omitting multiple occurences of the same signer.
127 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
128 Bugzilla 985, 986. Patch by Simon Arlott
130 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
131 Patch by Simon Arlott
133 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
134 MBX locking. Notification from Dan Rosenberg.
140 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
142 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
144 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
146 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
148 NM/04 Bugzilla 915: Flag broken perl installation during build.
154 TK/01 Added patch by Johannes Berg that expands the main option
155 "spamd_address" if it starts with a dollar sign.
157 TK/02 Write list of recipients to X-Envelope-Sender header when building
158 the mbox-format spool file for content scanning (suggested by Jakob
161 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
162 (http://www.dcc-servers.net/) support via dccifd. Activated by
163 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
165 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
166 by Mark Daniel Reidel <mr@df.eu>.
168 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
169 When building exim an external PCRE library is now needed -
170 PCRE is a system library on the majority of modern systems.
171 See entry on PCRE_LIBS in EDITME file.
173 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
174 conversation. Added nologin parameter to request.
175 Patch contributed by Kirill Miazine.
177 TF/01 Do not log submission mode rewrites if they do not change the address.
179 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
181 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
182 log files in place. Contributed by Roberto Lima.
184 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
186 TF/03 Bugzilla 615: When checking the local_parts router precondition
187 after a local_part_suffix or local_part_prefix option, Exim now
188 does not use the address's named list lookup cache, since this
189 contains cached lookups for the whole local part.
191 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
192 Robert Millan. Documentation is in experimental-spec.txt.
194 TF/04 Bugzilla 668: Fix parallel build (make -j).
196 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
198 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
199 Patch provided by Jan Srzednicki.
201 TF/05 Leading white space used to be stripped from $spam_report which
202 wrecked the formatting. Now it is preserved.
204 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
205 that they are available at delivery time.
207 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
209 TF/08 TLS error reporting now respects the incoming_interface and
210 incoming_port log selectors.
212 TF/09 Produce a more useful error message if an SMTP transport's hosts
213 setting expands to an empty string.
215 NM/06 Bugzilla 744: EXPN did not work under TLS.
216 Patch provided by Phil Pennock.
218 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
219 Patch provided by Richard Godbee.
221 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
222 acl_smtp_notquit, added index entry.
224 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
225 Patch provided by Eugene Bujak.
227 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
228 accept(). Patch provided by Maxim Dounin.
230 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
231 Patch provided by Phil Pennock.
233 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
235 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
236 Patch provided by Brad "anomie" Jorsch.
238 NM/14 Bugzilla 622: Added timeout setting to transport filter.
239 Patch provided by Dean Brooks.
241 TK/05 Add native DKIM support (does not depend on external libraries).
243 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
244 Patch provided by Graeme Fowler.
246 NM/16 Bugzilla 851: Documentation example syntax fix.
248 NM/17 Changed NOTICE file to remove references to embedded PCRE.
250 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
253 NM/19 Bugzilla 745: TLS version reporting.
254 Patch provided by Phil Pennock.
256 NM/20 Bugzilla 167: bool: condition support.
257 Patch provided by Phil Pennock.
259 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
260 clients. Patch provided by Phil Pennock.
262 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
263 Patch provided by Brad "anomie" Jorsch.
265 NM/23 Bugzilla 687: Fix misparses in eximstats.
266 Patch provided by Heiko Schlittermann.
268 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
269 Patch provided by Heiko Schlittermann.
271 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
272 plus update to original patch.
274 NM/26 Bugzilla 799: Documentation correction for ratelimit.
276 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
277 Patch provided by David Brownlee.
279 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
281 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
283 NM/30 Bugzilla 888: TLS documentation bugfixes.
285 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
287 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
288 Unlike the original bugzilla I have changed all shell scripts in src tree.
290 NM/33 Bugzilla 898: Transport filter timeout fix.
291 Patch by Todd Rinaldo.
293 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
294 Patch by Serge Demonchaux.
296 NM/35 Bugzilla 39: Base64 decode bug fixes.
297 Patch by Jakob Hirsch.
299 NM/36 Bugzilla 909: Correct connect() call in dcc code.
301 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
303 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
305 NM/39 Bugzilla 911: Fixed MakeLinks build script.
311 TK/01 Add preliminary DKIM support. Currently requires a forked version of
312 ALT-N's libdkim that I have put here:
313 http://duncanthrax.net/exim-experimental/
315 Note to Michael Haardt: I had to rename some vars in sieve.c. They
316 were called 'true' and it seems that C99 defines that as a reserved
317 keyword to be used with 'bool' variable types. That means you could
318 not include C99-style headers which use bools without triggering
319 build errors in sieve.c.
321 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
322 as mailq or other aliases. Changed the --help handling significantly
323 to do whats expected. exim_usage() emits usage/help information.
325 SC/01 Added the -bylocaldomain option to eximstats.
327 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
329 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
331 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
337 PH/01 Another patch from the Sieve maintainer.
339 PH/02 When an IPv6 address is converted to a string for single-key lookup
340 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
341 dots are used instead of colons so that keys in lsearch files need not
342 contain colons. This was done some time before quoting was made available
343 in lsearch files. However, iplsearch files do require colons in IPv6 keys
344 (notated using the quote facility) so as to distinguish them from IPv4
345 keys. This meant that lookups for IP addresses in host lists did not work
346 for iplsearch lookups.
348 This has been fixed by arranging for IPv6 addresses to be expressed with
349 colons if the lookup type is iplsearch. This is not incompatible, because
350 previously such lookups could never work.
352 The situation is now rather anomolous, since one *can* have colons in
353 ordinary lsearch keys. However, making the change in all cases is
354 incompatible and would probably break a number of configurations.
356 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
359 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
360 conversion specification without a maximum field width, thereby enabling
361 a rogue spamd server to cause a buffer overflow. While nobody in their
362 right mind would setup Exim to query an untrusted spamd server, an
363 attacker that gains access to a server running spamd could potentially
364 exploit this vulnerability to run arbitrary code as the Exim user.
366 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
367 $primary_hostname instead of what libspf2 thinks the hosts name is.
369 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
370 a directory entry by the name of the lookup key. Previously, if a
371 symlink pointed to a non-existing file or a file in a directory that
372 Exim lacked permissions to read, a lookup for a key matching that
373 symlink would fail. Now it is enough that a matching directory entry
374 exists, symlink or not. (Bugzilla 503.)
376 PH/03 The body_linecount and body_zerocount variables are now exported in the
379 PH/04 Added the $dnslist_matched variable.
381 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
382 This means they are set thereafter only if the connection becomes
385 PH/06 Added the client_condition to authenticators so that some can be skipped
386 by clients under certain conditions.
388 PH/07 The error message for a badly-placed control=no_multiline_responses left
389 "_responses" off the end of the name.
391 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
393 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
394 (without spaces) instead of just copying the configuration text.
396 PH/10 Added the /noupdate option to the ratelimit ACL condition.
398 PH/11 Added $max_received_linelength.
400 PH/12 Added +ignore_defer and +include_defer to host lists.
402 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
403 way in which PCRE > 7.0 is built.
405 PH/14 Implemented queue_only_load_latch.
407 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
408 MAIL command. The effect was to mangle the value on 64-bit systems.
410 PH/16 Another patch from the Sieve maintainer.
412 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
414 PH/18 If a system quota error occurred while trying to create the file for
415 a maildir delivery, the message "Mailbox is full" was not appended to the
416 bounce if the delivery eventually timed out. Change 4.67/27 below applied
417 only to a quota excession during the actual writing of the file.
419 PH/19 It seems that peer DN values may contain newlines (and other non-printing
420 characters?) which causes problems in log lines. The DN values are now
421 passed through string_printing() before being added to log lines.
423 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
424 and InterBase are left for another time.)
426 PH/21 Added message_body_newlines option.
428 PH/22 Guard against possible overflow in moan_check_errorcopy().
430 PH/23 POSIX allows open() to be a macro; guard against that.
432 PH/24 If the recipient of an error message contained an @ in the local part
433 (suitably quoted, of course), incorrect values were put in $domain and
434 $local_part during the evaluation of errors_copy.
440 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
441 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
444 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
445 issue a MAIL command.
447 PH/02 In an ACL statement such as
449 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
451 if a client was not listed at all, or was listed with a value other than
452 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
453 the condition was not true (as it should be), so access was not denied.
454 The bug was that the ! inversion was incorrectly passed on to the second
455 item. This has been fixed.
457 PH/03 Added additional dnslists conditions == and =& which are different from
458 = and & when the dns lookup returns more than one IP address.
460 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
461 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
463 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
464 FSYNC, which compiles an option called disable_fsync that allows for
465 bypassing fsync(). The documentation is heavily laced with warnings.
467 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
469 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
470 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
471 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
472 including adding "make clean"; (3) Added -fPIC when compiling the test
473 dynamically loaded module, to get rid of a warning.
475 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
476 message fails, move_frozen_messages = true and ignore_bounce_errors_after
477 = 0s. The bug is otherwise harmless.
479 PH/07 There was a bug in the dovecot authenticator such that the value of
480 $auth1 could be overwritten, and so not correctly preserved, after a
481 successful authentication. This usually meant that the value preserved by
482 the server_setid option was incorrect.
484 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
486 PH/09 Installed PCRE release 7.0.
488 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
489 run for batched SMTP input. It is now run at the start of every message
490 in the batch. While fixing this I discovered that the process information
491 (output by running exiwhat) was not always getting set for -bs and -bS
492 input. This is fixed, and it now also says "batched" for BSMTP.
494 PH/11 Added control=no_pipelining.
496 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
497 patch, slightly modified), and move the expansion of helo_data till after
498 the connection is made in the smtp transport (so it can use these
501 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
503 PH/14 Added log_selector = +pid.
505 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
507 PH/16 Add ${if forany and ${if forall.
509 PH/17 Added dsn_from option to vary the From: line in DSNs.
511 PH/18 Flush SMTP output before performing a callout, unless control =
512 no_callout_flush is set.
514 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
515 was true (the default) a successful delivery failed to delete the retry
516 item, thus causing premature timeout of the address. The bug is now
519 PH/20 Added hosts_avoid_pipelining to the smtp transport.
521 PH/21 Long custom messages for fakedefer and fakereject are now split up
522 into multiline reponses in the same way that messages for "deny" and
523 other ACL rejections are.
525 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
526 with slight modification.
528 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
529 draft, changing the syntax and factoring some duplicate code".
531 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
532 for deliveries of the second and subsequent messages over the same SMTP
535 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
536 ${reduce, with only minor "tidies".
538 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
540 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
541 expansion side effects.
543 PH/27 When a message times out after an over-quota error from an Exim-imposed
544 quota, the bounce message says "mailbox is full". This message was not
545 being given when it was a system quota that was exceeded. It now should
548 MH/03 Made $recipients available in local_scan(). local_scan() already has
549 better access to the recipient list through recipients_list[], but
550 $recipients can be useful in postmaster-provided expansion strings.
552 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
553 in the case of a MAIL command with additional options following the
554 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
555 were accidentally chopped off.
557 PH/29 SMTP synchronization checks are implemented when a command is read -
558 there is a check that no more input is waiting when there shouldn't be
559 any. However, for some commands, a delay in an ACL can mean that it is
560 some time before the response is written. In this time, more input might
561 arrive, invalidly. So now there are extra checks after an ACL has run for
562 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
563 pipelining has not been advertised.
565 PH/30 MH's patch to allow iscntrl() characters to be list separators.
567 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
568 returned in the SMTP response when smtp_return_error_details was false.
571 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
572 instead of the C I/O that was originally supplied, because problems were
575 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
576 Exim which did not show up earlier: it was assuming that a call to
577 SSL_CTX_set_info_callback() might give an error value. In fact, there is
578 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
579 was a macro that became an assignment, so it seemed to work. This has
580 changed to a proper function call with a void return, hence the compile
581 error. Exim's code has been fixed.
583 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
586 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
588 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
590 PH/37 If a message is not accepted after it has had an id assigned (e.g.
591 because it turns out to be too big or there is a timeout) there is no
592 "Completed" line in the log. When some messages of this type were
593 selected by exigrep, they were listed as "not completed". Others were
594 picked up by some special patterns. I have improved the selection
595 criteria to be more general.
597 PH/38 The host_find_failed option in the manualroute router can now be set
598 to "ignore", to completely ignore a host whose IP address cannot be
599 found. If all hosts are ignored, the behaviour is controlled by the new
600 host_all_ignored option.
602 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
603 homing or because of multiple MX records with /mx) generated more than
604 one IP address, and the following item turned out to be the local host,
605 all the secondary addresses of the first item were incorrectly removed
606 from the list, along with the local host and any following hosts (which
607 is what is supposed to happen).
609 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
610 whoever called Exim into the -H file. In the case of the daemon it was
611 behaving confusingly. When first started, it used values for whoever
612 started the daemon, but after a SIGHUP it used the Exim user (because it
613 calls itself on a restart). I have changed the code so that it now always
616 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
617 message are rejected with the same error (e.g. no authentication or bad
618 sender address), and a DATA command is nevertheless sent (as can happen
619 with PIPELINING or a stupid MUA), the error message that was given to the
620 RCPT commands is included in the rejection of the DATA command. This is
621 intended to be helpful for MUAs that show only the final error to their
624 PH/42 Another patch from the Sieve maintainer.
626 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
627 Eximstats - Fixed some broken HTML links and added missing column headers
629 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
630 columns for Rejects, Temp Rejects, Ham, and Spam rows.
632 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
634 PH/43 Yet another patch from the Sieve maintainer.
636 PH/44 I found a way to check for a TCP/IP connection going away before sending
637 the response to the final '.' that terminates a message, but only in the
638 case where the client has not sent further data following the '.'
639 (unfortunately, this is allowed). However, in many cases there won't be
640 any further data because there won't be any more messages to send. A call
641 to select() can be used: if it shows that the input is "ready", there is
642 either input waiting, or the socket has been closed. An attempt to read
643 the next input character can distinguish the two cases. Previously, Exim
644 would have sent an OK response which the client would never have see.
645 This could lead to message repetition. This fix should cure that, at
646 least in a lot of common cases.
648 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
649 advertised in response to EHLO.
655 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
656 fixed by 4.65/MH/01 (is this a record?) are fixed:
658 (i) An empty string was always treated as zero by the numeric comparison
659 operators. This behaviour has been restored.
661 (ii) It is documented that the numeric comparison operators always treat
662 their arguments as decimal numbers. This was broken in that numbers
663 starting with 0 were being interpreted as octal.
665 While fixing these problems I realized that there was another issue that
666 hadn't been noticed. Values of message_size_limit (both the global option
667 and the transport option) were treated as octal if they started with 0.
668 The documentation was vague. These values are now always treated as
669 decimal, and I will make that clear in the documentation.
675 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
676 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
679 MH/01 Don't check that the operands of numeric comparison operators are
680 integers when their expansion is in "skipping" mode (fixes bug
681 introduced by 4.64-PH/07).
683 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
684 child addresses, Exim now panics and dies. Previously, because the count
685 is held in a short int, deliveries were likely to be lost. As such a
686 large number of recipients for a single message is ridiculous
687 (performance will be very, very poor), I have chosen to impose a limit
688 rather than extend the field.
694 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
695 leftover -K file (the existence of which was triggered by #402).
696 While we were at it, introduced process PID as part of the -K
697 filename. This should rule out race conditions when creating
700 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
701 processing considerably. Previous code took too long for large mails,
702 triggering a timeout which in turn triggers #401.
704 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
705 in the DK code in transports.c. sendfile() is not really portable,
706 hence the _LINUX specificness.
708 TF/01 In the add_headers option to the mail command in an Exim filter,
709 there was a bug that Exim would claim a syntax error in any
710 header after the first one which had an odd number of characters
713 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
714 callout verification, Exim cached a "reject" for the entire domain. This
715 is correct for most verifications, but it is not correct for a recipient
716 verification with use_sender or use_postmaster set, because in that case
717 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
718 case of MAIL FROM:<> rejection from other early rejections (e.g.
719 rejection of HELO). When verifying a recipient using a non-null MAIL
720 address, the cache is ignored if it shows MAIL FROM:<> rejection.
721 Whatever the result of the callout, the value of the domain cache is
722 left unchanged (for any other kind of callout, getting as far as trying
723 RCPT means that the domain itself is ok).
725 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
728 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
729 manifest itself as EPIPE rather than ECONNECT. When tidying away a
730 session, the daemon ignores ECONNECT errors and logs others; it now
731 ignores EPIPE as well.
733 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
734 (quoted-printable decoding).
736 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
737 later the small subsequent patch to fix an introduced bug.
739 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
741 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
743 PH/08 An error is now given if message_size_limit is specified negative.
745 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
746 to be given (somewhat) arbitrary names.
748 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
751 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
752 miscellaneous code fixes
754 PH/10 Added the log_reject_target ACL modifier to specify where to log
757 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
758 hostname. This is wrong, because it relates to the incoming message (and
759 probably the interface on which it is arriving) and not to the outgoing
760 callout (which could be using a different interface). This has been
761 changed to use the value of the helo_data option from the smtp transport
762 instead - this is what is used when a message is actually being sent. If
763 there is no remote transport (possible with a router that sets up host
764 addresses), $smtp_active_hostname is used.
766 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
767 tweaks were necessary in order to get it to work (see also 21 below):
768 (a) The code assumed that strncpy() returns a negative number on buffer
769 overflow, which isn't the case. Replaced with Exim's string_format()
771 (b) There were several signed/unsigned issues. I just did the minimum
772 hacking in of casts. There is scope for a larger refactoring.
773 (c) The code used strcasecmp() which is not a standard C function.
774 Replaced with Exim's strcmpic() function.
775 (d) The code set only $1; it now sets $auth1 as well.
776 (e) A simple test gave the error "authentication client didn't specify
777 service in request". It would seem that Dovecot has changed its
778 interface. Fortunately there's a specification; I followed it and
779 changed what the client sends and it appears to be working now.
781 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
784 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
785 address A is aliased to B and C, where B exists and C does not. Without
786 -v the output is "A verified" because verification stops after a
787 successful redirection if more than one address is generated. However,
788 with -v the child addresses are also verified. Exim was outputting "A
789 failed to verify" and then showing the successful verification for C,
790 with its parentage. It now outputs "B failed to verify", showing B's
791 parentage before showing the successful verification of C.
793 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
794 look up a TXT record in a specific list after matching in a combined
797 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
798 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
799 they consult the DNS. I had assumed they would set it the way they
800 wanted; and indeed my experiments on Linux seem to show that in some
801 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
802 To be on the safe side, however, I have now made the interface to
803 host_find_byname() similar to host_find_bydns(), with an argument
804 containing the DNS resolver options. The host_find_byname() function now
805 sets these options at its start, just as host_find_bydns() does. The smtp
806 transport options dns_qualify_single and dns_search_parents are passed to
807 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
808 of host_find_byname() use the default settings of RES_DEFNAMES
809 (qualify_single) but not RES_DNSRCH (search_parents).
811 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
812 spool_read_header() do less string testing, by means of a preliminary
813 switch on the second character of optional "-foo" lines. (This is
814 overdue, caused by the large number of possibilities that now exist.
815 Originally there were few.) While I was there, I also converted the
816 str(n)cmp tests so they don't re-test the leading "-" and the first
817 character, in the hope this might squeeze out yet more improvement.
819 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
820 flag allowing group syntax was set by the header_syntax check but not
821 turned off, possible causing trouble later; (2) The flag was not being
822 set at all for the header_verify test, causing "group"-style headers to
823 be rejected. I have now set it in this case, and also caused header_
824 verify to ignore an empty address taken from a group. While doing this, I
825 came across some other cases where the code for allowing group syntax
826 while scanning a header line wasn't quite right (mostly, not resetting
827 the flag correctly in the right place). These bugs could have caused
828 trouble for malformed header lines. I hope it is now all correct.
830 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
831 with the "reply" argument non-NULL. The code, however (which originally
832 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
833 but it didn't always do it. This confused somebody who was copying the
834 code for some other use. I have removed all the tests.
836 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
837 feature that was used to support insecure browsers during the U.S. crypto
838 embargo. It requires special client support, and Exim is probably the
839 only MTA that supported it -- and would never use it because real RSA is
840 always available. This code has been removed, because it had the bad
841 effect of slowing Exim down by computing (never used) parameters for the
842 RSA_EXPORT functionality.
844 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
845 authenticator to fail if there's a tab character in the incoming data
846 (there should never be unless someone is messing about, as it's supposed
847 to be base64-encoded). Also added, on Timo's advice, the "secured" option
848 if the connection is using TLS or if the remote IP is the same as the
849 local IP, and the "valid-client-cert option" if a client certificate has
852 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
853 authenticators. This can be used for authorization after authentication
854 succeeds. (In the case of plaintext, it servers for both authentication
857 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
858 if any retry times were supplied.
860 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
861 connection, where there is no client IP address to check. In this
862 situation, the verify now always succeeds.
864 PH/25 Applied John Jetmore's -Mset patch.
866 PH/26 Added -bem to be like -Mset, but loading a message from a file.
868 PH/27 In a string expansion for a processed (not raw) header when multiple
869 headers of the same name were present, leading whitespace was being
870 removed from all of them, but trailing whitespace was being removed only
871 from the last one. Now trailing whitespace is removed from each header
872 before concatenation. Completely empty headers in a concatenation (as
875 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
876 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
878 PH/29 [Removed. This was a change that I later backed out, and forgot to
879 correct the ChangeLog entry (that I had efficiently created) before
880 committing the later change.]
882 PH/30 Exim was sometimes attempting to deliver messages that had suffered
883 address errors (4xx response to RCPT) over the same connection as other
884 messages routed to the same hosts. Such deliveries are always "forced",
885 so retry times are not inspected. This resulted in far too many retries
886 for the affected addresses. The effect occurred only when there were more
887 hosts than the hosts_max_try setting in the smtp transport when it had
888 the 4xx errors. Those hosts that it had tried were not added to the list
889 of hosts for which the message was waiting, so if all were tried, there
890 was no problem. Two fixes have been applied:
892 (i) If there are any address or message errors in an SMTP delivery, none
893 of the hosts (tried or untried) are now added to the list of hosts
894 for which the message is waiting, so the message should not be a
895 candidate for sending over the same connection that was used for a
896 successful delivery of some other message. This seems entirely
897 reasonable: after all the message is NOT "waiting for some host".
898 This is so "obvious" that I'm not sure why it wasn't done
899 previously. Hope I haven't missed anything, but it can't do any
900 harm, as the worst effect is to miss an optimization.
902 (ii) If, despite (i), such a delivery is accidentally attempted, the
903 routing retry time is respected, so at least it doesn't keep
904 hammering the server.
906 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
907 in ${readsocket because some servers need this prod.
909 PH/32 Added some extra debug output when updating a wait-xxx database.
911 PH/33 The hint "could be header name not terminated by colon", which has been
912 given for certain expansion errors for a long time, was not being given
913 for the ${if def:h_colon_omitted{... case.
915 PH/34 The spec says: "With one important exception, whenever a domain list is
916 being scanned, $domain contains the subject domain." There was at least
917 one case where this was not true.
919 PH/35 The error "getsockname() failed: connection reset by peer" was being
920 written to the panic log as well as the main log, but it isn't really
921 panic-worthy as it just means the connection died rather early on. I have
922 removed the panic log writing for the ECONNRESET error when getsockname()
925 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
926 runs only) independently of the message's sender address. This meant
927 that, if the 4xx error was in fact related to the sender, a different
928 message to the same recipient with a different sender could confuse
929 things. In particualar, this can happen when sending to a greylisting
930 server, but other circumstances could also provoke similar problems.
931 I have changed the default so that the retry time for these errors is now
932 based a combination of the sender and recipient addresses. This change
933 can be overridden by setting address_retry_include_sender=false in the
936 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
937 remote server are returned as part of bounce messages. This was not
938 happening for LMTP over a pipe (the lmtp transport), but now it is the
939 same for both kinds of LMTP.
941 PH/38 Despite being documented as not happening, Exim was rewriting addresses
942 in header lines that were in fact CNAMEs. This is no longer the case.
944 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
945 and queue runs started by the daemon processed all messages. This has
946 been fixed so that -R and -S can now usefully be given with -q<time>.
948 PH/40 Import PCRE release 6.7 (fixes some bugs).
950 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
952 PH/42 Give an error if -q is specified more than once.
954 PH/43 Renamed the variables $interface_address and $interface_port as
955 $received_ip_address and $received_port, to make it clear that these
956 values apply to message reception, and not to the outgoing interface when
957 a message is delivered. (The old names remain recognized, of course.)
959 PH/44 There was no timeout on the connect() call when using a Unix domain
960 socket in the ${readsocket expansion. There now is.
962 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
963 be meaningful with "accept".
965 SC/01 Eximstats V1.43
966 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
968 SC/02 Eximstats V1.44
969 Use a glob alias rather than an array ref in the generated
970 parser. This improves both readability and performance.
972 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
973 Collect SpamAssassin and rejection statistics.
974 Don't display local sender or destination tables unless
975 there is data to show.
976 Added average volumes into the top table text output.
978 SC/04 Eximstats V1.46
979 Collect data on the number of addresses (recipients)
980 as well as the number of messages.
982 SC/05 Eximstats V1.47
983 Added 'Message too big' to the list of mail rejection
984 reasons (thanks to Marco Gaiarin).
986 SC/06 Eximstats V1.48
987 Mainlog lines which have GMT offsets and are too short to
988 have a flag are now skipped.
990 SC/07 Eximstats V1.49 (Alain Williams)
991 Added the -emptyok flag.
993 SC/08 Eximstats V1.50
994 Fixes for obtaining the IP address from reject messages.
996 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
997 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
998 whitesspace changes from 4.64-PH/27
1000 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1003 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1004 are found, allow negative numbers in numeric criteria)
1006 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1008 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1009 to match changes made in 4.64-PH/43
1011 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1013 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1014 "domain" is actually an IP address, and force a failure. This locks out
1015 those revolvers/nameservers that support "A-for-A" lookups, in
1016 contravention of the specifications.
1018 PH/48 When a host name was looked up from an IP address, and the subsequent
1019 forward lookup of the name timed out, the host name was left in
1020 $sender_host_name, contrary to the specification.
1022 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1023 restricted to single-key lookups, Exim was not diagnosing an error if
1024 * or *@ was used with a query-style lookup.
1026 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1028 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1029 long ago, but noone interested enough thought of it. Let's just say that
1030 the "1.1" means that there are some new functions that weren't there at
1031 some point in the past.
1033 PH/51 Error processing for expansion failure of helo_data from an smtp
1034 transport during callout processing was broken.
1036 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1037 tested/used via the -bh/-bhc/-bs options.
1039 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1040 bug, fixed in subsequent PCRE releases).
1042 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1043 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1045 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1051 SC/01 Use a glob alias rather than an array ref in eximstats generated
1052 parser. This improves both readability and performance.
1054 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1055 Don't display local sender or destination tables in eximstats unless
1056 there is data to show.
1057 Added average volumes into the eximstats top table text output.
1059 SC/03 Collect data on the number of addresses (recipients) as well
1060 as the number of messages in eximstats.
1062 TF/01 Correct an error in the documentation for the redirect router. Exim
1063 does (usually) call initgroups() when daemonizing.
1065 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1066 with consistent privilege compared to when running as a daemon.
1068 TF/03 Note in the spec that $authenticated_id is not set for local
1069 submissions from trusted users.
1071 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1072 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1074 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1075 by adding some example configuration directives to the default
1076 configuration file. A little bit of work is required to uncomment the
1077 directives and define how usernames and passwords are checked, but
1078 there is now a framework to start from.
1080 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1081 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1082 without this. I don't know how relevant this is to other LDAP libraries.
1084 PH/02 Add the verb name to the "unknown ACL verb" error.
1086 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1088 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1090 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1091 directory not expanded when it should be if an expanded home directory
1092 was set for the address (which is overridden by the transport).
1094 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1097 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1098 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1099 because it is too late at that time, and has no effect.
1101 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1102 security issue with \' (bugzilla #107). I could not use the
1103 PQescapeStringConn() function, because it needs a PGconn value as one of
1106 PH/08 When testing addresses using -bt, indicate those final addresses that
1107 are duplicates that would not cause an additional delivery. At least one
1108 person was confused, thinking that -bt output corresponded to deliveries.
1109 (Suppressing duplicates isn't a good idea as you lose the information
1110 about possibly different redirections that led to the duplicates.)
1112 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1113 systems where poll() doesn't work, in particular OS X.
1115 PH/10 Added more information to debugging output for retry time not reached.
1117 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1118 operations in malware.c.
1120 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1123 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1124 syslog_duplication set false, log lines that would normally be written
1125 both the the main log and to the reject log were not written to syslog at
1128 PH/14 In the default configuration, change the use of "message" in ACL warn
1129 statements to "add_header".
1131 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1132 not followed by a command (e.g. "seen endif").
1134 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1135 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1138 PH/17 Added extra conditions to the default value of delay_warning_condition
1142 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1143 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1144 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1147 The Auto-Submitted: and various List- headers are standardised, whereas I
1148 don't think Precedence: ever was.
1150 PH/18 Refactored debugging code in route_finduser() to show more information,
1151 in particular, the error code if getpwnam() issues one.
1153 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1154 This is apparently needed in addition to the PH/07 change above to avoid
1155 any possible encoding problems.
1157 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1158 but not after initializing Perl.
1160 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1161 output them only if debugging. By default they are written stderr,
1162 apparently, which is not desirable.
1164 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1167 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1170 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1172 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1173 authenticated or an ident call has been made. Suppress the default
1174 values for $authenticated_id and $authenticated_sender (but permit -oMai
1175 and -oMas) when testing with -bh.
1177 PH/24 Re-jigged the order of the tests in the default configuration so that the
1178 tests for valid domains and recipients precede the DNS black list and CSA
1179 tests, on the grounds that those ones are more expensive.
1181 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1182 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1183 This bug exists in every version of Exim that I still have, right back to
1186 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1187 However, an attempt to turn on case-sensitivity in a regex key by
1188 including (?-i) didn't work because the subject string was already
1189 lowercased, and the effects were non-intuitive. It turns out that a
1190 one-line patch can be used to allow (?-i) to work as expected.
1196 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1197 other effects) broke the use of negated acl sub-conditions.
1199 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1202 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1203 "Deny" causes Exim to reject the incoming connection with a 554 error.
1204 Unfortunately, if there is a major crisis, such as a disk failure,
1205 tcp-wrappers gives "deny", whereas what one would like would be some
1206 kind of temporary error. A kludge has been added to help with this.
1207 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1208 554 error is used if errno is still zero or contains ENOENT (which occurs
1209 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1212 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1214 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1215 errors. Otherwise a message that provokes a temporary error (when other
1216 messages do not) can cause a whole host to time out.
1218 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1219 addresses were routed directly to files or pipes from a redirect router.
1220 File deliveries just didn't batch; pipe deliveries might have suffered
1223 PH/06 A failure to get a lock for a hints database would erroneously always say
1224 "Failed to get write lock", even when it was really a read lock.
1226 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1227 of 0600. This has been changed to use the value of the lockfile_mode
1228 option (which defaults to 0600).
1230 PH/08 Applied small patch from the Sieve maintainer.
1232 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1233 folder from quota calculations, a direct delivery into this folder messed
1234 up the contents of the maildirsize file. This was because the regex was
1235 used only to exclude .Trash (or whatever) when the size of the mailbox
1236 was calculated. There was no check that a delivery was happening into an
1237 excluded directory. This bug has been fixed by ignoring all quota
1238 processing for deliveries into excluded directories.
1240 PH/10 Added the maildirfolder_create_regex option to appendfile.
1246 PH/01 The code for finding all the local interface addresses on a FreeBSD
1247 system running IPv6 was broken. This may well have applied to all BSD
1248 systems, as well as to others that have similar system calls. The broken
1249 code found IPv4 interfaces correctly, but gave incorrect values for the
1250 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1251 that it would not match correctly against @[] and not recognize the IPv6
1254 PH/02 The ipliteral router was not recognizing addresses of the form user@
1255 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1257 PH/03 Added disable_ipv6.
1259 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1260 decoded form, because it is most often used to construct To: headers
1261 lines in autoreplies, and the decoded form may well be syntactically
1262 invalid. However, $reply_address has leading white space removed, and all
1263 newlines turned into spaces so that the autoreply transport does not
1266 PH/05 If group was specified without a user on a router, and no group or user
1267 was specified on a transport, the group from the router was ignored.
1269 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1270 for visible compile-time settings that can be used to change these
1271 numbers, for those that want even more. Backwards compatibility with old
1272 spool files has been maintained. However, going back to a previous Exim
1273 release will lost any variables that are in spool files.
1275 PH/07 Two small changes when running in the test harness: increase delay when
1276 passing a TCP/IP connection to a new process, in case the original
1277 process has to generate a bounce, and remove special handling of
1278 127.0.0.2 (sic), which is no longer necessary.
1280 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1281 be the same on different OS.
1283 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1286 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1287 whether --show-vars was specified or not
1289 JJ/02 exipick: Added support for new ACL variable spool format introduced
1292 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1293 syntactically invalid From: or Reply-to: line, and a filter used this to
1294 generate an autoreply, and therefore failed to obtain an address for the
1295 autoreply, Exim could try to deliver to a non-existent relative file
1296 name, causing unrelated and misleading errors. What now happens is that
1297 it logs this as a hard delivery error, but does not attempt to create a
1300 PH/11 The exinext utility has a -C option for testing purposes, but although
1301 the given file was scanned by exinext itself; it wasn't being passed on
1302 when Exim was called.
1304 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1305 an end-of-file indication when reading a command response.
1307 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1308 compiled. In many other places in Exim, IPv6 addresses are always
1309 recognized, so I have changed this. It also means that IPv4 domain
1310 literals of the form [IPV4:n.n.n.n] are now always recognized.
1312 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1313 used if the router is not running as root, for example, when verifying at
1314 ACL time, or when using -bh. The debugging output from this situation was
1315 non-existent - all you got was a failure to exec. I have made two
1318 (a) Failures to set uid/gid, the current directory, or a process leader
1319 in a subprocess such as that created by queryprogram now generate
1320 suitable debugging ouput when -d is set.
1322 (b) The queryprogram router detects when it is not running as root,
1323 outputs suitable debugging information if -d is set, and then runs
1324 the subprocess without attempting to change uid/gid.
1326 PH/15 Minor change to Makefile for building test_host (undocumented testing
1329 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1330 additional section of a DNS packet that returns MX or SRV records.
1331 Instead, it always explicitly searches for A/AAAA records. This avoids
1332 major problems that occur when a DNS server includes only records of one
1333 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1334 fixed another bug: if SRV records were looked up and the corresponding
1335 address records were *not* found in the additional section, the port
1336 values from the SRV records were lost.
1338 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1339 using the correct key (the original address) when searching the retry
1340 rules in order to find which one to use for generating the retry hint.
1342 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1343 adding the default one. Similarly, if it contains a Reply-To: header, the
1344 errors_reply_to option, if set, is not used.
1346 PH/19 When calculating a retry time, Exim used to measure the "time since
1347 failure" by looking at the "first failed" field in the retry record. Now
1348 it does not use this if it is later than than the arrival time of the
1349 message. Instead it uses the arrival time. This makes for better
1350 behaviour in cases where some deliveries succeed, thus re-setting the
1351 "first failed" field. An example is a quota failure for a huge message
1352 when small messages continue to be delivered. Without this change, the
1353 "time since failure" will always be short, possible causing more frequent
1354 delivery attempts for the huge message than are intended.
1355 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1357 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1358 $1, $2, $3) because the numerical variables can be reset during some
1359 expansion items (e.g. "match"), thereby losing the authentication data.
1361 PH/21 Make -bV show the size of off_t variables so that the test suite can
1362 decide whether to run tests for quotas > 2G.
1364 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1365 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1366 filecount value is greater than 2G or if a quota value is greater than 2G
1367 on a system where the size of off_t is not greater than 4, a panic error
1370 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1371 never match. The debug and -bh output now contains an explicit error
1372 message indicating a malformed IPv4 address or mask.
1374 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1375 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1376 PH/23 above applies.
1378 PH/25 Do not write to syslog when running in the test harness. The only
1379 occasion when this arises is a failure to open the main or panic logs
1380 (for which there is an explicit test).
1382 PH/26 Added the /no_tell option to "control=freeze".
1384 PH/27 If a host name lookup failed very early in a connection, for example, if
1385 the IP address matched host_lookup and the reverse lookup yielded a name
1386 that did not have a forward lookup, an error message of the form "no IP
1387 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1388 could be logged. Now it outputs the IP address instead of "NULL".
1390 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1391 allows the sender and the authenticated sender to be set when
1392 submitting a message from within Exim. Since child_open_exim() is
1393 documented for local_scan(), the new function should be too.
1395 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1396 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1397 results in an empty string is now treated as unset.
1399 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1401 PH/31 Added sender_verify_fail logging option.
1403 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1404 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1405 tidied the source and removed it altogether.
1407 PH/33 When a queue run was abandoned because the load average was too high, a
1408 log line was always written; now it is written only if the queue_run log
1409 selector is set. In addition, the log line for abandonment now contains
1410 information about the queue run such as the pid. This is always present
1411 in "start" and "stop" lines but was omitted from the "abandon" line.
1413 PH/34 Omit spaces between a header name and the colon in the error message that
1414 is given when verify = headers_syntax fails (if there are lots of them,
1415 the message gets confusing).
1417 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1418 names, as there are now some PTR records that contain slashes. This check
1419 is only to protect against broken name servers that fall over on strange
1420 characters, so the fact that it applies to all lookups doesn't matter.
1422 PH/36 Now that the new test suite is complete, we can remove some of the
1423 special code in Exim that was needed for the old test suite. For example,
1424 sorting DNS records because real resolvers return them in an arbitrary
1425 order. The new test suite's fake resolver always returns records in the
1428 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1429 bounces) except when queue_only is set, to avoid logging races between
1430 the different processes.
1432 PH/38 Panic-die if .include specifies a non-absolute path.
1434 PH/39 A tweak to the "H" retry rule from its user.
1436 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1437 a label. They prevented compilation on older perls.
1439 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1440 a warning to be raised on newish perls.
1442 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1443 on queue. Changes to match documented behaviour of showing count of
1444 messages matching specified criteria.
1446 PH/40 Changed the default ident timeout from 30s to 5s.
1448 PH/41 Added support for the use of login_cap features, on those BSD systems
1449 that have them, for controlling the resources used by pipe deliveries.
1451 PH/42 The content-scanning code uses fopen() to create files in which to put
1452 message data. Previously it was not paying any attention to the mode of
1453 the files. Exim runs with umask(0) because the rest of the code creates
1454 files with open(), and sets the required mode explicitly. Thus, these
1455 files were ending up world-writeable. This was not a big issue, because,
1456 being within the spool directory, they were not world-accessible. I have
1457 created a function called modefopen, which takes an additional mode
1458 argument. It sets umask(777), creates the file, chmods it to the required
1459 mode, then resets the umask. All the relevant calls to fopen() in the
1460 content scanning code have been changed to use this function.
1462 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1463 to 24 hours. This avoids potential overflow problems when processing G
1464 and H retry rules. I suspect nobody ever tinkers with this value.
1466 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1468 PH/45 When the plaintext authenticator is running as a client, the server's
1469 challenges are checked to ensure they are valid base64 strings. By
1470 default, the authentication attempt is cancelled if an invalid string is
1471 received. Setting client_ignore_invalid_base64 true ignores these errors.
1472 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1473 they are received. Thus, the responses can be made to depend on the
1474 challenges. If an invalid string is ignored, an empty string is placed in
1477 PH/46 Messages that are created by the autoreply transport now contains a
1478 References: header, in accordance with RFCs 2822 and 3834.
1480 PH/47 Added authenticated_sender_force to the smtp transport.
1482 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1484 PH/49 Installed latest patch from the Sieve maintainer.
1486 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1487 was also set, the appendfile transport was unnecessarily scanning a
1488 directory of message files (e.g. for maildir delivery) to find the count
1489 of files (along with the size), even though it did not need this
1490 information. It now does the scan only if it needs to find either the
1491 size of the count of files.
1493 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1495 PH/52 Two bugs concerned with error handling when the smtp transport is
1498 (i) Exim was not creating retry information for temporary errors given
1499 for individual recipients after the DATA command when the smtp transport
1500 was used in LMTP mode. This meant that they could be retried too
1501 frequently, and not timed out correctly.
1503 (ii) Exim was setting the flag that allows error details to be returned
1504 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1505 recipients that were returned after the DATA command.
1507 PH/53 This is related to PH/52, but is more general: for any failing address,
1508 when detailed error information was permitted to be returned to the
1509 sender, but the error was temporary, then after the final timeout, only
1510 "retry timeout exceeded" was returned. Now it returns the full error as
1511 well as "retry timeout exceeded".
1513 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1514 do this, and (what is worse) MTAs that accept it.
1516 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1517 will now be deprecated.
1519 PH/56 New os.c-cygwin from the Cygwin maintainer.
1521 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1522 formats (previously only available in exim formats via -bpr, -bpru,
1523 and -bpra. Now also available in native and exiqgrep formats)
1525 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1526 with very large, slow to parse queues
1528 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1530 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1532 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1533 responses to authentication challenges, though it was showing the
1534 challenges; (ii) I've removed the CR characters from the debug output for
1537 PH/58 Allow for the insertion of a newline as well as a space when a string
1538 is turned into more than one encoded-word during RFC 2047 encoding. The
1539 Sieve code now uses this.
1541 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1542 data_4xx, lost_connection, tls_required.
1544 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1545 message was being sent as an SMTP response.
1547 PH/61 Add -l and -k options to exicyclog.
1549 PH/62 When verifying, if an address was redirected to one new address, so that
1550 verification continued, and the new address failed or deferred after
1551 having set something in $address_data, the value of $address_data was not
1552 passed back to the ACL. This was different to the case when no
1553 redirection occurred. The value is now passed back in both cases.
1555 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1556 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1557 use login_cap.h, so on its own it isn't the distinguishing feature. The
1558 new name refers directly to the setclassresources() function.
1560 PH/65 Added configuration files for NetBSD3.
1562 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1564 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1565 is preferred over IPv4.
1567 PH/68 The bounce_return_message and bounce_return_body options were not being
1568 honoured for bounces generated during the reception of non-SMTP messages.
1569 In particular, this applied to messages rejected by the ACL. This bug has
1570 been fixed. However, if bounce_return_message is true and bounce_return_
1571 body is false, the headers that are returned for a non-SMTP message
1572 include only those that have been read before the error was detected.
1573 (In the case of an ACL rejection, they have all been read.)
1575 PH/69 The HTML version of the specification is now built in a directory called
1576 spec_html instead of spec.html, because the latter looks like a path with
1577 a MIME-type, and this confuses some software.
1579 PH/70 Catch two compiler warnings in sieve.c.
1581 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1582 function verify_get_ident() calls ip_connect() to connect a socket, but
1583 if the "connect()" function timed out, ip_connect() used to close the
1584 socket. However, verify_get_ident() also closes the socket later, and in
1585 between Exim writes to the log, which may get opened at this point. When
1586 the socket was closed in ip_connect(), the log could get the same file
1587 descriptor number as the socket. This naturally causes chaos. The fix is
1588 not to close the socket in ip_connect(); the socket should be closed by
1589 the function that creates it. There was only one place in the code where
1590 this was missing, in the iplookup router, which I don't think anybody now
1591 uses, but I've fixed it anyway.
1593 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1594 well as to direct DNS lookups. Otherwise the handling of names in host
1595 lists is inconsistent and therefore confusing.
1601 PH/01 Two changes to the default runtime configuration:
1603 (1) Move the checks for relay_from_hosts and authenticated clients from
1604 after to before the (commented out) DNS black list checks.
1606 (2) Add control=submission to the relay_from_hosts and authenticated
1607 clients checks, on the grounds that messages accepted by these
1608 statements are most likely to be submissions.
1610 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1612 (1) Generate an error if the third argument for the ${prvs expansion is
1615 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1618 (3) Reset the variables that are obtained from the first argument of
1619 ${prvscheck and used in the second argument before leaving the code,
1620 because their memory is reclaimed, so using them afterwards may do
1623 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1624 one (it's much easier than Tom thought :-).
1626 (5) Because of (4), we can now allow for the use of $prvscheck_result
1627 inside the third argument.
1629 PH/03 For some reason, the default setting of PATH when running a command from
1630 a pipe transport was just "/usr/bin". I have changed it to
1633 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1634 anything to be listed in the output from -bV.
1636 PH/05 When a filter generated an autoreply, the entire To: header line was
1637 quoted in the delivery log line, like this:
1639 => >A.N.Other <ano@some.domain> <original@ddress> ...
1641 This has been changed so that it extracts the operative address. There
1642 may be more than one such address. If so, they are comma-separated, like
1645 => >ano@some.domain,ona@other.domain <original@ddress> ...
1647 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1648 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1649 not being looked up in the rDNS to get a host name, Exim was showing the
1650 IP address twice in Received: lines, even though the IP addresses were
1651 identical. For example:
1653 Received: from [1.2.3.4] (helo=[1.2.3.4])
1655 However, if the real host name was known, it was omitting the HELO data
1656 if it matched the actual IP address. This has been tidied up so that it
1657 doesn't show the same IP address twice.
1659 PH/07 When both +timestamp and +memory debugging was on, the value given by
1660 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1661 called by the debug printing, thereby overwriting the timestamp buffer.
1662 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1664 PH/08 When the original message was included in an autoreply transport, it
1665 always said "this is a copy of the message, including all the headers",
1666 even if body_only or headers_only was set. It now gives an appropriate
1669 PH/09 Applied a patch from the Sieve maintainer which:
1671 o fixes some comments
1672 o adds the (disabled) notify extension core
1673 o adds some debug output for the result of if/elsif tests
1674 o points to the current vacation draft in the documentation
1675 and documents the missing references header update
1679 o fixes a bug in processing the envelope test (when testing
1680 multiple envelope elements, the last element determinted the
1683 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1684 Electronic Mail") by including:
1686 Auto-submitted: auto-generated
1688 in the messages that it generates (bounce messages and others, such as
1689 warnings). In the case of bounce messages for non-SMTP mesages, there was
1690 also a typo: it was using "Auto_submitted" (underscore instead of
1691 hyphen). Since every message generated by Exim is necessarily in response
1692 to another message, thes have all been changed to:
1694 Auto-Submitted: auto-replied
1696 in accordance with these statements in the RFC:
1698 The auto-replied keyword:
1700 - SHOULD be used on messages sent in direct response to another
1701 message by an automatic process,
1703 - MUST NOT be used on manually-generated messages,
1705 - MAY be used on Delivery Status Notifications (DSNs) and Message
1706 Disposition Notifications (MDNs),
1708 - MUST NOT be used on messages generated by automatic or periodic
1709 processes, except for messages which are automatic responses to
1712 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1713 to the default Received: header definition.
1715 PH/12 Added log selector acl_warn_skipped (default on).
1717 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1718 variables because (a) they are in the wrong storage pool and (b) even if
1719 they were copied, it wouldn't work properly because of the caching.
1721 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1722 checking when decoding. Apparently there are clients that generate
1723 overlong encoded strings. Why am I not surprised?
1725 PH/15 If the first argument of "${if match_address" was not empty, but did not
1726 contain an "@" character, Exim crashed. Now it writes a panic log message
1727 and treats the condition as false.
1729 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1731 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1732 extension "envelope-auth". The code is finished and in agreement with
1733 other implementations, but there is no documentation so far and in fact,
1734 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1735 not changing the active code.
1737 Print executed "if" and "elsif" statements when debugging is used. This
1738 helps a great deal to understand what a filter does.
1740 Document more things not specified clearly in RFC3028. I had all this
1741 sorted out, when out of a sudden new issues came to my mind. Oops."
1743 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1746 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1747 canonical form (as documented). However, after a host name lookup from
1748 the IP address, check_host() was doing a simple string comparison with
1749 addresses acquired from the DNS when checking that the found name did
1750 have the original IP as one of its addresses. Since any found IPv6
1751 addresses are likely to be in abbreviated form, the comparison could
1752 fail. Luckily, there already exists a function for doing the comparison
1753 by converting both addresses to binary, so now that is used instead of
1754 the text comparison.
1756 PH/20 There was another similar case to PH/19, when a complete host name was
1757 given in a host list; looking up its IP address could give an abbreviated
1758 form, whereas the current host's name might or might not be abbreviated.
1759 The same fix has been applied.
1765 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1766 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1769 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1770 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1772 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1774 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1775 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1776 Also, for "Auto-Submitted", treat anything other than "no" as
1777 non-personal, in accordance with RFC 3834. (Previously it treated
1778 anything starting "auto-" as non-personal.)
1780 TF/01 The control=submission/name=... option had a problem with syntax
1781 errors if the name included a slash character. The /name= option
1782 now slurps the rest of the string, so it can include any characters
1783 but it must come last in the list of options (after /sender_retain
1786 PH/05 Some modifications to the interface to the fake nameserver for the new
1794 TK/01 Added the "success_on_redirect" address verification option. See
1795 NewStuff for rationale and an example.
1797 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1799 PH/02 Patch to exigrep to allow it to work on syslog lines.
1801 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1802 fread() to skip over the body file's header line, because in Cygwin the
1803 header line is locked and is inaccessible.
1805 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1806 co-exist for some time) to make it clear that it is the Exim ID that is
1807 referenced, not the Message-ID: header line.
1809 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1810 string_format() function, because snprintf() does not exist on all
1813 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1814 ${stat: expansion item.
1816 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1817 protocol synchronization error", to keep the pedants happy.
1819 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1820 well as for IRIX systems, when gcc is being used. See the host.c source
1823 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1825 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1828 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1829 transport and to the smtp transport in LMTP mode.
1831 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1833 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1834 run a filter in a subprocess. This could lead to confusion in subsequent
1835 lookups in the parent process. There should also be a search_tidyup() at
1836 the end of the subprocess.
1838 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1839 only if the host matched helo_try_verify_hosts, which caused the
1840 verification to occur when the EHLO/HELO command was issued. The ACL just
1841 tested the remembered result. Now, if a previous verification attempt has
1842 not happened, "verify = helo" does it there and then.
1844 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1846 TK/03 Fix log output including CR from clamd.
1848 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1849 exist provoked a memory error which could cause a segfault.
1851 PH/15 Installed PCRE 6.2
1853 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1855 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1856 of the problem. Specifically, suggested +O2 rather than +O1 for the
1859 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1861 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1862 record, the port was not correctly logged when the outgoing_port log
1863 selector was set (it logged the transort's default port).
1865 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1866 fallback_hosts, and "hosts" in the smtp transport.
1868 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1869 host errors such as "Connection refused".
1871 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1872 authentication with radiusclient 0.4.9:
1874 - Error returned from rc_read_config was caught wrongly
1875 - Username/password not passed on to radius server due to wrong length.
1877 The presumption is that some radiusclient API changes for 4.51/PH/17
1878 were not taken care of correctly. The code is still untested by me (my
1879 Linux distribution still has 0.3.2 of radiusclient), but it was
1880 contributed by a Radius user.
1882 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1883 expanding the "port" option of the smtp transport.
1885 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1886 while reading a MIME header. Thanks to Tom Hughes for a patch.
1888 PH/24 Include config.h inside local_scan.h so that configuration settings are
1891 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1892 that in an ACL for RCPT (for example), you can examine exactly what was
1895 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1896 commands, but it was not correctly comparing the address with the actual
1897 client host address. Thus, it would show the EHLO address in Received:
1898 header lines when this was not necessary.
1900 PH/27 Added the % operator to ${eval:}.
1902 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1903 it should be ignoring failures (because with -C, for example, it has lost
1904 privilege). It wasn't ignoring creation failures other than "already
1907 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1910 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1911 because an input line was too long, either on its own, or by virtue of
1912 too many continuations, the temporary file was not being removed, and the
1913 return code was incorrect.
1915 PH/31 Missing "BOOL" in function definition in filtertest.c.
1917 PH/32 Applied Sieve patches from the maintainer.
1919 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1921 PH/33 Added "verify = not_blind".
1923 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1924 Local/Makefile (with some defaults set). These are used in built scripts
1925 such as exicyclog, but they have never been used in the exim_install
1926 script (though there are many overriding facilities there). I have
1927 arranged that the exim_install script now takes note of these two
1930 PH/35 Installed configuration files for Dragonfly.
1932 PH/36 When a locally submitted message by a trusted user did not contain a
1933 From: header, and the sender address was obtained from -f or from an SMTP
1934 MAIL command, and the trusted user did not use -F to supply a sender
1935 name, $originator_name was incorrectly used when constructing a From:
1936 header. Furthermore, $originator_name was used for submission mode
1937 messages from external hosts without From: headers in a similar way,
1938 which is clearly wrong.
1940 PH/37 Added control=suppress_local_fixups.
1942 PH/38 When log_selector = +received_sender was set, and the addition of the
1943 sender made the log line's construction buffer exactly full, or one byte
1944 less than full, an overflow happened when the terminating "\n" was
1947 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1948 when the result of a list match is failure because a DNS lookup failed.
1950 PH/40 RM_COMMAND is now used in the building process.
1952 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1953 the "build-* directories that it finds.
1955 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1956 address in a domain literal was a prefix of an interface address.
1958 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1959 when verifying a sender address, unless rewrite_headers is false.
1961 PH/44 Wrote a long comment about why errors_to addresses are verified as
1962 recipients, not senders.
1964 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1965 the ratelimit ACL was added.
1967 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1969 PH/46 Added extra information about PostgreSQL errors to the error string.
1971 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1972 avoiding the need to install special zones in a real server. This is
1973 backwards compatible; if it can't find the fake resolver, it drops back.
1974 Thus, both old and new test suites can be run.
1976 TF/02 Added util/ratelimit.pl
1978 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1979 clock is set back in time.
1981 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1982 Candler <B.Candler@pobox.com>.
1984 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1985 OK for addresses that are the result of redirections.
1987 PH/48 A number of further additions for the benefit of the new test suite,
1988 including a fake gethostbyname() that interfaces to the fake DNS resolver
1991 TF/06 The fix for widen_domains has also been applied to qualify_single and
1992 search_parents which are the other dnslookup options that can cause
1995 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1998 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2000 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2001 service" if it encountered an MX record with an empty target hostname.
2002 The message is now "an MX or SRV record indicated no SMTP service".
2004 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2005 if the DNS of the sending site is misconfigured. This is quite a
2006 common situation. This change restores the behaviour of treating a
2007 helo verification defer as a failure.
2009 PH/51 If self=fail was set on a router, the bounce message did not include the
2010 actual error message.
2016 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2018 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2019 command itself ended in error, the underlying message about the transport
2020 filter timeout was being overwritten with the pipe command error. Now the
2021 underlying error message should be appended to the second error message.
2023 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2025 PH/02 Reduce the amount of output that "make" produces by default. Full output
2026 can still be requested.
2028 PH/03 The warning log line about a condition test deferring for a "warn" verb
2029 was being output only once per connection, rather than after each
2030 occurrence (because it was using the same function as for successful
2031 "warn" verbs). This seems wrong, so I have changed it.
2033 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2034 it should not have, which might have caused a crash in the right
2035 circumstances, but probably never did.
2037 PH/04 Installed a modified version of Tony Finch's patch to make submission
2038 mode fix the return path as well as the Sender: header line, and to
2039 add a /name= option so that you can make the user's friendly name appear
2042 TF/03 Added the control = fakedefer ACL modifier.
2044 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2045 Mark Lowes for thorough testing.
2047 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2049 TK/03 Merged latest SRS patch from Miles Wilton.
2051 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2052 with the definition in sysexits.h (which is #included earlier).
2053 Fortunately, Exim does not actually use EX_OK. The code used to try to
2054 preserve the sysexits.h value, by assumimg that macro definitions were
2055 scanned for macro replacements. I have been disabused of this notion,
2056 so now the code just undefines EX_OK before #including unistd.h.
2058 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2059 in the smtp transport. When a block could not be written in a single
2060 write() function, the timeout was being re-applied to each part-write.
2061 This seems wrong - if the receiver was accepting one byte at a time it
2062 would take for ever. The timeout is now adjusted when this happens. It
2063 doesn't have to be particularly precise.
2065 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2066 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2068 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2069 without a domain if the check to <postmaster@domain> fails.
2071 SC/01 Eximstats: added -xls and the ability to specify output files
2072 (patch written by Frank Heydlauf).
2074 SC/02 Eximstats: use FileHandles for outputing results.
2076 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2078 SC/04 Eximstats: fixed display of large numbers with -nvr option
2080 SC/05 Eximstats: fixed merging of reports with empty tables.
2082 SC/06 Eximstats: added the -include_original_destination flag
2084 SC/07 Eximstats: removed tabs and trailing whitespace.
2086 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2088 TK/06 MBOX spool code: Add real "From " MBOX separator line
2089 so the .eml file is really in mbox format (even though
2090 most programs do not really care). Patch from Alex Miller.
2092 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2093 The latter is generated from $received_to and is only set if the
2094 message has one envelope recipient. SA can use these headers,
2095 obviously out-of-the-box. Patch from Alex Miller.
2097 PH/08 The ${def test on a variable was returning false if the variable's
2098 value was "0", contrary to what the specification has always said!
2099 The result should be true unless the variable is empty.
2101 PH/09 The syntax error of a character other than { following "${if
2102 def:variable_name" (after optional whitespace) was not being diagnosed.
2103 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2104 accidental colon was present, for example, could give incorrect results.
2106 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2107 result is used (not including appendfile, where other changes are about
2110 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2111 This involved changing a lot of size variables from int to off_t. It
2112 should work with maildirs and everything.
2114 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2115 spamd dying while we are connected to it.
2117 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2120 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2121 function for BDB 4.3.
2123 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2125 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2126 the sources for PCRE have been re-arranged and more files are now
2129 PH/15 The code I had for printing potentially long long variables in PH/11
2130 above was not the best (it lost precision). The length of off_t variables
2131 is now inspected at build time, and an appropriate printing format (%ld
2132 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2133 to be "long long int" or "long int". This is needed for the internal
2134 formatting function string_vformat().
2136 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2137 the configuration file to be ":syslog", then the script "guesses" where
2138 the logs files are, rather than using the compiled in default. In our
2139 case the guess is not the same as the compiled default, so the script
2140 suddenly stopped working when I started to use syslog. The patch checks
2141 to see if log_file_path is "". If so, it attempts to read it from exim
2142 with no configuration file to get the compiled in version, before it
2143 falls back to the previous guessing code."
2145 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2146 implementing BATV in an Exim configuration. See NewStuff for the gory
2149 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2150 Makefile that are specific to HP-UX.
2152 PH/18 If the "use_postmaster" option was set for a recipient callout together
2153 with the "random" option, the postmaster address was used as the MAIL
2154 FROM address for the random test, but not for the subsequent recipient
2155 test. It is now used for both.
2157 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2158 patch removes a few documentation additions to RFC 3028, because the
2159 latest draft now contains them. It adds the new en;ascii-case comparator
2160 and a new error check for 8bit text in MIME parts. Comparator and
2161 require names are now matched exactly. I enabled the subaddress
2162 extension, but it is not well tested yet (read: it works for me)."
2164 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2165 rework some of the code of TK/09 above to avoid the hardwired use of
2166 "%lld" and "long long". Replaced the call to snprintf() with a call to
2169 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2170 records point to non-existent hosts", "retry timeout exceeded", and
2171 "retry time not reached for any host after a long failure period".
2173 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2174 experimental DomainKeys support:
2176 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2177 (2) On an error such as an illegally used "control", the wrong name for
2178 the control was given.
2180 These problems did NOT occur unless DomainKeys support was compiled.
2182 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2184 PH/24 Added ${if match_ip condition.
2186 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2187 databases so that it will be absolutely obvious if a crash occurs in the
2188 DB library. This is a regular occurrence (often caused by mis-matched
2191 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2192 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2193 picked up on a user's system that detects such things. There doesn't seem
2194 to be a gcc warning option for this - only an attribute that has to be
2195 put on the function's prototype. It seems that in Fedora Core 4 they have
2196 set this on a number of new functions. No doubt there will be more in due
2199 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2200 specify a transport. However, if an address that was verified by such a
2201 router was the subject of a callout, Exim crashed because it tried to
2202 read the rcpt_include_affixes from the non-existent transport. Now it
2203 just assumes that the setting of that option is false. This bug was
2204 introduced by 4.51/PH/31.
2206 PH/28 Changed -d+all to exclude +memory, because that information is very
2207 rarely of interest, but it makes the output a lot bigger. People tend to
2208 do -d+all out of habit.
2210 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2211 code in os-type was giving problems when libc.so lives in lib64, like on
2214 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2215 aren't the modern standard, and it seems that some systems' include files
2216 don't always have them. Exim was already checking for some of the newer
2217 ones like T_AAAA, and defining it itself. I've added checks for all the
2218 record types that Exim uses.
2220 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2221 not automatically generating a new one, as it is supposed to. This
2222 prevented TLS from working. If the file did exist, but contained invalid
2223 data, a new version was generated, as expected. It was only the case of a
2224 non-existent file that was broken.
2226 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2227 with a change in libdomainkeys > 0.64.
2229 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2230 from DNS. If the selector record carries the flag, it now has
2231 precedence over the domain-wide flag.
2233 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2235 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2236 the use of a port name that isn't defined in /etc/services) occurred, the
2237 message was deferred as in a normal delivery, and thus remained on the
2238 spool, instead of being failed because of the mua_wrapper setting. This
2239 is now fixed, and I tidied up some of the mua_wrapper messages at the
2242 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2243 the messages in a hash of arrays rather than using individual hashes.
2244 This is a bit cleaner and results in dramatic memory savings, albeit
2245 at a slight CPU cost.
2247 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2248 as requested by Marc Sherman.
2250 SC/10 Eximstats: added histograms for user specified patterns as requested
2253 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2255 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2256 fopen() in the content-scanning modules that did not already have it.
2262 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2263 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2265 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2267 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2269 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2270 to test Sieve filters that use "vacation".
2272 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2273 that changes the way the GnuTLS parameters are stored in the cache file.
2274 The new format can be generated externally. For backward compatibility,
2275 if the data in the cache doesn't make sense, Exim assumes it has read an
2276 old-format file, and it generates new data and writes a new file. This
2277 means that you can't go back to an older release without removing the
2280 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2281 work if there are any delivery delays because "one_time" forces the
2282 parent to be marked "delivered", so its unseen clone is never tried
2283 again. For this reason, Exim now forbids the simultaneous setting of
2286 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2287 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2288 entry if you want to know the details. The fix, however, neglected to
2289 consider the case when local delivery batching is involved. The test for
2290 "previously delivered" was not happening when checking to see if an
2291 address could be batched with a previous (undelivered) one; under
2292 certain circumstances this could lead to multiple deliveries to the same
2295 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2296 in its include files, and this causes problems building Exim.
2298 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2299 header_syntax) but Exim was just ignoring anything given after a slash.
2300 In particular, this caused confusion with an attempt to use "verify =
2301 reverse_host_lookup/defer_ok". An error is now given when options are
2302 supplied for verify items that do not have them. (Maybe reverse_host_
2303 lookup should have a defer_ok option, but that's a different point.)
2305 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2306 defined by RFC 821) to 2048, because there were problems with some AUTH
2307 commands, and RFC 1869 says the size should be increased for extended
2308 SMTP commands that take arguments.
2310 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2313 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2314 "unknown" error; now it says that the functionality isn't in the binary.
2316 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2317 an address' error message when a string expansion fails (syntax or
2318 whatever). Otherwise the password may appear in the log. Following change
2319 PH/42 below, there is no longer a chance of it appearing in a bounce
2322 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2324 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2325 of its addresses was ever tried. (Bugzilla bug #2.)
2327 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2328 the result incorrectly in the debug output. (It correctly added a newline
2329 to what was transported.)
2331 TF/01 Added $received_time.
2333 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2334 commented out examples of how to interface to a virus scanner and to
2335 SpamAssassin. Also added commented examples of av_scanner and
2336 spamd_address settings.
2338 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2339 and controls are allowed in which ACLs. There were a couple of minor
2340 errors. Some of the entries in the conditions table (which is a table of
2341 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2342 negation of where the condition IS allowed.
2344 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2346 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2347 header file does not have a version number, so I've had to invent a new
2348 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2349 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2350 radiusclient), but it was contributed by a Radius user.
2352 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2353 files or directories, for OpenSSL.
2355 PH/19 When an Exim process that is running as root has to create an Exim log
2356 file, it does so in a subprocess that runs as exim:exim so as to get the
2357 ownership right at creation (otherwise, other Exim processes might see
2358 the file with the wrong ownership). There was no test for failure of this
2359 fork() call, which would lead to the process getting stuck as it waited
2360 for a non-existent subprocess. Forks do occasionally fail when resources
2361 run out. I reviewed all the other calls to fork(); they all seem to check
2364 PH/20 When checking for unexpected SMTP input at connect time (before writing
2365 the banner), Exim was not dealing correctly with a non-positive return
2366 from the read() function. If the client had disconnected by this time,
2367 the result was a log entry for a synchronization error with an empty
2368 string after "input=" when read() returned zero. If read() returned -1
2369 (an event I could not check), uninitialized data bytes were printed.
2370 There were reports of junk text (parts of files, etc) appearing after
2373 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2375 PH/22 Added support for macro redefinition, and (re)definition in between
2376 driver and ACL definitions.
2378 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2379 forgetting to use the resulting value; it was using the unexpanded value.
2381 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2382 hadn't been configured. The fix is from Juergen Kreileder, who
2383 understands it better than I do:
2385 "Here's what I see happening with three configured cyrus_sasl
2386 authenticators configured (plain, login, cram-md5):
2388 On startup auth_cyrus_sasl_init() gets called for each of these.
2389 This means three calls to sasl_listmech() without a specified mech_list.
2390 => SASL tests which mechs of all available mechs actually work
2391 => three warnings about OTP not working
2392 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2394 With the patch, sasl_listmech() also gets called three times. But now
2395 SASL's mech_list option is set to the server_mech specified in the the
2396 authenticator. Or in other words, the answer from sasl_listmech()
2397 gets limited to just the mech you're testing for (which is different
2399 => the return list contains just 'plain' or 'login', 'cram-md5' or
2400 nothing depending on the value of ob->server_mech.
2402 I've just tested the patch: Authentication still works fine,
2403 unavailable mechs specified in the exim configuration are still
2404 caught, and the auth.log warnings about OTP are gone."
2406 PH/25 When debugging is enabled, the contents of the command line are added
2407 to the debugging output, even when log_selector=+arguments is not
2410 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2411 answer is "GNU", and only if the return is "GNU/something" is the answer
2414 PH/27 $acl_verify_message is now set immediately after the failure of a
2415 verification in an ACL, and so is available in subsequent modifiers. In
2416 particular, the message can be preserved by coding like this:
2418 warn !verify = sender
2419 set acl_m0 = $acl_verify_message
2421 Previously, $acl_verify_message was set only while expanding "message"
2422 and "log_message" when a very denied access.
2424 PH/28 Modified OS/os.c-Linux with
2426 -#ifndef OS_LOAD_AVERAGE
2427 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2429 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2430 nomenclature these days.)
2432 PH/29 Installed patch from the Sieve maintainer that adds the options
2433 sieve_useraddress and sieve_subaddress to the redirect router.
2435 PH/30 In these circumstances:
2436 . Two addresses routed to the same list of hosts;
2437 . First host does not offer TLS;
2438 . First host accepts first address;
2439 . First host gives temporary error to second address;
2440 . Second host offers TLS and a TLS session is established;
2441 . Second host accepts second address.
2442 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2443 and peerdn, if requested) that were in fact used only for the second
2446 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2447 attention to any local part prefix or suffix that was matched by the
2448 router that accepted the address. It now behaves in the same way as it
2449 does for delivery: the affixes are removed from the local part unless
2450 rcpt_include_affixes is set on the transport.
2452 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2453 timeout during the DATA phase of an incoming message.
2455 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2456 applied a patch sanctioned by the Sieve maintainer.
2458 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2459 the uid or gid is negative. A case of a negative gid caused this to be
2460 noticed. The fix allows for either to be negative.
2462 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2463 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2466 PH/36 The change PH/12 above was broken. Fixed it.
2468 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2469 the grounds that routing the same address twice would always produce the
2470 same answer. This might have been true once, but it is certainly no
2471 longer true now. Routing a child address may depend on the previous
2472 routing that produced that child. Some complicated redirection strategies
2473 went wrong when messages had multiple recipients, and made Exim's
2474 behaviour dependent on the order in which the addresses were given.
2476 I have moved the duplicate checking until after the routing is complete.
2477 Exim scans the addresses that are assigned to local and remote
2478 transports, and removes any duplicates. This means that more work will be
2479 done, as duplicates will always all be routed, but duplicates are
2480 presumably rare, so I don't expect this is of any significance.
2482 For deliveries to pipes, files, and autoreplies, the duplicate checking
2483 still happens during the routing process, since they are not going to be
2486 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2487 It corrects a timeout issue with spamd. This is Ian's comment: "The
2488 background is that sometimes spamd either never reads data from a
2489 connection it has accepted, or it never writes response data. The exiscan
2490 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2491 blindly assumes that writes won't block so it may never time out."
2493 PH/39 Allow G after quota size as well as K and M.
2495 PH/40 The value set for $authenticated_id in an authenticator may not contain
2496 binary zeroes or newlines because the value is written to log lines and
2497 to spool files. There was no check on this. Now the value is run through
2498 the string_printing() function so that such characters are converted to
2499 printable escape sequences.
2501 PH/41 $message_linecount is a new variable that contains the total number of
2502 lines in the message. Compare $body_linecount, which is the count for the
2505 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2506 bounce and delay warning messages, except in certain special cases, which
2509 (a) An SMTP error message from a remote host;
2510 (b) A message specified in a :fail: redirection;
2511 (c) A message specified in a "fail" command in a system filter;
2512 (d) A message specified in a FAIL return from the queryprogram router;
2513 (e) A message specified by the cannot_route_message router option.
2515 In these cases only, Exim does include the error details in bounce and
2516 warning messages. There are also a few cases where bland messages such
2517 as "unrouteable address" or "local delivery error" are given.
2519 PH/43 $value is now also set for the "else" part of a ${run expansion.
2521 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2522 being worked on, but at least Exim now implements the latest version to
2525 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2526 process to complete was treated as a delivery failure, a timeout while
2527 writing the message to the pipe was logged, but erroneously treated as a
2528 successful delivery. Such timeouts include transport filter timeouts. For
2529 consistency with the overall process timeout, these timeouts are now
2530 treated as errors, giving rise to delivery failures by default. However,
2531 there is now a new Boolean option for the pipe transport called
2532 timeout_defer, which, if set TRUE, converts the failures into defers for
2533 both kinds of timeout. A transport filter timeout is now identified in
2536 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2537 systems where "make" and "gmake" are different, calling "gmake" at top
2538 level broke things. I've arranged for the value of $(MAKE) to be passed
2539 from the Makefile to this script so that it can call the same version of
2543 A note about Exim versions 4.44 and 4.50
2544 ----------------------------------------
2546 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2547 changes of various kinds. As a consequence, a big documentation update was
2548 needed. This delayed the release for rather longer than seemed good, especially
2549 in the light of a couple of (minor) security issues. Therefore, the changes
2550 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2551 release. So 4.44 and 4.50 are in effect two different branches that both start
2554 I have left the 4.50 change log unchanged; it contains all the changes since
2555 4.43. The change log for 4.44 is below; many of its items are identical to
2556 those for 4.50. This seems to be the most sensible way to preserve the
2557 historical information.
2563 1. Minor wording change to the doc/README.SIEVE file.
2565 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2566 computation of the current number of files was incorrect.
2568 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2569 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2572 4. Give more explanation in the error message when the command for a transport
2573 filter fails to execute.
2575 5. There are several places where Exim runs a non-Exim command in a
2576 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2577 was being done only for the command run by the queryprogram router. It is
2578 now done for all such subprocesses. The other cases are: ${run, transport
2579 filters, and the commands run by the lmtp and pipe transports.
2581 6. Added CONFIGURE_GROUP build-time option.
2583 7. Some older OS have a limit of 256 on the maximum number of file
2584 descriptors. Exim was using setrlimit() to set 1000 as a large value
2585 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2586 systems. I've change it so that if it can't get 1000, it tries for 256.
2588 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2589 was an oversight, and furthermore, ever since the addition of extra
2590 controls (e.g. 4.43/32), the checks on when to allow different forms of
2591 "control" were broken. There should now be diagnostics for all cases when a
2592 control that does not make sense is encountered.
2594 9. Added the /retain_sender option to "control=submission".
2596 10. $recipients is now available in the predata ACL (oversight).
2598 11. Tidy the search cache before the fork to do a delivery from a message
2599 received from the command line. Otherwise the child will trigger a lookup
2600 failure and thereby defer the delivery if it tries to use (for example) a
2601 cached ldap connection that the parent has called unbind on.
2603 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2604 of $address_data from the recipient verification was clobbered by the
2605 sender verification.
2607 13. The value of address_data from a sender verification is now available in
2608 $sender_address_data in subsequent conditions in the ACL statement.
2610 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2612 15. Added a new option "connect=<time>" to callout options, to set a different
2615 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2616 was its contents. (It was OK if the option was not defined at all.)
2618 17. A "Completed" log line is now written for messages that are removed from
2619 the spool by the -Mrm option.
2621 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2622 information about exactly what failed.
2624 19. Added -dd to debug only the daemon process.
2626 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2627 handles timeouts, both on the server side and network timeouts. Renamed the
2628 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2630 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2631 It is now set to "smtps".
2633 22. $host_address is now set to the target address during the checking of
2634 ignore_target_hosts.
2636 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2637 being passed; this would have caused $sender_host_name to have been used if
2638 matching the list had actually called for a host name (not very likely,
2639 since this list is usually IP addresses). A host name is now passed as
2642 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2643 code that specifies a non-restarting handler (typically sigaction() in
2644 modern systems) in an attempt to fix a rare and obscure crash bug.
2646 25. Narrowed the window for a race in the daemon that could cause it to ignore
2647 SIGCHLD signals. This is not a major problem, because they are used only to
2648 wake it up if nothing else does.
2650 26. A malformed maildirsize file could cause Exim to calculate negative values
2651 for the mailbox size or file count. Odd effects could occur as a result.
2652 The maildirsize information is now recalculated if the size or filecount
2655 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2656 support for a long time. Removed HAVE_SYS_VFS_H.
2658 28. Installed the latest version of exipick from John Jetmore.
2660 29. In an address list, if the pattern was not a regular expression, an empty
2661 subject address (from a bounce message) matched only if the pattern was an
2662 empty string. Non-empty patterns were not even tested. This was the wrong
2663 because it is perfectly reasonable to use an empty address as part of a
2664 database query. An empty address is now tested by patterns that are
2665 lookups. However, all the other forms of pattern expect the subject to
2666 contain a local part and a domain, and therefore, for them, an empty
2667 address still always fails if the pattern is not itself empty.
2669 30. Exim went into a mad DNS loop when attempting to do a callout where the
2670 host was specified on an smtp transport, and looking it up yielded more
2671 than one IP address.
2673 31. Re-factored the code for checking spool and log partition space into a
2674 function that finds that data and another that does the check. The former
2675 is then used to implement four new variables: $spool_space, $log_space,
2676 $spool_inodes, and $log_inodes.
2678 32. The RFC2047 encoding function was originally intended for short strings
2679 such as real names; it was not keeping to the 75-character limit for
2680 encoded words that the RFC imposes. It now respects the limit, and
2681 generates multiple encoded words if necessary. To be on the safe side, I
2682 have increased the buffer size for the ${rfc2047: expansion operator from
2685 33. It is now permitted to omit both strings after an "if" condition; if the
2686 condition is true, the result is "true". As before, when the second string
2687 is omitted, a false condition yields an empty string. This makes it less
2688 cumbersome to write custom ACL and router conditions.
2690 34. Failure to deliver a bounce message always caused it to be frozen, even if
2691 there was an errors_to setting on the router. The errors_to setting is now
2694 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2695 canonical form (fully expanded) before being placed in
2696 $sender_host_address.
2698 36. The table in the code that translates DNS record types into text (T_A to
2699 "A" for instance) was missing entries for NS and CNAME. It is just possible
2700 that this could have caused confusion if both these types were looked up
2701 for the same domain, because the text type is used as part of Exim's
2702 per-process caching. But the chance of anyone hitting this buglet seems
2705 37. The dnsdb lookup has been extended in a number of ways.
2707 (1) There is a new type, "zns", which walks up the domain tree until it
2708 finds some nameserver records. It should be used with care.
2710 (2) There is a new type, "mxh", which is like "mx" except that it returns
2711 just the host names, not the priorities.
2713 (3) It is now possible to give a list of domains (or IP addresses) to be
2714 looked up. The behaviour when one of the lookups defers can be
2715 controlled by a keyword.
2717 (4) It is now possible to specify the separator character for use when
2718 multiple records are returned.
2720 38. The dnslists ACL condition has been extended: it is now possible to supply
2721 a list of IP addresses and/or domains to be looked up in a particular DNS
2724 39. Added log_selector=+queue_time_overall.
2726 40. When running the queue in the test harness, wait just a tad after forking a
2727 delivery process, to get repeatability of debugging output.
2729 41. Include certificate and key file names in error message when GnuTLS fails
2730 to set them up, because the GnuTLS error message doesn't include the name
2731 of the failing file when there is a problem reading it.
2733 42. Allow both -bf and -bF in the same test run.
2735 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2737 44. The "Exiscan patch" is now merged into the mainline Exim source.
2739 45. Sometimes the final signoff response after QUIT could fail to get
2740 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2741 < 0 before doing a fflush(). This bug looks as though it goes back to the
2742 introduction of TLS in release 3.20, but "sometimes" must have been rare
2743 because the tests only now provoked it.
2745 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2746 (this can affect the format of dates).
2748 47. exim_tidydb, when checking for the continued existence of a message for
2749 which it has found a message-specific retry record, was not finding
2750 messages that were in split spool directories. Consequently, it was
2751 deleting retry records that should have stayed in existence.
2753 48. Steve fixed some bugs in eximstats.
2755 49. The SPA authentication driver was not abandoning authentication and moving
2756 on to the next authenticator when an expansion was forced to fail,
2757 contradicting the general specification for all authenticators. Instead it
2758 was generating a temporary error. It now behaves as specified.
2760 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2761 (the order specifies the preference for clients). The order is now AES256,
2762 AES128, 3DES, ARCFOUR128.
2764 51. Small patch to Sieve code - explicitly set From: when generating an
2767 52. Exim crashed if a remote delivery caused a very long error message to be
2768 recorded - for instance if somebody sent an entire SpamAssassin report back
2769 as a large number of 550 error lines. This bug was coincidentally fixed by
2770 increasing the size of one of Exim's internal buffers (big_buffer) that
2771 happened as part of the Exiscan merge. However, to be on the safe side, I
2772 have made the code more robust (and fixed the comments that describe what
2775 53. Now that there can be additional text after "Completed" in log lines (if
2776 the queue_time_overall log selector is set), a one-byte patch to exigrep
2777 was needed to allow it to recognize "Completed" as not the last thing in
2780 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2781 patch that reportedly fixes this has been added. I am not expert enough to
2782 create a test for it. This is what the patch creator wrote:
2784 "I found a little strange behaviour of ldap code when working with
2785 Windows 2003 AD Domain, where users was placed in more than one
2786 Organization Units. When I tried to give exim partial DN, the exit code
2787 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2788 But simultaneously result of request was absolutely normal ldap result,
2789 so I produce this patch..."
2791 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2792 REFERENCE, so I have modified the code to exclude the patch when that macro
2795 55. Some experimental protocols are using DNS PTR records for new purposes. The
2796 keys for these records are domain names, not reversed IP addresses. The
2797 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2798 leaves it alone. Component reversal etc. now happens only for IP addresses.
2801 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2803 57. Double the size of the debug message buffer (to 2048) so that more of very
2804 long debug lines gets shown.
2806 58. The exicyclog utility now does better if the number of log files to keep
2807 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2809 59. Two changes related to the smtp_active_hostname option:
2811 (1) $smtp_active_hostname is now available as a variable.
2812 (2) The default for smtp_banner uses $smtp_active_hostname instead
2813 of $primary_hostname.
2815 60. The host_aton() function is supposed to be passed a string that is known
2816 to be a valid IP address. However, in the case of IPv6 addresses, it was
2817 not checking this. This is a hostage to fortune. Exim now panics and dies
2818 if the condition is not met. A case was found where this could be provoked
2819 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2820 components; fortuitously, this particular loophole had already been fixed
2821 by change 4.50/55 above.
2823 If there are any other similar loopholes, the new check in host_aton()
2824 itself should stop them being exploited. The report I received stated that
2825 data on the command line could provoke the exploit when Exim was running as
2826 exim, but did not say which command line option was involved. All I could
2827 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2828 running as the user.
2831 61. There was a buffer overflow vulnerability in the SPA authentication code
2832 (which came originally from the Samba project). I have added a test to the
2833 spa_base64_to_bits() function which I hope fixes it.
2836 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2837 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2839 63. The daemon start-up calls getloadavg() while still root for those OS that
2840 need the first call to be done as root, but it missed one case: when
2841 deliver_queue_load_max is set with deliver_drop_privilege. This is
2842 necessary for the benefit of the queue runner, because there is no re-exec
2843 when deliver_drop_privilege is set.
2845 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2846 This has been fixed.
2848 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2849 was in use, was not putting the data itself into the right store pool;
2850 consequently, it could be overwritten for a subsequent message in the same
2851 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2854 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2856 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2857 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2858 the function were treating the return as a boolean value, which happened to
2859 work because 0=false and not-0=true, but is not correct code.
2861 68. The host_aton() function was not handling scoped IPv6 addresses (those
2862 with, for example, "%eth0" on the end) correctly.
2864 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2865 negated items (that is, ~something) in unsigned ints. Some compilers
2866 apparently mutter when there is no cast.
2868 70. If an address verification called from an ACL failed, and did not produce a
2869 user-specific message (i.e. there was only a "system" message), nothing was
2870 put in $acl_verify_message. In this situation, it now puts the system
2873 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2874 session; change 4.31/43 added the unwanted input to the log line - except
2875 that it did not do this in the start of session case. It now does.
2877 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2878 This is wrong and can cause the other end to generate a synchronization
2879 error if it is another Exim or anything else that does the synchronization
2880 check. A QUIT command is no longer sent after a timeout.
2882 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2883 during host lookups.
2885 74. The defer_ok option of callout verification was not working if it was used
2886 when verifying addresses in header lines, that is, for this case:
2888 verify = header_sender/callout=defer_ok
2890 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2891 those file descriptors could be used for SMTP connections. If anything
2892 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2893 could be sent to the SMTP client, causing chaos. The daemon now opens
2894 stdin, stdout, and stderr to /dev/null when it puts itself into the
2897 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2898 log by default. The user can override this with suitable Perl magic.
2900 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2901 the log message when discard triggers, was not working for the DATA ACL or
2902 for the non-SMTP ACL.
2904 78. Error message wording change in sieve.c.
2906 79. If smtp_accept_max_per_host was set, the number of connections could be
2907 restricted to fewer than expected, because the daemon was trying to set up
2908 a new connection before checking whether the processes handling previous
2909 connections had finished. The check for completed processes is now done
2910 earlier. On busy systems, this bug wouldn't be noticed because something
2911 else would have woken the daemon, and it would have reaped the completed
2914 80. If a message was submitted locally by a user whose login name contained one
2915 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2916 It caused a spool format error. I have fixed the spool reading code. A
2917 related problem was that the "from" clause in the Received: line became
2918 illegal because of the space(s). It is now covered by ${quote_local_part.
2920 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2923 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2925 83. Patch from Sieve maintainer to fix unterminated string problem in
2926 "vacation" handling.
2928 84. Some minor changes to the Linux configuration files to help with other
2929 OS variants using glibc.
2931 85. One more patch for Sieve to update vacation handling to latest spec.
2934 ----------------------------------------------------
2935 See the note above about the 4.44 and 4.50 releases.
2936 ----------------------------------------------------
2942 1. Change 4.43/35 introduced a bug that caused file counts to be
2943 incorrectly computed when quota_filecount was set in an appendfile
2946 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2947 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2950 3. Give more explanation in the error message when the command for a transport
2951 filter fails to execute.
2953 4. There are several places where Exim runs a non-Exim command in a
2954 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2955 was being done only for the command run by the queryprogram router. It is
2956 now done for all such subprocesses. The other cases are: ${run, transport
2957 filters, and the commands run by the lmtp and pipe transports.
2959 5. Some older OS have a limit of 256 on the maximum number of file
2960 descriptors. Exim was using setrlimit() to set 1000 as a large value
2961 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2962 systems. I've change it so that if it can't get 1000, it tries for 256.
2964 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2965 was an oversight, and furthermore, ever since the addition of extra
2966 controls (e.g. 4.43/32), the checks on when to allow different forms of
2967 "control" were broken. There should now be diagnostics for all cases when a
2968 control that does not make sense is encountered.
2970 7. $recipients is now available in the predata ACL (oversight).
2972 8. Tidy the search cache before the fork to do a delivery from a message
2973 received from the command line. Otherwise the child will trigger a lookup
2974 failure and thereby defer the delivery if it tries to use (for example) a
2975 cached ldap connection that the parent has called unbind on.
2977 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2978 of $address_data from the recipient verification was clobbered by the
2979 sender verification.
2981 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2982 was its contents. (It was OK if the option was not defined at all.)
2984 11. A "Completed" log line is now written for messages that are removed from
2985 the spool by the -Mrm option.
2987 12. $host_address is now set to the target address during the checking of
2988 ignore_target_hosts.
2990 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2991 being passed; this would have caused $sender_host_name to have been used if
2992 matching the list had actually called for a host name (not very likely,
2993 since this list is usually IP addresses). A host name is now passed as
2996 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2997 code that specifies a non-restarting handler (typically sigaction() in
2998 modern systems) in an attempt to fix a rare and obscure crash bug.
3000 15. Narrowed the window for a race in the daemon that could cause it to ignore
3001 SIGCHLD signals. This is not a major problem, because they are used only to
3002 wake it up if nothing else does.
3004 16. A malformed maildirsize file could cause Exim to calculate negative values
3005 for the mailbox size or file count. Odd effects could occur as a result.
3006 The maildirsize information is now recalculated if the size or filecount
3009 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3010 support for a long time. Removed HAVE_SYS_VFS_H.
3012 18. Updated exipick to current release from John Jetmore.
3014 19. Allow an empty sender to be matched against a lookup in an address list.
3015 Previously the only cases considered were a regular expression, or an
3018 20. Exim went into a mad DNS lookup loop when doing a callout where the
3019 host was specified on the transport, if the DNS lookup yielded more than
3022 21. The RFC2047 encoding function was originally intended for short strings
3023 such as real names; it was not keeping to the 75-character limit for
3024 encoded words that the RFC imposes. It now respects the limit, and
3025 generates multiple encoded words if necessary. To be on the safe side, I
3026 have increased the buffer size for the ${rfc2047: expansion operator from
3029 22. Failure to deliver a bounce message always caused it to be frozen, even if
3030 there was an errors_to setting on the router. The errors_to setting is now
3033 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3034 canonical form (fully expanded) before being placed in
3035 $sender_host_address.
3037 24. Updated eximstats to version 1.33
3039 25. Include certificate and key file names in error message when GnuTLS fails
3040 to set them up, because the GnuTLS error message doesn't include the name
3041 of the failing file when there is a problem reading it.
3043 26. Expand error message when OpenSSL has problems setting up cert/key files.
3046 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3047 (this can affect the format of dates).
3049 28. exim_tidydb, when checking for the continued existence of a message for
3050 which it has found a message-specific retry record, was not finding
3051 messages that were in split spool directories. Consequently, it was
3052 deleting retry records that should have stayed in existence.
3054 29. eximstats updated to version 1.35
3055 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3056 1.35 - bugfix such that pie charts by volume are generated correctly
3058 30. The SPA authentication driver was not abandoning authentication and moving
3059 on to the next authenticator when an expansion was forced to fail,
3060 contradicting the general specification for all authenticators. Instead it
3061 was generating a temporary error. It now behaves as specified.
3063 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3064 (the order specifies the preference for clients). The order is now AES256,
3065 AES128, 3DES, ARCFOUR128.
3067 31. Small patch to Sieve code - explicitly set From: when generating an
3070 32. Exim crashed if a remote delivery caused a very long error message to be
3071 recorded - for instance if somebody sent an entire SpamAssassin report back
3072 as a large number of 550 error lines. This bug was coincidentally fixed by
3073 increasing the size of one of Exim's internal buffers (big_buffer) that
3074 happened as part of the Exiscan merge. However, to be on the safe side, I
3075 have made the code more robust (and fixed the comments that describe what
3078 33. Some experimental protocols are using DNS PTR records for new purposes. The
3079 keys for these records are domain names, not reversed IP addresses. The
3080 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3081 leaves it alone. Component reversal etc. now happens only for IP addresses.
3084 34. The host_aton() function is supposed to be passed a string that is known
3085 to be a valid IP address. However, in the case of IPv6 addresses, it was
3086 not checking this. This is a hostage to fortune. Exim now panics and dies
3087 if the condition is not met. A case was found where this could be provoked
3088 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3089 components; fortuitously, this particular loophole had already been fixed
3090 by change 4.50/55 or 4.44/33 above.
3092 If there are any other similar loopholes, the new check in host_aton()
3093 itself should stop them being exploited. The report I received stated that
3094 data on the command line could provoke the exploit when Exim was running as
3095 exim, but did not say which command line option was involved. All I could
3096 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3097 running as the user.
3100 35. There was a buffer overflow vulnerability in the SPA authentication code
3101 (which came originally from the Samba project). I have added a test to the
3102 spa_base64_to_bits() function which I hope fixes it.
3105 36. The daemon start-up calls getloadavg() while still root for those OS that
3106 need the first call to be done as root, but it missed one case: when
3107 deliver_queue_load_max is set with deliver_drop_privilege. This is
3108 necessary for the benefit of the queue runner, because there is no re-exec
3109 when deliver_drop_privilege is set.
3111 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3112 was in use, was not putting the data itself into the right store pool;
3113 consequently, it could be overwritten for a subsequent message in the same
3114 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3117 38. Sometimes the final signoff response after QUIT could fail to get
3118 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3119 < 0 before doing a fflush(). This bug looks as though it goes back to the
3120 introduction of TLS in release 3.20, but "sometimes" must have been rare
3121 because the tests only now provoked it.
3127 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3128 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3129 Now it returns an integer. A number of calls were still expecting a T/F
3130 return. Fortuitously, in all cases, the tests worked in OK situations,
3131 which is the norm. However, things would have gone wrong on any write
3132 failures on the smtp file descriptor. This function is used when sending
3133 messages over SMTP and also when doing verify callouts.
3135 2. When Exim is called to do synchronous delivery of a locally submitted
3136 message (the -odf or -odi options), it no longer closes stderr before doing
3139 3. Implemented the mua_wrapper option.
3141 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3143 5. Implemented the functions header_remove(), header_testname(),
3144 header_add_at_position(), and receive_remove_recipient(), and exported them
3147 6. If an ACL "warn" statement specified the addition of headers, Exim already
3148 inserted X-ACL-Warn: at the start if there was no header name. However, it
3149 was not making this test for the second and subsequent header lines if
3150 there were newlines in the string. This meant that an invalid header could
3151 be inserted if Exim was badly configured.
3153 7. Allow an ACL "warn" statement to add header lines at the start or after all
3154 the Received: headers, as well as at the end.
3156 8. Added the rcpt_4xx retry error code.
3158 9. Added postmaster_mailfrom=xxx to callout verification option.
3160 10. Added mailfrom=xxxx to the callout verification option, for verify=
3163 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3164 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3166 12. Inserted some casts to stop certain compilers warning when using pointer
3167 differences as field lengths or precisions in printf-type calls (mostly
3168 affecting debugging statements).
3170 13. Added optional readline() support for -be (dynamically loaded).
3172 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3173 same clock tick as a message's arrival, so that its received time was the
3174 same as the "first fail" time on the retry record, and that message
3175 remained on the queue past the ultimate address timeout, every queue runner
3176 would try a delivery (because it was past the ultimate address timeout) but
3177 after another failure, the ultimate address timeout, which should have then
3178 bounced the address, did not kick in. This was a "< instead of <=" error;
3179 in most cases the first failure would have been in the next clock tick
3180 after the received time, and all would be well.
3182 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3183 being recognized when the domain list was tested by the match_domain
3184 condition in an expansion string.
3186 16. Added the ${str2b64: operator.
3188 17. Exim was always calling setrlimit() to set a large limit for the number of
3189 processes, without checking whether the existing limit was already
3190 adequate. (It did check for the limit on file descriptors.) Furthermore,
3191 errors from getrlimit() and setrlimit() were being ignored. Now they are
3192 logged to the main and panic logs, but Exim does carry on, to try to do its
3193 job under whatever limits there are.
3195 18. Imported PCRE 5.0.
3197 19. Trivial typo in log message " temporarily refused connection" (the leading
3200 20. If the log selector return_path_on_delivery was set and an address was
3201 redirected to /dev/null, the delivery process crashed because it assumed
3202 that a return path would always be set for a "successful" delivery. In this
3203 case, the whole delivery is bypassed as an optimization, and therefore no
3206 21. Internal re-arrangement: the function for sending a challenge and reading
3207 a response while authentication was assuming a zero-terminated challenge
3208 string. It's now changed to take a pointer and a length, to allow for
3209 binary data in such strings.
3211 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3213 23. Exim was not respecting finduser_retries when seeking the login of the
3214 uid under which it was called; it was always trying 10 times. (The default
3215 setting of finduser_retries is zero.) Also, it was sleeping after the final
3216 failure, which is pointless.
3218 24. Implemented tls_on_connect_ports.
3220 25. Implemented acl_smtp_predata.
3222 26. If the domain in control=submission is set empty, Exim assumes that the
3223 authenticated id is a complete email address when it generates From: or
3224 Sender: header lines.
3226 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3227 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3228 chown and chgrp in /bin and hostname in /usr/bin.
3230 28. Exim was keeping the "process log" file open after each use, just as it
3231 does for the main log. This opens the possibility of it remaining open for
3232 long periods when the USR1 signal hits a daemon. Occasional processlog
3233 errors were reported, that could have been caused by this. Anyway, it seems
3234 much more sensible not to leave this file open at all, so that is what now
3237 29. The long-running daemon process does not normally write to the log once it
3238 has entered its main loop, and it closes the log before doing so. This is
3239 so that log files can straightforwardly be renamed and moved. However,
3240 there are a couple of unusual error situations where the daemon does write
3241 log entries, and I had neglected to close the log afterwards.
3243 30. The text of an SMTP error response that was received during a remote
3244 delivery was being truncated at 512 bytes. This is too short for some of
3245 the long messages that one sometimes sees. I've increased the limit to
3248 31. It is now possible to make retry rules that apply only when a message has a
3249 specific sender, in particular, an empty sender.
3251 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3252 it possible to be selective about when SMTP synchronization is enforced.
3254 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3256 32. Implemented hosts_connection_nolog.
3258 33. Added an ACL for QUIT.
3260 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3263 35. Added mailbox_size and mailbox_filecount to appendfile.
3265 36. Added control = no_multiline_responses to ACLs.
3267 37. There was a bug in the logic of the code that waits for the clock to tick
3268 in the case where the clock went backwards by a substantial amount such
3269 that the microsecond fraction of "now" was more than the microsecond
3270 fraction of "then" (but the whole seconds number was less).
3272 38. Added support for the libradius Radius client library this is found on
3273 FreeBSD (previously only the radiusclient library was supported).
3279 1. When certain lookups returned multiple values in the form name=value, the
3280 quoting of the values was not always being done properly. Specifically:
3281 (a) If the value started with a double quote, but contained no whitespace,
3283 (b) If the value contained whitespace other than a space character (i.e.
3284 tabs or newlines or carriage returns) it was not quoted.
3285 This fix has been applied to the mysql and pgsql lookups by writing a
3286 separate quoting function and calling it from the lookup code. The fix
3287 should probably also be applied to nisplus, ibase and oracle lookups, but
3288 since I cannot test any of those, I have not disturbed their existing code.
3290 2. A hit in the callout cache for a specific address caused a log line with no
3291 reason for rejecting RCPT. Now it says "Previous (cached) callout
3292 verification failure".
3294 3. There was an off-by-one bug in the queryprogram router. An over-long
3295 return line was truncated at 256 instead of 255 characters, thereby
3296 overflowing its buffer with the terminating zero. As well as fixing this, I
3297 have increased the buffer size to 1024 (and made a note to document this).
3299 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3300 when Exim is waiting for an SMTP response from a remote server, Exim
3301 restarts its select() call on the socket, thereby resetting its timeout.
3302 This is not a problem when such interrupts are rare. Somebody set up a cron
3303 job to run exiwhat every 2 minutes, which is less than the normal select()
3304 timeout (5 or 10 minutes). This meant that the select() timeout never
3305 kicked in because it was always reset. I have fixed this by comparing the
3306 time when an interrupt arrives with the time at the start of the first call
3307 to select(). If more time than the timeout has elapsed, the interrupt is
3308 treated as a timeout.
3310 5. Some internal re-factoring in preparation for the addition of Sieve
3311 extensions (by MH). In particular, the "personal" test is moved to a
3312 separate function, and given an option for scanning Cc: and Bcc: (which is
3313 not set for Exim filters).
3315 6. When Exim created an email address using the login of the caller as the
3316 local part (e.g. when creating a From: or Sender: header line), it was not
3317 quoting the local part when it contained special characters such as @.
3319 7. Installed new OpenBSD configuration files.
3321 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3322 try to make them clearer.
3324 9. Callout options, other than the timeout value, were being ignored when
3325 verifying sender addresses in header lines. For example, when using
3327 verify = header_sender/callout=no_cache
3329 the cache was (incorrectly) being used.
3331 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3332 only the Cygwin environment.
3334 11. When return_path_on_delivery was set as a log selector, if different remote
3335 addresses in the same message used different return paths and parallel
3336 remote delivery occurred, the wrong values would sometimes be logged.
3337 (Whenever a remote delivery process finished, the return path value from
3338 the most recently started remote delivery process was logged.)
3340 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3341 lines when AUTH and/or TLS are in use. This is the "received protocol"
3342 field. Exim used to use "asmtp" for authenticated SMTP, without any
3343 indication (in the protocol name) for TLS use. Now it follows the RFC and
3344 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3345 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3346 names appear in log lines as well as in Received: header lines.
3348 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3349 extensions, and comparison tests, and to fix some bugs.
3351 14. Changes to the "personal" filter test:
3353 (1) The test was buggy in that it was just doing the equivalent of
3354 "contains" tests on header lines. For example, if a user's address was
3355 anne@some.where, the "personal" test would incorrectly be true for
3357 To: susanne@some.where
3359 This test is now done by extracting each address from the header in turn,
3360 and checking the entire address. Other tests that are part of "personal"
3361 are now done using regular expressions (for example, to check local parts
3362 of addresses in From: header lines).
3364 (2) The list of non-personal local parts in From: addresses has been
3365 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3366 taken from the Sieve specification recommendations.
3368 (3) If the message contains any header line starting with "List-" it is
3369 treated as non-personal.
3371 (4) The test for "circular" in the Subject: header line has been removed
3372 because it now seems ill-conceived.
3374 15. Minor typos in src/EDITME comments corrected.
3376 16. Installed latest exipick from John Jetmore.
3378 17. If headers_add on a router specified a text string that was too long for
3379 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3380 of string_sprintf() is now avoided.
3382 18. $message_body_size was not set (it was always zero) when running the DATA
3383 ACL and the local_scan() function.
3385 19. For the "mail" command in an Exim filter, no default was being set for
3386 the once_repeat time, causing a random time value to be used if "once" was
3387 specified. (If the value happened to be <= 0, no repeat happened.) The
3388 default is now 0s, meaning "never repeat". The "vacation" command was OK
3389 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3390 (I found it when inspecting the code).
3392 20. There is now an overall timeout for performing a callout verification. It
3393 defaults to 4 times the callout timeout, which applies to individual SMTP
3394 commands during the callout. The overall timeout applies when there is more
3395 than one host that can be tried. The timeout is checked before trying the
3396 next host. This prevents very long delays if there are a large number of
3397 hosts and all are timing out (e.g. when the network connections are timing
3398 out). The value of the overall timeout can be changed by specifying an
3399 additional sub-option for "callout", called "maxwait". For example:
3401 verify = sender/callout=5s,maxwait=20s
3403 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3404 to the end before writing, but this should make it even safer).
3406 22. Exim was forgetting that it had advertised PIPELINING for the second and
3407 subsequent messages on an SMTP connection. It was also not resetting its
3408 memory on STARTTLS and an internal HELO.
3410 23. When Exim logs an SMTP synchronization error within a session, it now
3411 records whether PIPELINING has been advertised or not.
3413 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3414 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3415 rather than long int.
3417 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3419 26. Added the never_mail option to autoreply.
3425 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3426 crash if the getsockname() call failed; this can happen if a connection is
3427 closed very soon after it is established. The problem was simply in the
3428 order in which certain operations were done, causing Exim to try to write
3429 to the SMTP stream before it had set up the file descriptor. The bug has
3430 been fixed by making things happen in the correct order.
3436 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3437 before the connection was closed, thus losing the rejection response.
3439 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3440 some early Solaris releases, but causes trouble in current releases where
3441 socklen_t is defined.
3443 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3446 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3449 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3450 that releasing the top of it at the end releases what was used for sub-
3451 expansions (unless the block got too big). However, discard this block if
3452 the first thing is a variable or header, so that we can use its block when
3453 it is dynamic (useful for very large $message_headers, for example).
3455 6. Lookups now cache *every* query, not just the most recent. A new, separate
3456 store pool is used for this. It can be recovered when all lookup caches are
3457 flushed. Lookups now release memory at the end of their result strings.
3458 This has involved some general refactoring of the lookup sources.
3460 7. Some code has been added to the store_xxx() functions to reduce the amount
3461 of flapping under certain conditions.
3463 8. log_incoming_interface used to affect only the <= reception log lines. Now
3464 it causes the local interface and port to be added to several more SMTP log
3465 lines, for example "SMTP connection from", and rejection lines.
3467 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3469 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3471 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3472 could be overwritten at the end of the current message (or the start of a
3473 new message if it was set in a HELO ACL). The value is now preserved for
3474 the duration of the SMTP connection.
3476 12. If a transport had a headers_rewrite setting, and a matching header line
3477 contained an unqualified address, that address was qualified, even if it
3478 did not match any rewriting rules. The underlying bug was that the values
3479 of the flags that permit the existence of unqualified sender and recipient
3480 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3481 non-local messages, and by -bnq for local messages) were not being
3482 preserved with the message after it was received.
3484 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3485 "next input=" as part of the text comprising the host identity instead of
3486 the correct text. The code was using the same buffer for two different
3487 strings. However, depending on which order the printing function evaluated
3488 its arguments, the bug did not always show up. Under Linux, for example, my
3489 test suite worked just fine.
3491 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3492 doesn't work with some older versions of Perl. It has been changed to "my",
3493 which in any case is probably the better facility to use.
3495 15. A really picky compiler found some instances of statements for creating
3496 error messages that either had too many or two few arguments for the format
3499 16. The size of the buffer for calls to the DNS resolver has been increased
3500 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3501 for addresses that have a lot of PTR records. This alleviates a problem; it
3502 does not fully solve it.
3504 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3505 buffer now truncates the list and logs the incident, which is the same
3506 action as happens when Exim is looking up a host name and its aliases.
3507 Previously in this situation something unpredictable would happen;
3508 sometimes it was "internal error: store_reset failed".
3510 18. If a server dropped the connection unexpectedly when an Exim client was
3511 using GnuTLS and trying to read a response, the client delivery process
3512 crashed while trying to generate an error log message.
3514 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3515 string, for example:
3517 warn message = H1: something\nH2: something
3519 the text was added as a single header line from Exim's point of view
3520 though it ended up OK in the delivered message. However, searching for the
3521 second and subsequent header lines using $h_h2: did not work. This has been
3522 fixed. Similarly, if a system filter added multiple headers in this way,
3523 the routers could not see them.
3525 20. Expanded the error message when iplsearch is called with an invalid key to
3526 suggest using net-iplsearch in a host list.
3528 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3529 ACLs are no longer actually imposed (and a message to that effect is
3532 22. If a "gecos" field in a passwd entry contained escaped characters, in
3533 particular, if it contained a \" sequence, Exim got it wrong when building
3534 a From: or a Sender: header from that name. A second bug also caused
3535 incorrect handling when an unquoted " was present following a character
3536 that needed quoting.
3538 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3539 was not being matched caselessly.
3541 24. Arranged for all hyphens in the exim.8 source to be escaped with
3544 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3545 properly. Recipient callout cache records were still being keyed to include
3546 the sender, even when use_sender was set false. This led to far more
3547 callouts that were necessary. The sender is no longer included in the key
3548 when use_sender is false.
3550 26. Added "control = submission" modifier to ACLs.
3552 27. Added the ${base62d: operator to decode base 62 numbers.
3554 28. dnsdb lookups can now access SRV records.
3556 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3557 the configuration file.
3559 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3560 (-v) mode. This makes the output for a verbose queue run more intelligible.
3562 31. Added a use_postmaster feature to recipient callouts.
3564 32. Added the $body_zerocount variable, containing the number of binary zero
3565 bytes in the message body.
3567 33. The time of last modification of the "new" subdirectory is now used as the
3568 "mailbox time last read" when there is a quota error for a maildir
3571 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3573 35. Added +ignore_unknown as a special item in host lists.
3575 36. Code for decoding IPv6 addresses in host lists is now included, even if
3576 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3577 address was recognized as an IP address, but was then not correctly decoded
3578 into binary, causing unexpected and incorrect effects when compared with
3585 1. Very minor rewording of debugging text in manualroute to say "list of
3586 hosts" instead of "hostlist".
3588 2. If verify=header_syntax was set, and a header line with an unqualified
3589 address (no domain) and a large number of spaces between the end of the
3590 name and the colon was received, the reception process suffered a buffer
3591 overflow, and (when I tested it) crashed. This was caused by some obsolete
3592 code that should have been removed. The fix is to remove it!
3594 3. When running in the test harness, delay a bit after writing a bounce
3595 message to get a bit more predictability in the log output.
3597 4. Added a call to search_tidyup() just before forking a reception process. In
3598 theory, someone could use a lookup in the expansion of smtp_accept_max_
3599 per_host which, without the tidyup, could leave open a database connection.
3601 5. Added the variables $recipient_data and $sender_data which get set from a
3602 lookup success in an ACL "recipients" or "senders" condition, or a router
3603 "senders" option, similar to $domain_data and $local_part_data.
3605 6. Moved the writing of debug_print from before to after the "senders" test
3608 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3609 problems for message scanning, either using a data ACL, or using
3610 local_scan() because the Received: header was not generated till after they
3611 were called (in order to set the time as the time of reception completion).
3612 I have revised the way this works. The header is now generated after the
3613 body is received, but before the ACL or local_scan() are called. After they
3614 are run, the timestamp in the header is updated.
3620 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3621 before starting a queue runner without re-exec. This happened only when
3622 deliver_drop_privilege was set or when the Exim user was set to root. The
3623 effect of the bug was that timeouts during subsequent deliveries caused
3624 crashes instead of being properly handled. The handler is now left at its
3625 default (and expected) setting.
3627 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3628 message, again when deliver_drop_privilege is set or Exim is run as root.
3629 The bug described in (1) was not present in this case, but the tidying up
3630 of the other signals was missing. I have made the two cases consistent.
3632 3. The ignore_target_hosts setting on a manualroute router was being ignored
3633 for hosts that were looked up using the /MX notation.
3635 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3638 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3639 operated on the sender address. After changing the $sender_address to <>
3640 for the sender address verify, Exim was re-instated it as the original
3641 (before rewriting) address, but remembering that it had rewritten it, so it
3642 wasn't rewriting it again. This bug also had the effect of breaking the
3643 sender address verification caching when the sender address was rewritten.
3645 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3646 This has been changed so that if the ip literal address matches
3647 ignore_target_hosts, the router declines.
3649 7. Added expansion conditions match_domain, match_address, and match_local_
3650 part (NOT match_host).
3652 8. The placeholder for the Received: header didn't have a length field set.
3654 9. Added code to Exim itself and to exim_lock to test for a specific race
3655 condition that could lead to file corruption when using MBX delivery. The
3656 issue is with the lockfile that is created in /tmp. If this file is removed
3657 after a process has opened it but before that process has acquired a lock,
3658 there is the potential for a second process to recreate the file and also
3659 acquire a lock. This could lead to two Exim processes writing to the file
3660 at the same time. The added code performs the same test as UW imapd; it
3661 checks after acquiring the lock that its file descriptor still refers to
3662 the same named file.
3664 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3665 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3668 11. Added the smtp_active_hostname option. If used, this will typically be made
3669 to depend on the incoming interface address. Because $interface_address is
3670 not set up until the daemon has forked a reception process, error responses
3671 that can happen earlier (such as "too many connections") no longer contain
3674 12. If an expansion in a condition on a "warn" statement fails because a lookup
3675 defers, the "warn" statement is abandoned, and the next ACL statement is
3676 processed. Previously this caused the whole ACL to be aborted.
3678 13. Added the iplsearch lookup type.
3680 14. Added ident_timeout as a log selector.
3682 15. Added tls_certificate_verified as a log selector.
3684 16. Added a global option tls_require_ciphers (compare the smtp transport
3685 option of the same name). This controls incoming TLS connections.
3687 17. I finally figured out how to make tls_require_ciphers do a similar thing
3688 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3689 before starting the TLS session.
3691 18. Tabs are now shown as \t in -bP output.
3693 19. If the log selector return_path_on_delivery was set, Exim crashed when
3694 bouncing a message because it had too many Received: header lines.
3696 20. If two routers both had headers_remove settings, and the first one included
3697 a superfluous trailing colon, the final name in the first list and the
3698 first name in the second list were incorrectly joined into one item (with a
3699 colon in the middle).
3705 1. Added -C and -D options to the exinext utility, mainly to make it easier
3706 to include in the automated testing, but these could be helpful when
3707 multiple configurations are in use.
3709 2. The exinext utility was not formatting the output nicely when there was
3710 an alternate port involved in the retry record key, nor when there was a
3711 message id as well (for retries that were specific to a specific message
3712 and a specific host). It was also confused by IPv6 addresses, because of
3713 the additional colons they contain. I have fixed the IPv4 problem, and
3714 patched it up to do a reasonable job for IPv6.
3716 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3717 delivery, the log line now contains "pipelined" if PIPELINING was used.
3719 4. An SMTP transport process used to panic and die if the bind() call to set
3720 an explicit outgoing interface failed. This has been changed; it is now
3721 treated in the same way as a connect() failure.
3723 5. A reference to $sender_host_name in the part of a conditional expansion
3724 that was being skipped was still causing a DNS lookup. This no longer
3727 6. The def: expansion condition was not recognizing references to header lines
3728 that used bh_ and bheader_.
3730 7. Added the _cache feature to named lists.
3732 8. The code for checking quota_filecount in the appendfile transport was
3733 allowing one more file than it should have been.
3735 9. For compatibility with Sendmail, the command line option
3743 and sets the incoming protocol and host name (for trusted callers). The
3744 host name and its colon can be omitted when only the protocol is to be set.
3745 Note the Exim already has two private options, -pd and -ps, that refer to
3746 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3747 "s", but I don't think that's a major issue.
3749 10. A number of refactoring changes to the code, none of which should affect
3752 (a) The number of logging options was getting close to filling up the
3753 32-bit word that was used as a bit map. I have split them into two classes:
3754 those that are passed in the argument to log_write(), and those that are
3755 only ever tested independently outside of that function. These are now in
3756 separate 32-bit words, so there is plenty of room for expansion again.
3757 There is no change in the user interface or the logging behaviour.
3759 (b) When building, for example, log lines, the code previously used a
3760 macro that called string_cat() twice, in order to add two strings. This is
3761 not really sufficiently general. Furthermore, there was one instance where
3762 it was actually wrong because one of the argument was used twice, and in
3763 one call a function was used. (As it happened, calling the function twice
3764 did not affect the overall behaviour.) The macro has been replaced by a
3765 function that can join an arbitrary number of extra strings onto a growing
3768 (c) The code for expansion conditions now uses a table and a binary chop
3769 instead of a serial search (which was left over from when there were very
3770 few conditions). Also, it now recognizes conditions like "pam" even when
3771 the relevant support is not compiled in: a suitably worded error message is
3772 given if an attempt is made to use such a condition.
3774 11. Added ${time_interval:xxxxx}.
3776 12. A bug was causing one of the ddress fields not to be passed back correctly
3777 from remote delivery subprocesses. The field in question was not being
3778 subsequently used, so this caused to problems in practice.
3780 13. Added new log selectors queue_time and deliver_time.
3782 14. Might have fixed a bug in maildirsizefile handling that threw up
3783 "unexpected character" debug warnings, and recalculated the data
3784 unnecessarily. In any case, I expanded the warning message to give more
3787 15. Added the message "Restricted characters in address" to the statements in
3788 the default ACL that block characters like @ and % in local parts.
3790 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3791 Three changes have been made:
3793 (a) There was a serious bug; a negative response to MAIL caused the whole
3794 recipient domain to be cached as invalid, thereby blocking all messages
3795 to all local parts at the same domain, from all senders. This bug has
3796 been fixed. The domain is no longer cached after a negative response to
3797 MAIL if the sender used is not empty.
3799 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3802 (c) A new callout option, "use_sender" has been added for people who want
3803 the modified behaviour.
3809 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3812 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3813 indeed breaks things for older releases.
3815 3. Added additional logging to the case where there is a problem reading data
3816 from a filter that is running in a subprocess using a pipe, in order to
3817 try to track down a specific problem.
3819 4. Testing facility fudge: when running in the test harness and attempting
3820 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3821 getting "No route to host". Convert this to a timeout.
3823 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3826 6. Some OS don't have socklen_t but use size_t instead. This affects the
3827 fifth argument of getsockopt() amongst other things. This is now
3828 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3829 can be set for individual OS. I have set it for SunOS5, OSF1, and
3830 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3831 some earlier ones do not.
3833 7. Change 4.30/15 was not doing the test caselessly.
3835 8. The standard form for an IPv6 address literal was being rejected by address
3836 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3837 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3838 this, as well as the form without the "IPv6" on the front (but only when
3839 address literals are enabled, of course).
3841 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3843 10. Exim crashed if a message with an empty sender address specified by -f
3844 encountered a router with an errors_to setting. This could be provoked only
3845 by a command such as
3849 where an empty string was supplied; "<>" did not hit this bug.
3851 11. Installed PCRE release 4.5.
3853 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3854 remained set. It is now erased.
3856 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3857 times from message ids (which are base 36 rather than the normal 62).
3859 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3860 were being counted as actual protocol errors, and logged if the log
3861 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3862 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3863 and DATA following a set of rejected RCPTs do not count as protocol errors.
3864 In other words, Exim assumes they were pipelined, though this may not
3865 actually be the case. Of course, in all cases the client gets an
3866 appropriate error code.
3868 15. If a lookup fails in an ACL condition, a message about the failure may
3869 be available; it is used if testing the ACL cannot continue, because most
3870 such messages specify what the cause of the deferral is. However, some
3871 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3872 that caused an old message to be retained and used if a later statement
3873 caused a defer, replacing the real cause of the deferral.
3875 16. If an IP address had so many PTR records that the DNS lookup buffer
3876 was not large enough to hold them, Exim could crash while trying to process
3877 the truncated data. It now detects and logs this case.
3879 17. Further to 4.21/58, another change has been made: if (and only if) the
3880 first line of a message (the first header line) ends with CRLF, a bare LF
3881 in a subsequent header line has a space inserted after it, so as not to
3882 terminate the header.
3884 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3885 unnecessarily, used atoi() instead of strtol(), and didn't check the
3886 termination when getting file sizes from file names by regex.
3888 19. Completely re-implemented the support for maildirsize files, in the light
3889 of a number of problems with the previous contributed implementation
3890 (4.30/29). In particular:
3892 . If the quota is zero, the maildirsize file is maintained, but no quota is
3895 . If the maildir directory does not exist, it is created before any attempt
3896 to write a maildirsize file.
3898 . The quota value in the file is just a cache; if the quota is changed in
3899 the transport, the new value overrides.
3901 . A regular expression is available for excluding directories from the
3904 20. The autoreply transport checks the characters in options that define the
3905 message's headers; it allows continued headers, but it was checking with
3906 isspace() after an embedded newline instead of explicitly looking for a
3909 21. If all the "regular" hosts to which an address was routed had passed their
3910 expiry times, and had not reached their retry times, the address was
3911 bounced, even if fallback hosts were defined. Now Exim should go on to try
3914 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3915 equivalent code in the SMTP transport. Some hosts send humungous responses
3916 to HELO/EHLO, more than 1024 it seems.
3918 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3919 gives compiler warnings in some environments. I've now done it "properly",
3922 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3923 (because of problems with the built-in one) was declared to return uschar *
3924 instead of char *, causing compiler failure.
3926 25. Fixed a file descriptor leak when processing alias/forward files.
3928 26. Fixed a minor format string issue in dbfn.c.
3930 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3932 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3933 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3936 29. When a sender address is verified, it is cached, to save repeating the test
3937 when there is more than one recipient in a message. However, when the
3938 verification involves a callout, it is possible for different callout
3939 options to be set for different recipients. It is too complicated to keep
3940 track of this in the cache, so now Exim always runs a verification when a
3941 callout is required, relying on the callout cache for the optimization.
3942 The overhead is duplication of the address routing, but this should not be
3945 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3946 to be verified with callout=postmaster, and the main callout worked but the
3947 postmaster check failed, the verification correctly failed. However, if a
3948 subsequent RCPT command asked for sender verification *without* the
3949 postmaster check, incorrect caching caused this verification also to fail,
3952 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3953 it was not caching the DNS options (qualify_single, search_parents) that
3954 were used when the lookup failed. A subsequent lookup with different
3955 options therefore always gave the same answer, though there were cases
3956 where it should not have. (Example: a "domains = !$mx_any" option on a
3957 dnslookup router: the "domains" option is always processed without any
3958 widening, but the router might have qualify_single set.) Now Exim uses the
3959 cached value only when the same options are set.
3961 32. Added John Jetmore's "exipick" utility to the distribution.
3963 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3964 than a timeout (e.g. a certificate is required, and is not provided), an
3965 Exim server now closes the connection immediately. Previously it waited for
3966 the client to close - but if the client is SSL, it seems that they each
3967 wait for each other, leading to a delay before one of them times out.
3969 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3970 maintained 0.8.x compatibility because I don't think many are using it, and
3971 it is clearly obsolete.
3973 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3976 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3977 client certificate was expired. A simple patch fixes this, though I don't
3978 understand the full logic of why the verify callback is called multiple
3981 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3982 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3983 which causes problems with some clients (such as the Certicom SSL Plus
3984 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3985 disables the coutermeasure allowing Eudora to connect."
3987 38. Exim was not checking that a write() to a log file succeeded. This could
3988 lead to Bad Things if a log got too big, in particular if it hit a file
3989 size limit. Exim now panics and dies if it cannot write to a log file, just
3990 as it does if it cannot open a log file.
3992 39. Modified OS/Makefile-Linux so that it now contains
3994 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3996 The two -D definitions ensure that Exim is compiled with large file
3997 support, which makes it possible to handle log files that are bigger than
4000 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4001 instance) a domain was checked against a named list that involved a lookup,
4002 causing $domain_data to be set, then another domain was checked against the
4003 same list, then the first domain was re-checked, the value of $domain_data
4004 after the final check could be wrong. In particular, if the second check
4005 failed, it could be set empty. This bug probably also applied to
4008 41. The strip_trailing_dot option was not being applied to the address given
4009 with the -f command-line option.
4011 42. The code for reading a message's header from the spool was incrementing
4012 $received_count, but never initializing it. This meant that the value was
4013 incorrect (doubled) while delivering a message in the same process in which
4014 it was received. In the most common configuration of Exim, this never
4015 happens - a fresh exec is done - but it can happen when
4016 deliver_drop_privilege is set.
4018 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4019 it now includes up to 150 characters of the unexpected data in the log
4022 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4023 and building data strings. The size of both of these buffers was 10 000
4024 bytes - far larger than anybody would *ever* want, thought I. Needless to
4025 say, somebody hit the limit. I have increased the maximum line length to
4026 20 000 and the maximum data length of concatenated lines to 100 000. I have
4027 also fixed two bugs, because there was no checking on these buffers. Tsk,
4028 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4029 buffer is too small.
4031 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4032 lsearch lookups. Now it does.
4034 46. When parsing a route_list item in a manualroute router, a fixed-length
4035 buffer was used for the list of hosts. I made this 1024 bytes long,
4036 thinking that nobody would ever have a list of hosts that long. Wrong.
4037 Somebody had a whole pile of complicated expansion conditions, and the
4038 string was silently truncated, leading to an expansion error. It turns out
4039 that it is easier to change to an unlimited length (owing to other changes
4040 that have happened since this code was originally written) than to build
4041 structure for giving a limitation error. The length of the item that
4042 expands into the list of hosts is now unlimited.
4044 47. The lsearch lookup could not handle data where the length of text line was
4045 more than 4095 characters. Such lines were truncated, leading to shortened
4046 data being returned. It should now handle lines of any length.
4048 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4049 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4052 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4053 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4054 no longer happens in comment lines. A list of replacements is now placed
4055 at the head of all of the source files, except those whose only change is
4056 to replace PERL_COMMAND in the very first #! line.
4058 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4059 messages on the queue, with a bottom-up merge sort, using code contributed
4060 by Michael Haardt. This should make operations like -bp somewhat faster on
4061 large queues. It won't affect queue runners, except when queue_run_in_order
4064 51. Installed eximstats 1.31 in the distribution.
4066 52. Added support for SRV lookups to the dnslookup router.
4068 53. If an ACL referred to $message_body or $message_body_end, the value was not
4069 reset for any messages that followed in the same SMTP session.
4071 54. The store-handling optimization for building very long strings was not
4072 differentiating between the different store pools. I don't think this
4073 actually made any difference in practice, but I've tidied it.
4075 55. While running the routers to verify a sender address, $sender_address
4076 was still set to the sender address. This is wrong, because when routing to
4077 send a bounce to the sender, it would be empty. Therefore, I have changed
4078 it so that, while verifying a sender address, $sender_address is set to <>.
4079 (There is no change to what happens when verifying a recipient address.)
4081 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4082 target A or AAAA records (if not already returned) without resetting the
4083 qualify_single or search_parents options of the DNS resolver. These are
4084 inappropriate in this case because the targets of MX and SRV records must
4085 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4086 target that, when qualified, matched something in the local domain. These
4087 two options are now turned off when doing these lookups.
4089 57. It seems that at least some releases of Reiserfs (which does not have the
4090 concept of a fixed number of inodes) returns zero and not -1 for the
4091 number of available inodes. This interacted badly with check_spool_inodes,
4092 which assumed that -1 was the "no such thing" setting. What I have done is
4093 to check that the total number of inodes is greater than zero before doing
4094 the test of how many are available.
4096 58. When a "warn" ACL statement has a log_message modifier, the message is
4097 remembered, and not repeated. This is to avoid a lot of repetition when a
4098 message has many recipients that cause the same warning to be written.
4099 Howewer, Exim was preserving the list of already written lines for an
4100 entire SMTP session, which doesn't seem right. The memory is now reset if a
4101 new message is started.
4103 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4104 debugging output unless log_rewrite was also set.
4106 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4107 of (int)(handle) when we know that handle contains (void *)(-1).
4109 61. The Exim daemon panic-logs an error return when it closes the incoming
4110 connection. However "connection reset by peer" seems to be common, and
4111 isn't really an error worthy of noting specially, so that particular error
4114 62. When Exim is trying to find all the local interfaces, it used to panic and
4115 die if the ioctl to get the interface flags failed. However, it seems that
4116 on at least one OS (Solaris 9) it is possible to have an interface that is
4117 included in the list of interfaces, but for which you get a failure error
4118 for this call. This happens when the interface is not "plumbed" into a
4119 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4120 failure of the "get flags" call assumes that the interface is down.
4122 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4123 makes life easier for people who are doing arithmetic on fields extracted
4124 from dates, where you often get leading zeros that should not be
4125 interpreted as octal.
4127 64. Added qualify_domain to the redirect router, to override the global
4130 65. If a pathologically long header line contained very many addresses (the
4131 report of this problem mentioned 10 000) and each of them was rewritten,
4132 Exim could use up a very large amount of memory. (It kept on making new
4133 copies of the header line as it rewrote, and never released the old ones.)
4134 At the expense of a bit more processing, the header rewriting function has
4135 been changed so that it no longer eats memory in this way.
4137 66. The generation of the Received: header has been moved from the time that a
4138 message starts to be received, to the time that it finishes. The timestamp
4139 in the Received: header should now be very close to that of the <= log
4140 line. There are two side-effects of this change:
4142 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4143 logged header lines no longer include the local Received: line, because
4144 it has not yet been created. The same applies to a copy of the message
4145 that is returned to a non-SMTP sender when a message is rejected.
4147 (b) When a filter file is tested using -bf, no additional Received: header
4148 is added to the test message. After some thought, I decided that this
4151 This change does not affect the value of $received_for. It is still set
4152 after address rewriting, but before local_scan() is called.
4154 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4156 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4157 gave an unhelpful panic error message, and a defer error. I have managed to
4158 change this behaviour so that it now rejects any supplied certificate,
4159 which seems right, as the list of acceptable certificates is empty.
4161 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4162 gave an unhelpful defer error. I have not managed to make this reject any
4163 supplied certificates, but the error message it gives is "no certificate
4164 supplied", which is not helpful.
4166 70. exigrep's output now also includes lines that are not associated with any
4167 message, but which match the given pattern. Implemented by a patch from
4168 Martin Sluka, which also tidied up the Perl a bit.
4170 71. Recipient callout verification, like sender verification, was using <> in
4171 the MAIL FROM command. This isn't really the right thing, since the actual
4172 sender may affect whether the remote host accepts the recipient or not. I
4173 have changed it to use the actual sender in the callout; this means that
4174 the cache record is now keyed on a recipient/sender pair, not just the
4175 recipient address. There doesn't seem to be a real danger of callout loops,
4176 since a callout by the remote host to check the sender would use <>.
4177 [SEE ABOVE: changed after hitting problems.]
4179 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4180 temporary errors. However, in the case of such a code being given after
4181 the end of a data transmission (i.e. after ".") Exim was failing to write
4182 a retry record for the message. (Yes, there was some broken host that was
4183 actually sending 8xx at this point.)
4185 73. An unknown lookup type in a host list could cause Exim to panic-die when
4186 the list was checked. (An example that provoked this was putting <; in the
4187 middle of a list instead of at the start.) If this happened during a DATA
4188 ACL check, a -D file could be left lying around. This kind of configuration
4189 error no longer causes Exim to die; instead it causes a defer errror. The
4190 incident is still logged to the main and panic logs.
4192 74. Buglet left over from Exim 3 conversion. The message "too many messages
4193 in one connection" was written to the rejectlog but not the mainlog, except
4194 when address rewriting (yes!) was being logged.
4196 75. Added write_rejectlog option.
4198 76. When a system filter was run not as root (that is, when system_filter_user
4199 was set), the values of the $n variables were not being returned to the
4200 main process; thus, they were not subsequently available in the $sn
4203 77. Added +return_path_on_delivery log selector.
4205 78. A connection timeout was being treated differently from recipients deferred
4206 when testing hosts_max_try with a message that was older than the host's
4207 retry timeout. (The host should not be counted, thus allowing all hosts to
4208 be tried at least once before bouncing.) This may have been the cause of an
4209 occasionally reported bug whereby a message would remain on the queue
4210 longer than the retry timeout, but would be bounced if a delivery was
4211 forced. I say "may" because I never totally pinned down the problem;
4212 setting up timeout/retry tests is difficult. See also the next item.
4214 79. The ultimate address timeout was not being applied to errors that involved
4215 a combination of host plus message (for example, a timeout on a MAIL
4216 command). When an address resolved to a number of possible hosts, and they
4217 were not all tried for each delivery (e.g. because of hosts_max_try), a
4218 message could remain on the queue longer than the retry timeout.
4220 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4223 81. Fixed an obscure SMTP outgoing bug which required at least the following
4224 conditions: (a) there was another message waiting for the same server;
4225 (b) the server returned 5xx to all RCPT commands in the first message so
4226 that the message was not completed; (c) the server dropped the connection
4227 or gave a negative response to the RSET that Exim sends to abort the
4228 transaction. The observed case was a dropped connection after DATA that had
4229 been sent in pipelining mode. That is, the server had advertised PIPELINING
4230 but was not implementing it correctly. The effect of the bug was incorrect
4231 behaviour, such as trying another host, and this could lead to a crash.
4237 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4238 and daemon.c were passed as pointers to ints; they should have been
4239 pointers to socklen_t variables (which are typically unsigned ints).
4241 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4244 3. Fixed a really odd bug that affected only the testing scheme; patching a
4245 certain fixed string in the binary changed the value of another string that
4246 happened to be identical to the end of the original first string.
4248 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4249 name", it returns that address as the IP address. On some operating
4250 systems (e.g. Solaris), it also passes back the IP address string as the
4251 "host name". However, on others (e.g. Linux), it passes back an empty
4252 string. Exim wasn't checking for this, and was changing the host name to an
4253 empty string, assuming it had been canonicized.
4255 5. Although rare, it is permitted to have more than one PTR record for a given
4256 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4257 all the names associated with an address, because they do in Solaris.
4258 However, it seems that they do not in Linux for data that comes from the
4259 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4260 I found this out when I moved to a new Linux workstation and tried to run
4261 the Exim test suite.
4263 To get round this problem I have changed the code so that it now does its
4264 own call to the DNS to look up PTR records when searching for a host name.
4265 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4266 addresses that are only in /etc/hosts are still found.
4268 This behaviour is, however, controlled by an option called host_lookup_
4269 order, which defaults to "bydns:byaddr". If people want to use the other
4270 order, or indeed, just use one or the other means of lookup, they can
4271 specify it in this variable.
4273 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4274 some operating systems, this comes back from gethostbyaddr() as an empty
4275 string, and this is what Exim used to test for. However, it seems that in
4276 other systems, "." is yielded. Exim now tests for this case too.
4278 7. The values of check_spool_space and check_log_space are now held internally
4279 as a number of kilobytes instead of an absolute number of bytes. If a
4280 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4281 kilobyte. This means that much larger values can be stored.
4283 8. Exim monitor: an attempt to get the action menu when not actually pointing
4284 at a message produces an empty menu entitled "No message selected". This
4285 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4286 no entries in it ("Shell widget menu has zero width and/or height"). So I
4287 have added a single, blank menu entry in this case.
4289 9. Added ${quote_local_part.
4291 10. MIME decoding is now applied to the contents of Subject: header lines when
4294 11. Now that a reference to $sender_host_address automatically causes a reverse
4295 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4296 host lookup before query-style lookups in lists that might use this
4297 variable. This has therefore been abolished, and the "net-" prefix is no
4298 longer necessary for query-style lookups.
4300 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4301 have been a typo for LFLAGS, so it has been changed.
4303 13. The install script calls Exim with "-C /dev/null" in order to find the
4304 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4305 to be output. Howeve, since Exim outputs its version number before the
4306 error, it didn't break the script. It just looked ugly. I fixed this by
4307 always allowing "-C /dev/null" if the caller is root.
4309 14. Ignore overlarge ACL variable number when reading spool file - insurance
4310 against a later release with more variables having written the file.
4312 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4313 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4314 this, as well as the form without the "IPv6" on the front.
4316 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4317 OS/Makefile-Darwin file.
4319 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4320 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4322 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4323 variable. However, some OS use ints for inodes. I've added cast to long int
4324 to get rid of the compiler warning.
4326 19. I had forgotten to lock out "/../" in configuration file names when
4327 ALT_CONFIG_PREFIX was set.
4329 20. Routers used for verification do not need to specify transports. However,
4330 if such a router generated a host list, and callout was configured, Exim
4331 crashed, because it could not find a port number from the (non-existent)
4332 transport. It now assumes port 25 in this circumstance.
4334 21. Added the -t option to exigrep.
4336 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4337 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4338 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4339 them is set, LOOKUP_LSEARCH is forced.
4341 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4343 24. Added sender and host information to the "rejected by local_scan()" log
4344 line; previously there was no indication of these.
4346 25. Added .include_if_exists.
4348 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4349 Linux. It turns out that not all file systems support this. Apparently some
4350 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4351 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4352 not supported on the file descriptor, is now ignored when Exim is trying to
4353 sync a directory. This applies only to Linux.
4355 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4357 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4358 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4361 29. Added support for maildirsize files from supplied patch (modified a bit).
4363 30. The use of :fail: followed by an empty string could lead Exim to respond to
4364 sender verification failures with (e.g.):
4366 550 Verification failed for <xxx>
4367 550 Sender verify failed
4369 where the first response line was missing the '-' that indicates it is not
4370 the final line of the response.
4372 31. The loop for finding the name of the user that called Exim had a hardwired
4373 limit of 10; it now uses the value of finduser_retries, which is used for
4374 all other user lookups.
4376 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4379 33. Exim was neglecting to zero errno before one call of strtol() when
4380 expanding a string and expecting an integer value. On some systems this
4381 resulted in spurious "integer overflow" errors. Also, it was casting the
4382 result into an int without checking.
4384 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4385 did not work. The code looks as if it has *never* worked, though it appears
4386 to have been documented since at least releast 1.62. I have made it work.
4388 35. The "timeout_DNS" error in retry rules, also documented since at least
4389 1.62, also never worked. As it isn't clear exactly what this means, and
4390 clearly it isn't a major issue, I have abolished the feature by treating it
4391 as "timeout", and writing a warning to the main and panic logs.
4393 36. The display of retry rules for -brt wasn't always showing the error code
4396 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4397 timeout_connect_A, timeout_connect_MX.
4399 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4400 to the empty sender.
4402 39. The daemon was not analysing the content of -oX till after it had closed
4403 stderr and disconnected from the controlling terminal. This meant that any
4404 syntax errors were only noted on the panic log, and the return code from
4405 the command was 0. By re-arranging the code a little, I've made the
4406 decoding happen first, so such errors now appear on stderr, and the return
4407 code is 1. However, the actual setting up of the sockets still happens in
4408 the disconnected process, so errors there are still only recorded on the
4411 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4412 connections (as happens on some IP stacks) was logged at start up time as
4413 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4414 it from "IPv6 and IPv4", which means that two separate sockets are being
4417 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4418 says whether AF_INET or AF_INET6 was passed as an argument.
4420 42. Exiwhat output was messed up when time zones were included in log
4423 43. Exiwhat now gives more information about the daemon's listening ports,
4424 and whether -tls-on-connect was used.
4426 44. The "port" option of the smtp transport is now expanded.
4428 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4429 silently ignored. Now an error message is written to the main and panic
4432 46. There's a new ACL modifier called "logwrite" which writes to a log file
4433 as soon as it is encountered.
4435 47. Added $local_user_uid and $local_user_gid at routing time.
4437 48. Exim crashed when trying to verify a sender address that was being
4440 49. Exim was recognizing only a space character after ".include". It now also
4441 recognizes a tab character.
4443 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4444 extracting the relevant information from the specification. The man page no
4445 longer contains scrambled data for the -d option, and I've added a section
4446 at the front about calling Exim under different names.
4448 51. Added "extra_headers" argument to the "mail" command in filter files.
4450 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4453 53. Installed eximstats 1.29.
4455 54. Added transport_filter_timeout as a generic transport option.
4457 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4458 Cc: header lines. This was required by RFC 822, but it not required by RFC
4461 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4462 incoming messages that did not have them. Now it does so only if the
4463 message originates locally, that is, if there is no associated remote host
4464 address. When Resent- header lines are present, this applies to the Resent-
4465 lines rather than the non-Resent- lines.
4467 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4468 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4470 58. Messages for configuration errors now include the name of the main
4471 configuration file - useful now that there may be more than one file in a
4472 list (.included file names were always shown).
4474 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4475 for those rare installations that do not start the daemon as root or run it
4476 setuid root. I've cut out the call to initgroups() if the daemon is not
4479 60. The Exim user and group can now be bound into the binary as text strings
4480 that are looked up at the start of Exim's processing.
4482 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4484 62. Added $mailstore_basename variable.
4486 63. Installed patch to sieve.c from Michael Haardt.
4488 64. When Exim failed to open the panic log after failing to open the main log,
4489 the original message it was trying to log was written to stderr and debug
4490 output, but if they were not available (the usual case in production), it
4491 was lost. Now it is written to syslog before the two lines that record the
4492 failures to open the logs.
4494 65. Users' Exim filters run in subprocesses under the user's uid. It is
4495 possible for a "deliver" command or an alias in a "personal" command to
4496 provoke an address rewrite. If logging of address rewriting is configured,
4497 this fails because the process is not running as root or exim. There may be
4498 a better way of dealing with this, but for the moment (because 4.30 needs
4499 to be released), I have disabled address rewrite logging when running a
4500 filter in a non-root, non-exim process.
4506 1. The buildconfig auxiliary program wasn't quoting the value set for
4507 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4508 not defined. This bug was masked in 4.22 by the effect that was fixed in
4511 2. Some messages that were rejected after a message id was allocated were
4512 shown as "incomplete" by exigrep. It no longer does this for messages that
4513 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4515 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4516 have allow_domain_literals set, the ID did not get logged in the <= line.
4517 Domain literals are now always recognized in Message-ID: header lines.
4519 4. The first argument for a ${extract expansion item is the key name or field
4520 number. Leading and trailing spaces in this item were not being ignored,
4521 causing some misleading effects.
4523 5. When deliver_drop_privilege was set, single queue runner processes started
4524 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4525 same command in the process it spins off) were not dropping privilege.
4527 6. When the daemon running as "exim" started a queue runner, it always
4528 re-executed Exim in the spun-off process. This is a waste of effort when
4529 deliver_drop_privilege is set. The new process now just calls the
4530 queue-runner function directly.
4536 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4539 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4540 ignored. Though the use of -oP was forcing the writing of a pid file, it
4541 was always written to the default place.
4543 3. If the message "no IP address found for host xxxx" is generated during
4544 incoming verification, it is now followed by identification of the incoming
4545 connection (so you can more easily find what provoked it).
4547 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4549 5. Added some features to "harden" Exim a bit more against certain attacks:
4551 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4552 be put in Local/Makefile. This is like the never_users runtime option,
4553 but it cannot be overridden. The default setting is "root".
4555 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4556 prefix string with which any file named in a -C command line option
4559 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4560 is retained for -C and -D only if the caller of Exim is root. Without
4561 it, the exim user may also use -C and -D and retain privilege.
4563 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4564 command line option is disabled.
4566 6. Macro names set by the -D option must start with an upper case letter, just
4567 like macro names defined in the configuration file.
4569 7. Added "dereference=" facility to LDAP.
4571 8. Two instances of the typo "uknown" in the source files are fixed.
4573 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4574 the Configure-Makefile script screwed up while processing it.
4576 10. Incorporated PCRE 4.4.
4578 11. The SMTP synchronization check was not operating right at the start of an
4579 SMTP session. For example, it could not catch a HELO sent before the client
4580 waited for the greeting. There is now a check for outstanding input at the
4581 point when the greeting is written. Because of the duplex, asynchronous
4582 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4583 way, but not yet received, when the check is performed.
4585 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4586 on TCP/IP sockets, because this apparently causes some broken clients to
4589 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4590 unchanged) from the Cygwin maintainer.
4592 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4593 was supported instead of testing for mbx. Effectively a typo.
4595 15. The spa authenticator server code was not checking that the input it
4596 received was valid base64.
4598 16. The debug output line for the "set" modifier in ACLs was not showing the
4599 name of the variable that was being set.
4601 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4603 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4604 lookup on its own. Something else was needed to trigger the lookup. For
4605 example, a match in host_lookup or the need for a host name in a host list.
4606 Now, if $sender_host_name is referenced and the host name has not yet been
4607 looked up, a lookup is performed. If the lookup fails, the variable remains
4608 empty, and $host_lookup_failed is set to "1".
4610 19. Added "eqi" as a case-independent comparison operator.
4612 20. The saslauthd authentication condition could segfault if neither service
4613 nor realm was specified.
4615 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4616 error message that was logged was misleading, and incoming SMTP
4617 connections were dropped. The message is now more accurate, and temporary
4618 errors are given to SMTP connections.
4620 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4621 (except RSET) with a 421 error, until QUIT is received. However, it was
4622 failing to send a response to QUIT.
4624 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4625 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4626 ACL is now run after the helo_try_verify_hosts code.
4628 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4629 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4630 case-independent, so other case variants are also recognized). Apparently
4631 some systems use these upper case variants.
4633 25. If more than two messages were waiting for the same host, and a transport
4634 filter was specified for the transport, Exim sent two messages over the
4635 same TCP/IP connection, and then failed with "socket operation on non-
4636 socket" when it tried to send the third.
4638 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4640 27. The extern definition of crypt16() in expand.c was not being excluded when
4641 the OS had its own crypt16() function.
4643 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4644 as a preferred synonym for return_size_limit, both as an option and as an
4647 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4649 30. Changed the default configuration ACL to relax the local part checking rule
4650 for addresses that are not in any local domains. For these addresses,
4651 slashes and pipe symbols are allowed within local parts, but the sequence
4652 /../ is explicitly forbidden.
4654 31. SPA server authentication was not clearing the challenge buffer before
4657 32. log_message in a "warn" ACL statement was writing to the reject log as
4658 well as to the main log, which contradicts the documentation and doesn't
4659 seem right (because no rejection is happening). So I have stopped it.
4661 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4662 However, I am unable to do any testing of this.
4664 34. Fixed an infelicity in the appendfile transport. When checking directories
4665 for a mailbox, to see if any needed to be created, it was accidentally
4666 using path names with one or more superfluous leading slashes; tracing
4667 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4669 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4670 contents are added to the log line that is written for every discarded
4671 recipient. (Previously a log_message setting was ignored.)
4673 36. The ${quote: operator now quotes the string if it is empty.
4675 37. The install script runs exim in order to find its version number. If for
4676 some reason other than non-existence or emptiness, which it checks, it
4677 could not run './exim', it was installing it with an empty version number,
4678 i.e. as "exim-". This error state is now caught, and the installation is
4681 38. An argument was missing from the function that creates an error message
4682 when Exim fails to connect to the socket for saslauthd authentication.
4683 This could cause Exim to crash, or give a corrupted message.
4685 39. Added isip, isip4, and isip6 to ${if conditions.
4687 40. The ACL variables $acl_xx are now saved with the message, and can be
4688 accessed later in routers, transports, and filters.
4690 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4691 strings in the file are not string-expanded.
4693 42. If a MAIL command specified a SIZE value that was too large to fit into an
4694 int variable, the check against message_size_limit failed. Such values are
4695 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4696 day this will have to be increased, but I don't think I want to be around
4697 when emails are that large.
4704 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4705 iconv() is not standard in FreeBSD.
4707 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4708 IPv6 enabled. The observed symptom was a segmentation fault on return from
4709 the function os_common_find_running_interfaces() in src/os.c.
4711 3. In the check_special_case() function in daemon.c I had used "errno" as an
4712 argument name, which causes warnings on some systems. This was basically a
4713 typo, since it was named "eno" in the comments!
4715 4. The code that waits for the clock to tick (at a resolution of some fraction
4716 of a second) so as to ensure message-id uniqueness was always waiting for
4717 at least one whole tick, when it could have waited for less. [This is
4718 almost certainly not relevant at current processor speeds, where it is
4719 unlikely to ever wait at all. But we try to future-proof.]
4721 5. The function that sleeps for a time interval that includes fractions of a
4722 second contained a race. It did not block SIGALRM between setting the
4723 timer, and suspending (a couple of lines later). If the interval was short
4724 and the sigsuspend() was delayed until after it had expired, the suspension
4725 never ended. On busy systems this could lead to processes getting stuck for
4728 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4729 process, before it forks any delivery processes. The open lookup caching
4730 mechanism meant that the open file or database connection was passed into
4731 the delivery process. The problem was that delivery processes always tidy
4732 up cached lookup data. This could cause a problem for the next delivery
4733 process started by the queue runner, because the external queue runner
4734 process does not know about the closure. So the next delivery process
4735 still has data in the lookup cache. In the case of a file lookup, there was
4736 no problem because closing a file descriptor in a subprocess doesn't affect
4737 the parent. However, if the lookup was caching a connection to a database,
4738 the connection was closed, and the second delivery process was likely to
4739 see errors such as "PGSQL: query failed: server closed the connection
4740 unexpectedly". The problem has been fixed by closing all cached lookups
4741 in a queue runner before running a delivery process.
4743 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4744 seem to have the "const" qualifier which it has on other OS. I've
4747 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4748 *of the same type* (client or server) with the same public name that an
4749 error should be diagnosed.
4751 9. When Exim looked up a host name for an IP address, but failed to find the
4752 original IP address when looking up the host name (a safety check), it
4753 output the message "<ip address> does not match any IP for NULL", which was
4754 confusing, to say the least. The bug was that the host name should have
4755 appeared instead of "NULL".
4757 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4758 user that is built into the binary, and the -C or -D options is used, root
4759 privilege is dropped before the configuration file is read. In addition,
4760 logging is switched to stderr instead of the normal log files. If the
4761 configuration then re-defines the Exim user, the unprivileged environment
4762 is probably not what is expected, so Exim logs a panic warning message (but
4765 However, if deliver_drop_privilege is set, the unprivileged state may well
4766 be exactly what is intended, so the warning has been cut out in that case,
4767 and Exim is allowed to try to write to its normal log files.
4773 1. smtp_return_error_details was not giving details for temporary sender
4774 or receiver verification errors.
4776 2. Diagnose a configuration error if two authenticators have the same public
4779 3. Exim used not to create the message log file for a message until the first
4780 delivery attempt. This could be confusing when incoming messages were held
4781 for policy or load reasons. The message log file is now created at the time
4782 the message is received, and an initial "Received" line is written to it.
4784 4. The automatically generated man page for command line options had a minor
4785 bug that caused no ill effects; however, a more serious problem was that
4786 the procedure for building the man page automatically didn't always
4787 operate. Consequently, release 4.20 contains an out-of-date version. This
4788 shouldn't happen again.
4790 5. When building Exim with embedded Perl support, the script that builds the
4791 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4792 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4794 6. The freeze_tell option was not being used for messages that were frozen on
4795 arrival, either by an ACL or by local_scan().
4797 7. Added the smtp_incomplete_transaction log selector.
4799 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4800 was accepting AUTH without a new EHLO.
4802 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4803 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4806 10. Logging of TCP/IP connections (when configured) now happens in the main
4807 daemon process instead of the child process, so that the TCP/IP connection
4808 count is more accurate (but it can never be perfect).
4810 11. The use of "drop" in a nested ACL was not being handled correctly in the
4811 outer ACL. Now, if condition failure induced by the nested "drop" causes
4812 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4813 or "require"), the connection is dropped.
4815 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4816 that yield "discard" can now be used with an "accept" or a "discard" verb,
4817 but an error is generated for any others (because I can't see a useful way
4818 to define what should happen).
4820 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4821 are now processed in the same way as lines in the Exim configuration file.
4822 In particular, continuation lines are supported.
4824 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4826 15. Added -ti meaning -t -i.
4828 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4829 domains, and warn by logging if others are found.
4831 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4832 returned from GIFCONF when Exim is trying to find the list of interfaces on
4833 a host. The code in os.c has been modified to copy each ifreq to an aligned
4834 structure in all cases.
4836 Also, in some cases, the returned ifreq's were being copied to a 'struct
4837 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4838 means the last couple of bytes of an IPv6 address could be chopped if the
4839 ifreq contained only a normal sockaddr (14 bytes storage).
4841 18. Named domain lists were not supported in the hosts_treat_as_local option.
4842 An entry such as +xxxx was not recognized, and was treated as a literal
4845 19. Ensure that header lines added by a DATA ACL are included in the reject log
4846 if the ACL subsequently rejects the message.
4848 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4849 MD5 (which is deprecated).
4851 21. When testing a filter file using -bf, Exim was writing a message when it
4852 took the sender from a "From " line in the message, but it was not doing so
4853 when it took $return_path from a Return-Path: header line. It now does.
4855 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4856 with a valid header line field name (a series of printing characters
4857 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4859 23. Changed "disc" in the source to "disk" to conform to the documentation and
4860 the book and for uniformity.
4862 24. Ignore Sendmail's -Ooption=value command line item.
4864 25. When execve() failed while trying to run a command in a pipe transport,
4865 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4866 could be confused with a return value of 69 from the command itself. This
4867 has been changed to 127, the value the shell returns if it is asked to run
4868 a non-existent command. The wording for the related log line suggests a
4869 non-existent command as the problem.
4871 26. If received_header_text expands to an empty string, do not add a Received:
4872 header line to the message. (Well, it adds a token one on the spool, but
4873 marks it "old" so that it doesn't get used or transmitted.)
4875 27. Installed eximstats 1.28 (addition of -nt option).
4877 28. There was no check for failure on the call to getsockname() in the daemon
4878 code. This can fail if there is a shortage of resources on the system, with
4879 ENOMEM, for example. A temporary error is now given on failure.
4881 29. Contrary to the C standard, it seems that in some environments, the
4882 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4883 program. Exim now does this explicitly; it affects the formatting of
4884 timestamps using strftime().
4886 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4887 complaints. I've now initialized all the variables, to avoid this.
4889 32. Header lines added by a system filter were not being "seen" during
4890 transport-time rewrites.
4892 33. The info_callback() function passed to OpenSSL is set up with type void
4893 (*)(SSL *, int, int), as described somewhere. However, when calling the
4894 function (actually a macro) that sets it up, the type void(*)() is
4895 expected. I've put in a cast to prevent warnings from picky compilers.
4897 34. If a DNS black list lookup found a CNAME record, but there were no A
4898 records associated with the domain it pointed at, Exim crashed.
4900 35. If a DNS black list lookup returned more than one A record, Exim ignored
4901 all but the first. It now scans all returned addresses if a particular IP
4902 value is being sought. In this situation, the contents of the
4903 $dnslist_value variable are a list of all the addresses, separated by a
4906 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4907 to accept a host name and all its aliases if the forward lookup for any of
4908 them yielded the IP address of the incoming connection. Now it accepts only
4909 those names whose forward lookup yields the correct IP address. Any other
4910 names are discarded. This closes a loophole whereby a rogue DNS
4911 administrator could create reverse DNS records to break through a
4912 wildcarded host restriction in an ACL.
4914 37. If a user filter or a system filter that ran in a subprocess used any of
4915 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4916 the wrong values were passed to the pipe command ($thisaddress had the
4917 value of $0, $0 had the value of $1, etc). This bug was introduced by
4918 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4920 38. Improved the line breaking for long SMTP error messages from ACLs.
4921 Previously, if there was no break point between 40 and 75 characters, Exim
4922 left the rest of the message alone. Two changes have been made: (a) I've
4923 reduced the minimum length to 35 characters; (b) if it can't find a break
4924 point between 35 and 75 characters, it looks ahead and uses the first one
4925 that it finds. This may give the occasional overlong line, but at least the
4926 remaining text gets split now.
4928 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4929 file descriptors might be low, and that setting 1000 would always raise it.
4930 It turns out that in some environments, the limit is already over 1000 and
4931 that lowering it causes trouble. So now Exim takes care not to decrease it.
4933 40. When delivering a message, the value of $return_path is set to $sender_
4934 address at the start of routing (routers may change the value). By an
4935 oversight, this default was not being set up when an address was tested by
4936 -bt or -bv, which affected the outcome if any router or filter referred to
4939 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4940 log only when "message" or "log_message" are set. However, if one of the
4941 conditions was an address verification, or a call to a nested ACL, the
4942 messages generated by the underlying test were being passed through. This
4943 no longer happens. The underlying message is available in $acl_verify_
4944 message for both "message" and "log_message" expansions, so it can be
4945 passed through if needed.
4947 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4948 new expansion $bh_ to give the encoded byte string without charset
4949 translation. Translation happens only if iconv() is available; HAVE_ICONV
4950 indicates this at build time. HEADERS_CHARSET gives the charset to
4951 translate to; headers_charset can change it in the configuration, and
4952 "headers charset" can change it in an individual filter file.
4954 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4955 that creates RFC 2047 encoded "words" labels them as that charset instead
4956 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4957 expansion operator; (ii) when Exim creates a From: line for a local
4958 message; (iii) when a header line is rewritten to include a "phrase" part.
4960 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4961 buggy, causing it to skip the first lines of messages whose message ID
4962 ended in 'D'. This would not have bitten before Exim release 4.14, because
4963 message IDs were unlikely to end in 'D' before then. The effect was to have
4964 incorrect size information for certain domains.
4966 45. #include "config.h" was missing at the start of the crypt16.c module. This
4967 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4970 46. If there was a timeout during a "random" callout check, Exim treated it as
4971 a failure of the random address, and carried on sending RSET and the real
4972 address. If the delay was just some slowness somewhere, the response to the
4973 original RCPT would be taken as a response to RSET and so on, causing
4974 mayhem of various kinds.
4976 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4977 when I implemented it. It didn't allow for the fact that some option values
4978 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4979 the right test for positive values.
4981 48. Domain names in DNS records are case-independent. Exim always looks them up
4982 in lower case. Some resolvers return domain names in exactly the case they
4983 appear in the zone file, that is, they may contain uppercase letters. Not
4984 all resolvers do this - some return always lower case. Exim was treating a
4985 change of case by a resolver as a change of domain, similar to a widening
4986 of a domain abbreviation. This triggered its re-routing code and so it was
4987 trying to route what was effectively the same domain again. This normally
4988 caused routing to fail (because the router wouldn't handle the domain
4989 twice). Now Exim checks for this case specially, and just changes the
4990 casing of the domain that it ultimately uses when it transmits the message
4993 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4996 50. If a filter generated a file delivery with a non-absolute name (possible if
4997 no home directory exists for the router), the forbid_file option was not
5000 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5001 the existing equality matching.
5003 52. Exim was using ints instead of ino_t variables in some places where it was
5004 dealing with inode numbers.
5006 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5007 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5008 called TMPDIR, and if it finds it is different, it changes its value.
5010 54. The smtp_printf() function is now made available to local_scan() so
5011 additional output lines can be written before returning. There is also an
5012 smtp_fflush() function to enable the detection of a dropped connection.
5013 The variables smtp_input and smtp_batched_input are exported to
5016 55. Changed the default runtime configuration: the message "Unknown user"
5017 has been removed from the ACL, and instead placed on the localuser router,
5018 using the cannot_route_message feature. This means that any verification
5019 failures that generate their own messages won't get overridden. Similarly,
5020 the "Unrouteable address" message that was in the ACL for unverifiable
5021 relay addresses has also been removed.
5023 56. Added hosts_avoid_esmtp to the smtp transport.
5025 57. The exicyclog script was not checking for the esoteric option
5026 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5027 will work only if exicyclog is run under the appropriate euid.
5029 58. Following a discussion on the list, the rules by which Exim recognises line
5030 endings on incoming messages have been changed. The -dropcr and drop_cr
5031 options are now no-ops, retained only for backwards compatibility. The
5032 following line terminators are recognized: LF CRLF CR. However, special
5033 processing applies to CR:
5035 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5036 nor a local message in the state where . is a terminator.
5038 (ii) If a bare CR is encountered in a header line, an extra space is added
5039 after the line terminator so as not to end the header. The reasoning
5040 behind this is that bare CRs in header lines are most likely either
5041 to be mistakes, or people trying to play silly games.
5043 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5044 was being incorrectly given as 18 bytes larger than it should have been.
5045 This is a VOB (very old bug).
5047 60. This may never have affected anything current, but just in case it has:
5048 When the local host is found other than at the start of a list of hosts,
5049 the local host, those with the same MX, and any that follow, are discarded.
5050 When the list in question was part of a longer list of hosts, the following
5051 hosts (not currently being processed) were also being discarded. This no
5052 longer happens. I'm not sure if this situation could ever has previously
5055 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5058 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5059 header. This is something that is recommended in a new Internet Draft, and
5060 is something that is documented as being done by Sendmail. There are two
5061 possible values. For messages generated by the autoreply transport, Exim
5064 Auto-Submitted: auto-replied
5066 whereas for all other generated messages (e.g. bounces) it adds
5068 Auto-Submitted: auto-generated
5070 63. The "personal" condition in filters now includes a test for the
5071 Auto-Submitted: header. If it contains the string "auto-" the message it
5072 not considered personal.
5074 64. Added rcpt_include_affixes as a generic transport option.
5076 65. Added queue_only_override (default true).
5078 66. Added the syslog_duplication option.
5080 67. If what should have been the first header line of a message consisted of
5081 a space followed by a colon, Exim was mis-interpreting it as a header line.
5082 It isn't of course - it is syntactically invalid and should therefore be
5083 treated as the start of the message body. The misbehaviour could have
5084 caused a number of strange effects, including loss of data in subsequent
5085 header lines, and spool format errors.
5087 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5088 client host had authenticated. This control can now be exercised by an ACL
5089 for more flexibility.
5091 69. By default, callouts do not happen when testing with -bh. There is now a
5092 variant, -bhc, which does actually run the callout code, including
5093 consulting and updating the callout cache.
5095 70. Added support for saslauthd authentication, courtesy of Alexander
5098 71. If statvfs() failed on the spool or log directories while checking their
5099 size for availability, Exim confusingly gave the error "space shortage".
5100 Furthermore, in debugging mode it crashed with a floating point exception.
5101 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5102 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5103 serious problem, Exim now writes to the main and panic logs when this
5104 happens, with details of the failure. It then refuses to accept the
5105 incoming message, giving the message "spool directory problem" or "log
5106 directory problem" with a 421 code for SMTP messages.
5108 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5109 0, 1, and 2 exist, because some OS complain for execs without them (see
5110 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5111 descriptors. However, the code omitted to check that the open succeeded,
5112 causing mysterious errors if for some reason the permissions on /dev/null
5113 got screwed. Now Exim writes a message to the main and panic logs, and
5114 bombs out if it can't open /dev/null.
5116 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5117 interact so that it is all more flexible. It is supposed to remain
5118 backwards compatible. Also added extra_local_interfaces.
5120 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5121 to bomb out with an assertion failure - to the client this appears as a
5122 connection drop. This problem occurs in the part of the code that was taken
5123 from the Samba project. Fortunately, the assertion is in a very simple
5124 function, so I have fixed this by reproducing the function inline in the
5125 one place where it is called, and arranging for authentication to fail
5126 instead of killing the process with assert().
5128 75. The SPA client code was not working when the server requested OEM rather
5129 than Unicode encoding.
5131 76. Added code to make require_files with a specific uid setting more usable in
5132 the case where statting the file as root fails - usually a non-root-mounted
5133 NFS file system. When this happens and the failure is EACCES, Exim now
5134 forks a subprocess and does the per-uid checking as the relevant uid.
5136 77. Added process_log_path.
5138 78. If log_file_path was not explicitly set, a setting of check_log_space or
5139 check_log_inodes was ignored.
5141 79. If a space check for the spool or log partitions fails, the incident is now
5142 logged. Of course, in the latter case the data may get lost...
5144 80. Added the %p formatting code to string_format() so that it can be used to
5145 print addresses in debug_print(). Adjusted all the address printing in the
5146 debugging in store.c to use %p rather than %d.
5148 81. There was a concern that a line of code in smtp_in.c could overflow a
5149 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5150 initially expressed, the concern was not well-founded, because trailing
5151 spaces are removed early. However, if the trailing spaces were followed by
5152 a NULL, they did not get removed, so the overflow was possible. Two fixes
5155 (a) I re-wrote the offending code in a cleaner fashion.
5156 (b) If an incoming SMTP command contains a NULL character, it is rejected
5159 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5160 runs initgroups(), so that if the Exim user is in any additional groups,
5161 they will be used during message reception.
5167 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob