1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
74 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
77 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
78 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
79 configuration file which is writeable by the Exim user or group.
81 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
82 of configuration files to cover files specified with the -C option if
83 they are going to be used with root privileges, not just the default
86 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
87 option (effectively making it always true).
89 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
90 files to be used while preserving root privileges.
92 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
93 that rogue child processes cannot use them.
95 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
96 run-time user, instead of root.
98 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
99 Exim run-time user without dropping privileges.
105 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
106 $data_path, and $header_path variables; fixed documentation bugs and
109 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
110 exipick to access non-standard spools, including the "frozen" queue
113 NM/01 Bugzilla 965: Support mysql stored procedures.
114 Patch from Alain Williams
116 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
118 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
119 Patch from Andreas Metzler
121 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
122 Patch from Kirill Miazine
124 NM/05 Bugzilla 671: Added umask to procmail example.
126 JJ/03 installed exipick 20100323.0, fixing doc bug
128 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
129 directory. Notification and patch from Dan Rosenberg.
131 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
133 TK/02 Improve log output when DKIM signing operation fails.
135 MH/01 Treat the transport option dkim_domain as a colon separated
136 list, not as a single string, and sign the message with each element,
137 omitting multiple occurences of the same signer.
139 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
140 Bugzilla 985, 986. Patch by Simon Arlott
142 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
143 Patch by Simon Arlott
145 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
146 MBX locking. Notification from Dan Rosenberg.
152 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
154 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
156 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
158 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
160 NM/04 Bugzilla 915: Flag broken perl installation during build.
166 TK/01 Added patch by Johannes Berg that expands the main option
167 "spamd_address" if it starts with a dollar sign.
169 TK/02 Write list of recipients to X-Envelope-Sender header when building
170 the mbox-format spool file for content scanning (suggested by Jakob
173 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
174 (http://www.dcc-servers.net/) support via dccifd. Activated by
175 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
177 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
178 by Mark Daniel Reidel <mr@df.eu>.
180 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
181 When building exim an external PCRE library is now needed -
182 PCRE is a system library on the majority of modern systems.
183 See entry on PCRE_LIBS in EDITME file.
185 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
186 conversation. Added nologin parameter to request.
187 Patch contributed by Kirill Miazine.
189 TF/01 Do not log submission mode rewrites if they do not change the address.
191 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
193 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
194 log files in place. Contributed by Roberto Lima.
196 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
198 TF/03 Bugzilla 615: When checking the local_parts router precondition
199 after a local_part_suffix or local_part_prefix option, Exim now
200 does not use the address's named list lookup cache, since this
201 contains cached lookups for the whole local part.
203 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
204 Robert Millan. Documentation is in experimental-spec.txt.
206 TF/04 Bugzilla 668: Fix parallel build (make -j).
208 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
210 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
211 Patch provided by Jan Srzednicki.
213 TF/05 Leading white space used to be stripped from $spam_report which
214 wrecked the formatting. Now it is preserved.
216 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
217 that they are available at delivery time.
219 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
221 TF/08 TLS error reporting now respects the incoming_interface and
222 incoming_port log selectors.
224 TF/09 Produce a more useful error message if an SMTP transport's hosts
225 setting expands to an empty string.
227 NM/06 Bugzilla 744: EXPN did not work under TLS.
228 Patch provided by Phil Pennock.
230 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
231 Patch provided by Richard Godbee.
233 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
234 acl_smtp_notquit, added index entry.
236 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
237 Patch provided by Eugene Bujak.
239 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
240 accept(). Patch provided by Maxim Dounin.
242 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
243 Patch provided by Phil Pennock.
245 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
247 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
248 Patch provided by Brad "anomie" Jorsch.
250 NM/14 Bugzilla 622: Added timeout setting to transport filter.
251 Patch provided by Dean Brooks.
253 TK/05 Add native DKIM support (does not depend on external libraries).
255 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
256 Patch provided by Graeme Fowler.
258 NM/16 Bugzilla 851: Documentation example syntax fix.
260 NM/17 Changed NOTICE file to remove references to embedded PCRE.
262 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
265 NM/19 Bugzilla 745: TLS version reporting.
266 Patch provided by Phil Pennock.
268 NM/20 Bugzilla 167: bool: condition support.
269 Patch provided by Phil Pennock.
271 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
272 clients. Patch provided by Phil Pennock.
274 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
275 Patch provided by Brad "anomie" Jorsch.
277 NM/23 Bugzilla 687: Fix misparses in eximstats.
278 Patch provided by Heiko Schlittermann.
280 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
281 Patch provided by Heiko Schlittermann.
283 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
284 plus update to original patch.
286 NM/26 Bugzilla 799: Documentation correction for ratelimit.
288 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
289 Patch provided by David Brownlee.
291 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
293 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
295 NM/30 Bugzilla 888: TLS documentation bugfixes.
297 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
299 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
300 Unlike the original bugzilla I have changed all shell scripts in src tree.
302 NM/33 Bugzilla 898: Transport filter timeout fix.
303 Patch by Todd Rinaldo.
305 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
306 Patch by Serge Demonchaux.
308 NM/35 Bugzilla 39: Base64 decode bug fixes.
309 Patch by Jakob Hirsch.
311 NM/36 Bugzilla 909: Correct connect() call in dcc code.
313 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
315 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
317 NM/39 Bugzilla 911: Fixed MakeLinks build script.
323 TK/01 Add preliminary DKIM support. Currently requires a forked version of
324 ALT-N's libdkim that I have put here:
325 http://duncanthrax.net/exim-experimental/
327 Note to Michael Haardt: I had to rename some vars in sieve.c. They
328 were called 'true' and it seems that C99 defines that as a reserved
329 keyword to be used with 'bool' variable types. That means you could
330 not include C99-style headers which use bools without triggering
331 build errors in sieve.c.
333 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
334 as mailq or other aliases. Changed the --help handling significantly
335 to do whats expected. exim_usage() emits usage/help information.
337 SC/01 Added the -bylocaldomain option to eximstats.
339 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
341 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
343 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
349 PH/01 Another patch from the Sieve maintainer.
351 PH/02 When an IPv6 address is converted to a string for single-key lookup
352 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
353 dots are used instead of colons so that keys in lsearch files need not
354 contain colons. This was done some time before quoting was made available
355 in lsearch files. However, iplsearch files do require colons in IPv6 keys
356 (notated using the quote facility) so as to distinguish them from IPv4
357 keys. This meant that lookups for IP addresses in host lists did not work
358 for iplsearch lookups.
360 This has been fixed by arranging for IPv6 addresses to be expressed with
361 colons if the lookup type is iplsearch. This is not incompatible, because
362 previously such lookups could never work.
364 The situation is now rather anomolous, since one *can* have colons in
365 ordinary lsearch keys. However, making the change in all cases is
366 incompatible and would probably break a number of configurations.
368 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
371 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
372 conversion specification without a maximum field width, thereby enabling
373 a rogue spamd server to cause a buffer overflow. While nobody in their
374 right mind would setup Exim to query an untrusted spamd server, an
375 attacker that gains access to a server running spamd could potentially
376 exploit this vulnerability to run arbitrary code as the Exim user.
378 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
379 $primary_hostname instead of what libspf2 thinks the hosts name is.
381 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
382 a directory entry by the name of the lookup key. Previously, if a
383 symlink pointed to a non-existing file or a file in a directory that
384 Exim lacked permissions to read, a lookup for a key matching that
385 symlink would fail. Now it is enough that a matching directory entry
386 exists, symlink or not. (Bugzilla 503.)
388 PH/03 The body_linecount and body_zerocount variables are now exported in the
391 PH/04 Added the $dnslist_matched variable.
393 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
394 This means they are set thereafter only if the connection becomes
397 PH/06 Added the client_condition to authenticators so that some can be skipped
398 by clients under certain conditions.
400 PH/07 The error message for a badly-placed control=no_multiline_responses left
401 "_responses" off the end of the name.
403 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
405 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
406 (without spaces) instead of just copying the configuration text.
408 PH/10 Added the /noupdate option to the ratelimit ACL condition.
410 PH/11 Added $max_received_linelength.
412 PH/12 Added +ignore_defer and +include_defer to host lists.
414 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
415 way in which PCRE > 7.0 is built.
417 PH/14 Implemented queue_only_load_latch.
419 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
420 MAIL command. The effect was to mangle the value on 64-bit systems.
422 PH/16 Another patch from the Sieve maintainer.
424 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
426 PH/18 If a system quota error occurred while trying to create the file for
427 a maildir delivery, the message "Mailbox is full" was not appended to the
428 bounce if the delivery eventually timed out. Change 4.67/27 below applied
429 only to a quota excession during the actual writing of the file.
431 PH/19 It seems that peer DN values may contain newlines (and other non-printing
432 characters?) which causes problems in log lines. The DN values are now
433 passed through string_printing() before being added to log lines.
435 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
436 and InterBase are left for another time.)
438 PH/21 Added message_body_newlines option.
440 PH/22 Guard against possible overflow in moan_check_errorcopy().
442 PH/23 POSIX allows open() to be a macro; guard against that.
444 PH/24 If the recipient of an error message contained an @ in the local part
445 (suitably quoted, of course), incorrect values were put in $domain and
446 $local_part during the evaluation of errors_copy.
452 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
453 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
456 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
457 issue a MAIL command.
459 PH/02 In an ACL statement such as
461 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
463 if a client was not listed at all, or was listed with a value other than
464 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
465 the condition was not true (as it should be), so access was not denied.
466 The bug was that the ! inversion was incorrectly passed on to the second
467 item. This has been fixed.
469 PH/03 Added additional dnslists conditions == and =& which are different from
470 = and & when the dns lookup returns more than one IP address.
472 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
473 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
475 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
476 FSYNC, which compiles an option called disable_fsync that allows for
477 bypassing fsync(). The documentation is heavily laced with warnings.
479 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
481 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
482 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
483 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
484 including adding "make clean"; (3) Added -fPIC when compiling the test
485 dynamically loaded module, to get rid of a warning.
487 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
488 message fails, move_frozen_messages = true and ignore_bounce_errors_after
489 = 0s. The bug is otherwise harmless.
491 PH/07 There was a bug in the dovecot authenticator such that the value of
492 $auth1 could be overwritten, and so not correctly preserved, after a
493 successful authentication. This usually meant that the value preserved by
494 the server_setid option was incorrect.
496 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
498 PH/09 Installed PCRE release 7.0.
500 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
501 run for batched SMTP input. It is now run at the start of every message
502 in the batch. While fixing this I discovered that the process information
503 (output by running exiwhat) was not always getting set for -bs and -bS
504 input. This is fixed, and it now also says "batched" for BSMTP.
506 PH/11 Added control=no_pipelining.
508 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
509 patch, slightly modified), and move the expansion of helo_data till after
510 the connection is made in the smtp transport (so it can use these
513 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
515 PH/14 Added log_selector = +pid.
517 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
519 PH/16 Add ${if forany and ${if forall.
521 PH/17 Added dsn_from option to vary the From: line in DSNs.
523 PH/18 Flush SMTP output before performing a callout, unless control =
524 no_callout_flush is set.
526 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
527 was true (the default) a successful delivery failed to delete the retry
528 item, thus causing premature timeout of the address. The bug is now
531 PH/20 Added hosts_avoid_pipelining to the smtp transport.
533 PH/21 Long custom messages for fakedefer and fakereject are now split up
534 into multiline reponses in the same way that messages for "deny" and
535 other ACL rejections are.
537 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
538 with slight modification.
540 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
541 draft, changing the syntax and factoring some duplicate code".
543 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
544 for deliveries of the second and subsequent messages over the same SMTP
547 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
548 ${reduce, with only minor "tidies".
550 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
552 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
553 expansion side effects.
555 PH/27 When a message times out after an over-quota error from an Exim-imposed
556 quota, the bounce message says "mailbox is full". This message was not
557 being given when it was a system quota that was exceeded. It now should
560 MH/03 Made $recipients available in local_scan(). local_scan() already has
561 better access to the recipient list through recipients_list[], but
562 $recipients can be useful in postmaster-provided expansion strings.
564 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
565 in the case of a MAIL command with additional options following the
566 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
567 were accidentally chopped off.
569 PH/29 SMTP synchronization checks are implemented when a command is read -
570 there is a check that no more input is waiting when there shouldn't be
571 any. However, for some commands, a delay in an ACL can mean that it is
572 some time before the response is written. In this time, more input might
573 arrive, invalidly. So now there are extra checks after an ACL has run for
574 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
575 pipelining has not been advertised.
577 PH/30 MH's patch to allow iscntrl() characters to be list separators.
579 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
580 returned in the SMTP response when smtp_return_error_details was false.
583 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
584 instead of the C I/O that was originally supplied, because problems were
587 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
588 Exim which did not show up earlier: it was assuming that a call to
589 SSL_CTX_set_info_callback() might give an error value. In fact, there is
590 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
591 was a macro that became an assignment, so it seemed to work. This has
592 changed to a proper function call with a void return, hence the compile
593 error. Exim's code has been fixed.
595 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
598 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
600 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
602 PH/37 If a message is not accepted after it has had an id assigned (e.g.
603 because it turns out to be too big or there is a timeout) there is no
604 "Completed" line in the log. When some messages of this type were
605 selected by exigrep, they were listed as "not completed". Others were
606 picked up by some special patterns. I have improved the selection
607 criteria to be more general.
609 PH/38 The host_find_failed option in the manualroute router can now be set
610 to "ignore", to completely ignore a host whose IP address cannot be
611 found. If all hosts are ignored, the behaviour is controlled by the new
612 host_all_ignored option.
614 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
615 homing or because of multiple MX records with /mx) generated more than
616 one IP address, and the following item turned out to be the local host,
617 all the secondary addresses of the first item were incorrectly removed
618 from the list, along with the local host and any following hosts (which
619 is what is supposed to happen).
621 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
622 whoever called Exim into the -H file. In the case of the daemon it was
623 behaving confusingly. When first started, it used values for whoever
624 started the daemon, but after a SIGHUP it used the Exim user (because it
625 calls itself on a restart). I have changed the code so that it now always
628 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
629 message are rejected with the same error (e.g. no authentication or bad
630 sender address), and a DATA command is nevertheless sent (as can happen
631 with PIPELINING or a stupid MUA), the error message that was given to the
632 RCPT commands is included in the rejection of the DATA command. This is
633 intended to be helpful for MUAs that show only the final error to their
636 PH/42 Another patch from the Sieve maintainer.
638 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
639 Eximstats - Fixed some broken HTML links and added missing column headers
641 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
642 columns for Rejects, Temp Rejects, Ham, and Spam rows.
644 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
646 PH/43 Yet another patch from the Sieve maintainer.
648 PH/44 I found a way to check for a TCP/IP connection going away before sending
649 the response to the final '.' that terminates a message, but only in the
650 case where the client has not sent further data following the '.'
651 (unfortunately, this is allowed). However, in many cases there won't be
652 any further data because there won't be any more messages to send. A call
653 to select() can be used: if it shows that the input is "ready", there is
654 either input waiting, or the socket has been closed. An attempt to read
655 the next input character can distinguish the two cases. Previously, Exim
656 would have sent an OK response which the client would never have see.
657 This could lead to message repetition. This fix should cure that, at
658 least in a lot of common cases.
660 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
661 advertised in response to EHLO.
667 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
668 fixed by 4.65/MH/01 (is this a record?) are fixed:
670 (i) An empty string was always treated as zero by the numeric comparison
671 operators. This behaviour has been restored.
673 (ii) It is documented that the numeric comparison operators always treat
674 their arguments as decimal numbers. This was broken in that numbers
675 starting with 0 were being interpreted as octal.
677 While fixing these problems I realized that there was another issue that
678 hadn't been noticed. Values of message_size_limit (both the global option
679 and the transport option) were treated as octal if they started with 0.
680 The documentation was vague. These values are now always treated as
681 decimal, and I will make that clear in the documentation.
687 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
688 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
691 MH/01 Don't check that the operands of numeric comparison operators are
692 integers when their expansion is in "skipping" mode (fixes bug
693 introduced by 4.64-PH/07).
695 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
696 child addresses, Exim now panics and dies. Previously, because the count
697 is held in a short int, deliveries were likely to be lost. As such a
698 large number of recipients for a single message is ridiculous
699 (performance will be very, very poor), I have chosen to impose a limit
700 rather than extend the field.
706 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
707 leftover -K file (the existence of which was triggered by #402).
708 While we were at it, introduced process PID as part of the -K
709 filename. This should rule out race conditions when creating
712 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
713 processing considerably. Previous code took too long for large mails,
714 triggering a timeout which in turn triggers #401.
716 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
717 in the DK code in transports.c. sendfile() is not really portable,
718 hence the _LINUX specificness.
720 TF/01 In the add_headers option to the mail command in an Exim filter,
721 there was a bug that Exim would claim a syntax error in any
722 header after the first one which had an odd number of characters
725 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
726 callout verification, Exim cached a "reject" for the entire domain. This
727 is correct for most verifications, but it is not correct for a recipient
728 verification with use_sender or use_postmaster set, because in that case
729 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
730 case of MAIL FROM:<> rejection from other early rejections (e.g.
731 rejection of HELO). When verifying a recipient using a non-null MAIL
732 address, the cache is ignored if it shows MAIL FROM:<> rejection.
733 Whatever the result of the callout, the value of the domain cache is
734 left unchanged (for any other kind of callout, getting as far as trying
735 RCPT means that the domain itself is ok).
737 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
740 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
741 manifest itself as EPIPE rather than ECONNECT. When tidying away a
742 session, the daemon ignores ECONNECT errors and logs others; it now
743 ignores EPIPE as well.
745 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
746 (quoted-printable decoding).
748 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
749 later the small subsequent patch to fix an introduced bug.
751 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
753 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
755 PH/08 An error is now given if message_size_limit is specified negative.
757 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
758 to be given (somewhat) arbitrary names.
760 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
763 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
764 miscellaneous code fixes
766 PH/10 Added the log_reject_target ACL modifier to specify where to log
769 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
770 hostname. This is wrong, because it relates to the incoming message (and
771 probably the interface on which it is arriving) and not to the outgoing
772 callout (which could be using a different interface). This has been
773 changed to use the value of the helo_data option from the smtp transport
774 instead - this is what is used when a message is actually being sent. If
775 there is no remote transport (possible with a router that sets up host
776 addresses), $smtp_active_hostname is used.
778 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
779 tweaks were necessary in order to get it to work (see also 21 below):
780 (a) The code assumed that strncpy() returns a negative number on buffer
781 overflow, which isn't the case. Replaced with Exim's string_format()
783 (b) There were several signed/unsigned issues. I just did the minimum
784 hacking in of casts. There is scope for a larger refactoring.
785 (c) The code used strcasecmp() which is not a standard C function.
786 Replaced with Exim's strcmpic() function.
787 (d) The code set only $1; it now sets $auth1 as well.
788 (e) A simple test gave the error "authentication client didn't specify
789 service in request". It would seem that Dovecot has changed its
790 interface. Fortunately there's a specification; I followed it and
791 changed what the client sends and it appears to be working now.
793 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
796 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
797 address A is aliased to B and C, where B exists and C does not. Without
798 -v the output is "A verified" because verification stops after a
799 successful redirection if more than one address is generated. However,
800 with -v the child addresses are also verified. Exim was outputting "A
801 failed to verify" and then showing the successful verification for C,
802 with its parentage. It now outputs "B failed to verify", showing B's
803 parentage before showing the successful verification of C.
805 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
806 look up a TXT record in a specific list after matching in a combined
809 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
810 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
811 they consult the DNS. I had assumed they would set it the way they
812 wanted; and indeed my experiments on Linux seem to show that in some
813 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
814 To be on the safe side, however, I have now made the interface to
815 host_find_byname() similar to host_find_bydns(), with an argument
816 containing the DNS resolver options. The host_find_byname() function now
817 sets these options at its start, just as host_find_bydns() does. The smtp
818 transport options dns_qualify_single and dns_search_parents are passed to
819 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
820 of host_find_byname() use the default settings of RES_DEFNAMES
821 (qualify_single) but not RES_DNSRCH (search_parents).
823 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
824 spool_read_header() do less string testing, by means of a preliminary
825 switch on the second character of optional "-foo" lines. (This is
826 overdue, caused by the large number of possibilities that now exist.
827 Originally there were few.) While I was there, I also converted the
828 str(n)cmp tests so they don't re-test the leading "-" and the first
829 character, in the hope this might squeeze out yet more improvement.
831 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
832 flag allowing group syntax was set by the header_syntax check but not
833 turned off, possible causing trouble later; (2) The flag was not being
834 set at all for the header_verify test, causing "group"-style headers to
835 be rejected. I have now set it in this case, and also caused header_
836 verify to ignore an empty address taken from a group. While doing this, I
837 came across some other cases where the code for allowing group syntax
838 while scanning a header line wasn't quite right (mostly, not resetting
839 the flag correctly in the right place). These bugs could have caused
840 trouble for malformed header lines. I hope it is now all correct.
842 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
843 with the "reply" argument non-NULL. The code, however (which originally
844 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
845 but it didn't always do it. This confused somebody who was copying the
846 code for some other use. I have removed all the tests.
848 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
849 feature that was used to support insecure browsers during the U.S. crypto
850 embargo. It requires special client support, and Exim is probably the
851 only MTA that supported it -- and would never use it because real RSA is
852 always available. This code has been removed, because it had the bad
853 effect of slowing Exim down by computing (never used) parameters for the
854 RSA_EXPORT functionality.
856 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
857 authenticator to fail if there's a tab character in the incoming data
858 (there should never be unless someone is messing about, as it's supposed
859 to be base64-encoded). Also added, on Timo's advice, the "secured" option
860 if the connection is using TLS or if the remote IP is the same as the
861 local IP, and the "valid-client-cert option" if a client certificate has
864 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
865 authenticators. This can be used for authorization after authentication
866 succeeds. (In the case of plaintext, it servers for both authentication
869 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
870 if any retry times were supplied.
872 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
873 connection, where there is no client IP address to check. In this
874 situation, the verify now always succeeds.
876 PH/25 Applied John Jetmore's -Mset patch.
878 PH/26 Added -bem to be like -Mset, but loading a message from a file.
880 PH/27 In a string expansion for a processed (not raw) header when multiple
881 headers of the same name were present, leading whitespace was being
882 removed from all of them, but trailing whitespace was being removed only
883 from the last one. Now trailing whitespace is removed from each header
884 before concatenation. Completely empty headers in a concatenation (as
887 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
888 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
890 PH/29 [Removed. This was a change that I later backed out, and forgot to
891 correct the ChangeLog entry (that I had efficiently created) before
892 committing the later change.]
894 PH/30 Exim was sometimes attempting to deliver messages that had suffered
895 address errors (4xx response to RCPT) over the same connection as other
896 messages routed to the same hosts. Such deliveries are always "forced",
897 so retry times are not inspected. This resulted in far too many retries
898 for the affected addresses. The effect occurred only when there were more
899 hosts than the hosts_max_try setting in the smtp transport when it had
900 the 4xx errors. Those hosts that it had tried were not added to the list
901 of hosts for which the message was waiting, so if all were tried, there
902 was no problem. Two fixes have been applied:
904 (i) If there are any address or message errors in an SMTP delivery, none
905 of the hosts (tried or untried) are now added to the list of hosts
906 for which the message is waiting, so the message should not be a
907 candidate for sending over the same connection that was used for a
908 successful delivery of some other message. This seems entirely
909 reasonable: after all the message is NOT "waiting for some host".
910 This is so "obvious" that I'm not sure why it wasn't done
911 previously. Hope I haven't missed anything, but it can't do any
912 harm, as the worst effect is to miss an optimization.
914 (ii) If, despite (i), such a delivery is accidentally attempted, the
915 routing retry time is respected, so at least it doesn't keep
916 hammering the server.
918 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
919 in ${readsocket because some servers need this prod.
921 PH/32 Added some extra debug output when updating a wait-xxx database.
923 PH/33 The hint "could be header name not terminated by colon", which has been
924 given for certain expansion errors for a long time, was not being given
925 for the ${if def:h_colon_omitted{... case.
927 PH/34 The spec says: "With one important exception, whenever a domain list is
928 being scanned, $domain contains the subject domain." There was at least
929 one case where this was not true.
931 PH/35 The error "getsockname() failed: connection reset by peer" was being
932 written to the panic log as well as the main log, but it isn't really
933 panic-worthy as it just means the connection died rather early on. I have
934 removed the panic log writing for the ECONNRESET error when getsockname()
937 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
938 runs only) independently of the message's sender address. This meant
939 that, if the 4xx error was in fact related to the sender, a different
940 message to the same recipient with a different sender could confuse
941 things. In particualar, this can happen when sending to a greylisting
942 server, but other circumstances could also provoke similar problems.
943 I have changed the default so that the retry time for these errors is now
944 based a combination of the sender and recipient addresses. This change
945 can be overridden by setting address_retry_include_sender=false in the
948 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
949 remote server are returned as part of bounce messages. This was not
950 happening for LMTP over a pipe (the lmtp transport), but now it is the
951 same for both kinds of LMTP.
953 PH/38 Despite being documented as not happening, Exim was rewriting addresses
954 in header lines that were in fact CNAMEs. This is no longer the case.
956 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
957 and queue runs started by the daemon processed all messages. This has
958 been fixed so that -R and -S can now usefully be given with -q<time>.
960 PH/40 Import PCRE release 6.7 (fixes some bugs).
962 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
964 PH/42 Give an error if -q is specified more than once.
966 PH/43 Renamed the variables $interface_address and $interface_port as
967 $received_ip_address and $received_port, to make it clear that these
968 values apply to message reception, and not to the outgoing interface when
969 a message is delivered. (The old names remain recognized, of course.)
971 PH/44 There was no timeout on the connect() call when using a Unix domain
972 socket in the ${readsocket expansion. There now is.
974 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
975 be meaningful with "accept".
977 SC/01 Eximstats V1.43
978 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
980 SC/02 Eximstats V1.44
981 Use a glob alias rather than an array ref in the generated
982 parser. This improves both readability and performance.
984 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
985 Collect SpamAssassin and rejection statistics.
986 Don't display local sender or destination tables unless
987 there is data to show.
988 Added average volumes into the top table text output.
990 SC/04 Eximstats V1.46
991 Collect data on the number of addresses (recipients)
992 as well as the number of messages.
994 SC/05 Eximstats V1.47
995 Added 'Message too big' to the list of mail rejection
996 reasons (thanks to Marco Gaiarin).
998 SC/06 Eximstats V1.48
999 Mainlog lines which have GMT offsets and are too short to
1000 have a flag are now skipped.
1002 SC/07 Eximstats V1.49 (Alain Williams)
1003 Added the -emptyok flag.
1005 SC/08 Eximstats V1.50
1006 Fixes for obtaining the IP address from reject messages.
1008 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1009 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1010 whitesspace changes from 4.64-PH/27
1012 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1015 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1016 are found, allow negative numbers in numeric criteria)
1018 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1020 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1021 to match changes made in 4.64-PH/43
1023 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1025 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1026 "domain" is actually an IP address, and force a failure. This locks out
1027 those revolvers/nameservers that support "A-for-A" lookups, in
1028 contravention of the specifications.
1030 PH/48 When a host name was looked up from an IP address, and the subsequent
1031 forward lookup of the name timed out, the host name was left in
1032 $sender_host_name, contrary to the specification.
1034 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1035 restricted to single-key lookups, Exim was not diagnosing an error if
1036 * or *@ was used with a query-style lookup.
1038 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1040 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1041 long ago, but noone interested enough thought of it. Let's just say that
1042 the "1.1" means that there are some new functions that weren't there at
1043 some point in the past.
1045 PH/51 Error processing for expansion failure of helo_data from an smtp
1046 transport during callout processing was broken.
1048 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1049 tested/used via the -bh/-bhc/-bs options.
1051 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1052 bug, fixed in subsequent PCRE releases).
1054 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1055 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1057 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1063 SC/01 Use a glob alias rather than an array ref in eximstats generated
1064 parser. This improves both readability and performance.
1066 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1067 Don't display local sender or destination tables in eximstats unless
1068 there is data to show.
1069 Added average volumes into the eximstats top table text output.
1071 SC/03 Collect data on the number of addresses (recipients) as well
1072 as the number of messages in eximstats.
1074 TF/01 Correct an error in the documentation for the redirect router. Exim
1075 does (usually) call initgroups() when daemonizing.
1077 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1078 with consistent privilege compared to when running as a daemon.
1080 TF/03 Note in the spec that $authenticated_id is not set for local
1081 submissions from trusted users.
1083 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1084 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1086 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1087 by adding some example configuration directives to the default
1088 configuration file. A little bit of work is required to uncomment the
1089 directives and define how usernames and passwords are checked, but
1090 there is now a framework to start from.
1092 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1093 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1094 without this. I don't know how relevant this is to other LDAP libraries.
1096 PH/02 Add the verb name to the "unknown ACL verb" error.
1098 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1100 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1102 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1103 directory not expanded when it should be if an expanded home directory
1104 was set for the address (which is overridden by the transport).
1106 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1109 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1110 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1111 because it is too late at that time, and has no effect.
1113 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1114 security issue with \' (bugzilla #107). I could not use the
1115 PQescapeStringConn() function, because it needs a PGconn value as one of
1118 PH/08 When testing addresses using -bt, indicate those final addresses that
1119 are duplicates that would not cause an additional delivery. At least one
1120 person was confused, thinking that -bt output corresponded to deliveries.
1121 (Suppressing duplicates isn't a good idea as you lose the information
1122 about possibly different redirections that led to the duplicates.)
1124 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1125 systems where poll() doesn't work, in particular OS X.
1127 PH/10 Added more information to debugging output for retry time not reached.
1129 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1130 operations in malware.c.
1132 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1135 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1136 syslog_duplication set false, log lines that would normally be written
1137 both the the main log and to the reject log were not written to syslog at
1140 PH/14 In the default configuration, change the use of "message" in ACL warn
1141 statements to "add_header".
1143 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1144 not followed by a command (e.g. "seen endif").
1146 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1147 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1150 PH/17 Added extra conditions to the default value of delay_warning_condition
1154 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1155 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1156 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1159 The Auto-Submitted: and various List- headers are standardised, whereas I
1160 don't think Precedence: ever was.
1162 PH/18 Refactored debugging code in route_finduser() to show more information,
1163 in particular, the error code if getpwnam() issues one.
1165 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1166 This is apparently needed in addition to the PH/07 change above to avoid
1167 any possible encoding problems.
1169 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1170 but not after initializing Perl.
1172 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1173 output them only if debugging. By default they are written stderr,
1174 apparently, which is not desirable.
1176 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1179 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1182 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1184 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1185 authenticated or an ident call has been made. Suppress the default
1186 values for $authenticated_id and $authenticated_sender (but permit -oMai
1187 and -oMas) when testing with -bh.
1189 PH/24 Re-jigged the order of the tests in the default configuration so that the
1190 tests for valid domains and recipients precede the DNS black list and CSA
1191 tests, on the grounds that those ones are more expensive.
1193 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1194 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1195 This bug exists in every version of Exim that I still have, right back to
1198 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1199 However, an attempt to turn on case-sensitivity in a regex key by
1200 including (?-i) didn't work because the subject string was already
1201 lowercased, and the effects were non-intuitive. It turns out that a
1202 one-line patch can be used to allow (?-i) to work as expected.
1208 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1209 other effects) broke the use of negated acl sub-conditions.
1211 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1214 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1215 "Deny" causes Exim to reject the incoming connection with a 554 error.
1216 Unfortunately, if there is a major crisis, such as a disk failure,
1217 tcp-wrappers gives "deny", whereas what one would like would be some
1218 kind of temporary error. A kludge has been added to help with this.
1219 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1220 554 error is used if errno is still zero or contains ENOENT (which occurs
1221 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1224 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1226 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1227 errors. Otherwise a message that provokes a temporary error (when other
1228 messages do not) can cause a whole host to time out.
1230 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1231 addresses were routed directly to files or pipes from a redirect router.
1232 File deliveries just didn't batch; pipe deliveries might have suffered
1235 PH/06 A failure to get a lock for a hints database would erroneously always say
1236 "Failed to get write lock", even when it was really a read lock.
1238 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1239 of 0600. This has been changed to use the value of the lockfile_mode
1240 option (which defaults to 0600).
1242 PH/08 Applied small patch from the Sieve maintainer.
1244 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1245 folder from quota calculations, a direct delivery into this folder messed
1246 up the contents of the maildirsize file. This was because the regex was
1247 used only to exclude .Trash (or whatever) when the size of the mailbox
1248 was calculated. There was no check that a delivery was happening into an
1249 excluded directory. This bug has been fixed by ignoring all quota
1250 processing for deliveries into excluded directories.
1252 PH/10 Added the maildirfolder_create_regex option to appendfile.
1258 PH/01 The code for finding all the local interface addresses on a FreeBSD
1259 system running IPv6 was broken. This may well have applied to all BSD
1260 systems, as well as to others that have similar system calls. The broken
1261 code found IPv4 interfaces correctly, but gave incorrect values for the
1262 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1263 that it would not match correctly against @[] and not recognize the IPv6
1266 PH/02 The ipliteral router was not recognizing addresses of the form user@
1267 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1269 PH/03 Added disable_ipv6.
1271 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1272 decoded form, because it is most often used to construct To: headers
1273 lines in autoreplies, and the decoded form may well be syntactically
1274 invalid. However, $reply_address has leading white space removed, and all
1275 newlines turned into spaces so that the autoreply transport does not
1278 PH/05 If group was specified without a user on a router, and no group or user
1279 was specified on a transport, the group from the router was ignored.
1281 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1282 for visible compile-time settings that can be used to change these
1283 numbers, for those that want even more. Backwards compatibility with old
1284 spool files has been maintained. However, going back to a previous Exim
1285 release will lost any variables that are in spool files.
1287 PH/07 Two small changes when running in the test harness: increase delay when
1288 passing a TCP/IP connection to a new process, in case the original
1289 process has to generate a bounce, and remove special handling of
1290 127.0.0.2 (sic), which is no longer necessary.
1292 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1293 be the same on different OS.
1295 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1298 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1299 whether --show-vars was specified or not
1301 JJ/02 exipick: Added support for new ACL variable spool format introduced
1304 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1305 syntactically invalid From: or Reply-to: line, and a filter used this to
1306 generate an autoreply, and therefore failed to obtain an address for the
1307 autoreply, Exim could try to deliver to a non-existent relative file
1308 name, causing unrelated and misleading errors. What now happens is that
1309 it logs this as a hard delivery error, but does not attempt to create a
1312 PH/11 The exinext utility has a -C option for testing purposes, but although
1313 the given file was scanned by exinext itself; it wasn't being passed on
1314 when Exim was called.
1316 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1317 an end-of-file indication when reading a command response.
1319 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1320 compiled. In many other places in Exim, IPv6 addresses are always
1321 recognized, so I have changed this. It also means that IPv4 domain
1322 literals of the form [IPV4:n.n.n.n] are now always recognized.
1324 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1325 used if the router is not running as root, for example, when verifying at
1326 ACL time, or when using -bh. The debugging output from this situation was
1327 non-existent - all you got was a failure to exec. I have made two
1330 (a) Failures to set uid/gid, the current directory, or a process leader
1331 in a subprocess such as that created by queryprogram now generate
1332 suitable debugging ouput when -d is set.
1334 (b) The queryprogram router detects when it is not running as root,
1335 outputs suitable debugging information if -d is set, and then runs
1336 the subprocess without attempting to change uid/gid.
1338 PH/15 Minor change to Makefile for building test_host (undocumented testing
1341 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1342 additional section of a DNS packet that returns MX or SRV records.
1343 Instead, it always explicitly searches for A/AAAA records. This avoids
1344 major problems that occur when a DNS server includes only records of one
1345 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1346 fixed another bug: if SRV records were looked up and the corresponding
1347 address records were *not* found in the additional section, the port
1348 values from the SRV records were lost.
1350 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1351 using the correct key (the original address) when searching the retry
1352 rules in order to find which one to use for generating the retry hint.
1354 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1355 adding the default one. Similarly, if it contains a Reply-To: header, the
1356 errors_reply_to option, if set, is not used.
1358 PH/19 When calculating a retry time, Exim used to measure the "time since
1359 failure" by looking at the "first failed" field in the retry record. Now
1360 it does not use this if it is later than than the arrival time of the
1361 message. Instead it uses the arrival time. This makes for better
1362 behaviour in cases where some deliveries succeed, thus re-setting the
1363 "first failed" field. An example is a quota failure for a huge message
1364 when small messages continue to be delivered. Without this change, the
1365 "time since failure" will always be short, possible causing more frequent
1366 delivery attempts for the huge message than are intended.
1367 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1369 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1370 $1, $2, $3) because the numerical variables can be reset during some
1371 expansion items (e.g. "match"), thereby losing the authentication data.
1373 PH/21 Make -bV show the size of off_t variables so that the test suite can
1374 decide whether to run tests for quotas > 2G.
1376 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1377 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1378 filecount value is greater than 2G or if a quota value is greater than 2G
1379 on a system where the size of off_t is not greater than 4, a panic error
1382 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1383 never match. The debug and -bh output now contains an explicit error
1384 message indicating a malformed IPv4 address or mask.
1386 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1387 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1388 PH/23 above applies.
1390 PH/25 Do not write to syslog when running in the test harness. The only
1391 occasion when this arises is a failure to open the main or panic logs
1392 (for which there is an explicit test).
1394 PH/26 Added the /no_tell option to "control=freeze".
1396 PH/27 If a host name lookup failed very early in a connection, for example, if
1397 the IP address matched host_lookup and the reverse lookup yielded a name
1398 that did not have a forward lookup, an error message of the form "no IP
1399 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1400 could be logged. Now it outputs the IP address instead of "NULL".
1402 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1403 allows the sender and the authenticated sender to be set when
1404 submitting a message from within Exim. Since child_open_exim() is
1405 documented for local_scan(), the new function should be too.
1407 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1408 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1409 results in an empty string is now treated as unset.
1411 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1413 PH/31 Added sender_verify_fail logging option.
1415 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1416 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1417 tidied the source and removed it altogether.
1419 PH/33 When a queue run was abandoned because the load average was too high, a
1420 log line was always written; now it is written only if the queue_run log
1421 selector is set. In addition, the log line for abandonment now contains
1422 information about the queue run such as the pid. This is always present
1423 in "start" and "stop" lines but was omitted from the "abandon" line.
1425 PH/34 Omit spaces between a header name and the colon in the error message that
1426 is given when verify = headers_syntax fails (if there are lots of them,
1427 the message gets confusing).
1429 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1430 names, as there are now some PTR records that contain slashes. This check
1431 is only to protect against broken name servers that fall over on strange
1432 characters, so the fact that it applies to all lookups doesn't matter.
1434 PH/36 Now that the new test suite is complete, we can remove some of the
1435 special code in Exim that was needed for the old test suite. For example,
1436 sorting DNS records because real resolvers return them in an arbitrary
1437 order. The new test suite's fake resolver always returns records in the
1440 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1441 bounces) except when queue_only is set, to avoid logging races between
1442 the different processes.
1444 PH/38 Panic-die if .include specifies a non-absolute path.
1446 PH/39 A tweak to the "H" retry rule from its user.
1448 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1449 a label. They prevented compilation on older perls.
1451 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1452 a warning to be raised on newish perls.
1454 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1455 on queue. Changes to match documented behaviour of showing count of
1456 messages matching specified criteria.
1458 PH/40 Changed the default ident timeout from 30s to 5s.
1460 PH/41 Added support for the use of login_cap features, on those BSD systems
1461 that have them, for controlling the resources used by pipe deliveries.
1463 PH/42 The content-scanning code uses fopen() to create files in which to put
1464 message data. Previously it was not paying any attention to the mode of
1465 the files. Exim runs with umask(0) because the rest of the code creates
1466 files with open(), and sets the required mode explicitly. Thus, these
1467 files were ending up world-writeable. This was not a big issue, because,
1468 being within the spool directory, they were not world-accessible. I have
1469 created a function called modefopen, which takes an additional mode
1470 argument. It sets umask(777), creates the file, chmods it to the required
1471 mode, then resets the umask. All the relevant calls to fopen() in the
1472 content scanning code have been changed to use this function.
1474 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1475 to 24 hours. This avoids potential overflow problems when processing G
1476 and H retry rules. I suspect nobody ever tinkers with this value.
1478 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1480 PH/45 When the plaintext authenticator is running as a client, the server's
1481 challenges are checked to ensure they are valid base64 strings. By
1482 default, the authentication attempt is cancelled if an invalid string is
1483 received. Setting client_ignore_invalid_base64 true ignores these errors.
1484 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1485 they are received. Thus, the responses can be made to depend on the
1486 challenges. If an invalid string is ignored, an empty string is placed in
1489 PH/46 Messages that are created by the autoreply transport now contains a
1490 References: header, in accordance with RFCs 2822 and 3834.
1492 PH/47 Added authenticated_sender_force to the smtp transport.
1494 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1496 PH/49 Installed latest patch from the Sieve maintainer.
1498 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1499 was also set, the appendfile transport was unnecessarily scanning a
1500 directory of message files (e.g. for maildir delivery) to find the count
1501 of files (along with the size), even though it did not need this
1502 information. It now does the scan only if it needs to find either the
1503 size of the count of files.
1505 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1507 PH/52 Two bugs concerned with error handling when the smtp transport is
1510 (i) Exim was not creating retry information for temporary errors given
1511 for individual recipients after the DATA command when the smtp transport
1512 was used in LMTP mode. This meant that they could be retried too
1513 frequently, and not timed out correctly.
1515 (ii) Exim was setting the flag that allows error details to be returned
1516 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1517 recipients that were returned after the DATA command.
1519 PH/53 This is related to PH/52, but is more general: for any failing address,
1520 when detailed error information was permitted to be returned to the
1521 sender, but the error was temporary, then after the final timeout, only
1522 "retry timeout exceeded" was returned. Now it returns the full error as
1523 well as "retry timeout exceeded".
1525 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1526 do this, and (what is worse) MTAs that accept it.
1528 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1529 will now be deprecated.
1531 PH/56 New os.c-cygwin from the Cygwin maintainer.
1533 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1534 formats (previously only available in exim formats via -bpr, -bpru,
1535 and -bpra. Now also available in native and exiqgrep formats)
1537 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1538 with very large, slow to parse queues
1540 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1542 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1544 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1545 responses to authentication challenges, though it was showing the
1546 challenges; (ii) I've removed the CR characters from the debug output for
1549 PH/58 Allow for the insertion of a newline as well as a space when a string
1550 is turned into more than one encoded-word during RFC 2047 encoding. The
1551 Sieve code now uses this.
1553 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1554 data_4xx, lost_connection, tls_required.
1556 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1557 message was being sent as an SMTP response.
1559 PH/61 Add -l and -k options to exicyclog.
1561 PH/62 When verifying, if an address was redirected to one new address, so that
1562 verification continued, and the new address failed or deferred after
1563 having set something in $address_data, the value of $address_data was not
1564 passed back to the ACL. This was different to the case when no
1565 redirection occurred. The value is now passed back in both cases.
1567 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1568 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1569 use login_cap.h, so on its own it isn't the distinguishing feature. The
1570 new name refers directly to the setclassresources() function.
1572 PH/65 Added configuration files for NetBSD3.
1574 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1576 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1577 is preferred over IPv4.
1579 PH/68 The bounce_return_message and bounce_return_body options were not being
1580 honoured for bounces generated during the reception of non-SMTP messages.
1581 In particular, this applied to messages rejected by the ACL. This bug has
1582 been fixed. However, if bounce_return_message is true and bounce_return_
1583 body is false, the headers that are returned for a non-SMTP message
1584 include only those that have been read before the error was detected.
1585 (In the case of an ACL rejection, they have all been read.)
1587 PH/69 The HTML version of the specification is now built in a directory called
1588 spec_html instead of spec.html, because the latter looks like a path with
1589 a MIME-type, and this confuses some software.
1591 PH/70 Catch two compiler warnings in sieve.c.
1593 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1594 function verify_get_ident() calls ip_connect() to connect a socket, but
1595 if the "connect()" function timed out, ip_connect() used to close the
1596 socket. However, verify_get_ident() also closes the socket later, and in
1597 between Exim writes to the log, which may get opened at this point. When
1598 the socket was closed in ip_connect(), the log could get the same file
1599 descriptor number as the socket. This naturally causes chaos. The fix is
1600 not to close the socket in ip_connect(); the socket should be closed by
1601 the function that creates it. There was only one place in the code where
1602 this was missing, in the iplookup router, which I don't think anybody now
1603 uses, but I've fixed it anyway.
1605 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1606 well as to direct DNS lookups. Otherwise the handling of names in host
1607 lists is inconsistent and therefore confusing.
1613 PH/01 Two changes to the default runtime configuration:
1615 (1) Move the checks for relay_from_hosts and authenticated clients from
1616 after to before the (commented out) DNS black list checks.
1618 (2) Add control=submission to the relay_from_hosts and authenticated
1619 clients checks, on the grounds that messages accepted by these
1620 statements are most likely to be submissions.
1622 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1624 (1) Generate an error if the third argument for the ${prvs expansion is
1627 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1630 (3) Reset the variables that are obtained from the first argument of
1631 ${prvscheck and used in the second argument before leaving the code,
1632 because their memory is reclaimed, so using them afterwards may do
1635 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1636 one (it's much easier than Tom thought :-).
1638 (5) Because of (4), we can now allow for the use of $prvscheck_result
1639 inside the third argument.
1641 PH/03 For some reason, the default setting of PATH when running a command from
1642 a pipe transport was just "/usr/bin". I have changed it to
1645 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1646 anything to be listed in the output from -bV.
1648 PH/05 When a filter generated an autoreply, the entire To: header line was
1649 quoted in the delivery log line, like this:
1651 => >A.N.Other <ano@some.domain> <original@ddress> ...
1653 This has been changed so that it extracts the operative address. There
1654 may be more than one such address. If so, they are comma-separated, like
1657 => >ano@some.domain,ona@other.domain <original@ddress> ...
1659 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1660 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1661 not being looked up in the rDNS to get a host name, Exim was showing the
1662 IP address twice in Received: lines, even though the IP addresses were
1663 identical. For example:
1665 Received: from [1.2.3.4] (helo=[1.2.3.4])
1667 However, if the real host name was known, it was omitting the HELO data
1668 if it matched the actual IP address. This has been tidied up so that it
1669 doesn't show the same IP address twice.
1671 PH/07 When both +timestamp and +memory debugging was on, the value given by
1672 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1673 called by the debug printing, thereby overwriting the timestamp buffer.
1674 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1676 PH/08 When the original message was included in an autoreply transport, it
1677 always said "this is a copy of the message, including all the headers",
1678 even if body_only or headers_only was set. It now gives an appropriate
1681 PH/09 Applied a patch from the Sieve maintainer which:
1683 o fixes some comments
1684 o adds the (disabled) notify extension core
1685 o adds some debug output for the result of if/elsif tests
1686 o points to the current vacation draft in the documentation
1687 and documents the missing references header update
1691 o fixes a bug in processing the envelope test (when testing
1692 multiple envelope elements, the last element determinted the
1695 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1696 Electronic Mail") by including:
1698 Auto-submitted: auto-generated
1700 in the messages that it generates (bounce messages and others, such as
1701 warnings). In the case of bounce messages for non-SMTP mesages, there was
1702 also a typo: it was using "Auto_submitted" (underscore instead of
1703 hyphen). Since every message generated by Exim is necessarily in response
1704 to another message, thes have all been changed to:
1706 Auto-Submitted: auto-replied
1708 in accordance with these statements in the RFC:
1710 The auto-replied keyword:
1712 - SHOULD be used on messages sent in direct response to another
1713 message by an automatic process,
1715 - MUST NOT be used on manually-generated messages,
1717 - MAY be used on Delivery Status Notifications (DSNs) and Message
1718 Disposition Notifications (MDNs),
1720 - MUST NOT be used on messages generated by automatic or periodic
1721 processes, except for messages which are automatic responses to
1724 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1725 to the default Received: header definition.
1727 PH/12 Added log selector acl_warn_skipped (default on).
1729 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1730 variables because (a) they are in the wrong storage pool and (b) even if
1731 they were copied, it wouldn't work properly because of the caching.
1733 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1734 checking when decoding. Apparently there are clients that generate
1735 overlong encoded strings. Why am I not surprised?
1737 PH/15 If the first argument of "${if match_address" was not empty, but did not
1738 contain an "@" character, Exim crashed. Now it writes a panic log message
1739 and treats the condition as false.
1741 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1743 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1744 extension "envelope-auth". The code is finished and in agreement with
1745 other implementations, but there is no documentation so far and in fact,
1746 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1747 not changing the active code.
1749 Print executed "if" and "elsif" statements when debugging is used. This
1750 helps a great deal to understand what a filter does.
1752 Document more things not specified clearly in RFC3028. I had all this
1753 sorted out, when out of a sudden new issues came to my mind. Oops."
1755 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1758 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1759 canonical form (as documented). However, after a host name lookup from
1760 the IP address, check_host() was doing a simple string comparison with
1761 addresses acquired from the DNS when checking that the found name did
1762 have the original IP as one of its addresses. Since any found IPv6
1763 addresses are likely to be in abbreviated form, the comparison could
1764 fail. Luckily, there already exists a function for doing the comparison
1765 by converting both addresses to binary, so now that is used instead of
1766 the text comparison.
1768 PH/20 There was another similar case to PH/19, when a complete host name was
1769 given in a host list; looking up its IP address could give an abbreviated
1770 form, whereas the current host's name might or might not be abbreviated.
1771 The same fix has been applied.
1777 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1778 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1781 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1782 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1784 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1786 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1787 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1788 Also, for "Auto-Submitted", treat anything other than "no" as
1789 non-personal, in accordance with RFC 3834. (Previously it treated
1790 anything starting "auto-" as non-personal.)
1792 TF/01 The control=submission/name=... option had a problem with syntax
1793 errors if the name included a slash character. The /name= option
1794 now slurps the rest of the string, so it can include any characters
1795 but it must come last in the list of options (after /sender_retain
1798 PH/05 Some modifications to the interface to the fake nameserver for the new
1806 TK/01 Added the "success_on_redirect" address verification option. See
1807 NewStuff for rationale and an example.
1809 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1811 PH/02 Patch to exigrep to allow it to work on syslog lines.
1813 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1814 fread() to skip over the body file's header line, because in Cygwin the
1815 header line is locked and is inaccessible.
1817 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1818 co-exist for some time) to make it clear that it is the Exim ID that is
1819 referenced, not the Message-ID: header line.
1821 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1822 string_format() function, because snprintf() does not exist on all
1825 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1826 ${stat: expansion item.
1828 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1829 protocol synchronization error", to keep the pedants happy.
1831 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1832 well as for IRIX systems, when gcc is being used. See the host.c source
1835 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1837 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1840 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1841 transport and to the smtp transport in LMTP mode.
1843 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1845 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1846 run a filter in a subprocess. This could lead to confusion in subsequent
1847 lookups in the parent process. There should also be a search_tidyup() at
1848 the end of the subprocess.
1850 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1851 only if the host matched helo_try_verify_hosts, which caused the
1852 verification to occur when the EHLO/HELO command was issued. The ACL just
1853 tested the remembered result. Now, if a previous verification attempt has
1854 not happened, "verify = helo" does it there and then.
1856 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1858 TK/03 Fix log output including CR from clamd.
1860 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1861 exist provoked a memory error which could cause a segfault.
1863 PH/15 Installed PCRE 6.2
1865 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1867 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1868 of the problem. Specifically, suggested +O2 rather than +O1 for the
1871 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1873 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1874 record, the port was not correctly logged when the outgoing_port log
1875 selector was set (it logged the transort's default port).
1877 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1878 fallback_hosts, and "hosts" in the smtp transport.
1880 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1881 host errors such as "Connection refused".
1883 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1884 authentication with radiusclient 0.4.9:
1886 - Error returned from rc_read_config was caught wrongly
1887 - Username/password not passed on to radius server due to wrong length.
1889 The presumption is that some radiusclient API changes for 4.51/PH/17
1890 were not taken care of correctly. The code is still untested by me (my
1891 Linux distribution still has 0.3.2 of radiusclient), but it was
1892 contributed by a Radius user.
1894 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1895 expanding the "port" option of the smtp transport.
1897 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1898 while reading a MIME header. Thanks to Tom Hughes for a patch.
1900 PH/24 Include config.h inside local_scan.h so that configuration settings are
1903 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1904 that in an ACL for RCPT (for example), you can examine exactly what was
1907 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1908 commands, but it was not correctly comparing the address with the actual
1909 client host address. Thus, it would show the EHLO address in Received:
1910 header lines when this was not necessary.
1912 PH/27 Added the % operator to ${eval:}.
1914 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1915 it should be ignoring failures (because with -C, for example, it has lost
1916 privilege). It wasn't ignoring creation failures other than "already
1919 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1922 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1923 because an input line was too long, either on its own, or by virtue of
1924 too many continuations, the temporary file was not being removed, and the
1925 return code was incorrect.
1927 PH/31 Missing "BOOL" in function definition in filtertest.c.
1929 PH/32 Applied Sieve patches from the maintainer.
1931 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1933 PH/33 Added "verify = not_blind".
1935 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1936 Local/Makefile (with some defaults set). These are used in built scripts
1937 such as exicyclog, but they have never been used in the exim_install
1938 script (though there are many overriding facilities there). I have
1939 arranged that the exim_install script now takes note of these two
1942 PH/35 Installed configuration files for Dragonfly.
1944 PH/36 When a locally submitted message by a trusted user did not contain a
1945 From: header, and the sender address was obtained from -f or from an SMTP
1946 MAIL command, and the trusted user did not use -F to supply a sender
1947 name, $originator_name was incorrectly used when constructing a From:
1948 header. Furthermore, $originator_name was used for submission mode
1949 messages from external hosts without From: headers in a similar way,
1950 which is clearly wrong.
1952 PH/37 Added control=suppress_local_fixups.
1954 PH/38 When log_selector = +received_sender was set, and the addition of the
1955 sender made the log line's construction buffer exactly full, or one byte
1956 less than full, an overflow happened when the terminating "\n" was
1959 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1960 when the result of a list match is failure because a DNS lookup failed.
1962 PH/40 RM_COMMAND is now used in the building process.
1964 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1965 the "build-* directories that it finds.
1967 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1968 address in a domain literal was a prefix of an interface address.
1970 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1971 when verifying a sender address, unless rewrite_headers is false.
1973 PH/44 Wrote a long comment about why errors_to addresses are verified as
1974 recipients, not senders.
1976 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1977 the ratelimit ACL was added.
1979 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1981 PH/46 Added extra information about PostgreSQL errors to the error string.
1983 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1984 avoiding the need to install special zones in a real server. This is
1985 backwards compatible; if it can't find the fake resolver, it drops back.
1986 Thus, both old and new test suites can be run.
1988 TF/02 Added util/ratelimit.pl
1990 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1991 clock is set back in time.
1993 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1994 Candler <B.Candler@pobox.com>.
1996 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1997 OK for addresses that are the result of redirections.
1999 PH/48 A number of further additions for the benefit of the new test suite,
2000 including a fake gethostbyname() that interfaces to the fake DNS resolver
2003 TF/06 The fix for widen_domains has also been applied to qualify_single and
2004 search_parents which are the other dnslookup options that can cause
2007 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2010 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2012 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2013 service" if it encountered an MX record with an empty target hostname.
2014 The message is now "an MX or SRV record indicated no SMTP service".
2016 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2017 if the DNS of the sending site is misconfigured. This is quite a
2018 common situation. This change restores the behaviour of treating a
2019 helo verification defer as a failure.
2021 PH/51 If self=fail was set on a router, the bounce message did not include the
2022 actual error message.
2028 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2030 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2031 command itself ended in error, the underlying message about the transport
2032 filter timeout was being overwritten with the pipe command error. Now the
2033 underlying error message should be appended to the second error message.
2035 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2037 PH/02 Reduce the amount of output that "make" produces by default. Full output
2038 can still be requested.
2040 PH/03 The warning log line about a condition test deferring for a "warn" verb
2041 was being output only once per connection, rather than after each
2042 occurrence (because it was using the same function as for successful
2043 "warn" verbs). This seems wrong, so I have changed it.
2045 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2046 it should not have, which might have caused a crash in the right
2047 circumstances, but probably never did.
2049 PH/04 Installed a modified version of Tony Finch's patch to make submission
2050 mode fix the return path as well as the Sender: header line, and to
2051 add a /name= option so that you can make the user's friendly name appear
2054 TF/03 Added the control = fakedefer ACL modifier.
2056 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2057 Mark Lowes for thorough testing.
2059 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2061 TK/03 Merged latest SRS patch from Miles Wilton.
2063 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2064 with the definition in sysexits.h (which is #included earlier).
2065 Fortunately, Exim does not actually use EX_OK. The code used to try to
2066 preserve the sysexits.h value, by assumimg that macro definitions were
2067 scanned for macro replacements. I have been disabused of this notion,
2068 so now the code just undefines EX_OK before #including unistd.h.
2070 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2071 in the smtp transport. When a block could not be written in a single
2072 write() function, the timeout was being re-applied to each part-write.
2073 This seems wrong - if the receiver was accepting one byte at a time it
2074 would take for ever. The timeout is now adjusted when this happens. It
2075 doesn't have to be particularly precise.
2077 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2078 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2080 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2081 without a domain if the check to <postmaster@domain> fails.
2083 SC/01 Eximstats: added -xls and the ability to specify output files
2084 (patch written by Frank Heydlauf).
2086 SC/02 Eximstats: use FileHandles for outputing results.
2088 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2090 SC/04 Eximstats: fixed display of large numbers with -nvr option
2092 SC/05 Eximstats: fixed merging of reports with empty tables.
2094 SC/06 Eximstats: added the -include_original_destination flag
2096 SC/07 Eximstats: removed tabs and trailing whitespace.
2098 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2100 TK/06 MBOX spool code: Add real "From " MBOX separator line
2101 so the .eml file is really in mbox format (even though
2102 most programs do not really care). Patch from Alex Miller.
2104 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2105 The latter is generated from $received_to and is only set if the
2106 message has one envelope recipient. SA can use these headers,
2107 obviously out-of-the-box. Patch from Alex Miller.
2109 PH/08 The ${def test on a variable was returning false if the variable's
2110 value was "0", contrary to what the specification has always said!
2111 The result should be true unless the variable is empty.
2113 PH/09 The syntax error of a character other than { following "${if
2114 def:variable_name" (after optional whitespace) was not being diagnosed.
2115 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2116 accidental colon was present, for example, could give incorrect results.
2118 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2119 result is used (not including appendfile, where other changes are about
2122 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2123 This involved changing a lot of size variables from int to off_t. It
2124 should work with maildirs and everything.
2126 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2127 spamd dying while we are connected to it.
2129 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2132 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2133 function for BDB 4.3.
2135 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2137 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2138 the sources for PCRE have been re-arranged and more files are now
2141 PH/15 The code I had for printing potentially long long variables in PH/11
2142 above was not the best (it lost precision). The length of off_t variables
2143 is now inspected at build time, and an appropriate printing format (%ld
2144 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2145 to be "long long int" or "long int". This is needed for the internal
2146 formatting function string_vformat().
2148 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2149 the configuration file to be ":syslog", then the script "guesses" where
2150 the logs files are, rather than using the compiled in default. In our
2151 case the guess is not the same as the compiled default, so the script
2152 suddenly stopped working when I started to use syslog. The patch checks
2153 to see if log_file_path is "". If so, it attempts to read it from exim
2154 with no configuration file to get the compiled in version, before it
2155 falls back to the previous guessing code."
2157 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2158 implementing BATV in an Exim configuration. See NewStuff for the gory
2161 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2162 Makefile that are specific to HP-UX.
2164 PH/18 If the "use_postmaster" option was set for a recipient callout together
2165 with the "random" option, the postmaster address was used as the MAIL
2166 FROM address for the random test, but not for the subsequent recipient
2167 test. It is now used for both.
2169 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2170 patch removes a few documentation additions to RFC 3028, because the
2171 latest draft now contains them. It adds the new en;ascii-case comparator
2172 and a new error check for 8bit text in MIME parts. Comparator and
2173 require names are now matched exactly. I enabled the subaddress
2174 extension, but it is not well tested yet (read: it works for me)."
2176 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2177 rework some of the code of TK/09 above to avoid the hardwired use of
2178 "%lld" and "long long". Replaced the call to snprintf() with a call to
2181 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2182 records point to non-existent hosts", "retry timeout exceeded", and
2183 "retry time not reached for any host after a long failure period".
2185 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2186 experimental DomainKeys support:
2188 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2189 (2) On an error such as an illegally used "control", the wrong name for
2190 the control was given.
2192 These problems did NOT occur unless DomainKeys support was compiled.
2194 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2196 PH/24 Added ${if match_ip condition.
2198 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2199 databases so that it will be absolutely obvious if a crash occurs in the
2200 DB library. This is a regular occurrence (often caused by mis-matched
2203 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2204 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2205 picked up on a user's system that detects such things. There doesn't seem
2206 to be a gcc warning option for this - only an attribute that has to be
2207 put on the function's prototype. It seems that in Fedora Core 4 they have
2208 set this on a number of new functions. No doubt there will be more in due
2211 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2212 specify a transport. However, if an address that was verified by such a
2213 router was the subject of a callout, Exim crashed because it tried to
2214 read the rcpt_include_affixes from the non-existent transport. Now it
2215 just assumes that the setting of that option is false. This bug was
2216 introduced by 4.51/PH/31.
2218 PH/28 Changed -d+all to exclude +memory, because that information is very
2219 rarely of interest, but it makes the output a lot bigger. People tend to
2220 do -d+all out of habit.
2222 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2223 code in os-type was giving problems when libc.so lives in lib64, like on
2226 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2227 aren't the modern standard, and it seems that some systems' include files
2228 don't always have them. Exim was already checking for some of the newer
2229 ones like T_AAAA, and defining it itself. I've added checks for all the
2230 record types that Exim uses.
2232 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2233 not automatically generating a new one, as it is supposed to. This
2234 prevented TLS from working. If the file did exist, but contained invalid
2235 data, a new version was generated, as expected. It was only the case of a
2236 non-existent file that was broken.
2238 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2239 with a change in libdomainkeys > 0.64.
2241 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2242 from DNS. If the selector record carries the flag, it now has
2243 precedence over the domain-wide flag.
2245 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2247 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2248 the use of a port name that isn't defined in /etc/services) occurred, the
2249 message was deferred as in a normal delivery, and thus remained on the
2250 spool, instead of being failed because of the mua_wrapper setting. This
2251 is now fixed, and I tidied up some of the mua_wrapper messages at the
2254 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2255 the messages in a hash of arrays rather than using individual hashes.
2256 This is a bit cleaner and results in dramatic memory savings, albeit
2257 at a slight CPU cost.
2259 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2260 as requested by Marc Sherman.
2262 SC/10 Eximstats: added histograms for user specified patterns as requested
2265 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2267 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2268 fopen() in the content-scanning modules that did not already have it.
2274 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2275 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2277 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2279 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2281 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2282 to test Sieve filters that use "vacation".
2284 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2285 that changes the way the GnuTLS parameters are stored in the cache file.
2286 The new format can be generated externally. For backward compatibility,
2287 if the data in the cache doesn't make sense, Exim assumes it has read an
2288 old-format file, and it generates new data and writes a new file. This
2289 means that you can't go back to an older release without removing the
2292 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2293 work if there are any delivery delays because "one_time" forces the
2294 parent to be marked "delivered", so its unseen clone is never tried
2295 again. For this reason, Exim now forbids the simultaneous setting of
2298 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2299 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2300 entry if you want to know the details. The fix, however, neglected to
2301 consider the case when local delivery batching is involved. The test for
2302 "previously delivered" was not happening when checking to see if an
2303 address could be batched with a previous (undelivered) one; under
2304 certain circumstances this could lead to multiple deliveries to the same
2307 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2308 in its include files, and this causes problems building Exim.
2310 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2311 header_syntax) but Exim was just ignoring anything given after a slash.
2312 In particular, this caused confusion with an attempt to use "verify =
2313 reverse_host_lookup/defer_ok". An error is now given when options are
2314 supplied for verify items that do not have them. (Maybe reverse_host_
2315 lookup should have a defer_ok option, but that's a different point.)
2317 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2318 defined by RFC 821) to 2048, because there were problems with some AUTH
2319 commands, and RFC 1869 says the size should be increased for extended
2320 SMTP commands that take arguments.
2322 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2325 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2326 "unknown" error; now it says that the functionality isn't in the binary.
2328 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2329 an address' error message when a string expansion fails (syntax or
2330 whatever). Otherwise the password may appear in the log. Following change
2331 PH/42 below, there is no longer a chance of it appearing in a bounce
2334 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2336 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2337 of its addresses was ever tried. (Bugzilla bug #2.)
2339 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2340 the result incorrectly in the debug output. (It correctly added a newline
2341 to what was transported.)
2343 TF/01 Added $received_time.
2345 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2346 commented out examples of how to interface to a virus scanner and to
2347 SpamAssassin. Also added commented examples of av_scanner and
2348 spamd_address settings.
2350 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2351 and controls are allowed in which ACLs. There were a couple of minor
2352 errors. Some of the entries in the conditions table (which is a table of
2353 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2354 negation of where the condition IS allowed.
2356 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2358 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2359 header file does not have a version number, so I've had to invent a new
2360 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2361 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2362 radiusclient), but it was contributed by a Radius user.
2364 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2365 files or directories, for OpenSSL.
2367 PH/19 When an Exim process that is running as root has to create an Exim log
2368 file, it does so in a subprocess that runs as exim:exim so as to get the
2369 ownership right at creation (otherwise, other Exim processes might see
2370 the file with the wrong ownership). There was no test for failure of this
2371 fork() call, which would lead to the process getting stuck as it waited
2372 for a non-existent subprocess. Forks do occasionally fail when resources
2373 run out. I reviewed all the other calls to fork(); they all seem to check
2376 PH/20 When checking for unexpected SMTP input at connect time (before writing
2377 the banner), Exim was not dealing correctly with a non-positive return
2378 from the read() function. If the client had disconnected by this time,
2379 the result was a log entry for a synchronization error with an empty
2380 string after "input=" when read() returned zero. If read() returned -1
2381 (an event I could not check), uninitialized data bytes were printed.
2382 There were reports of junk text (parts of files, etc) appearing after
2385 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2387 PH/22 Added support for macro redefinition, and (re)definition in between
2388 driver and ACL definitions.
2390 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2391 forgetting to use the resulting value; it was using the unexpanded value.
2393 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2394 hadn't been configured. The fix is from Juergen Kreileder, who
2395 understands it better than I do:
2397 "Here's what I see happening with three configured cyrus_sasl
2398 authenticators configured (plain, login, cram-md5):
2400 On startup auth_cyrus_sasl_init() gets called for each of these.
2401 This means three calls to sasl_listmech() without a specified mech_list.
2402 => SASL tests which mechs of all available mechs actually work
2403 => three warnings about OTP not working
2404 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2406 With the patch, sasl_listmech() also gets called three times. But now
2407 SASL's mech_list option is set to the server_mech specified in the the
2408 authenticator. Or in other words, the answer from sasl_listmech()
2409 gets limited to just the mech you're testing for (which is different
2411 => the return list contains just 'plain' or 'login', 'cram-md5' or
2412 nothing depending on the value of ob->server_mech.
2414 I've just tested the patch: Authentication still works fine,
2415 unavailable mechs specified in the exim configuration are still
2416 caught, and the auth.log warnings about OTP are gone."
2418 PH/25 When debugging is enabled, the contents of the command line are added
2419 to the debugging output, even when log_selector=+arguments is not
2422 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2423 answer is "GNU", and only if the return is "GNU/something" is the answer
2426 PH/27 $acl_verify_message is now set immediately after the failure of a
2427 verification in an ACL, and so is available in subsequent modifiers. In
2428 particular, the message can be preserved by coding like this:
2430 warn !verify = sender
2431 set acl_m0 = $acl_verify_message
2433 Previously, $acl_verify_message was set only while expanding "message"
2434 and "log_message" when a very denied access.
2436 PH/28 Modified OS/os.c-Linux with
2438 -#ifndef OS_LOAD_AVERAGE
2439 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2441 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2442 nomenclature these days.)
2444 PH/29 Installed patch from the Sieve maintainer that adds the options
2445 sieve_useraddress and sieve_subaddress to the redirect router.
2447 PH/30 In these circumstances:
2448 . Two addresses routed to the same list of hosts;
2449 . First host does not offer TLS;
2450 . First host accepts first address;
2451 . First host gives temporary error to second address;
2452 . Second host offers TLS and a TLS session is established;
2453 . Second host accepts second address.
2454 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2455 and peerdn, if requested) that were in fact used only for the second
2458 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2459 attention to any local part prefix or suffix that was matched by the
2460 router that accepted the address. It now behaves in the same way as it
2461 does for delivery: the affixes are removed from the local part unless
2462 rcpt_include_affixes is set on the transport.
2464 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2465 timeout during the DATA phase of an incoming message.
2467 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2468 applied a patch sanctioned by the Sieve maintainer.
2470 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2471 the uid or gid is negative. A case of a negative gid caused this to be
2472 noticed. The fix allows for either to be negative.
2474 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2475 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2478 PH/36 The change PH/12 above was broken. Fixed it.
2480 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2481 the grounds that routing the same address twice would always produce the
2482 same answer. This might have been true once, but it is certainly no
2483 longer true now. Routing a child address may depend on the previous
2484 routing that produced that child. Some complicated redirection strategies
2485 went wrong when messages had multiple recipients, and made Exim's
2486 behaviour dependent on the order in which the addresses were given.
2488 I have moved the duplicate checking until after the routing is complete.
2489 Exim scans the addresses that are assigned to local and remote
2490 transports, and removes any duplicates. This means that more work will be
2491 done, as duplicates will always all be routed, but duplicates are
2492 presumably rare, so I don't expect this is of any significance.
2494 For deliveries to pipes, files, and autoreplies, the duplicate checking
2495 still happens during the routing process, since they are not going to be
2498 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2499 It corrects a timeout issue with spamd. This is Ian's comment: "The
2500 background is that sometimes spamd either never reads data from a
2501 connection it has accepted, or it never writes response data. The exiscan
2502 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2503 blindly assumes that writes won't block so it may never time out."
2505 PH/39 Allow G after quota size as well as K and M.
2507 PH/40 The value set for $authenticated_id in an authenticator may not contain
2508 binary zeroes or newlines because the value is written to log lines and
2509 to spool files. There was no check on this. Now the value is run through
2510 the string_printing() function so that such characters are converted to
2511 printable escape sequences.
2513 PH/41 $message_linecount is a new variable that contains the total number of
2514 lines in the message. Compare $body_linecount, which is the count for the
2517 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2518 bounce and delay warning messages, except in certain special cases, which
2521 (a) An SMTP error message from a remote host;
2522 (b) A message specified in a :fail: redirection;
2523 (c) A message specified in a "fail" command in a system filter;
2524 (d) A message specified in a FAIL return from the queryprogram router;
2525 (e) A message specified by the cannot_route_message router option.
2527 In these cases only, Exim does include the error details in bounce and
2528 warning messages. There are also a few cases where bland messages such
2529 as "unrouteable address" or "local delivery error" are given.
2531 PH/43 $value is now also set for the "else" part of a ${run expansion.
2533 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2534 being worked on, but at least Exim now implements the latest version to
2537 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2538 process to complete was treated as a delivery failure, a timeout while
2539 writing the message to the pipe was logged, but erroneously treated as a
2540 successful delivery. Such timeouts include transport filter timeouts. For
2541 consistency with the overall process timeout, these timeouts are now
2542 treated as errors, giving rise to delivery failures by default. However,
2543 there is now a new Boolean option for the pipe transport called
2544 timeout_defer, which, if set TRUE, converts the failures into defers for
2545 both kinds of timeout. A transport filter timeout is now identified in
2548 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2549 systems where "make" and "gmake" are different, calling "gmake" at top
2550 level broke things. I've arranged for the value of $(MAKE) to be passed
2551 from the Makefile to this script so that it can call the same version of
2555 A note about Exim versions 4.44 and 4.50
2556 ----------------------------------------
2558 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2559 changes of various kinds. As a consequence, a big documentation update was
2560 needed. This delayed the release for rather longer than seemed good, especially
2561 in the light of a couple of (minor) security issues. Therefore, the changes
2562 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2563 release. So 4.44 and 4.50 are in effect two different branches that both start
2566 I have left the 4.50 change log unchanged; it contains all the changes since
2567 4.43. The change log for 4.44 is below; many of its items are identical to
2568 those for 4.50. This seems to be the most sensible way to preserve the
2569 historical information.
2575 1. Minor wording change to the doc/README.SIEVE file.
2577 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2578 computation of the current number of files was incorrect.
2580 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2581 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2584 4. Give more explanation in the error message when the command for a transport
2585 filter fails to execute.
2587 5. There are several places where Exim runs a non-Exim command in a
2588 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2589 was being done only for the command run by the queryprogram router. It is
2590 now done for all such subprocesses. The other cases are: ${run, transport
2591 filters, and the commands run by the lmtp and pipe transports.
2593 6. Added CONFIGURE_GROUP build-time option.
2595 7. Some older OS have a limit of 256 on the maximum number of file
2596 descriptors. Exim was using setrlimit() to set 1000 as a large value
2597 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2598 systems. I've change it so that if it can't get 1000, it tries for 256.
2600 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2601 was an oversight, and furthermore, ever since the addition of extra
2602 controls (e.g. 4.43/32), the checks on when to allow different forms of
2603 "control" were broken. There should now be diagnostics for all cases when a
2604 control that does not make sense is encountered.
2606 9. Added the /retain_sender option to "control=submission".
2608 10. $recipients is now available in the predata ACL (oversight).
2610 11. Tidy the search cache before the fork to do a delivery from a message
2611 received from the command line. Otherwise the child will trigger a lookup
2612 failure and thereby defer the delivery if it tries to use (for example) a
2613 cached ldap connection that the parent has called unbind on.
2615 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2616 of $address_data from the recipient verification was clobbered by the
2617 sender verification.
2619 13. The value of address_data from a sender verification is now available in
2620 $sender_address_data in subsequent conditions in the ACL statement.
2622 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2624 15. Added a new option "connect=<time>" to callout options, to set a different
2627 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2628 was its contents. (It was OK if the option was not defined at all.)
2630 17. A "Completed" log line is now written for messages that are removed from
2631 the spool by the -Mrm option.
2633 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2634 information about exactly what failed.
2636 19. Added -dd to debug only the daemon process.
2638 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2639 handles timeouts, both on the server side and network timeouts. Renamed the
2640 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2642 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2643 It is now set to "smtps".
2645 22. $host_address is now set to the target address during the checking of
2646 ignore_target_hosts.
2648 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2649 being passed; this would have caused $sender_host_name to have been used if
2650 matching the list had actually called for a host name (not very likely,
2651 since this list is usually IP addresses). A host name is now passed as
2654 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2655 code that specifies a non-restarting handler (typically sigaction() in
2656 modern systems) in an attempt to fix a rare and obscure crash bug.
2658 25. Narrowed the window for a race in the daemon that could cause it to ignore
2659 SIGCHLD signals. This is not a major problem, because they are used only to
2660 wake it up if nothing else does.
2662 26. A malformed maildirsize file could cause Exim to calculate negative values
2663 for the mailbox size or file count. Odd effects could occur as a result.
2664 The maildirsize information is now recalculated if the size or filecount
2667 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2668 support for a long time. Removed HAVE_SYS_VFS_H.
2670 28. Installed the latest version of exipick from John Jetmore.
2672 29. In an address list, if the pattern was not a regular expression, an empty
2673 subject address (from a bounce message) matched only if the pattern was an
2674 empty string. Non-empty patterns were not even tested. This was the wrong
2675 because it is perfectly reasonable to use an empty address as part of a
2676 database query. An empty address is now tested by patterns that are
2677 lookups. However, all the other forms of pattern expect the subject to
2678 contain a local part and a domain, and therefore, for them, an empty
2679 address still always fails if the pattern is not itself empty.
2681 30. Exim went into a mad DNS loop when attempting to do a callout where the
2682 host was specified on an smtp transport, and looking it up yielded more
2683 than one IP address.
2685 31. Re-factored the code for checking spool and log partition space into a
2686 function that finds that data and another that does the check. The former
2687 is then used to implement four new variables: $spool_space, $log_space,
2688 $spool_inodes, and $log_inodes.
2690 32. The RFC2047 encoding function was originally intended for short strings
2691 such as real names; it was not keeping to the 75-character limit for
2692 encoded words that the RFC imposes. It now respects the limit, and
2693 generates multiple encoded words if necessary. To be on the safe side, I
2694 have increased the buffer size for the ${rfc2047: expansion operator from
2697 33. It is now permitted to omit both strings after an "if" condition; if the
2698 condition is true, the result is "true". As before, when the second string
2699 is omitted, a false condition yields an empty string. This makes it less
2700 cumbersome to write custom ACL and router conditions.
2702 34. Failure to deliver a bounce message always caused it to be frozen, even if
2703 there was an errors_to setting on the router. The errors_to setting is now
2706 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2707 canonical form (fully expanded) before being placed in
2708 $sender_host_address.
2710 36. The table in the code that translates DNS record types into text (T_A to
2711 "A" for instance) was missing entries for NS and CNAME. It is just possible
2712 that this could have caused confusion if both these types were looked up
2713 for the same domain, because the text type is used as part of Exim's
2714 per-process caching. But the chance of anyone hitting this buglet seems
2717 37. The dnsdb lookup has been extended in a number of ways.
2719 (1) There is a new type, "zns", which walks up the domain tree until it
2720 finds some nameserver records. It should be used with care.
2722 (2) There is a new type, "mxh", which is like "mx" except that it returns
2723 just the host names, not the priorities.
2725 (3) It is now possible to give a list of domains (or IP addresses) to be
2726 looked up. The behaviour when one of the lookups defers can be
2727 controlled by a keyword.
2729 (4) It is now possible to specify the separator character for use when
2730 multiple records are returned.
2732 38. The dnslists ACL condition has been extended: it is now possible to supply
2733 a list of IP addresses and/or domains to be looked up in a particular DNS
2736 39. Added log_selector=+queue_time_overall.
2738 40. When running the queue in the test harness, wait just a tad after forking a
2739 delivery process, to get repeatability of debugging output.
2741 41. Include certificate and key file names in error message when GnuTLS fails
2742 to set them up, because the GnuTLS error message doesn't include the name
2743 of the failing file when there is a problem reading it.
2745 42. Allow both -bf and -bF in the same test run.
2747 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2749 44. The "Exiscan patch" is now merged into the mainline Exim source.
2751 45. Sometimes the final signoff response after QUIT could fail to get
2752 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2753 < 0 before doing a fflush(). This bug looks as though it goes back to the
2754 introduction of TLS in release 3.20, but "sometimes" must have been rare
2755 because the tests only now provoked it.
2757 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2758 (this can affect the format of dates).
2760 47. exim_tidydb, when checking for the continued existence of a message for
2761 which it has found a message-specific retry record, was not finding
2762 messages that were in split spool directories. Consequently, it was
2763 deleting retry records that should have stayed in existence.
2765 48. Steve fixed some bugs in eximstats.
2767 49. The SPA authentication driver was not abandoning authentication and moving
2768 on to the next authenticator when an expansion was forced to fail,
2769 contradicting the general specification for all authenticators. Instead it
2770 was generating a temporary error. It now behaves as specified.
2772 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2773 (the order specifies the preference for clients). The order is now AES256,
2774 AES128, 3DES, ARCFOUR128.
2776 51. Small patch to Sieve code - explicitly set From: when generating an
2779 52. Exim crashed if a remote delivery caused a very long error message to be
2780 recorded - for instance if somebody sent an entire SpamAssassin report back
2781 as a large number of 550 error lines. This bug was coincidentally fixed by
2782 increasing the size of one of Exim's internal buffers (big_buffer) that
2783 happened as part of the Exiscan merge. However, to be on the safe side, I
2784 have made the code more robust (and fixed the comments that describe what
2787 53. Now that there can be additional text after "Completed" in log lines (if
2788 the queue_time_overall log selector is set), a one-byte patch to exigrep
2789 was needed to allow it to recognize "Completed" as not the last thing in
2792 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2793 patch that reportedly fixes this has been added. I am not expert enough to
2794 create a test for it. This is what the patch creator wrote:
2796 "I found a little strange behaviour of ldap code when working with
2797 Windows 2003 AD Domain, where users was placed in more than one
2798 Organization Units. When I tried to give exim partial DN, the exit code
2799 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2800 But simultaneously result of request was absolutely normal ldap result,
2801 so I produce this patch..."
2803 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2804 REFERENCE, so I have modified the code to exclude the patch when that macro
2807 55. Some experimental protocols are using DNS PTR records for new purposes. The
2808 keys for these records are domain names, not reversed IP addresses. The
2809 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2810 leaves it alone. Component reversal etc. now happens only for IP addresses.
2813 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2815 57. Double the size of the debug message buffer (to 2048) so that more of very
2816 long debug lines gets shown.
2818 58. The exicyclog utility now does better if the number of log files to keep
2819 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2821 59. Two changes related to the smtp_active_hostname option:
2823 (1) $smtp_active_hostname is now available as a variable.
2824 (2) The default for smtp_banner uses $smtp_active_hostname instead
2825 of $primary_hostname.
2827 60. The host_aton() function is supposed to be passed a string that is known
2828 to be a valid IP address. However, in the case of IPv6 addresses, it was
2829 not checking this. This is a hostage to fortune. Exim now panics and dies
2830 if the condition is not met. A case was found where this could be provoked
2831 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2832 components; fortuitously, this particular loophole had already been fixed
2833 by change 4.50/55 above.
2835 If there are any other similar loopholes, the new check in host_aton()
2836 itself should stop them being exploited. The report I received stated that
2837 data on the command line could provoke the exploit when Exim was running as
2838 exim, but did not say which command line option was involved. All I could
2839 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2840 running as the user.
2843 61. There was a buffer overflow vulnerability in the SPA authentication code
2844 (which came originally from the Samba project). I have added a test to the
2845 spa_base64_to_bits() function which I hope fixes it.
2848 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2849 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2851 63. The daemon start-up calls getloadavg() while still root for those OS that
2852 need the first call to be done as root, but it missed one case: when
2853 deliver_queue_load_max is set with deliver_drop_privilege. This is
2854 necessary for the benefit of the queue runner, because there is no re-exec
2855 when deliver_drop_privilege is set.
2857 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2858 This has been fixed.
2860 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2861 was in use, was not putting the data itself into the right store pool;
2862 consequently, it could be overwritten for a subsequent message in the same
2863 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2866 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2868 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2869 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2870 the function were treating the return as a boolean value, which happened to
2871 work because 0=false and not-0=true, but is not correct code.
2873 68. The host_aton() function was not handling scoped IPv6 addresses (those
2874 with, for example, "%eth0" on the end) correctly.
2876 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2877 negated items (that is, ~something) in unsigned ints. Some compilers
2878 apparently mutter when there is no cast.
2880 70. If an address verification called from an ACL failed, and did not produce a
2881 user-specific message (i.e. there was only a "system" message), nothing was
2882 put in $acl_verify_message. In this situation, it now puts the system
2885 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2886 session; change 4.31/43 added the unwanted input to the log line - except
2887 that it did not do this in the start of session case. It now does.
2889 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2890 This is wrong and can cause the other end to generate a synchronization
2891 error if it is another Exim or anything else that does the synchronization
2892 check. A QUIT command is no longer sent after a timeout.
2894 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2895 during host lookups.
2897 74. The defer_ok option of callout verification was not working if it was used
2898 when verifying addresses in header lines, that is, for this case:
2900 verify = header_sender/callout=defer_ok
2902 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2903 those file descriptors could be used for SMTP connections. If anything
2904 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2905 could be sent to the SMTP client, causing chaos. The daemon now opens
2906 stdin, stdout, and stderr to /dev/null when it puts itself into the
2909 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2910 log by default. The user can override this with suitable Perl magic.
2912 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2913 the log message when discard triggers, was not working for the DATA ACL or
2914 for the non-SMTP ACL.
2916 78. Error message wording change in sieve.c.
2918 79. If smtp_accept_max_per_host was set, the number of connections could be
2919 restricted to fewer than expected, because the daemon was trying to set up
2920 a new connection before checking whether the processes handling previous
2921 connections had finished. The check for completed processes is now done
2922 earlier. On busy systems, this bug wouldn't be noticed because something
2923 else would have woken the daemon, and it would have reaped the completed
2926 80. If a message was submitted locally by a user whose login name contained one
2927 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2928 It caused a spool format error. I have fixed the spool reading code. A
2929 related problem was that the "from" clause in the Received: line became
2930 illegal because of the space(s). It is now covered by ${quote_local_part.
2932 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2935 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2937 83. Patch from Sieve maintainer to fix unterminated string problem in
2938 "vacation" handling.
2940 84. Some minor changes to the Linux configuration files to help with other
2941 OS variants using glibc.
2943 85. One more patch for Sieve to update vacation handling to latest spec.
2946 ----------------------------------------------------
2947 See the note above about the 4.44 and 4.50 releases.
2948 ----------------------------------------------------
2954 1. Change 4.43/35 introduced a bug that caused file counts to be
2955 incorrectly computed when quota_filecount was set in an appendfile
2958 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2959 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2962 3. Give more explanation in the error message when the command for a transport
2963 filter fails to execute.
2965 4. There are several places where Exim runs a non-Exim command in a
2966 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2967 was being done only for the command run by the queryprogram router. It is
2968 now done for all such subprocesses. The other cases are: ${run, transport
2969 filters, and the commands run by the lmtp and pipe transports.
2971 5. Some older OS have a limit of 256 on the maximum number of file
2972 descriptors. Exim was using setrlimit() to set 1000 as a large value
2973 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2974 systems. I've change it so that if it can't get 1000, it tries for 256.
2976 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2977 was an oversight, and furthermore, ever since the addition of extra
2978 controls (e.g. 4.43/32), the checks on when to allow different forms of
2979 "control" were broken. There should now be diagnostics for all cases when a
2980 control that does not make sense is encountered.
2982 7. $recipients is now available in the predata ACL (oversight).
2984 8. Tidy the search cache before the fork to do a delivery from a message
2985 received from the command line. Otherwise the child will trigger a lookup
2986 failure and thereby defer the delivery if it tries to use (for example) a
2987 cached ldap connection that the parent has called unbind on.
2989 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2990 of $address_data from the recipient verification was clobbered by the
2991 sender verification.
2993 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2994 was its contents. (It was OK if the option was not defined at all.)
2996 11. A "Completed" log line is now written for messages that are removed from
2997 the spool by the -Mrm option.
2999 12. $host_address is now set to the target address during the checking of
3000 ignore_target_hosts.
3002 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3003 being passed; this would have caused $sender_host_name to have been used if
3004 matching the list had actually called for a host name (not very likely,
3005 since this list is usually IP addresses). A host name is now passed as
3008 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3009 code that specifies a non-restarting handler (typically sigaction() in
3010 modern systems) in an attempt to fix a rare and obscure crash bug.
3012 15. Narrowed the window for a race in the daemon that could cause it to ignore
3013 SIGCHLD signals. This is not a major problem, because they are used only to
3014 wake it up if nothing else does.
3016 16. A malformed maildirsize file could cause Exim to calculate negative values
3017 for the mailbox size or file count. Odd effects could occur as a result.
3018 The maildirsize information is now recalculated if the size or filecount
3021 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3022 support for a long time. Removed HAVE_SYS_VFS_H.
3024 18. Updated exipick to current release from John Jetmore.
3026 19. Allow an empty sender to be matched against a lookup in an address list.
3027 Previously the only cases considered were a regular expression, or an
3030 20. Exim went into a mad DNS lookup loop when doing a callout where the
3031 host was specified on the transport, if the DNS lookup yielded more than
3034 21. The RFC2047 encoding function was originally intended for short strings
3035 such as real names; it was not keeping to the 75-character limit for
3036 encoded words that the RFC imposes. It now respects the limit, and
3037 generates multiple encoded words if necessary. To be on the safe side, I
3038 have increased the buffer size for the ${rfc2047: expansion operator from
3041 22. Failure to deliver a bounce message always caused it to be frozen, even if
3042 there was an errors_to setting on the router. The errors_to setting is now
3045 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3046 canonical form (fully expanded) before being placed in
3047 $sender_host_address.
3049 24. Updated eximstats to version 1.33
3051 25. Include certificate and key file names in error message when GnuTLS fails
3052 to set them up, because the GnuTLS error message doesn't include the name
3053 of the failing file when there is a problem reading it.
3055 26. Expand error message when OpenSSL has problems setting up cert/key files.
3058 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3059 (this can affect the format of dates).
3061 28. exim_tidydb, when checking for the continued existence of a message for
3062 which it has found a message-specific retry record, was not finding
3063 messages that were in split spool directories. Consequently, it was
3064 deleting retry records that should have stayed in existence.
3066 29. eximstats updated to version 1.35
3067 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3068 1.35 - bugfix such that pie charts by volume are generated correctly
3070 30. The SPA authentication driver was not abandoning authentication and moving
3071 on to the next authenticator when an expansion was forced to fail,
3072 contradicting the general specification for all authenticators. Instead it
3073 was generating a temporary error. It now behaves as specified.
3075 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3076 (the order specifies the preference for clients). The order is now AES256,
3077 AES128, 3DES, ARCFOUR128.
3079 31. Small patch to Sieve code - explicitly set From: when generating an
3082 32. Exim crashed if a remote delivery caused a very long error message to be
3083 recorded - for instance if somebody sent an entire SpamAssassin report back
3084 as a large number of 550 error lines. This bug was coincidentally fixed by
3085 increasing the size of one of Exim's internal buffers (big_buffer) that
3086 happened as part of the Exiscan merge. However, to be on the safe side, I
3087 have made the code more robust (and fixed the comments that describe what
3090 33. Some experimental protocols are using DNS PTR records for new purposes. The
3091 keys for these records are domain names, not reversed IP addresses. The
3092 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3093 leaves it alone. Component reversal etc. now happens only for IP addresses.
3096 34. The host_aton() function is supposed to be passed a string that is known
3097 to be a valid IP address. However, in the case of IPv6 addresses, it was
3098 not checking this. This is a hostage to fortune. Exim now panics and dies
3099 if the condition is not met. A case was found where this could be provoked
3100 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3101 components; fortuitously, this particular loophole had already been fixed
3102 by change 4.50/55 or 4.44/33 above.
3104 If there are any other similar loopholes, the new check in host_aton()
3105 itself should stop them being exploited. The report I received stated that
3106 data on the command line could provoke the exploit when Exim was running as
3107 exim, but did not say which command line option was involved. All I could
3108 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3109 running as the user.
3112 35. There was a buffer overflow vulnerability in the SPA authentication code
3113 (which came originally from the Samba project). I have added a test to the
3114 spa_base64_to_bits() function which I hope fixes it.
3117 36. The daemon start-up calls getloadavg() while still root for those OS that
3118 need the first call to be done as root, but it missed one case: when
3119 deliver_queue_load_max is set with deliver_drop_privilege. This is
3120 necessary for the benefit of the queue runner, because there is no re-exec
3121 when deliver_drop_privilege is set.
3123 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3124 was in use, was not putting the data itself into the right store pool;
3125 consequently, it could be overwritten for a subsequent message in the same
3126 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3129 38. Sometimes the final signoff response after QUIT could fail to get
3130 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3131 < 0 before doing a fflush(). This bug looks as though it goes back to the
3132 introduction of TLS in release 3.20, but "sometimes" must have been rare
3133 because the tests only now provoked it.
3139 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3140 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3141 Now it returns an integer. A number of calls were still expecting a T/F
3142 return. Fortuitously, in all cases, the tests worked in OK situations,
3143 which is the norm. However, things would have gone wrong on any write
3144 failures on the smtp file descriptor. This function is used when sending
3145 messages over SMTP and also when doing verify callouts.
3147 2. When Exim is called to do synchronous delivery of a locally submitted
3148 message (the -odf or -odi options), it no longer closes stderr before doing
3151 3. Implemented the mua_wrapper option.
3153 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3155 5. Implemented the functions header_remove(), header_testname(),
3156 header_add_at_position(), and receive_remove_recipient(), and exported them
3159 6. If an ACL "warn" statement specified the addition of headers, Exim already
3160 inserted X-ACL-Warn: at the start if there was no header name. However, it
3161 was not making this test for the second and subsequent header lines if
3162 there were newlines in the string. This meant that an invalid header could
3163 be inserted if Exim was badly configured.
3165 7. Allow an ACL "warn" statement to add header lines at the start or after all
3166 the Received: headers, as well as at the end.
3168 8. Added the rcpt_4xx retry error code.
3170 9. Added postmaster_mailfrom=xxx to callout verification option.
3172 10. Added mailfrom=xxxx to the callout verification option, for verify=
3175 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3176 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3178 12. Inserted some casts to stop certain compilers warning when using pointer
3179 differences as field lengths or precisions in printf-type calls (mostly
3180 affecting debugging statements).
3182 13. Added optional readline() support for -be (dynamically loaded).
3184 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3185 same clock tick as a message's arrival, so that its received time was the
3186 same as the "first fail" time on the retry record, and that message
3187 remained on the queue past the ultimate address timeout, every queue runner
3188 would try a delivery (because it was past the ultimate address timeout) but
3189 after another failure, the ultimate address timeout, which should have then
3190 bounced the address, did not kick in. This was a "< instead of <=" error;
3191 in most cases the first failure would have been in the next clock tick
3192 after the received time, and all would be well.
3194 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3195 being recognized when the domain list was tested by the match_domain
3196 condition in an expansion string.
3198 16. Added the ${str2b64: operator.
3200 17. Exim was always calling setrlimit() to set a large limit for the number of
3201 processes, without checking whether the existing limit was already
3202 adequate. (It did check for the limit on file descriptors.) Furthermore,
3203 errors from getrlimit() and setrlimit() were being ignored. Now they are
3204 logged to the main and panic logs, but Exim does carry on, to try to do its
3205 job under whatever limits there are.
3207 18. Imported PCRE 5.0.
3209 19. Trivial typo in log message " temporarily refused connection" (the leading
3212 20. If the log selector return_path_on_delivery was set and an address was
3213 redirected to /dev/null, the delivery process crashed because it assumed
3214 that a return path would always be set for a "successful" delivery. In this
3215 case, the whole delivery is bypassed as an optimization, and therefore no
3218 21. Internal re-arrangement: the function for sending a challenge and reading
3219 a response while authentication was assuming a zero-terminated challenge
3220 string. It's now changed to take a pointer and a length, to allow for
3221 binary data in such strings.
3223 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3225 23. Exim was not respecting finduser_retries when seeking the login of the
3226 uid under which it was called; it was always trying 10 times. (The default
3227 setting of finduser_retries is zero.) Also, it was sleeping after the final
3228 failure, which is pointless.
3230 24. Implemented tls_on_connect_ports.
3232 25. Implemented acl_smtp_predata.
3234 26. If the domain in control=submission is set empty, Exim assumes that the
3235 authenticated id is a complete email address when it generates From: or
3236 Sender: header lines.
3238 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3239 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3240 chown and chgrp in /bin and hostname in /usr/bin.
3242 28. Exim was keeping the "process log" file open after each use, just as it
3243 does for the main log. This opens the possibility of it remaining open for
3244 long periods when the USR1 signal hits a daemon. Occasional processlog
3245 errors were reported, that could have been caused by this. Anyway, it seems
3246 much more sensible not to leave this file open at all, so that is what now
3249 29. The long-running daemon process does not normally write to the log once it
3250 has entered its main loop, and it closes the log before doing so. This is
3251 so that log files can straightforwardly be renamed and moved. However,
3252 there are a couple of unusual error situations where the daemon does write
3253 log entries, and I had neglected to close the log afterwards.
3255 30. The text of an SMTP error response that was received during a remote
3256 delivery was being truncated at 512 bytes. This is too short for some of
3257 the long messages that one sometimes sees. I've increased the limit to
3260 31. It is now possible to make retry rules that apply only when a message has a
3261 specific sender, in particular, an empty sender.
3263 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3264 it possible to be selective about when SMTP synchronization is enforced.
3266 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3268 32. Implemented hosts_connection_nolog.
3270 33. Added an ACL for QUIT.
3272 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3275 35. Added mailbox_size and mailbox_filecount to appendfile.
3277 36. Added control = no_multiline_responses to ACLs.
3279 37. There was a bug in the logic of the code that waits for the clock to tick
3280 in the case where the clock went backwards by a substantial amount such
3281 that the microsecond fraction of "now" was more than the microsecond
3282 fraction of "then" (but the whole seconds number was less).
3284 38. Added support for the libradius Radius client library this is found on
3285 FreeBSD (previously only the radiusclient library was supported).
3291 1. When certain lookups returned multiple values in the form name=value, the
3292 quoting of the values was not always being done properly. Specifically:
3293 (a) If the value started with a double quote, but contained no whitespace,
3295 (b) If the value contained whitespace other than a space character (i.e.
3296 tabs or newlines or carriage returns) it was not quoted.
3297 This fix has been applied to the mysql and pgsql lookups by writing a
3298 separate quoting function and calling it from the lookup code. The fix
3299 should probably also be applied to nisplus, ibase and oracle lookups, but
3300 since I cannot test any of those, I have not disturbed their existing code.
3302 2. A hit in the callout cache for a specific address caused a log line with no
3303 reason for rejecting RCPT. Now it says "Previous (cached) callout
3304 verification failure".
3306 3. There was an off-by-one bug in the queryprogram router. An over-long
3307 return line was truncated at 256 instead of 255 characters, thereby
3308 overflowing its buffer with the terminating zero. As well as fixing this, I
3309 have increased the buffer size to 1024 (and made a note to document this).
3311 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3312 when Exim is waiting for an SMTP response from a remote server, Exim
3313 restarts its select() call on the socket, thereby resetting its timeout.
3314 This is not a problem when such interrupts are rare. Somebody set up a cron
3315 job to run exiwhat every 2 minutes, which is less than the normal select()
3316 timeout (5 or 10 minutes). This meant that the select() timeout never
3317 kicked in because it was always reset. I have fixed this by comparing the
3318 time when an interrupt arrives with the time at the start of the first call
3319 to select(). If more time than the timeout has elapsed, the interrupt is
3320 treated as a timeout.
3322 5. Some internal re-factoring in preparation for the addition of Sieve
3323 extensions (by MH). In particular, the "personal" test is moved to a
3324 separate function, and given an option for scanning Cc: and Bcc: (which is
3325 not set for Exim filters).
3327 6. When Exim created an email address using the login of the caller as the
3328 local part (e.g. when creating a From: or Sender: header line), it was not
3329 quoting the local part when it contained special characters such as @.
3331 7. Installed new OpenBSD configuration files.
3333 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3334 try to make them clearer.
3336 9. Callout options, other than the timeout value, were being ignored when
3337 verifying sender addresses in header lines. For example, when using
3339 verify = header_sender/callout=no_cache
3341 the cache was (incorrectly) being used.
3343 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3344 only the Cygwin environment.
3346 11. When return_path_on_delivery was set as a log selector, if different remote
3347 addresses in the same message used different return paths and parallel
3348 remote delivery occurred, the wrong values would sometimes be logged.
3349 (Whenever a remote delivery process finished, the return path value from
3350 the most recently started remote delivery process was logged.)
3352 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3353 lines when AUTH and/or TLS are in use. This is the "received protocol"
3354 field. Exim used to use "asmtp" for authenticated SMTP, without any
3355 indication (in the protocol name) for TLS use. Now it follows the RFC and
3356 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3357 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3358 names appear in log lines as well as in Received: header lines.
3360 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3361 extensions, and comparison tests, and to fix some bugs.
3363 14. Changes to the "personal" filter test:
3365 (1) The test was buggy in that it was just doing the equivalent of
3366 "contains" tests on header lines. For example, if a user's address was
3367 anne@some.where, the "personal" test would incorrectly be true for
3369 To: susanne@some.where
3371 This test is now done by extracting each address from the header in turn,
3372 and checking the entire address. Other tests that are part of "personal"
3373 are now done using regular expressions (for example, to check local parts
3374 of addresses in From: header lines).
3376 (2) The list of non-personal local parts in From: addresses has been
3377 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3378 taken from the Sieve specification recommendations.
3380 (3) If the message contains any header line starting with "List-" it is
3381 treated as non-personal.
3383 (4) The test for "circular" in the Subject: header line has been removed
3384 because it now seems ill-conceived.
3386 15. Minor typos in src/EDITME comments corrected.
3388 16. Installed latest exipick from John Jetmore.
3390 17. If headers_add on a router specified a text string that was too long for
3391 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3392 of string_sprintf() is now avoided.
3394 18. $message_body_size was not set (it was always zero) when running the DATA
3395 ACL and the local_scan() function.
3397 19. For the "mail" command in an Exim filter, no default was being set for
3398 the once_repeat time, causing a random time value to be used if "once" was
3399 specified. (If the value happened to be <= 0, no repeat happened.) The
3400 default is now 0s, meaning "never repeat". The "vacation" command was OK
3401 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3402 (I found it when inspecting the code).
3404 20. There is now an overall timeout for performing a callout verification. It
3405 defaults to 4 times the callout timeout, which applies to individual SMTP
3406 commands during the callout. The overall timeout applies when there is more
3407 than one host that can be tried. The timeout is checked before trying the
3408 next host. This prevents very long delays if there are a large number of
3409 hosts and all are timing out (e.g. when the network connections are timing
3410 out). The value of the overall timeout can be changed by specifying an
3411 additional sub-option for "callout", called "maxwait". For example:
3413 verify = sender/callout=5s,maxwait=20s
3415 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3416 to the end before writing, but this should make it even safer).
3418 22. Exim was forgetting that it had advertised PIPELINING for the second and
3419 subsequent messages on an SMTP connection. It was also not resetting its
3420 memory on STARTTLS and an internal HELO.
3422 23. When Exim logs an SMTP synchronization error within a session, it now
3423 records whether PIPELINING has been advertised or not.
3425 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3426 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3427 rather than long int.
3429 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3431 26. Added the never_mail option to autoreply.
3437 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3438 crash if the getsockname() call failed; this can happen if a connection is
3439 closed very soon after it is established. The problem was simply in the
3440 order in which certain operations were done, causing Exim to try to write
3441 to the SMTP stream before it had set up the file descriptor. The bug has
3442 been fixed by making things happen in the correct order.
3448 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3449 before the connection was closed, thus losing the rejection response.
3451 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3452 some early Solaris releases, but causes trouble in current releases where
3453 socklen_t is defined.
3455 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3458 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3461 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3462 that releasing the top of it at the end releases what was used for sub-
3463 expansions (unless the block got too big). However, discard this block if
3464 the first thing is a variable or header, so that we can use its block when
3465 it is dynamic (useful for very large $message_headers, for example).
3467 6. Lookups now cache *every* query, not just the most recent. A new, separate
3468 store pool is used for this. It can be recovered when all lookup caches are
3469 flushed. Lookups now release memory at the end of their result strings.
3470 This has involved some general refactoring of the lookup sources.
3472 7. Some code has been added to the store_xxx() functions to reduce the amount
3473 of flapping under certain conditions.
3475 8. log_incoming_interface used to affect only the <= reception log lines. Now
3476 it causes the local interface and port to be added to several more SMTP log
3477 lines, for example "SMTP connection from", and rejection lines.
3479 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3481 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3483 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3484 could be overwritten at the end of the current message (or the start of a
3485 new message if it was set in a HELO ACL). The value is now preserved for
3486 the duration of the SMTP connection.
3488 12. If a transport had a headers_rewrite setting, and a matching header line
3489 contained an unqualified address, that address was qualified, even if it
3490 did not match any rewriting rules. The underlying bug was that the values
3491 of the flags that permit the existence of unqualified sender and recipient
3492 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3493 non-local messages, and by -bnq for local messages) were not being
3494 preserved with the message after it was received.
3496 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3497 "next input=" as part of the text comprising the host identity instead of
3498 the correct text. The code was using the same buffer for two different
3499 strings. However, depending on which order the printing function evaluated
3500 its arguments, the bug did not always show up. Under Linux, for example, my
3501 test suite worked just fine.
3503 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3504 doesn't work with some older versions of Perl. It has been changed to "my",
3505 which in any case is probably the better facility to use.
3507 15. A really picky compiler found some instances of statements for creating
3508 error messages that either had too many or two few arguments for the format
3511 16. The size of the buffer for calls to the DNS resolver has been increased
3512 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3513 for addresses that have a lot of PTR records. This alleviates a problem; it
3514 does not fully solve it.
3516 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3517 buffer now truncates the list and logs the incident, which is the same
3518 action as happens when Exim is looking up a host name and its aliases.
3519 Previously in this situation something unpredictable would happen;
3520 sometimes it was "internal error: store_reset failed".
3522 18. If a server dropped the connection unexpectedly when an Exim client was
3523 using GnuTLS and trying to read a response, the client delivery process
3524 crashed while trying to generate an error log message.
3526 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3527 string, for example:
3529 warn message = H1: something\nH2: something
3531 the text was added as a single header line from Exim's point of view
3532 though it ended up OK in the delivered message. However, searching for the
3533 second and subsequent header lines using $h_h2: did not work. This has been
3534 fixed. Similarly, if a system filter added multiple headers in this way,
3535 the routers could not see them.
3537 20. Expanded the error message when iplsearch is called with an invalid key to
3538 suggest using net-iplsearch in a host list.
3540 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3541 ACLs are no longer actually imposed (and a message to that effect is
3544 22. If a "gecos" field in a passwd entry contained escaped characters, in
3545 particular, if it contained a \" sequence, Exim got it wrong when building
3546 a From: or a Sender: header from that name. A second bug also caused
3547 incorrect handling when an unquoted " was present following a character
3548 that needed quoting.
3550 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3551 was not being matched caselessly.
3553 24. Arranged for all hyphens in the exim.8 source to be escaped with
3556 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3557 properly. Recipient callout cache records were still being keyed to include
3558 the sender, even when use_sender was set false. This led to far more
3559 callouts that were necessary. The sender is no longer included in the key
3560 when use_sender is false.
3562 26. Added "control = submission" modifier to ACLs.
3564 27. Added the ${base62d: operator to decode base 62 numbers.
3566 28. dnsdb lookups can now access SRV records.
3568 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3569 the configuration file.
3571 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3572 (-v) mode. This makes the output for a verbose queue run more intelligible.
3574 31. Added a use_postmaster feature to recipient callouts.
3576 32. Added the $body_zerocount variable, containing the number of binary zero
3577 bytes in the message body.
3579 33. The time of last modification of the "new" subdirectory is now used as the
3580 "mailbox time last read" when there is a quota error for a maildir
3583 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3585 35. Added +ignore_unknown as a special item in host lists.
3587 36. Code for decoding IPv6 addresses in host lists is now included, even if
3588 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3589 address was recognized as an IP address, but was then not correctly decoded
3590 into binary, causing unexpected and incorrect effects when compared with
3597 1. Very minor rewording of debugging text in manualroute to say "list of
3598 hosts" instead of "hostlist".
3600 2. If verify=header_syntax was set, and a header line with an unqualified
3601 address (no domain) and a large number of spaces between the end of the
3602 name and the colon was received, the reception process suffered a buffer
3603 overflow, and (when I tested it) crashed. This was caused by some obsolete
3604 code that should have been removed. The fix is to remove it!
3606 3. When running in the test harness, delay a bit after writing a bounce
3607 message to get a bit more predictability in the log output.
3609 4. Added a call to search_tidyup() just before forking a reception process. In
3610 theory, someone could use a lookup in the expansion of smtp_accept_max_
3611 per_host which, without the tidyup, could leave open a database connection.
3613 5. Added the variables $recipient_data and $sender_data which get set from a
3614 lookup success in an ACL "recipients" or "senders" condition, or a router
3615 "senders" option, similar to $domain_data and $local_part_data.
3617 6. Moved the writing of debug_print from before to after the "senders" test
3620 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3621 problems for message scanning, either using a data ACL, or using
3622 local_scan() because the Received: header was not generated till after they
3623 were called (in order to set the time as the time of reception completion).
3624 I have revised the way this works. The header is now generated after the
3625 body is received, but before the ACL or local_scan() are called. After they
3626 are run, the timestamp in the header is updated.
3632 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3633 before starting a queue runner without re-exec. This happened only when
3634 deliver_drop_privilege was set or when the Exim user was set to root. The
3635 effect of the bug was that timeouts during subsequent deliveries caused
3636 crashes instead of being properly handled. The handler is now left at its
3637 default (and expected) setting.
3639 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3640 message, again when deliver_drop_privilege is set or Exim is run as root.
3641 The bug described in (1) was not present in this case, but the tidying up
3642 of the other signals was missing. I have made the two cases consistent.
3644 3. The ignore_target_hosts setting on a manualroute router was being ignored
3645 for hosts that were looked up using the /MX notation.
3647 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3650 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3651 operated on the sender address. After changing the $sender_address to <>
3652 for the sender address verify, Exim was re-instated it as the original
3653 (before rewriting) address, but remembering that it had rewritten it, so it
3654 wasn't rewriting it again. This bug also had the effect of breaking the
3655 sender address verification caching when the sender address was rewritten.
3657 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3658 This has been changed so that if the ip literal address matches
3659 ignore_target_hosts, the router declines.
3661 7. Added expansion conditions match_domain, match_address, and match_local_
3662 part (NOT match_host).
3664 8. The placeholder for the Received: header didn't have a length field set.
3666 9. Added code to Exim itself and to exim_lock to test for a specific race
3667 condition that could lead to file corruption when using MBX delivery. The
3668 issue is with the lockfile that is created in /tmp. If this file is removed
3669 after a process has opened it but before that process has acquired a lock,
3670 there is the potential for a second process to recreate the file and also
3671 acquire a lock. This could lead to two Exim processes writing to the file
3672 at the same time. The added code performs the same test as UW imapd; it
3673 checks after acquiring the lock that its file descriptor still refers to
3674 the same named file.
3676 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3677 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3680 11. Added the smtp_active_hostname option. If used, this will typically be made
3681 to depend on the incoming interface address. Because $interface_address is
3682 not set up until the daemon has forked a reception process, error responses
3683 that can happen earlier (such as "too many connections") no longer contain
3686 12. If an expansion in a condition on a "warn" statement fails because a lookup
3687 defers, the "warn" statement is abandoned, and the next ACL statement is
3688 processed. Previously this caused the whole ACL to be aborted.
3690 13. Added the iplsearch lookup type.
3692 14. Added ident_timeout as a log selector.
3694 15. Added tls_certificate_verified as a log selector.
3696 16. Added a global option tls_require_ciphers (compare the smtp transport
3697 option of the same name). This controls incoming TLS connections.
3699 17. I finally figured out how to make tls_require_ciphers do a similar thing
3700 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3701 before starting the TLS session.
3703 18. Tabs are now shown as \t in -bP output.
3705 19. If the log selector return_path_on_delivery was set, Exim crashed when
3706 bouncing a message because it had too many Received: header lines.
3708 20. If two routers both had headers_remove settings, and the first one included
3709 a superfluous trailing colon, the final name in the first list and the
3710 first name in the second list were incorrectly joined into one item (with a
3711 colon in the middle).
3717 1. Added -C and -D options to the exinext utility, mainly to make it easier
3718 to include in the automated testing, but these could be helpful when
3719 multiple configurations are in use.
3721 2. The exinext utility was not formatting the output nicely when there was
3722 an alternate port involved in the retry record key, nor when there was a
3723 message id as well (for retries that were specific to a specific message
3724 and a specific host). It was also confused by IPv6 addresses, because of
3725 the additional colons they contain. I have fixed the IPv4 problem, and
3726 patched it up to do a reasonable job for IPv6.
3728 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3729 delivery, the log line now contains "pipelined" if PIPELINING was used.
3731 4. An SMTP transport process used to panic and die if the bind() call to set
3732 an explicit outgoing interface failed. This has been changed; it is now
3733 treated in the same way as a connect() failure.
3735 5. A reference to $sender_host_name in the part of a conditional expansion
3736 that was being skipped was still causing a DNS lookup. This no longer
3739 6. The def: expansion condition was not recognizing references to header lines
3740 that used bh_ and bheader_.
3742 7. Added the _cache feature to named lists.
3744 8. The code for checking quota_filecount in the appendfile transport was
3745 allowing one more file than it should have been.
3747 9. For compatibility with Sendmail, the command line option
3755 and sets the incoming protocol and host name (for trusted callers). The
3756 host name and its colon can be omitted when only the protocol is to be set.
3757 Note the Exim already has two private options, -pd and -ps, that refer to
3758 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3759 "s", but I don't think that's a major issue.
3761 10. A number of refactoring changes to the code, none of which should affect
3764 (a) The number of logging options was getting close to filling up the
3765 32-bit word that was used as a bit map. I have split them into two classes:
3766 those that are passed in the argument to log_write(), and those that are
3767 only ever tested independently outside of that function. These are now in
3768 separate 32-bit words, so there is plenty of room for expansion again.
3769 There is no change in the user interface or the logging behaviour.
3771 (b) When building, for example, log lines, the code previously used a
3772 macro that called string_cat() twice, in order to add two strings. This is
3773 not really sufficiently general. Furthermore, there was one instance where
3774 it was actually wrong because one of the argument was used twice, and in
3775 one call a function was used. (As it happened, calling the function twice
3776 did not affect the overall behaviour.) The macro has been replaced by a
3777 function that can join an arbitrary number of extra strings onto a growing
3780 (c) The code for expansion conditions now uses a table and a binary chop
3781 instead of a serial search (which was left over from when there were very
3782 few conditions). Also, it now recognizes conditions like "pam" even when
3783 the relevant support is not compiled in: a suitably worded error message is
3784 given if an attempt is made to use such a condition.
3786 11. Added ${time_interval:xxxxx}.
3788 12. A bug was causing one of the ddress fields not to be passed back correctly
3789 from remote delivery subprocesses. The field in question was not being
3790 subsequently used, so this caused to problems in practice.
3792 13. Added new log selectors queue_time and deliver_time.
3794 14. Might have fixed a bug in maildirsizefile handling that threw up
3795 "unexpected character" debug warnings, and recalculated the data
3796 unnecessarily. In any case, I expanded the warning message to give more
3799 15. Added the message "Restricted characters in address" to the statements in
3800 the default ACL that block characters like @ and % in local parts.
3802 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3803 Three changes have been made:
3805 (a) There was a serious bug; a negative response to MAIL caused the whole
3806 recipient domain to be cached as invalid, thereby blocking all messages
3807 to all local parts at the same domain, from all senders. This bug has
3808 been fixed. The domain is no longer cached after a negative response to
3809 MAIL if the sender used is not empty.
3811 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3814 (c) A new callout option, "use_sender" has been added for people who want
3815 the modified behaviour.
3821 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3824 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3825 indeed breaks things for older releases.
3827 3. Added additional logging to the case where there is a problem reading data
3828 from a filter that is running in a subprocess using a pipe, in order to
3829 try to track down a specific problem.
3831 4. Testing facility fudge: when running in the test harness and attempting
3832 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3833 getting "No route to host". Convert this to a timeout.
3835 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3838 6. Some OS don't have socklen_t but use size_t instead. This affects the
3839 fifth argument of getsockopt() amongst other things. This is now
3840 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3841 can be set for individual OS. I have set it for SunOS5, OSF1, and
3842 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3843 some earlier ones do not.
3845 7. Change 4.30/15 was not doing the test caselessly.
3847 8. The standard form for an IPv6 address literal was being rejected by address
3848 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3849 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3850 this, as well as the form without the "IPv6" on the front (but only when
3851 address literals are enabled, of course).
3853 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3855 10. Exim crashed if a message with an empty sender address specified by -f
3856 encountered a router with an errors_to setting. This could be provoked only
3857 by a command such as
3861 where an empty string was supplied; "<>" did not hit this bug.
3863 11. Installed PCRE release 4.5.
3865 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3866 remained set. It is now erased.
3868 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3869 times from message ids (which are base 36 rather than the normal 62).
3871 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3872 were being counted as actual protocol errors, and logged if the log
3873 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3874 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3875 and DATA following a set of rejected RCPTs do not count as protocol errors.
3876 In other words, Exim assumes they were pipelined, though this may not
3877 actually be the case. Of course, in all cases the client gets an
3878 appropriate error code.
3880 15. If a lookup fails in an ACL condition, a message about the failure may
3881 be available; it is used if testing the ACL cannot continue, because most
3882 such messages specify what the cause of the deferral is. However, some
3883 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3884 that caused an old message to be retained and used if a later statement
3885 caused a defer, replacing the real cause of the deferral.
3887 16. If an IP address had so many PTR records that the DNS lookup buffer
3888 was not large enough to hold them, Exim could crash while trying to process
3889 the truncated data. It now detects and logs this case.
3891 17. Further to 4.21/58, another change has been made: if (and only if) the
3892 first line of a message (the first header line) ends with CRLF, a bare LF
3893 in a subsequent header line has a space inserted after it, so as not to
3894 terminate the header.
3896 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3897 unnecessarily, used atoi() instead of strtol(), and didn't check the
3898 termination when getting file sizes from file names by regex.
3900 19. Completely re-implemented the support for maildirsize files, in the light
3901 of a number of problems with the previous contributed implementation
3902 (4.30/29). In particular:
3904 . If the quota is zero, the maildirsize file is maintained, but no quota is
3907 . If the maildir directory does not exist, it is created before any attempt
3908 to write a maildirsize file.
3910 . The quota value in the file is just a cache; if the quota is changed in
3911 the transport, the new value overrides.
3913 . A regular expression is available for excluding directories from the
3916 20. The autoreply transport checks the characters in options that define the
3917 message's headers; it allows continued headers, but it was checking with
3918 isspace() after an embedded newline instead of explicitly looking for a
3921 21. If all the "regular" hosts to which an address was routed had passed their
3922 expiry times, and had not reached their retry times, the address was
3923 bounced, even if fallback hosts were defined. Now Exim should go on to try
3926 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3927 equivalent code in the SMTP transport. Some hosts send humungous responses
3928 to HELO/EHLO, more than 1024 it seems.
3930 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3931 gives compiler warnings in some environments. I've now done it "properly",
3934 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3935 (because of problems with the built-in one) was declared to return uschar *
3936 instead of char *, causing compiler failure.
3938 25. Fixed a file descriptor leak when processing alias/forward files.
3940 26. Fixed a minor format string issue in dbfn.c.
3942 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3944 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3945 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3948 29. When a sender address is verified, it is cached, to save repeating the test
3949 when there is more than one recipient in a message. However, when the
3950 verification involves a callout, it is possible for different callout
3951 options to be set for different recipients. It is too complicated to keep
3952 track of this in the cache, so now Exim always runs a verification when a
3953 callout is required, relying on the callout cache for the optimization.
3954 The overhead is duplication of the address routing, but this should not be
3957 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3958 to be verified with callout=postmaster, and the main callout worked but the
3959 postmaster check failed, the verification correctly failed. However, if a
3960 subsequent RCPT command asked for sender verification *without* the
3961 postmaster check, incorrect caching caused this verification also to fail,
3964 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3965 it was not caching the DNS options (qualify_single, search_parents) that
3966 were used when the lookup failed. A subsequent lookup with different
3967 options therefore always gave the same answer, though there were cases
3968 where it should not have. (Example: a "domains = !$mx_any" option on a
3969 dnslookup router: the "domains" option is always processed without any
3970 widening, but the router might have qualify_single set.) Now Exim uses the
3971 cached value only when the same options are set.
3973 32. Added John Jetmore's "exipick" utility to the distribution.
3975 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3976 than a timeout (e.g. a certificate is required, and is not provided), an
3977 Exim server now closes the connection immediately. Previously it waited for
3978 the client to close - but if the client is SSL, it seems that they each
3979 wait for each other, leading to a delay before one of them times out.
3981 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3982 maintained 0.8.x compatibility because I don't think many are using it, and
3983 it is clearly obsolete.
3985 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3988 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3989 client certificate was expired. A simple patch fixes this, though I don't
3990 understand the full logic of why the verify callback is called multiple
3993 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3994 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3995 which causes problems with some clients (such as the Certicom SSL Plus
3996 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3997 disables the coutermeasure allowing Eudora to connect."
3999 38. Exim was not checking that a write() to a log file succeeded. This could
4000 lead to Bad Things if a log got too big, in particular if it hit a file
4001 size limit. Exim now panics and dies if it cannot write to a log file, just
4002 as it does if it cannot open a log file.
4004 39. Modified OS/Makefile-Linux so that it now contains
4006 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4008 The two -D definitions ensure that Exim is compiled with large file
4009 support, which makes it possible to handle log files that are bigger than
4012 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4013 instance) a domain was checked against a named list that involved a lookup,
4014 causing $domain_data to be set, then another domain was checked against the
4015 same list, then the first domain was re-checked, the value of $domain_data
4016 after the final check could be wrong. In particular, if the second check
4017 failed, it could be set empty. This bug probably also applied to
4020 41. The strip_trailing_dot option was not being applied to the address given
4021 with the -f command-line option.
4023 42. The code for reading a message's header from the spool was incrementing
4024 $received_count, but never initializing it. This meant that the value was
4025 incorrect (doubled) while delivering a message in the same process in which
4026 it was received. In the most common configuration of Exim, this never
4027 happens - a fresh exec is done - but it can happen when
4028 deliver_drop_privilege is set.
4030 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4031 it now includes up to 150 characters of the unexpected data in the log
4034 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4035 and building data strings. The size of both of these buffers was 10 000
4036 bytes - far larger than anybody would *ever* want, thought I. Needless to
4037 say, somebody hit the limit. I have increased the maximum line length to
4038 20 000 and the maximum data length of concatenated lines to 100 000. I have
4039 also fixed two bugs, because there was no checking on these buffers. Tsk,
4040 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4041 buffer is too small.
4043 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4044 lsearch lookups. Now it does.
4046 46. When parsing a route_list item in a manualroute router, a fixed-length
4047 buffer was used for the list of hosts. I made this 1024 bytes long,
4048 thinking that nobody would ever have a list of hosts that long. Wrong.
4049 Somebody had a whole pile of complicated expansion conditions, and the
4050 string was silently truncated, leading to an expansion error. It turns out
4051 that it is easier to change to an unlimited length (owing to other changes
4052 that have happened since this code was originally written) than to build
4053 structure for giving a limitation error. The length of the item that
4054 expands into the list of hosts is now unlimited.
4056 47. The lsearch lookup could not handle data where the length of text line was
4057 more than 4095 characters. Such lines were truncated, leading to shortened
4058 data being returned. It should now handle lines of any length.
4060 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4061 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4064 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4065 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4066 no longer happens in comment lines. A list of replacements is now placed
4067 at the head of all of the source files, except those whose only change is
4068 to replace PERL_COMMAND in the very first #! line.
4070 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4071 messages on the queue, with a bottom-up merge sort, using code contributed
4072 by Michael Haardt. This should make operations like -bp somewhat faster on
4073 large queues. It won't affect queue runners, except when queue_run_in_order
4076 51. Installed eximstats 1.31 in the distribution.
4078 52. Added support for SRV lookups to the dnslookup router.
4080 53. If an ACL referred to $message_body or $message_body_end, the value was not
4081 reset for any messages that followed in the same SMTP session.
4083 54. The store-handling optimization for building very long strings was not
4084 differentiating between the different store pools. I don't think this
4085 actually made any difference in practice, but I've tidied it.
4087 55. While running the routers to verify a sender address, $sender_address
4088 was still set to the sender address. This is wrong, because when routing to
4089 send a bounce to the sender, it would be empty. Therefore, I have changed
4090 it so that, while verifying a sender address, $sender_address is set to <>.
4091 (There is no change to what happens when verifying a recipient address.)
4093 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4094 target A or AAAA records (if not already returned) without resetting the
4095 qualify_single or search_parents options of the DNS resolver. These are
4096 inappropriate in this case because the targets of MX and SRV records must
4097 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4098 target that, when qualified, matched something in the local domain. These
4099 two options are now turned off when doing these lookups.
4101 57. It seems that at least some releases of Reiserfs (which does not have the
4102 concept of a fixed number of inodes) returns zero and not -1 for the
4103 number of available inodes. This interacted badly with check_spool_inodes,
4104 which assumed that -1 was the "no such thing" setting. What I have done is
4105 to check that the total number of inodes is greater than zero before doing
4106 the test of how many are available.
4108 58. When a "warn" ACL statement has a log_message modifier, the message is
4109 remembered, and not repeated. This is to avoid a lot of repetition when a
4110 message has many recipients that cause the same warning to be written.
4111 Howewer, Exim was preserving the list of already written lines for an
4112 entire SMTP session, which doesn't seem right. The memory is now reset if a
4113 new message is started.
4115 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4116 debugging output unless log_rewrite was also set.
4118 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4119 of (int)(handle) when we know that handle contains (void *)(-1).
4121 61. The Exim daemon panic-logs an error return when it closes the incoming
4122 connection. However "connection reset by peer" seems to be common, and
4123 isn't really an error worthy of noting specially, so that particular error
4126 62. When Exim is trying to find all the local interfaces, it used to panic and
4127 die if the ioctl to get the interface flags failed. However, it seems that
4128 on at least one OS (Solaris 9) it is possible to have an interface that is
4129 included in the list of interfaces, but for which you get a failure error
4130 for this call. This happens when the interface is not "plumbed" into a
4131 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4132 failure of the "get flags" call assumes that the interface is down.
4134 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4135 makes life easier for people who are doing arithmetic on fields extracted
4136 from dates, where you often get leading zeros that should not be
4137 interpreted as octal.
4139 64. Added qualify_domain to the redirect router, to override the global
4142 65. If a pathologically long header line contained very many addresses (the
4143 report of this problem mentioned 10 000) and each of them was rewritten,
4144 Exim could use up a very large amount of memory. (It kept on making new
4145 copies of the header line as it rewrote, and never released the old ones.)
4146 At the expense of a bit more processing, the header rewriting function has
4147 been changed so that it no longer eats memory in this way.
4149 66. The generation of the Received: header has been moved from the time that a
4150 message starts to be received, to the time that it finishes. The timestamp
4151 in the Received: header should now be very close to that of the <= log
4152 line. There are two side-effects of this change:
4154 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4155 logged header lines no longer include the local Received: line, because
4156 it has not yet been created. The same applies to a copy of the message
4157 that is returned to a non-SMTP sender when a message is rejected.
4159 (b) When a filter file is tested using -bf, no additional Received: header
4160 is added to the test message. After some thought, I decided that this
4163 This change does not affect the value of $received_for. It is still set
4164 after address rewriting, but before local_scan() is called.
4166 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4168 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4169 gave an unhelpful panic error message, and a defer error. I have managed to
4170 change this behaviour so that it now rejects any supplied certificate,
4171 which seems right, as the list of acceptable certificates is empty.
4173 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4174 gave an unhelpful defer error. I have not managed to make this reject any
4175 supplied certificates, but the error message it gives is "no certificate
4176 supplied", which is not helpful.
4178 70. exigrep's output now also includes lines that are not associated with any
4179 message, but which match the given pattern. Implemented by a patch from
4180 Martin Sluka, which also tidied up the Perl a bit.
4182 71. Recipient callout verification, like sender verification, was using <> in
4183 the MAIL FROM command. This isn't really the right thing, since the actual
4184 sender may affect whether the remote host accepts the recipient or not. I
4185 have changed it to use the actual sender in the callout; this means that
4186 the cache record is now keyed on a recipient/sender pair, not just the
4187 recipient address. There doesn't seem to be a real danger of callout loops,
4188 since a callout by the remote host to check the sender would use <>.
4189 [SEE ABOVE: changed after hitting problems.]
4191 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4192 temporary errors. However, in the case of such a code being given after
4193 the end of a data transmission (i.e. after ".") Exim was failing to write
4194 a retry record for the message. (Yes, there was some broken host that was
4195 actually sending 8xx at this point.)
4197 73. An unknown lookup type in a host list could cause Exim to panic-die when
4198 the list was checked. (An example that provoked this was putting <; in the
4199 middle of a list instead of at the start.) If this happened during a DATA
4200 ACL check, a -D file could be left lying around. This kind of configuration
4201 error no longer causes Exim to die; instead it causes a defer errror. The
4202 incident is still logged to the main and panic logs.
4204 74. Buglet left over from Exim 3 conversion. The message "too many messages
4205 in one connection" was written to the rejectlog but not the mainlog, except
4206 when address rewriting (yes!) was being logged.
4208 75. Added write_rejectlog option.
4210 76. When a system filter was run not as root (that is, when system_filter_user
4211 was set), the values of the $n variables were not being returned to the
4212 main process; thus, they were not subsequently available in the $sn
4215 77. Added +return_path_on_delivery log selector.
4217 78. A connection timeout was being treated differently from recipients deferred
4218 when testing hosts_max_try with a message that was older than the host's
4219 retry timeout. (The host should not be counted, thus allowing all hosts to
4220 be tried at least once before bouncing.) This may have been the cause of an
4221 occasionally reported bug whereby a message would remain on the queue
4222 longer than the retry timeout, but would be bounced if a delivery was
4223 forced. I say "may" because I never totally pinned down the problem;
4224 setting up timeout/retry tests is difficult. See also the next item.
4226 79. The ultimate address timeout was not being applied to errors that involved
4227 a combination of host plus message (for example, a timeout on a MAIL
4228 command). When an address resolved to a number of possible hosts, and they
4229 were not all tried for each delivery (e.g. because of hosts_max_try), a
4230 message could remain on the queue longer than the retry timeout.
4232 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4235 81. Fixed an obscure SMTP outgoing bug which required at least the following
4236 conditions: (a) there was another message waiting for the same server;
4237 (b) the server returned 5xx to all RCPT commands in the first message so
4238 that the message was not completed; (c) the server dropped the connection
4239 or gave a negative response to the RSET that Exim sends to abort the
4240 transaction. The observed case was a dropped connection after DATA that had
4241 been sent in pipelining mode. That is, the server had advertised PIPELINING
4242 but was not implementing it correctly. The effect of the bug was incorrect
4243 behaviour, such as trying another host, and this could lead to a crash.
4249 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4250 and daemon.c were passed as pointers to ints; they should have been
4251 pointers to socklen_t variables (which are typically unsigned ints).
4253 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4256 3. Fixed a really odd bug that affected only the testing scheme; patching a
4257 certain fixed string in the binary changed the value of another string that
4258 happened to be identical to the end of the original first string.
4260 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4261 name", it returns that address as the IP address. On some operating
4262 systems (e.g. Solaris), it also passes back the IP address string as the
4263 "host name". However, on others (e.g. Linux), it passes back an empty
4264 string. Exim wasn't checking for this, and was changing the host name to an
4265 empty string, assuming it had been canonicized.
4267 5. Although rare, it is permitted to have more than one PTR record for a given
4268 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4269 all the names associated with an address, because they do in Solaris.
4270 However, it seems that they do not in Linux for data that comes from the
4271 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4272 I found this out when I moved to a new Linux workstation and tried to run
4273 the Exim test suite.
4275 To get round this problem I have changed the code so that it now does its
4276 own call to the DNS to look up PTR records when searching for a host name.
4277 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4278 addresses that are only in /etc/hosts are still found.
4280 This behaviour is, however, controlled by an option called host_lookup_
4281 order, which defaults to "bydns:byaddr". If people want to use the other
4282 order, or indeed, just use one or the other means of lookup, they can
4283 specify it in this variable.
4285 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4286 some operating systems, this comes back from gethostbyaddr() as an empty
4287 string, and this is what Exim used to test for. However, it seems that in
4288 other systems, "." is yielded. Exim now tests for this case too.
4290 7. The values of check_spool_space and check_log_space are now held internally
4291 as a number of kilobytes instead of an absolute number of bytes. If a
4292 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4293 kilobyte. This means that much larger values can be stored.
4295 8. Exim monitor: an attempt to get the action menu when not actually pointing
4296 at a message produces an empty menu entitled "No message selected". This
4297 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4298 no entries in it ("Shell widget menu has zero width and/or height"). So I
4299 have added a single, blank menu entry in this case.
4301 9. Added ${quote_local_part.
4303 10. MIME decoding is now applied to the contents of Subject: header lines when
4306 11. Now that a reference to $sender_host_address automatically causes a reverse
4307 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4308 host lookup before query-style lookups in lists that might use this
4309 variable. This has therefore been abolished, and the "net-" prefix is no
4310 longer necessary for query-style lookups.
4312 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4313 have been a typo for LFLAGS, so it has been changed.
4315 13. The install script calls Exim with "-C /dev/null" in order to find the
4316 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4317 to be output. Howeve, since Exim outputs its version number before the
4318 error, it didn't break the script. It just looked ugly. I fixed this by
4319 always allowing "-C /dev/null" if the caller is root.
4321 14. Ignore overlarge ACL variable number when reading spool file - insurance
4322 against a later release with more variables having written the file.
4324 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4325 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4326 this, as well as the form without the "IPv6" on the front.
4328 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4329 OS/Makefile-Darwin file.
4331 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4332 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4334 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4335 variable. However, some OS use ints for inodes. I've added cast to long int
4336 to get rid of the compiler warning.
4338 19. I had forgotten to lock out "/../" in configuration file names when
4339 ALT_CONFIG_PREFIX was set.
4341 20. Routers used for verification do not need to specify transports. However,
4342 if such a router generated a host list, and callout was configured, Exim
4343 crashed, because it could not find a port number from the (non-existent)
4344 transport. It now assumes port 25 in this circumstance.
4346 21. Added the -t option to exigrep.
4348 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4349 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4350 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4351 them is set, LOOKUP_LSEARCH is forced.
4353 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4355 24. Added sender and host information to the "rejected by local_scan()" log
4356 line; previously there was no indication of these.
4358 25. Added .include_if_exists.
4360 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4361 Linux. It turns out that not all file systems support this. Apparently some
4362 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4363 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4364 not supported on the file descriptor, is now ignored when Exim is trying to
4365 sync a directory. This applies only to Linux.
4367 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4369 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4370 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4373 29. Added support for maildirsize files from supplied patch (modified a bit).
4375 30. The use of :fail: followed by an empty string could lead Exim to respond to
4376 sender verification failures with (e.g.):
4378 550 Verification failed for <xxx>
4379 550 Sender verify failed
4381 where the first response line was missing the '-' that indicates it is not
4382 the final line of the response.
4384 31. The loop for finding the name of the user that called Exim had a hardwired
4385 limit of 10; it now uses the value of finduser_retries, which is used for
4386 all other user lookups.
4388 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4391 33. Exim was neglecting to zero errno before one call of strtol() when
4392 expanding a string and expecting an integer value. On some systems this
4393 resulted in spurious "integer overflow" errors. Also, it was casting the
4394 result into an int without checking.
4396 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4397 did not work. The code looks as if it has *never* worked, though it appears
4398 to have been documented since at least releast 1.62. I have made it work.
4400 35. The "timeout_DNS" error in retry rules, also documented since at least
4401 1.62, also never worked. As it isn't clear exactly what this means, and
4402 clearly it isn't a major issue, I have abolished the feature by treating it
4403 as "timeout", and writing a warning to the main and panic logs.
4405 36. The display of retry rules for -brt wasn't always showing the error code
4408 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4409 timeout_connect_A, timeout_connect_MX.
4411 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4412 to the empty sender.
4414 39. The daemon was not analysing the content of -oX till after it had closed
4415 stderr and disconnected from the controlling terminal. This meant that any
4416 syntax errors were only noted on the panic log, and the return code from
4417 the command was 0. By re-arranging the code a little, I've made the
4418 decoding happen first, so such errors now appear on stderr, and the return
4419 code is 1. However, the actual setting up of the sockets still happens in
4420 the disconnected process, so errors there are still only recorded on the
4423 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4424 connections (as happens on some IP stacks) was logged at start up time as
4425 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4426 it from "IPv6 and IPv4", which means that two separate sockets are being
4429 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4430 says whether AF_INET or AF_INET6 was passed as an argument.
4432 42. Exiwhat output was messed up when time zones were included in log
4435 43. Exiwhat now gives more information about the daemon's listening ports,
4436 and whether -tls-on-connect was used.
4438 44. The "port" option of the smtp transport is now expanded.
4440 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4441 silently ignored. Now an error message is written to the main and panic
4444 46. There's a new ACL modifier called "logwrite" which writes to a log file
4445 as soon as it is encountered.
4447 47. Added $local_user_uid and $local_user_gid at routing time.
4449 48. Exim crashed when trying to verify a sender address that was being
4452 49. Exim was recognizing only a space character after ".include". It now also
4453 recognizes a tab character.
4455 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4456 extracting the relevant information from the specification. The man page no
4457 longer contains scrambled data for the -d option, and I've added a section
4458 at the front about calling Exim under different names.
4460 51. Added "extra_headers" argument to the "mail" command in filter files.
4462 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4465 53. Installed eximstats 1.29.
4467 54. Added transport_filter_timeout as a generic transport option.
4469 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4470 Cc: header lines. This was required by RFC 822, but it not required by RFC
4473 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4474 incoming messages that did not have them. Now it does so only if the
4475 message originates locally, that is, if there is no associated remote host
4476 address. When Resent- header lines are present, this applies to the Resent-
4477 lines rather than the non-Resent- lines.
4479 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4480 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4482 58. Messages for configuration errors now include the name of the main
4483 configuration file - useful now that there may be more than one file in a
4484 list (.included file names were always shown).
4486 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4487 for those rare installations that do not start the daemon as root or run it
4488 setuid root. I've cut out the call to initgroups() if the daemon is not
4491 60. The Exim user and group can now be bound into the binary as text strings
4492 that are looked up at the start of Exim's processing.
4494 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4496 62. Added $mailstore_basename variable.
4498 63. Installed patch to sieve.c from Michael Haardt.
4500 64. When Exim failed to open the panic log after failing to open the main log,
4501 the original message it was trying to log was written to stderr and debug
4502 output, but if they were not available (the usual case in production), it
4503 was lost. Now it is written to syslog before the two lines that record the
4504 failures to open the logs.
4506 65. Users' Exim filters run in subprocesses under the user's uid. It is
4507 possible for a "deliver" command or an alias in a "personal" command to
4508 provoke an address rewrite. If logging of address rewriting is configured,
4509 this fails because the process is not running as root or exim. There may be
4510 a better way of dealing with this, but for the moment (because 4.30 needs
4511 to be released), I have disabled address rewrite logging when running a
4512 filter in a non-root, non-exim process.
4518 1. The buildconfig auxiliary program wasn't quoting the value set for
4519 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4520 not defined. This bug was masked in 4.22 by the effect that was fixed in
4523 2. Some messages that were rejected after a message id was allocated were
4524 shown as "incomplete" by exigrep. It no longer does this for messages that
4525 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4527 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4528 have allow_domain_literals set, the ID did not get logged in the <= line.
4529 Domain literals are now always recognized in Message-ID: header lines.
4531 4. The first argument for a ${extract expansion item is the key name or field
4532 number. Leading and trailing spaces in this item were not being ignored,
4533 causing some misleading effects.
4535 5. When deliver_drop_privilege was set, single queue runner processes started
4536 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4537 same command in the process it spins off) were not dropping privilege.
4539 6. When the daemon running as "exim" started a queue runner, it always
4540 re-executed Exim in the spun-off process. This is a waste of effort when
4541 deliver_drop_privilege is set. The new process now just calls the
4542 queue-runner function directly.
4548 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4551 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4552 ignored. Though the use of -oP was forcing the writing of a pid file, it
4553 was always written to the default place.
4555 3. If the message "no IP address found for host xxxx" is generated during
4556 incoming verification, it is now followed by identification of the incoming
4557 connection (so you can more easily find what provoked it).
4559 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4561 5. Added some features to "harden" Exim a bit more against certain attacks:
4563 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4564 be put in Local/Makefile. This is like the never_users runtime option,
4565 but it cannot be overridden. The default setting is "root".
4567 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4568 prefix string with which any file named in a -C command line option
4571 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4572 is retained for -C and -D only if the caller of Exim is root. Without
4573 it, the exim user may also use -C and -D and retain privilege.
4575 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4576 command line option is disabled.
4578 6. Macro names set by the -D option must start with an upper case letter, just
4579 like macro names defined in the configuration file.
4581 7. Added "dereference=" facility to LDAP.
4583 8. Two instances of the typo "uknown" in the source files are fixed.
4585 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4586 the Configure-Makefile script screwed up while processing it.
4588 10. Incorporated PCRE 4.4.
4590 11. The SMTP synchronization check was not operating right at the start of an
4591 SMTP session. For example, it could not catch a HELO sent before the client
4592 waited for the greeting. There is now a check for outstanding input at the
4593 point when the greeting is written. Because of the duplex, asynchronous
4594 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4595 way, but not yet received, when the check is performed.
4597 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4598 on TCP/IP sockets, because this apparently causes some broken clients to
4601 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4602 unchanged) from the Cygwin maintainer.
4604 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4605 was supported instead of testing for mbx. Effectively a typo.
4607 15. The spa authenticator server code was not checking that the input it
4608 received was valid base64.
4610 16. The debug output line for the "set" modifier in ACLs was not showing the
4611 name of the variable that was being set.
4613 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4615 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4616 lookup on its own. Something else was needed to trigger the lookup. For
4617 example, a match in host_lookup or the need for a host name in a host list.
4618 Now, if $sender_host_name is referenced and the host name has not yet been
4619 looked up, a lookup is performed. If the lookup fails, the variable remains
4620 empty, and $host_lookup_failed is set to "1".
4622 19. Added "eqi" as a case-independent comparison operator.
4624 20. The saslauthd authentication condition could segfault if neither service
4625 nor realm was specified.
4627 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4628 error message that was logged was misleading, and incoming SMTP
4629 connections were dropped. The message is now more accurate, and temporary
4630 errors are given to SMTP connections.
4632 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4633 (except RSET) with a 421 error, until QUIT is received. However, it was
4634 failing to send a response to QUIT.
4636 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4637 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4638 ACL is now run after the helo_try_verify_hosts code.
4640 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4641 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4642 case-independent, so other case variants are also recognized). Apparently
4643 some systems use these upper case variants.
4645 25. If more than two messages were waiting for the same host, and a transport
4646 filter was specified for the transport, Exim sent two messages over the
4647 same TCP/IP connection, and then failed with "socket operation on non-
4648 socket" when it tried to send the third.
4650 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4652 27. The extern definition of crypt16() in expand.c was not being excluded when
4653 the OS had its own crypt16() function.
4655 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4656 as a preferred synonym for return_size_limit, both as an option and as an
4659 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4661 30. Changed the default configuration ACL to relax the local part checking rule
4662 for addresses that are not in any local domains. For these addresses,
4663 slashes and pipe symbols are allowed within local parts, but the sequence
4664 /../ is explicitly forbidden.
4666 31. SPA server authentication was not clearing the challenge buffer before
4669 32. log_message in a "warn" ACL statement was writing to the reject log as
4670 well as to the main log, which contradicts the documentation and doesn't
4671 seem right (because no rejection is happening). So I have stopped it.
4673 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4674 However, I am unable to do any testing of this.
4676 34. Fixed an infelicity in the appendfile transport. When checking directories
4677 for a mailbox, to see if any needed to be created, it was accidentally
4678 using path names with one or more superfluous leading slashes; tracing
4679 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4681 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4682 contents are added to the log line that is written for every discarded
4683 recipient. (Previously a log_message setting was ignored.)
4685 36. The ${quote: operator now quotes the string if it is empty.
4687 37. The install script runs exim in order to find its version number. If for
4688 some reason other than non-existence or emptiness, which it checks, it
4689 could not run './exim', it was installing it with an empty version number,
4690 i.e. as "exim-". This error state is now caught, and the installation is
4693 38. An argument was missing from the function that creates an error message
4694 when Exim fails to connect to the socket for saslauthd authentication.
4695 This could cause Exim to crash, or give a corrupted message.
4697 39. Added isip, isip4, and isip6 to ${if conditions.
4699 40. The ACL variables $acl_xx are now saved with the message, and can be
4700 accessed later in routers, transports, and filters.
4702 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4703 strings in the file are not string-expanded.
4705 42. If a MAIL command specified a SIZE value that was too large to fit into an
4706 int variable, the check against message_size_limit failed. Such values are
4707 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4708 day this will have to be increased, but I don't think I want to be around
4709 when emails are that large.
4716 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4717 iconv() is not standard in FreeBSD.
4719 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4720 IPv6 enabled. The observed symptom was a segmentation fault on return from
4721 the function os_common_find_running_interfaces() in src/os.c.
4723 3. In the check_special_case() function in daemon.c I had used "errno" as an
4724 argument name, which causes warnings on some systems. This was basically a
4725 typo, since it was named "eno" in the comments!
4727 4. The code that waits for the clock to tick (at a resolution of some fraction
4728 of a second) so as to ensure message-id uniqueness was always waiting for
4729 at least one whole tick, when it could have waited for less. [This is
4730 almost certainly not relevant at current processor speeds, where it is
4731 unlikely to ever wait at all. But we try to future-proof.]
4733 5. The function that sleeps for a time interval that includes fractions of a
4734 second contained a race. It did not block SIGALRM between setting the
4735 timer, and suspending (a couple of lines later). If the interval was short
4736 and the sigsuspend() was delayed until after it had expired, the suspension
4737 never ended. On busy systems this could lead to processes getting stuck for
4740 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4741 process, before it forks any delivery processes. The open lookup caching
4742 mechanism meant that the open file or database connection was passed into
4743 the delivery process. The problem was that delivery processes always tidy
4744 up cached lookup data. This could cause a problem for the next delivery
4745 process started by the queue runner, because the external queue runner
4746 process does not know about the closure. So the next delivery process
4747 still has data in the lookup cache. In the case of a file lookup, there was
4748 no problem because closing a file descriptor in a subprocess doesn't affect
4749 the parent. However, if the lookup was caching a connection to a database,
4750 the connection was closed, and the second delivery process was likely to
4751 see errors such as "PGSQL: query failed: server closed the connection
4752 unexpectedly". The problem has been fixed by closing all cached lookups
4753 in a queue runner before running a delivery process.
4755 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4756 seem to have the "const" qualifier which it has on other OS. I've
4759 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4760 *of the same type* (client or server) with the same public name that an
4761 error should be diagnosed.
4763 9. When Exim looked up a host name for an IP address, but failed to find the
4764 original IP address when looking up the host name (a safety check), it
4765 output the message "<ip address> does not match any IP for NULL", which was
4766 confusing, to say the least. The bug was that the host name should have
4767 appeared instead of "NULL".
4769 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4770 user that is built into the binary, and the -C or -D options is used, root
4771 privilege is dropped before the configuration file is read. In addition,
4772 logging is switched to stderr instead of the normal log files. If the
4773 configuration then re-defines the Exim user, the unprivileged environment
4774 is probably not what is expected, so Exim logs a panic warning message (but
4777 However, if deliver_drop_privilege is set, the unprivileged state may well
4778 be exactly what is intended, so the warning has been cut out in that case,
4779 and Exim is allowed to try to write to its normal log files.
4785 1. smtp_return_error_details was not giving details for temporary sender
4786 or receiver verification errors.
4788 2. Diagnose a configuration error if two authenticators have the same public
4791 3. Exim used not to create the message log file for a message until the first
4792 delivery attempt. This could be confusing when incoming messages were held
4793 for policy or load reasons. The message log file is now created at the time
4794 the message is received, and an initial "Received" line is written to it.
4796 4. The automatically generated man page for command line options had a minor
4797 bug that caused no ill effects; however, a more serious problem was that
4798 the procedure for building the man page automatically didn't always
4799 operate. Consequently, release 4.20 contains an out-of-date version. This
4800 shouldn't happen again.
4802 5. When building Exim with embedded Perl support, the script that builds the
4803 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4804 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4806 6. The freeze_tell option was not being used for messages that were frozen on
4807 arrival, either by an ACL or by local_scan().
4809 7. Added the smtp_incomplete_transaction log selector.
4811 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4812 was accepting AUTH without a new EHLO.
4814 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4815 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4818 10. Logging of TCP/IP connections (when configured) now happens in the main
4819 daemon process instead of the child process, so that the TCP/IP connection
4820 count is more accurate (but it can never be perfect).
4822 11. The use of "drop" in a nested ACL was not being handled correctly in the
4823 outer ACL. Now, if condition failure induced by the nested "drop" causes
4824 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4825 or "require"), the connection is dropped.
4827 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4828 that yield "discard" can now be used with an "accept" or a "discard" verb,
4829 but an error is generated for any others (because I can't see a useful way
4830 to define what should happen).
4832 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4833 are now processed in the same way as lines in the Exim configuration file.
4834 In particular, continuation lines are supported.
4836 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4838 15. Added -ti meaning -t -i.
4840 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4841 domains, and warn by logging if others are found.
4843 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4844 returned from GIFCONF when Exim is trying to find the list of interfaces on
4845 a host. The code in os.c has been modified to copy each ifreq to an aligned
4846 structure in all cases.
4848 Also, in some cases, the returned ifreq's were being copied to a 'struct
4849 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4850 means the last couple of bytes of an IPv6 address could be chopped if the
4851 ifreq contained only a normal sockaddr (14 bytes storage).
4853 18. Named domain lists were not supported in the hosts_treat_as_local option.
4854 An entry such as +xxxx was not recognized, and was treated as a literal
4857 19. Ensure that header lines added by a DATA ACL are included in the reject log
4858 if the ACL subsequently rejects the message.
4860 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4861 MD5 (which is deprecated).
4863 21. When testing a filter file using -bf, Exim was writing a message when it
4864 took the sender from a "From " line in the message, but it was not doing so
4865 when it took $return_path from a Return-Path: header line. It now does.
4867 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4868 with a valid header line field name (a series of printing characters
4869 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4871 23. Changed "disc" in the source to "disk" to conform to the documentation and
4872 the book and for uniformity.
4874 24. Ignore Sendmail's -Ooption=value command line item.
4876 25. When execve() failed while trying to run a command in a pipe transport,
4877 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4878 could be confused with a return value of 69 from the command itself. This
4879 has been changed to 127, the value the shell returns if it is asked to run
4880 a non-existent command. The wording for the related log line suggests a
4881 non-existent command as the problem.
4883 26. If received_header_text expands to an empty string, do not add a Received:
4884 header line to the message. (Well, it adds a token one on the spool, but
4885 marks it "old" so that it doesn't get used or transmitted.)
4887 27. Installed eximstats 1.28 (addition of -nt option).
4889 28. There was no check for failure on the call to getsockname() in the daemon
4890 code. This can fail if there is a shortage of resources on the system, with
4891 ENOMEM, for example. A temporary error is now given on failure.
4893 29. Contrary to the C standard, it seems that in some environments, the
4894 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4895 program. Exim now does this explicitly; it affects the formatting of
4896 timestamps using strftime().
4898 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4899 complaints. I've now initialized all the variables, to avoid this.
4901 32. Header lines added by a system filter were not being "seen" during
4902 transport-time rewrites.
4904 33. The info_callback() function passed to OpenSSL is set up with type void
4905 (*)(SSL *, int, int), as described somewhere. However, when calling the
4906 function (actually a macro) that sets it up, the type void(*)() is
4907 expected. I've put in a cast to prevent warnings from picky compilers.
4909 34. If a DNS black list lookup found a CNAME record, but there were no A
4910 records associated with the domain it pointed at, Exim crashed.
4912 35. If a DNS black list lookup returned more than one A record, Exim ignored
4913 all but the first. It now scans all returned addresses if a particular IP
4914 value is being sought. In this situation, the contents of the
4915 $dnslist_value variable are a list of all the addresses, separated by a
4918 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4919 to accept a host name and all its aliases if the forward lookup for any of
4920 them yielded the IP address of the incoming connection. Now it accepts only
4921 those names whose forward lookup yields the correct IP address. Any other
4922 names are discarded. This closes a loophole whereby a rogue DNS
4923 administrator could create reverse DNS records to break through a
4924 wildcarded host restriction in an ACL.
4926 37. If a user filter or a system filter that ran in a subprocess used any of
4927 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4928 the wrong values were passed to the pipe command ($thisaddress had the
4929 value of $0, $0 had the value of $1, etc). This bug was introduced by
4930 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4932 38. Improved the line breaking for long SMTP error messages from ACLs.
4933 Previously, if there was no break point between 40 and 75 characters, Exim
4934 left the rest of the message alone. Two changes have been made: (a) I've
4935 reduced the minimum length to 35 characters; (b) if it can't find a break
4936 point between 35 and 75 characters, it looks ahead and uses the first one
4937 that it finds. This may give the occasional overlong line, but at least the
4938 remaining text gets split now.
4940 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4941 file descriptors might be low, and that setting 1000 would always raise it.
4942 It turns out that in some environments, the limit is already over 1000 and
4943 that lowering it causes trouble. So now Exim takes care not to decrease it.
4945 40. When delivering a message, the value of $return_path is set to $sender_
4946 address at the start of routing (routers may change the value). By an
4947 oversight, this default was not being set up when an address was tested by
4948 -bt or -bv, which affected the outcome if any router or filter referred to
4951 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4952 log only when "message" or "log_message" are set. However, if one of the
4953 conditions was an address verification, or a call to a nested ACL, the
4954 messages generated by the underlying test were being passed through. This
4955 no longer happens. The underlying message is available in $acl_verify_
4956 message for both "message" and "log_message" expansions, so it can be
4957 passed through if needed.
4959 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4960 new expansion $bh_ to give the encoded byte string without charset
4961 translation. Translation happens only if iconv() is available; HAVE_ICONV
4962 indicates this at build time. HEADERS_CHARSET gives the charset to
4963 translate to; headers_charset can change it in the configuration, and
4964 "headers charset" can change it in an individual filter file.
4966 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4967 that creates RFC 2047 encoded "words" labels them as that charset instead
4968 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4969 expansion operator; (ii) when Exim creates a From: line for a local
4970 message; (iii) when a header line is rewritten to include a "phrase" part.
4972 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4973 buggy, causing it to skip the first lines of messages whose message ID
4974 ended in 'D'. This would not have bitten before Exim release 4.14, because
4975 message IDs were unlikely to end in 'D' before then. The effect was to have
4976 incorrect size information for certain domains.
4978 45. #include "config.h" was missing at the start of the crypt16.c module. This
4979 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4982 46. If there was a timeout during a "random" callout check, Exim treated it as
4983 a failure of the random address, and carried on sending RSET and the real
4984 address. If the delay was just some slowness somewhere, the response to the
4985 original RCPT would be taken as a response to RSET and so on, causing
4986 mayhem of various kinds.
4988 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4989 when I implemented it. It didn't allow for the fact that some option values
4990 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4991 the right test for positive values.
4993 48. Domain names in DNS records are case-independent. Exim always looks them up
4994 in lower case. Some resolvers return domain names in exactly the case they
4995 appear in the zone file, that is, they may contain uppercase letters. Not
4996 all resolvers do this - some return always lower case. Exim was treating a
4997 change of case by a resolver as a change of domain, similar to a widening
4998 of a domain abbreviation. This triggered its re-routing code and so it was
4999 trying to route what was effectively the same domain again. This normally
5000 caused routing to fail (because the router wouldn't handle the domain
5001 twice). Now Exim checks for this case specially, and just changes the
5002 casing of the domain that it ultimately uses when it transmits the message
5005 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5008 50. If a filter generated a file delivery with a non-absolute name (possible if
5009 no home directory exists for the router), the forbid_file option was not
5012 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5013 the existing equality matching.
5015 52. Exim was using ints instead of ino_t variables in some places where it was
5016 dealing with inode numbers.
5018 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5019 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5020 called TMPDIR, and if it finds it is different, it changes its value.
5022 54. The smtp_printf() function is now made available to local_scan() so
5023 additional output lines can be written before returning. There is also an
5024 smtp_fflush() function to enable the detection of a dropped connection.
5025 The variables smtp_input and smtp_batched_input are exported to
5028 55. Changed the default runtime configuration: the message "Unknown user"
5029 has been removed from the ACL, and instead placed on the localuser router,
5030 using the cannot_route_message feature. This means that any verification
5031 failures that generate their own messages won't get overridden. Similarly,
5032 the "Unrouteable address" message that was in the ACL for unverifiable
5033 relay addresses has also been removed.
5035 56. Added hosts_avoid_esmtp to the smtp transport.
5037 57. The exicyclog script was not checking for the esoteric option
5038 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5039 will work only if exicyclog is run under the appropriate euid.
5041 58. Following a discussion on the list, the rules by which Exim recognises line
5042 endings on incoming messages have been changed. The -dropcr and drop_cr
5043 options are now no-ops, retained only for backwards compatibility. The
5044 following line terminators are recognized: LF CRLF CR. However, special
5045 processing applies to CR:
5047 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5048 nor a local message in the state where . is a terminator.
5050 (ii) If a bare CR is encountered in a header line, an extra space is added
5051 after the line terminator so as not to end the header. The reasoning
5052 behind this is that bare CRs in header lines are most likely either
5053 to be mistakes, or people trying to play silly games.
5055 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5056 was being incorrectly given as 18 bytes larger than it should have been.
5057 This is a VOB (very old bug).
5059 60. This may never have affected anything current, but just in case it has:
5060 When the local host is found other than at the start of a list of hosts,
5061 the local host, those with the same MX, and any that follow, are discarded.
5062 When the list in question was part of a longer list of hosts, the following
5063 hosts (not currently being processed) were also being discarded. This no
5064 longer happens. I'm not sure if this situation could ever has previously
5067 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5070 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5071 header. This is something that is recommended in a new Internet Draft, and
5072 is something that is documented as being done by Sendmail. There are two
5073 possible values. For messages generated by the autoreply transport, Exim
5076 Auto-Submitted: auto-replied
5078 whereas for all other generated messages (e.g. bounces) it adds
5080 Auto-Submitted: auto-generated
5082 63. The "personal" condition in filters now includes a test for the
5083 Auto-Submitted: header. If it contains the string "auto-" the message it
5084 not considered personal.
5086 64. Added rcpt_include_affixes as a generic transport option.
5088 65. Added queue_only_override (default true).
5090 66. Added the syslog_duplication option.
5092 67. If what should have been the first header line of a message consisted of
5093 a space followed by a colon, Exim was mis-interpreting it as a header line.
5094 It isn't of course - it is syntactically invalid and should therefore be
5095 treated as the start of the message body. The misbehaviour could have
5096 caused a number of strange effects, including loss of data in subsequent
5097 header lines, and spool format errors.
5099 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5100 client host had authenticated. This control can now be exercised by an ACL
5101 for more flexibility.
5103 69. By default, callouts do not happen when testing with -bh. There is now a
5104 variant, -bhc, which does actually run the callout code, including
5105 consulting and updating the callout cache.
5107 70. Added support for saslauthd authentication, courtesy of Alexander
5110 71. If statvfs() failed on the spool or log directories while checking their
5111 size for availability, Exim confusingly gave the error "space shortage".
5112 Furthermore, in debugging mode it crashed with a floating point exception.
5113 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5114 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5115 serious problem, Exim now writes to the main and panic logs when this
5116 happens, with details of the failure. It then refuses to accept the
5117 incoming message, giving the message "spool directory problem" or "log
5118 directory problem" with a 421 code for SMTP messages.
5120 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5121 0, 1, and 2 exist, because some OS complain for execs without them (see
5122 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5123 descriptors. However, the code omitted to check that the open succeeded,
5124 causing mysterious errors if for some reason the permissions on /dev/null
5125 got screwed. Now Exim writes a message to the main and panic logs, and
5126 bombs out if it can't open /dev/null.
5128 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5129 interact so that it is all more flexible. It is supposed to remain
5130 backwards compatible. Also added extra_local_interfaces.
5132 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5133 to bomb out with an assertion failure - to the client this appears as a
5134 connection drop. This problem occurs in the part of the code that was taken
5135 from the Samba project. Fortunately, the assertion is in a very simple
5136 function, so I have fixed this by reproducing the function inline in the
5137 one place where it is called, and arranging for authentication to fail
5138 instead of killing the process with assert().
5140 75. The SPA client code was not working when the server requested OEM rather
5141 than Unicode encoding.
5143 76. Added code to make require_files with a specific uid setting more usable in
5144 the case where statting the file as root fails - usually a non-root-mounted
5145 NFS file system. When this happens and the failure is EACCES, Exim now
5146 forks a subprocess and does the per-uid checking as the relevant uid.
5148 77. Added process_log_path.
5150 78. If log_file_path was not explicitly set, a setting of check_log_space or
5151 check_log_inodes was ignored.
5153 79. If a space check for the spool or log partitions fails, the incident is now
5154 logged. Of course, in the latter case the data may get lost...
5156 80. Added the %p formatting code to string_format() so that it can be used to
5157 print addresses in debug_print(). Adjusted all the address printing in the
5158 debugging in store.c to use %p rather than %d.
5160 81. There was a concern that a line of code in smtp_in.c could overflow a
5161 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5162 initially expressed, the concern was not well-founded, because trailing
5163 spaces are removed early. However, if the trailing spaces were followed by
5164 a NULL, they did not get removed, so the overflow was possible. Two fixes
5167 (a) I re-wrote the offending code in a cleaner fashion.
5168 (b) If an incoming SMTP command contains a NULL character, it is rejected
5171 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5172 runs initgroups(), so that if the Exim user is in any additional groups,
5173 they will be used during message reception.
5179 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob