1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
74 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
77 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
78 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
79 configuration file which is writeable by the Exim user or group.
81 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
82 of configuration files to cover files specified with the -C option if
83 they are going to be used with root privileges, not just the default
86 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
87 option (effectively making it always true).
89 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
90 files to be used while preserving root privileges.
92 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
93 that rogue child processes cannot use them.
95 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
96 run-time user, instead of root.
98 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
99 Exim run-time user without dropping privileges.
101 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
102 result string, instead of calling string_vformat() twice with the same
105 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
106 for other users. Others should always drop root privileges if they use
107 -C on the command line, even for a whitelisted configure file.
113 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
114 $data_path, and $header_path variables; fixed documentation bugs and
117 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
118 exipick to access non-standard spools, including the "frozen" queue
121 NM/01 Bugzilla 965: Support mysql stored procedures.
122 Patch from Alain Williams
124 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
126 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
127 Patch from Andreas Metzler
129 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
130 Patch from Kirill Miazine
132 NM/05 Bugzilla 671: Added umask to procmail example.
134 JJ/03 installed exipick 20100323.0, fixing doc bug
136 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
137 directory. Notification and patch from Dan Rosenberg.
139 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
141 TK/02 Improve log output when DKIM signing operation fails.
143 MH/01 Treat the transport option dkim_domain as a colon separated
144 list, not as a single string, and sign the message with each element,
145 omitting multiple occurences of the same signer.
147 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
148 Bugzilla 985, 986. Patch by Simon Arlott
150 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
151 Patch by Simon Arlott
153 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
154 MBX locking. Notification from Dan Rosenberg.
160 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
162 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
164 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
166 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
168 NM/04 Bugzilla 915: Flag broken perl installation during build.
174 TK/01 Added patch by Johannes Berg that expands the main option
175 "spamd_address" if it starts with a dollar sign.
177 TK/02 Write list of recipients to X-Envelope-Sender header when building
178 the mbox-format spool file for content scanning (suggested by Jakob
181 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
182 (http://www.dcc-servers.net/) support via dccifd. Activated by
183 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
185 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
186 by Mark Daniel Reidel <mr@df.eu>.
188 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
189 When building exim an external PCRE library is now needed -
190 PCRE is a system library on the majority of modern systems.
191 See entry on PCRE_LIBS in EDITME file.
193 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
194 conversation. Added nologin parameter to request.
195 Patch contributed by Kirill Miazine.
197 TF/01 Do not log submission mode rewrites if they do not change the address.
199 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
201 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
202 log files in place. Contributed by Roberto Lima.
204 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
206 TF/03 Bugzilla 615: When checking the local_parts router precondition
207 after a local_part_suffix or local_part_prefix option, Exim now
208 does not use the address's named list lookup cache, since this
209 contains cached lookups for the whole local part.
211 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
212 Robert Millan. Documentation is in experimental-spec.txt.
214 TF/04 Bugzilla 668: Fix parallel build (make -j).
216 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
218 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
219 Patch provided by Jan Srzednicki.
221 TF/05 Leading white space used to be stripped from $spam_report which
222 wrecked the formatting. Now it is preserved.
224 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
225 that they are available at delivery time.
227 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
229 TF/08 TLS error reporting now respects the incoming_interface and
230 incoming_port log selectors.
232 TF/09 Produce a more useful error message if an SMTP transport's hosts
233 setting expands to an empty string.
235 NM/06 Bugzilla 744: EXPN did not work under TLS.
236 Patch provided by Phil Pennock.
238 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
239 Patch provided by Richard Godbee.
241 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
242 acl_smtp_notquit, added index entry.
244 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
245 Patch provided by Eugene Bujak.
247 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
248 accept(). Patch provided by Maxim Dounin.
250 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
251 Patch provided by Phil Pennock.
253 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
255 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
256 Patch provided by Brad "anomie" Jorsch.
258 NM/14 Bugzilla 622: Added timeout setting to transport filter.
259 Patch provided by Dean Brooks.
261 TK/05 Add native DKIM support (does not depend on external libraries).
263 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
264 Patch provided by Graeme Fowler.
266 NM/16 Bugzilla 851: Documentation example syntax fix.
268 NM/17 Changed NOTICE file to remove references to embedded PCRE.
270 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
273 NM/19 Bugzilla 745: TLS version reporting.
274 Patch provided by Phil Pennock.
276 NM/20 Bugzilla 167: bool: condition support.
277 Patch provided by Phil Pennock.
279 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
280 clients. Patch provided by Phil Pennock.
282 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
283 Patch provided by Brad "anomie" Jorsch.
285 NM/23 Bugzilla 687: Fix misparses in eximstats.
286 Patch provided by Heiko Schlittermann.
288 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
289 Patch provided by Heiko Schlittermann.
291 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
292 plus update to original patch.
294 NM/26 Bugzilla 799: Documentation correction for ratelimit.
296 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
297 Patch provided by David Brownlee.
299 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
301 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
303 NM/30 Bugzilla 888: TLS documentation bugfixes.
305 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
307 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
308 Unlike the original bugzilla I have changed all shell scripts in src tree.
310 NM/33 Bugzilla 898: Transport filter timeout fix.
311 Patch by Todd Rinaldo.
313 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
314 Patch by Serge Demonchaux.
316 NM/35 Bugzilla 39: Base64 decode bug fixes.
317 Patch by Jakob Hirsch.
319 NM/36 Bugzilla 909: Correct connect() call in dcc code.
321 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
323 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
325 NM/39 Bugzilla 911: Fixed MakeLinks build script.
331 TK/01 Add preliminary DKIM support. Currently requires a forked version of
332 ALT-N's libdkim that I have put here:
333 http://duncanthrax.net/exim-experimental/
335 Note to Michael Haardt: I had to rename some vars in sieve.c. They
336 were called 'true' and it seems that C99 defines that as a reserved
337 keyword to be used with 'bool' variable types. That means you could
338 not include C99-style headers which use bools without triggering
339 build errors in sieve.c.
341 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
342 as mailq or other aliases. Changed the --help handling significantly
343 to do whats expected. exim_usage() emits usage/help information.
345 SC/01 Added the -bylocaldomain option to eximstats.
347 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
349 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
351 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
357 PH/01 Another patch from the Sieve maintainer.
359 PH/02 When an IPv6 address is converted to a string for single-key lookup
360 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
361 dots are used instead of colons so that keys in lsearch files need not
362 contain colons. This was done some time before quoting was made available
363 in lsearch files. However, iplsearch files do require colons in IPv6 keys
364 (notated using the quote facility) so as to distinguish them from IPv4
365 keys. This meant that lookups for IP addresses in host lists did not work
366 for iplsearch lookups.
368 This has been fixed by arranging for IPv6 addresses to be expressed with
369 colons if the lookup type is iplsearch. This is not incompatible, because
370 previously such lookups could never work.
372 The situation is now rather anomolous, since one *can* have colons in
373 ordinary lsearch keys. However, making the change in all cases is
374 incompatible and would probably break a number of configurations.
376 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
379 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
380 conversion specification without a maximum field width, thereby enabling
381 a rogue spamd server to cause a buffer overflow. While nobody in their
382 right mind would setup Exim to query an untrusted spamd server, an
383 attacker that gains access to a server running spamd could potentially
384 exploit this vulnerability to run arbitrary code as the Exim user.
386 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
387 $primary_hostname instead of what libspf2 thinks the hosts name is.
389 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
390 a directory entry by the name of the lookup key. Previously, if a
391 symlink pointed to a non-existing file or a file in a directory that
392 Exim lacked permissions to read, a lookup for a key matching that
393 symlink would fail. Now it is enough that a matching directory entry
394 exists, symlink or not. (Bugzilla 503.)
396 PH/03 The body_linecount and body_zerocount variables are now exported in the
399 PH/04 Added the $dnslist_matched variable.
401 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
402 This means they are set thereafter only if the connection becomes
405 PH/06 Added the client_condition to authenticators so that some can be skipped
406 by clients under certain conditions.
408 PH/07 The error message for a badly-placed control=no_multiline_responses left
409 "_responses" off the end of the name.
411 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
413 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
414 (without spaces) instead of just copying the configuration text.
416 PH/10 Added the /noupdate option to the ratelimit ACL condition.
418 PH/11 Added $max_received_linelength.
420 PH/12 Added +ignore_defer and +include_defer to host lists.
422 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
423 way in which PCRE > 7.0 is built.
425 PH/14 Implemented queue_only_load_latch.
427 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
428 MAIL command. The effect was to mangle the value on 64-bit systems.
430 PH/16 Another patch from the Sieve maintainer.
432 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
434 PH/18 If a system quota error occurred while trying to create the file for
435 a maildir delivery, the message "Mailbox is full" was not appended to the
436 bounce if the delivery eventually timed out. Change 4.67/27 below applied
437 only to a quota excession during the actual writing of the file.
439 PH/19 It seems that peer DN values may contain newlines (and other non-printing
440 characters?) which causes problems in log lines. The DN values are now
441 passed through string_printing() before being added to log lines.
443 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
444 and InterBase are left for another time.)
446 PH/21 Added message_body_newlines option.
448 PH/22 Guard against possible overflow in moan_check_errorcopy().
450 PH/23 POSIX allows open() to be a macro; guard against that.
452 PH/24 If the recipient of an error message contained an @ in the local part
453 (suitably quoted, of course), incorrect values were put in $domain and
454 $local_part during the evaluation of errors_copy.
460 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
461 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
464 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
465 issue a MAIL command.
467 PH/02 In an ACL statement such as
469 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
471 if a client was not listed at all, or was listed with a value other than
472 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
473 the condition was not true (as it should be), so access was not denied.
474 The bug was that the ! inversion was incorrectly passed on to the second
475 item. This has been fixed.
477 PH/03 Added additional dnslists conditions == and =& which are different from
478 = and & when the dns lookup returns more than one IP address.
480 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
481 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
483 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
484 FSYNC, which compiles an option called disable_fsync that allows for
485 bypassing fsync(). The documentation is heavily laced with warnings.
487 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
489 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
490 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
491 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
492 including adding "make clean"; (3) Added -fPIC when compiling the test
493 dynamically loaded module, to get rid of a warning.
495 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
496 message fails, move_frozen_messages = true and ignore_bounce_errors_after
497 = 0s. The bug is otherwise harmless.
499 PH/07 There was a bug in the dovecot authenticator such that the value of
500 $auth1 could be overwritten, and so not correctly preserved, after a
501 successful authentication. This usually meant that the value preserved by
502 the server_setid option was incorrect.
504 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
506 PH/09 Installed PCRE release 7.0.
508 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
509 run for batched SMTP input. It is now run at the start of every message
510 in the batch. While fixing this I discovered that the process information
511 (output by running exiwhat) was not always getting set for -bs and -bS
512 input. This is fixed, and it now also says "batched" for BSMTP.
514 PH/11 Added control=no_pipelining.
516 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
517 patch, slightly modified), and move the expansion of helo_data till after
518 the connection is made in the smtp transport (so it can use these
521 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
523 PH/14 Added log_selector = +pid.
525 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
527 PH/16 Add ${if forany and ${if forall.
529 PH/17 Added dsn_from option to vary the From: line in DSNs.
531 PH/18 Flush SMTP output before performing a callout, unless control =
532 no_callout_flush is set.
534 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
535 was true (the default) a successful delivery failed to delete the retry
536 item, thus causing premature timeout of the address. The bug is now
539 PH/20 Added hosts_avoid_pipelining to the smtp transport.
541 PH/21 Long custom messages for fakedefer and fakereject are now split up
542 into multiline reponses in the same way that messages for "deny" and
543 other ACL rejections are.
545 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
546 with slight modification.
548 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
549 draft, changing the syntax and factoring some duplicate code".
551 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
552 for deliveries of the second and subsequent messages over the same SMTP
555 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
556 ${reduce, with only minor "tidies".
558 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
560 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
561 expansion side effects.
563 PH/27 When a message times out after an over-quota error from an Exim-imposed
564 quota, the bounce message says "mailbox is full". This message was not
565 being given when it was a system quota that was exceeded. It now should
568 MH/03 Made $recipients available in local_scan(). local_scan() already has
569 better access to the recipient list through recipients_list[], but
570 $recipients can be useful in postmaster-provided expansion strings.
572 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
573 in the case of a MAIL command with additional options following the
574 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
575 were accidentally chopped off.
577 PH/29 SMTP synchronization checks are implemented when a command is read -
578 there is a check that no more input is waiting when there shouldn't be
579 any. However, for some commands, a delay in an ACL can mean that it is
580 some time before the response is written. In this time, more input might
581 arrive, invalidly. So now there are extra checks after an ACL has run for
582 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
583 pipelining has not been advertised.
585 PH/30 MH's patch to allow iscntrl() characters to be list separators.
587 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
588 returned in the SMTP response when smtp_return_error_details was false.
591 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
592 instead of the C I/O that was originally supplied, because problems were
595 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
596 Exim which did not show up earlier: it was assuming that a call to
597 SSL_CTX_set_info_callback() might give an error value. In fact, there is
598 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
599 was a macro that became an assignment, so it seemed to work. This has
600 changed to a proper function call with a void return, hence the compile
601 error. Exim's code has been fixed.
603 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
606 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
608 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
610 PH/37 If a message is not accepted after it has had an id assigned (e.g.
611 because it turns out to be too big or there is a timeout) there is no
612 "Completed" line in the log. When some messages of this type were
613 selected by exigrep, they were listed as "not completed". Others were
614 picked up by some special patterns. I have improved the selection
615 criteria to be more general.
617 PH/38 The host_find_failed option in the manualroute router can now be set
618 to "ignore", to completely ignore a host whose IP address cannot be
619 found. If all hosts are ignored, the behaviour is controlled by the new
620 host_all_ignored option.
622 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
623 homing or because of multiple MX records with /mx) generated more than
624 one IP address, and the following item turned out to be the local host,
625 all the secondary addresses of the first item were incorrectly removed
626 from the list, along with the local host and any following hosts (which
627 is what is supposed to happen).
629 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
630 whoever called Exim into the -H file. In the case of the daemon it was
631 behaving confusingly. When first started, it used values for whoever
632 started the daemon, but after a SIGHUP it used the Exim user (because it
633 calls itself on a restart). I have changed the code so that it now always
636 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
637 message are rejected with the same error (e.g. no authentication or bad
638 sender address), and a DATA command is nevertheless sent (as can happen
639 with PIPELINING or a stupid MUA), the error message that was given to the
640 RCPT commands is included in the rejection of the DATA command. This is
641 intended to be helpful for MUAs that show only the final error to their
644 PH/42 Another patch from the Sieve maintainer.
646 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
647 Eximstats - Fixed some broken HTML links and added missing column headers
649 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
650 columns for Rejects, Temp Rejects, Ham, and Spam rows.
652 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
654 PH/43 Yet another patch from the Sieve maintainer.
656 PH/44 I found a way to check for a TCP/IP connection going away before sending
657 the response to the final '.' that terminates a message, but only in the
658 case where the client has not sent further data following the '.'
659 (unfortunately, this is allowed). However, in many cases there won't be
660 any further data because there won't be any more messages to send. A call
661 to select() can be used: if it shows that the input is "ready", there is
662 either input waiting, or the socket has been closed. An attempt to read
663 the next input character can distinguish the two cases. Previously, Exim
664 would have sent an OK response which the client would never have see.
665 This could lead to message repetition. This fix should cure that, at
666 least in a lot of common cases.
668 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
669 advertised in response to EHLO.
675 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
676 fixed by 4.65/MH/01 (is this a record?) are fixed:
678 (i) An empty string was always treated as zero by the numeric comparison
679 operators. This behaviour has been restored.
681 (ii) It is documented that the numeric comparison operators always treat
682 their arguments as decimal numbers. This was broken in that numbers
683 starting with 0 were being interpreted as octal.
685 While fixing these problems I realized that there was another issue that
686 hadn't been noticed. Values of message_size_limit (both the global option
687 and the transport option) were treated as octal if they started with 0.
688 The documentation was vague. These values are now always treated as
689 decimal, and I will make that clear in the documentation.
695 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
696 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
699 MH/01 Don't check that the operands of numeric comparison operators are
700 integers when their expansion is in "skipping" mode (fixes bug
701 introduced by 4.64-PH/07).
703 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
704 child addresses, Exim now panics and dies. Previously, because the count
705 is held in a short int, deliveries were likely to be lost. As such a
706 large number of recipients for a single message is ridiculous
707 (performance will be very, very poor), I have chosen to impose a limit
708 rather than extend the field.
714 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
715 leftover -K file (the existence of which was triggered by #402).
716 While we were at it, introduced process PID as part of the -K
717 filename. This should rule out race conditions when creating
720 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
721 processing considerably. Previous code took too long for large mails,
722 triggering a timeout which in turn triggers #401.
724 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
725 in the DK code in transports.c. sendfile() is not really portable,
726 hence the _LINUX specificness.
728 TF/01 In the add_headers option to the mail command in an Exim filter,
729 there was a bug that Exim would claim a syntax error in any
730 header after the first one which had an odd number of characters
733 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
734 callout verification, Exim cached a "reject" for the entire domain. This
735 is correct for most verifications, but it is not correct for a recipient
736 verification with use_sender or use_postmaster set, because in that case
737 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
738 case of MAIL FROM:<> rejection from other early rejections (e.g.
739 rejection of HELO). When verifying a recipient using a non-null MAIL
740 address, the cache is ignored if it shows MAIL FROM:<> rejection.
741 Whatever the result of the callout, the value of the domain cache is
742 left unchanged (for any other kind of callout, getting as far as trying
743 RCPT means that the domain itself is ok).
745 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
748 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
749 manifest itself as EPIPE rather than ECONNECT. When tidying away a
750 session, the daemon ignores ECONNECT errors and logs others; it now
751 ignores EPIPE as well.
753 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
754 (quoted-printable decoding).
756 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
757 later the small subsequent patch to fix an introduced bug.
759 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
761 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
763 PH/08 An error is now given if message_size_limit is specified negative.
765 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
766 to be given (somewhat) arbitrary names.
768 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
771 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
772 miscellaneous code fixes
774 PH/10 Added the log_reject_target ACL modifier to specify where to log
777 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
778 hostname. This is wrong, because it relates to the incoming message (and
779 probably the interface on which it is arriving) and not to the outgoing
780 callout (which could be using a different interface). This has been
781 changed to use the value of the helo_data option from the smtp transport
782 instead - this is what is used when a message is actually being sent. If
783 there is no remote transport (possible with a router that sets up host
784 addresses), $smtp_active_hostname is used.
786 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
787 tweaks were necessary in order to get it to work (see also 21 below):
788 (a) The code assumed that strncpy() returns a negative number on buffer
789 overflow, which isn't the case. Replaced with Exim's string_format()
791 (b) There were several signed/unsigned issues. I just did the minimum
792 hacking in of casts. There is scope for a larger refactoring.
793 (c) The code used strcasecmp() which is not a standard C function.
794 Replaced with Exim's strcmpic() function.
795 (d) The code set only $1; it now sets $auth1 as well.
796 (e) A simple test gave the error "authentication client didn't specify
797 service in request". It would seem that Dovecot has changed its
798 interface. Fortunately there's a specification; I followed it and
799 changed what the client sends and it appears to be working now.
801 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
804 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
805 address A is aliased to B and C, where B exists and C does not. Without
806 -v the output is "A verified" because verification stops after a
807 successful redirection if more than one address is generated. However,
808 with -v the child addresses are also verified. Exim was outputting "A
809 failed to verify" and then showing the successful verification for C,
810 with its parentage. It now outputs "B failed to verify", showing B's
811 parentage before showing the successful verification of C.
813 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
814 look up a TXT record in a specific list after matching in a combined
817 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
818 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
819 they consult the DNS. I had assumed they would set it the way they
820 wanted; and indeed my experiments on Linux seem to show that in some
821 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
822 To be on the safe side, however, I have now made the interface to
823 host_find_byname() similar to host_find_bydns(), with an argument
824 containing the DNS resolver options. The host_find_byname() function now
825 sets these options at its start, just as host_find_bydns() does. The smtp
826 transport options dns_qualify_single and dns_search_parents are passed to
827 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
828 of host_find_byname() use the default settings of RES_DEFNAMES
829 (qualify_single) but not RES_DNSRCH (search_parents).
831 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
832 spool_read_header() do less string testing, by means of a preliminary
833 switch on the second character of optional "-foo" lines. (This is
834 overdue, caused by the large number of possibilities that now exist.
835 Originally there were few.) While I was there, I also converted the
836 str(n)cmp tests so they don't re-test the leading "-" and the first
837 character, in the hope this might squeeze out yet more improvement.
839 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
840 flag allowing group syntax was set by the header_syntax check but not
841 turned off, possible causing trouble later; (2) The flag was not being
842 set at all for the header_verify test, causing "group"-style headers to
843 be rejected. I have now set it in this case, and also caused header_
844 verify to ignore an empty address taken from a group. While doing this, I
845 came across some other cases where the code for allowing group syntax
846 while scanning a header line wasn't quite right (mostly, not resetting
847 the flag correctly in the right place). These bugs could have caused
848 trouble for malformed header lines. I hope it is now all correct.
850 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
851 with the "reply" argument non-NULL. The code, however (which originally
852 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
853 but it didn't always do it. This confused somebody who was copying the
854 code for some other use. I have removed all the tests.
856 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
857 feature that was used to support insecure browsers during the U.S. crypto
858 embargo. It requires special client support, and Exim is probably the
859 only MTA that supported it -- and would never use it because real RSA is
860 always available. This code has been removed, because it had the bad
861 effect of slowing Exim down by computing (never used) parameters for the
862 RSA_EXPORT functionality.
864 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
865 authenticator to fail if there's a tab character in the incoming data
866 (there should never be unless someone is messing about, as it's supposed
867 to be base64-encoded). Also added, on Timo's advice, the "secured" option
868 if the connection is using TLS or if the remote IP is the same as the
869 local IP, and the "valid-client-cert option" if a client certificate has
872 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
873 authenticators. This can be used for authorization after authentication
874 succeeds. (In the case of plaintext, it servers for both authentication
877 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
878 if any retry times were supplied.
880 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
881 connection, where there is no client IP address to check. In this
882 situation, the verify now always succeeds.
884 PH/25 Applied John Jetmore's -Mset patch.
886 PH/26 Added -bem to be like -Mset, but loading a message from a file.
888 PH/27 In a string expansion for a processed (not raw) header when multiple
889 headers of the same name were present, leading whitespace was being
890 removed from all of them, but trailing whitespace was being removed only
891 from the last one. Now trailing whitespace is removed from each header
892 before concatenation. Completely empty headers in a concatenation (as
895 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
896 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
898 PH/29 [Removed. This was a change that I later backed out, and forgot to
899 correct the ChangeLog entry (that I had efficiently created) before
900 committing the later change.]
902 PH/30 Exim was sometimes attempting to deliver messages that had suffered
903 address errors (4xx response to RCPT) over the same connection as other
904 messages routed to the same hosts. Such deliveries are always "forced",
905 so retry times are not inspected. This resulted in far too many retries
906 for the affected addresses. The effect occurred only when there were more
907 hosts than the hosts_max_try setting in the smtp transport when it had
908 the 4xx errors. Those hosts that it had tried were not added to the list
909 of hosts for which the message was waiting, so if all were tried, there
910 was no problem. Two fixes have been applied:
912 (i) If there are any address or message errors in an SMTP delivery, none
913 of the hosts (tried or untried) are now added to the list of hosts
914 for which the message is waiting, so the message should not be a
915 candidate for sending over the same connection that was used for a
916 successful delivery of some other message. This seems entirely
917 reasonable: after all the message is NOT "waiting for some host".
918 This is so "obvious" that I'm not sure why it wasn't done
919 previously. Hope I haven't missed anything, but it can't do any
920 harm, as the worst effect is to miss an optimization.
922 (ii) If, despite (i), such a delivery is accidentally attempted, the
923 routing retry time is respected, so at least it doesn't keep
924 hammering the server.
926 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
927 in ${readsocket because some servers need this prod.
929 PH/32 Added some extra debug output when updating a wait-xxx database.
931 PH/33 The hint "could be header name not terminated by colon", which has been
932 given for certain expansion errors for a long time, was not being given
933 for the ${if def:h_colon_omitted{... case.
935 PH/34 The spec says: "With one important exception, whenever a domain list is
936 being scanned, $domain contains the subject domain." There was at least
937 one case where this was not true.
939 PH/35 The error "getsockname() failed: connection reset by peer" was being
940 written to the panic log as well as the main log, but it isn't really
941 panic-worthy as it just means the connection died rather early on. I have
942 removed the panic log writing for the ECONNRESET error when getsockname()
945 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
946 runs only) independently of the message's sender address. This meant
947 that, if the 4xx error was in fact related to the sender, a different
948 message to the same recipient with a different sender could confuse
949 things. In particualar, this can happen when sending to a greylisting
950 server, but other circumstances could also provoke similar problems.
951 I have changed the default so that the retry time for these errors is now
952 based a combination of the sender and recipient addresses. This change
953 can be overridden by setting address_retry_include_sender=false in the
956 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
957 remote server are returned as part of bounce messages. This was not
958 happening for LMTP over a pipe (the lmtp transport), but now it is the
959 same for both kinds of LMTP.
961 PH/38 Despite being documented as not happening, Exim was rewriting addresses
962 in header lines that were in fact CNAMEs. This is no longer the case.
964 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
965 and queue runs started by the daemon processed all messages. This has
966 been fixed so that -R and -S can now usefully be given with -q<time>.
968 PH/40 Import PCRE release 6.7 (fixes some bugs).
970 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
972 PH/42 Give an error if -q is specified more than once.
974 PH/43 Renamed the variables $interface_address and $interface_port as
975 $received_ip_address and $received_port, to make it clear that these
976 values apply to message reception, and not to the outgoing interface when
977 a message is delivered. (The old names remain recognized, of course.)
979 PH/44 There was no timeout on the connect() call when using a Unix domain
980 socket in the ${readsocket expansion. There now is.
982 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
983 be meaningful with "accept".
985 SC/01 Eximstats V1.43
986 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
988 SC/02 Eximstats V1.44
989 Use a glob alias rather than an array ref in the generated
990 parser. This improves both readability and performance.
992 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
993 Collect SpamAssassin and rejection statistics.
994 Don't display local sender or destination tables unless
995 there is data to show.
996 Added average volumes into the top table text output.
998 SC/04 Eximstats V1.46
999 Collect data on the number of addresses (recipients)
1000 as well as the number of messages.
1002 SC/05 Eximstats V1.47
1003 Added 'Message too big' to the list of mail rejection
1004 reasons (thanks to Marco Gaiarin).
1006 SC/06 Eximstats V1.48
1007 Mainlog lines which have GMT offsets and are too short to
1008 have a flag are now skipped.
1010 SC/07 Eximstats V1.49 (Alain Williams)
1011 Added the -emptyok flag.
1013 SC/08 Eximstats V1.50
1014 Fixes for obtaining the IP address from reject messages.
1016 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1017 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1018 whitesspace changes from 4.64-PH/27
1020 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1023 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1024 are found, allow negative numbers in numeric criteria)
1026 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1028 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1029 to match changes made in 4.64-PH/43
1031 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1033 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1034 "domain" is actually an IP address, and force a failure. This locks out
1035 those revolvers/nameservers that support "A-for-A" lookups, in
1036 contravention of the specifications.
1038 PH/48 When a host name was looked up from an IP address, and the subsequent
1039 forward lookup of the name timed out, the host name was left in
1040 $sender_host_name, contrary to the specification.
1042 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1043 restricted to single-key lookups, Exim was not diagnosing an error if
1044 * or *@ was used with a query-style lookup.
1046 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1048 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1049 long ago, but noone interested enough thought of it. Let's just say that
1050 the "1.1" means that there are some new functions that weren't there at
1051 some point in the past.
1053 PH/51 Error processing for expansion failure of helo_data from an smtp
1054 transport during callout processing was broken.
1056 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1057 tested/used via the -bh/-bhc/-bs options.
1059 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1060 bug, fixed in subsequent PCRE releases).
1062 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1063 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1065 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1071 SC/01 Use a glob alias rather than an array ref in eximstats generated
1072 parser. This improves both readability and performance.
1074 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1075 Don't display local sender or destination tables in eximstats unless
1076 there is data to show.
1077 Added average volumes into the eximstats top table text output.
1079 SC/03 Collect data on the number of addresses (recipients) as well
1080 as the number of messages in eximstats.
1082 TF/01 Correct an error in the documentation for the redirect router. Exim
1083 does (usually) call initgroups() when daemonizing.
1085 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1086 with consistent privilege compared to when running as a daemon.
1088 TF/03 Note in the spec that $authenticated_id is not set for local
1089 submissions from trusted users.
1091 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1092 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1094 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1095 by adding some example configuration directives to the default
1096 configuration file. A little bit of work is required to uncomment the
1097 directives and define how usernames and passwords are checked, but
1098 there is now a framework to start from.
1100 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1101 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1102 without this. I don't know how relevant this is to other LDAP libraries.
1104 PH/02 Add the verb name to the "unknown ACL verb" error.
1106 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1108 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1110 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1111 directory not expanded when it should be if an expanded home directory
1112 was set for the address (which is overridden by the transport).
1114 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1117 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1118 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1119 because it is too late at that time, and has no effect.
1121 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1122 security issue with \' (bugzilla #107). I could not use the
1123 PQescapeStringConn() function, because it needs a PGconn value as one of
1126 PH/08 When testing addresses using -bt, indicate those final addresses that
1127 are duplicates that would not cause an additional delivery. At least one
1128 person was confused, thinking that -bt output corresponded to deliveries.
1129 (Suppressing duplicates isn't a good idea as you lose the information
1130 about possibly different redirections that led to the duplicates.)
1132 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1133 systems where poll() doesn't work, in particular OS X.
1135 PH/10 Added more information to debugging output for retry time not reached.
1137 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1138 operations in malware.c.
1140 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1143 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1144 syslog_duplication set false, log lines that would normally be written
1145 both the the main log and to the reject log were not written to syslog at
1148 PH/14 In the default configuration, change the use of "message" in ACL warn
1149 statements to "add_header".
1151 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1152 not followed by a command (e.g. "seen endif").
1154 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1155 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1158 PH/17 Added extra conditions to the default value of delay_warning_condition
1162 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1163 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1164 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1167 The Auto-Submitted: and various List- headers are standardised, whereas I
1168 don't think Precedence: ever was.
1170 PH/18 Refactored debugging code in route_finduser() to show more information,
1171 in particular, the error code if getpwnam() issues one.
1173 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1174 This is apparently needed in addition to the PH/07 change above to avoid
1175 any possible encoding problems.
1177 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1178 but not after initializing Perl.
1180 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1181 output them only if debugging. By default they are written stderr,
1182 apparently, which is not desirable.
1184 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1187 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1190 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1192 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1193 authenticated or an ident call has been made. Suppress the default
1194 values for $authenticated_id and $authenticated_sender (but permit -oMai
1195 and -oMas) when testing with -bh.
1197 PH/24 Re-jigged the order of the tests in the default configuration so that the
1198 tests for valid domains and recipients precede the DNS black list and CSA
1199 tests, on the grounds that those ones are more expensive.
1201 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1202 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1203 This bug exists in every version of Exim that I still have, right back to
1206 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1207 However, an attempt to turn on case-sensitivity in a regex key by
1208 including (?-i) didn't work because the subject string was already
1209 lowercased, and the effects were non-intuitive. It turns out that a
1210 one-line patch can be used to allow (?-i) to work as expected.
1216 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1217 other effects) broke the use of negated acl sub-conditions.
1219 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1222 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1223 "Deny" causes Exim to reject the incoming connection with a 554 error.
1224 Unfortunately, if there is a major crisis, such as a disk failure,
1225 tcp-wrappers gives "deny", whereas what one would like would be some
1226 kind of temporary error. A kludge has been added to help with this.
1227 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1228 554 error is used if errno is still zero or contains ENOENT (which occurs
1229 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1232 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1234 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1235 errors. Otherwise a message that provokes a temporary error (when other
1236 messages do not) can cause a whole host to time out.
1238 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1239 addresses were routed directly to files or pipes from a redirect router.
1240 File deliveries just didn't batch; pipe deliveries might have suffered
1243 PH/06 A failure to get a lock for a hints database would erroneously always say
1244 "Failed to get write lock", even when it was really a read lock.
1246 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1247 of 0600. This has been changed to use the value of the lockfile_mode
1248 option (which defaults to 0600).
1250 PH/08 Applied small patch from the Sieve maintainer.
1252 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1253 folder from quota calculations, a direct delivery into this folder messed
1254 up the contents of the maildirsize file. This was because the regex was
1255 used only to exclude .Trash (or whatever) when the size of the mailbox
1256 was calculated. There was no check that a delivery was happening into an
1257 excluded directory. This bug has been fixed by ignoring all quota
1258 processing for deliveries into excluded directories.
1260 PH/10 Added the maildirfolder_create_regex option to appendfile.
1266 PH/01 The code for finding all the local interface addresses on a FreeBSD
1267 system running IPv6 was broken. This may well have applied to all BSD
1268 systems, as well as to others that have similar system calls. The broken
1269 code found IPv4 interfaces correctly, but gave incorrect values for the
1270 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1271 that it would not match correctly against @[] and not recognize the IPv6
1274 PH/02 The ipliteral router was not recognizing addresses of the form user@
1275 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1277 PH/03 Added disable_ipv6.
1279 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1280 decoded form, because it is most often used to construct To: headers
1281 lines in autoreplies, and the decoded form may well be syntactically
1282 invalid. However, $reply_address has leading white space removed, and all
1283 newlines turned into spaces so that the autoreply transport does not
1286 PH/05 If group was specified without a user on a router, and no group or user
1287 was specified on a transport, the group from the router was ignored.
1289 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1290 for visible compile-time settings that can be used to change these
1291 numbers, for those that want even more. Backwards compatibility with old
1292 spool files has been maintained. However, going back to a previous Exim
1293 release will lost any variables that are in spool files.
1295 PH/07 Two small changes when running in the test harness: increase delay when
1296 passing a TCP/IP connection to a new process, in case the original
1297 process has to generate a bounce, and remove special handling of
1298 127.0.0.2 (sic), which is no longer necessary.
1300 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1301 be the same on different OS.
1303 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1306 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1307 whether --show-vars was specified or not
1309 JJ/02 exipick: Added support for new ACL variable spool format introduced
1312 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1313 syntactically invalid From: or Reply-to: line, and a filter used this to
1314 generate an autoreply, and therefore failed to obtain an address for the
1315 autoreply, Exim could try to deliver to a non-existent relative file
1316 name, causing unrelated and misleading errors. What now happens is that
1317 it logs this as a hard delivery error, but does not attempt to create a
1320 PH/11 The exinext utility has a -C option for testing purposes, but although
1321 the given file was scanned by exinext itself; it wasn't being passed on
1322 when Exim was called.
1324 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1325 an end-of-file indication when reading a command response.
1327 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1328 compiled. In many other places in Exim, IPv6 addresses are always
1329 recognized, so I have changed this. It also means that IPv4 domain
1330 literals of the form [IPV4:n.n.n.n] are now always recognized.
1332 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1333 used if the router is not running as root, for example, when verifying at
1334 ACL time, or when using -bh. The debugging output from this situation was
1335 non-existent - all you got was a failure to exec. I have made two
1338 (a) Failures to set uid/gid, the current directory, or a process leader
1339 in a subprocess such as that created by queryprogram now generate
1340 suitable debugging ouput when -d is set.
1342 (b) The queryprogram router detects when it is not running as root,
1343 outputs suitable debugging information if -d is set, and then runs
1344 the subprocess without attempting to change uid/gid.
1346 PH/15 Minor change to Makefile for building test_host (undocumented testing
1349 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1350 additional section of a DNS packet that returns MX or SRV records.
1351 Instead, it always explicitly searches for A/AAAA records. This avoids
1352 major problems that occur when a DNS server includes only records of one
1353 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1354 fixed another bug: if SRV records were looked up and the corresponding
1355 address records were *not* found in the additional section, the port
1356 values from the SRV records were lost.
1358 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1359 using the correct key (the original address) when searching the retry
1360 rules in order to find which one to use for generating the retry hint.
1362 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1363 adding the default one. Similarly, if it contains a Reply-To: header, the
1364 errors_reply_to option, if set, is not used.
1366 PH/19 When calculating a retry time, Exim used to measure the "time since
1367 failure" by looking at the "first failed" field in the retry record. Now
1368 it does not use this if it is later than than the arrival time of the
1369 message. Instead it uses the arrival time. This makes for better
1370 behaviour in cases where some deliveries succeed, thus re-setting the
1371 "first failed" field. An example is a quota failure for a huge message
1372 when small messages continue to be delivered. Without this change, the
1373 "time since failure" will always be short, possible causing more frequent
1374 delivery attempts for the huge message than are intended.
1375 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1377 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1378 $1, $2, $3) because the numerical variables can be reset during some
1379 expansion items (e.g. "match"), thereby losing the authentication data.
1381 PH/21 Make -bV show the size of off_t variables so that the test suite can
1382 decide whether to run tests for quotas > 2G.
1384 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1385 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1386 filecount value is greater than 2G or if a quota value is greater than 2G
1387 on a system where the size of off_t is not greater than 4, a panic error
1390 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1391 never match. The debug and -bh output now contains an explicit error
1392 message indicating a malformed IPv4 address or mask.
1394 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1395 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1396 PH/23 above applies.
1398 PH/25 Do not write to syslog when running in the test harness. The only
1399 occasion when this arises is a failure to open the main or panic logs
1400 (for which there is an explicit test).
1402 PH/26 Added the /no_tell option to "control=freeze".
1404 PH/27 If a host name lookup failed very early in a connection, for example, if
1405 the IP address matched host_lookup and the reverse lookup yielded a name
1406 that did not have a forward lookup, an error message of the form "no IP
1407 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1408 could be logged. Now it outputs the IP address instead of "NULL".
1410 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1411 allows the sender and the authenticated sender to be set when
1412 submitting a message from within Exim. Since child_open_exim() is
1413 documented for local_scan(), the new function should be too.
1415 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1416 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1417 results in an empty string is now treated as unset.
1419 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1421 PH/31 Added sender_verify_fail logging option.
1423 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1424 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1425 tidied the source and removed it altogether.
1427 PH/33 When a queue run was abandoned because the load average was too high, a
1428 log line was always written; now it is written only if the queue_run log
1429 selector is set. In addition, the log line for abandonment now contains
1430 information about the queue run such as the pid. This is always present
1431 in "start" and "stop" lines but was omitted from the "abandon" line.
1433 PH/34 Omit spaces between a header name and the colon in the error message that
1434 is given when verify = headers_syntax fails (if there are lots of them,
1435 the message gets confusing).
1437 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1438 names, as there are now some PTR records that contain slashes. This check
1439 is only to protect against broken name servers that fall over on strange
1440 characters, so the fact that it applies to all lookups doesn't matter.
1442 PH/36 Now that the new test suite is complete, we can remove some of the
1443 special code in Exim that was needed for the old test suite. For example,
1444 sorting DNS records because real resolvers return them in an arbitrary
1445 order. The new test suite's fake resolver always returns records in the
1448 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1449 bounces) except when queue_only is set, to avoid logging races between
1450 the different processes.
1452 PH/38 Panic-die if .include specifies a non-absolute path.
1454 PH/39 A tweak to the "H" retry rule from its user.
1456 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1457 a label. They prevented compilation on older perls.
1459 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1460 a warning to be raised on newish perls.
1462 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1463 on queue. Changes to match documented behaviour of showing count of
1464 messages matching specified criteria.
1466 PH/40 Changed the default ident timeout from 30s to 5s.
1468 PH/41 Added support for the use of login_cap features, on those BSD systems
1469 that have them, for controlling the resources used by pipe deliveries.
1471 PH/42 The content-scanning code uses fopen() to create files in which to put
1472 message data. Previously it was not paying any attention to the mode of
1473 the files. Exim runs with umask(0) because the rest of the code creates
1474 files with open(), and sets the required mode explicitly. Thus, these
1475 files were ending up world-writeable. This was not a big issue, because,
1476 being within the spool directory, they were not world-accessible. I have
1477 created a function called modefopen, which takes an additional mode
1478 argument. It sets umask(777), creates the file, chmods it to the required
1479 mode, then resets the umask. All the relevant calls to fopen() in the
1480 content scanning code have been changed to use this function.
1482 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1483 to 24 hours. This avoids potential overflow problems when processing G
1484 and H retry rules. I suspect nobody ever tinkers with this value.
1486 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1488 PH/45 When the plaintext authenticator is running as a client, the server's
1489 challenges are checked to ensure they are valid base64 strings. By
1490 default, the authentication attempt is cancelled if an invalid string is
1491 received. Setting client_ignore_invalid_base64 true ignores these errors.
1492 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1493 they are received. Thus, the responses can be made to depend on the
1494 challenges. If an invalid string is ignored, an empty string is placed in
1497 PH/46 Messages that are created by the autoreply transport now contains a
1498 References: header, in accordance with RFCs 2822 and 3834.
1500 PH/47 Added authenticated_sender_force to the smtp transport.
1502 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1504 PH/49 Installed latest patch from the Sieve maintainer.
1506 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1507 was also set, the appendfile transport was unnecessarily scanning a
1508 directory of message files (e.g. for maildir delivery) to find the count
1509 of files (along with the size), even though it did not need this
1510 information. It now does the scan only if it needs to find either the
1511 size of the count of files.
1513 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1515 PH/52 Two bugs concerned with error handling when the smtp transport is
1518 (i) Exim was not creating retry information for temporary errors given
1519 for individual recipients after the DATA command when the smtp transport
1520 was used in LMTP mode. This meant that they could be retried too
1521 frequently, and not timed out correctly.
1523 (ii) Exim was setting the flag that allows error details to be returned
1524 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1525 recipients that were returned after the DATA command.
1527 PH/53 This is related to PH/52, but is more general: for any failing address,
1528 when detailed error information was permitted to be returned to the
1529 sender, but the error was temporary, then after the final timeout, only
1530 "retry timeout exceeded" was returned. Now it returns the full error as
1531 well as "retry timeout exceeded".
1533 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1534 do this, and (what is worse) MTAs that accept it.
1536 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1537 will now be deprecated.
1539 PH/56 New os.c-cygwin from the Cygwin maintainer.
1541 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1542 formats (previously only available in exim formats via -bpr, -bpru,
1543 and -bpra. Now also available in native and exiqgrep formats)
1545 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1546 with very large, slow to parse queues
1548 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1550 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1552 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1553 responses to authentication challenges, though it was showing the
1554 challenges; (ii) I've removed the CR characters from the debug output for
1557 PH/58 Allow for the insertion of a newline as well as a space when a string
1558 is turned into more than one encoded-word during RFC 2047 encoding. The
1559 Sieve code now uses this.
1561 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1562 data_4xx, lost_connection, tls_required.
1564 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1565 message was being sent as an SMTP response.
1567 PH/61 Add -l and -k options to exicyclog.
1569 PH/62 When verifying, if an address was redirected to one new address, so that
1570 verification continued, and the new address failed or deferred after
1571 having set something in $address_data, the value of $address_data was not
1572 passed back to the ACL. This was different to the case when no
1573 redirection occurred. The value is now passed back in both cases.
1575 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1576 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1577 use login_cap.h, so on its own it isn't the distinguishing feature. The
1578 new name refers directly to the setclassresources() function.
1580 PH/65 Added configuration files for NetBSD3.
1582 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1584 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1585 is preferred over IPv4.
1587 PH/68 The bounce_return_message and bounce_return_body options were not being
1588 honoured for bounces generated during the reception of non-SMTP messages.
1589 In particular, this applied to messages rejected by the ACL. This bug has
1590 been fixed. However, if bounce_return_message is true and bounce_return_
1591 body is false, the headers that are returned for a non-SMTP message
1592 include only those that have been read before the error was detected.
1593 (In the case of an ACL rejection, they have all been read.)
1595 PH/69 The HTML version of the specification is now built in a directory called
1596 spec_html instead of spec.html, because the latter looks like a path with
1597 a MIME-type, and this confuses some software.
1599 PH/70 Catch two compiler warnings in sieve.c.
1601 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1602 function verify_get_ident() calls ip_connect() to connect a socket, but
1603 if the "connect()" function timed out, ip_connect() used to close the
1604 socket. However, verify_get_ident() also closes the socket later, and in
1605 between Exim writes to the log, which may get opened at this point. When
1606 the socket was closed in ip_connect(), the log could get the same file
1607 descriptor number as the socket. This naturally causes chaos. The fix is
1608 not to close the socket in ip_connect(); the socket should be closed by
1609 the function that creates it. There was only one place in the code where
1610 this was missing, in the iplookup router, which I don't think anybody now
1611 uses, but I've fixed it anyway.
1613 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1614 well as to direct DNS lookups. Otherwise the handling of names in host
1615 lists is inconsistent and therefore confusing.
1621 PH/01 Two changes to the default runtime configuration:
1623 (1) Move the checks for relay_from_hosts and authenticated clients from
1624 after to before the (commented out) DNS black list checks.
1626 (2) Add control=submission to the relay_from_hosts and authenticated
1627 clients checks, on the grounds that messages accepted by these
1628 statements are most likely to be submissions.
1630 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1632 (1) Generate an error if the third argument for the ${prvs expansion is
1635 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1638 (3) Reset the variables that are obtained from the first argument of
1639 ${prvscheck and used in the second argument before leaving the code,
1640 because their memory is reclaimed, so using them afterwards may do
1643 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1644 one (it's much easier than Tom thought :-).
1646 (5) Because of (4), we can now allow for the use of $prvscheck_result
1647 inside the third argument.
1649 PH/03 For some reason, the default setting of PATH when running a command from
1650 a pipe transport was just "/usr/bin". I have changed it to
1653 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1654 anything to be listed in the output from -bV.
1656 PH/05 When a filter generated an autoreply, the entire To: header line was
1657 quoted in the delivery log line, like this:
1659 => >A.N.Other <ano@some.domain> <original@ddress> ...
1661 This has been changed so that it extracts the operative address. There
1662 may be more than one such address. If so, they are comma-separated, like
1665 => >ano@some.domain,ona@other.domain <original@ddress> ...
1667 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1668 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1669 not being looked up in the rDNS to get a host name, Exim was showing the
1670 IP address twice in Received: lines, even though the IP addresses were
1671 identical. For example:
1673 Received: from [1.2.3.4] (helo=[1.2.3.4])
1675 However, if the real host name was known, it was omitting the HELO data
1676 if it matched the actual IP address. This has been tidied up so that it
1677 doesn't show the same IP address twice.
1679 PH/07 When both +timestamp and +memory debugging was on, the value given by
1680 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1681 called by the debug printing, thereby overwriting the timestamp buffer.
1682 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1684 PH/08 When the original message was included in an autoreply transport, it
1685 always said "this is a copy of the message, including all the headers",
1686 even if body_only or headers_only was set. It now gives an appropriate
1689 PH/09 Applied a patch from the Sieve maintainer which:
1691 o fixes some comments
1692 o adds the (disabled) notify extension core
1693 o adds some debug output for the result of if/elsif tests
1694 o points to the current vacation draft in the documentation
1695 and documents the missing references header update
1699 o fixes a bug in processing the envelope test (when testing
1700 multiple envelope elements, the last element determinted the
1703 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1704 Electronic Mail") by including:
1706 Auto-submitted: auto-generated
1708 in the messages that it generates (bounce messages and others, such as
1709 warnings). In the case of bounce messages for non-SMTP mesages, there was
1710 also a typo: it was using "Auto_submitted" (underscore instead of
1711 hyphen). Since every message generated by Exim is necessarily in response
1712 to another message, thes have all been changed to:
1714 Auto-Submitted: auto-replied
1716 in accordance with these statements in the RFC:
1718 The auto-replied keyword:
1720 - SHOULD be used on messages sent in direct response to another
1721 message by an automatic process,
1723 - MUST NOT be used on manually-generated messages,
1725 - MAY be used on Delivery Status Notifications (DSNs) and Message
1726 Disposition Notifications (MDNs),
1728 - MUST NOT be used on messages generated by automatic or periodic
1729 processes, except for messages which are automatic responses to
1732 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1733 to the default Received: header definition.
1735 PH/12 Added log selector acl_warn_skipped (default on).
1737 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1738 variables because (a) they are in the wrong storage pool and (b) even if
1739 they were copied, it wouldn't work properly because of the caching.
1741 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1742 checking when decoding. Apparently there are clients that generate
1743 overlong encoded strings. Why am I not surprised?
1745 PH/15 If the first argument of "${if match_address" was not empty, but did not
1746 contain an "@" character, Exim crashed. Now it writes a panic log message
1747 and treats the condition as false.
1749 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1751 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1752 extension "envelope-auth". The code is finished and in agreement with
1753 other implementations, but there is no documentation so far and in fact,
1754 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1755 not changing the active code.
1757 Print executed "if" and "elsif" statements when debugging is used. This
1758 helps a great deal to understand what a filter does.
1760 Document more things not specified clearly in RFC3028. I had all this
1761 sorted out, when out of a sudden new issues came to my mind. Oops."
1763 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1766 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1767 canonical form (as documented). However, after a host name lookup from
1768 the IP address, check_host() was doing a simple string comparison with
1769 addresses acquired from the DNS when checking that the found name did
1770 have the original IP as one of its addresses. Since any found IPv6
1771 addresses are likely to be in abbreviated form, the comparison could
1772 fail. Luckily, there already exists a function for doing the comparison
1773 by converting both addresses to binary, so now that is used instead of
1774 the text comparison.
1776 PH/20 There was another similar case to PH/19, when a complete host name was
1777 given in a host list; looking up its IP address could give an abbreviated
1778 form, whereas the current host's name might or might not be abbreviated.
1779 The same fix has been applied.
1785 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1786 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1789 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1790 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1792 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1794 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1795 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1796 Also, for "Auto-Submitted", treat anything other than "no" as
1797 non-personal, in accordance with RFC 3834. (Previously it treated
1798 anything starting "auto-" as non-personal.)
1800 TF/01 The control=submission/name=... option had a problem with syntax
1801 errors if the name included a slash character. The /name= option
1802 now slurps the rest of the string, so it can include any characters
1803 but it must come last in the list of options (after /sender_retain
1806 PH/05 Some modifications to the interface to the fake nameserver for the new
1814 TK/01 Added the "success_on_redirect" address verification option. See
1815 NewStuff for rationale and an example.
1817 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1819 PH/02 Patch to exigrep to allow it to work on syslog lines.
1821 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1822 fread() to skip over the body file's header line, because in Cygwin the
1823 header line is locked and is inaccessible.
1825 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1826 co-exist for some time) to make it clear that it is the Exim ID that is
1827 referenced, not the Message-ID: header line.
1829 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1830 string_format() function, because snprintf() does not exist on all
1833 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1834 ${stat: expansion item.
1836 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1837 protocol synchronization error", to keep the pedants happy.
1839 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1840 well as for IRIX systems, when gcc is being used. See the host.c source
1843 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1845 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1848 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1849 transport and to the smtp transport in LMTP mode.
1851 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1853 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1854 run a filter in a subprocess. This could lead to confusion in subsequent
1855 lookups in the parent process. There should also be a search_tidyup() at
1856 the end of the subprocess.
1858 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1859 only if the host matched helo_try_verify_hosts, which caused the
1860 verification to occur when the EHLO/HELO command was issued. The ACL just
1861 tested the remembered result. Now, if a previous verification attempt has
1862 not happened, "verify = helo" does it there and then.
1864 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1866 TK/03 Fix log output including CR from clamd.
1868 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1869 exist provoked a memory error which could cause a segfault.
1871 PH/15 Installed PCRE 6.2
1873 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1875 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1876 of the problem. Specifically, suggested +O2 rather than +O1 for the
1879 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1881 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1882 record, the port was not correctly logged when the outgoing_port log
1883 selector was set (it logged the transort's default port).
1885 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1886 fallback_hosts, and "hosts" in the smtp transport.
1888 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1889 host errors such as "Connection refused".
1891 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1892 authentication with radiusclient 0.4.9:
1894 - Error returned from rc_read_config was caught wrongly
1895 - Username/password not passed on to radius server due to wrong length.
1897 The presumption is that some radiusclient API changes for 4.51/PH/17
1898 were not taken care of correctly. The code is still untested by me (my
1899 Linux distribution still has 0.3.2 of radiusclient), but it was
1900 contributed by a Radius user.
1902 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1903 expanding the "port" option of the smtp transport.
1905 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1906 while reading a MIME header. Thanks to Tom Hughes for a patch.
1908 PH/24 Include config.h inside local_scan.h so that configuration settings are
1911 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1912 that in an ACL for RCPT (for example), you can examine exactly what was
1915 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1916 commands, but it was not correctly comparing the address with the actual
1917 client host address. Thus, it would show the EHLO address in Received:
1918 header lines when this was not necessary.
1920 PH/27 Added the % operator to ${eval:}.
1922 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1923 it should be ignoring failures (because with -C, for example, it has lost
1924 privilege). It wasn't ignoring creation failures other than "already
1927 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1930 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1931 because an input line was too long, either on its own, or by virtue of
1932 too many continuations, the temporary file was not being removed, and the
1933 return code was incorrect.
1935 PH/31 Missing "BOOL" in function definition in filtertest.c.
1937 PH/32 Applied Sieve patches from the maintainer.
1939 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1941 PH/33 Added "verify = not_blind".
1943 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1944 Local/Makefile (with some defaults set). These are used in built scripts
1945 such as exicyclog, but they have never been used in the exim_install
1946 script (though there are many overriding facilities there). I have
1947 arranged that the exim_install script now takes note of these two
1950 PH/35 Installed configuration files for Dragonfly.
1952 PH/36 When a locally submitted message by a trusted user did not contain a
1953 From: header, and the sender address was obtained from -f or from an SMTP
1954 MAIL command, and the trusted user did not use -F to supply a sender
1955 name, $originator_name was incorrectly used when constructing a From:
1956 header. Furthermore, $originator_name was used for submission mode
1957 messages from external hosts without From: headers in a similar way,
1958 which is clearly wrong.
1960 PH/37 Added control=suppress_local_fixups.
1962 PH/38 When log_selector = +received_sender was set, and the addition of the
1963 sender made the log line's construction buffer exactly full, or one byte
1964 less than full, an overflow happened when the terminating "\n" was
1967 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1968 when the result of a list match is failure because a DNS lookup failed.
1970 PH/40 RM_COMMAND is now used in the building process.
1972 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1973 the "build-* directories that it finds.
1975 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1976 address in a domain literal was a prefix of an interface address.
1978 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1979 when verifying a sender address, unless rewrite_headers is false.
1981 PH/44 Wrote a long comment about why errors_to addresses are verified as
1982 recipients, not senders.
1984 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1985 the ratelimit ACL was added.
1987 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1989 PH/46 Added extra information about PostgreSQL errors to the error string.
1991 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1992 avoiding the need to install special zones in a real server. This is
1993 backwards compatible; if it can't find the fake resolver, it drops back.
1994 Thus, both old and new test suites can be run.
1996 TF/02 Added util/ratelimit.pl
1998 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1999 clock is set back in time.
2001 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2002 Candler <B.Candler@pobox.com>.
2004 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2005 OK for addresses that are the result of redirections.
2007 PH/48 A number of further additions for the benefit of the new test suite,
2008 including a fake gethostbyname() that interfaces to the fake DNS resolver
2011 TF/06 The fix for widen_domains has also been applied to qualify_single and
2012 search_parents which are the other dnslookup options that can cause
2015 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2018 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2020 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2021 service" if it encountered an MX record with an empty target hostname.
2022 The message is now "an MX or SRV record indicated no SMTP service".
2024 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2025 if the DNS of the sending site is misconfigured. This is quite a
2026 common situation. This change restores the behaviour of treating a
2027 helo verification defer as a failure.
2029 PH/51 If self=fail was set on a router, the bounce message did not include the
2030 actual error message.
2036 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2038 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2039 command itself ended in error, the underlying message about the transport
2040 filter timeout was being overwritten with the pipe command error. Now the
2041 underlying error message should be appended to the second error message.
2043 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2045 PH/02 Reduce the amount of output that "make" produces by default. Full output
2046 can still be requested.
2048 PH/03 The warning log line about a condition test deferring for a "warn" verb
2049 was being output only once per connection, rather than after each
2050 occurrence (because it was using the same function as for successful
2051 "warn" verbs). This seems wrong, so I have changed it.
2053 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2054 it should not have, which might have caused a crash in the right
2055 circumstances, but probably never did.
2057 PH/04 Installed a modified version of Tony Finch's patch to make submission
2058 mode fix the return path as well as the Sender: header line, and to
2059 add a /name= option so that you can make the user's friendly name appear
2062 TF/03 Added the control = fakedefer ACL modifier.
2064 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2065 Mark Lowes for thorough testing.
2067 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2069 TK/03 Merged latest SRS patch from Miles Wilton.
2071 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2072 with the definition in sysexits.h (which is #included earlier).
2073 Fortunately, Exim does not actually use EX_OK. The code used to try to
2074 preserve the sysexits.h value, by assumimg that macro definitions were
2075 scanned for macro replacements. I have been disabused of this notion,
2076 so now the code just undefines EX_OK before #including unistd.h.
2078 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2079 in the smtp transport. When a block could not be written in a single
2080 write() function, the timeout was being re-applied to each part-write.
2081 This seems wrong - if the receiver was accepting one byte at a time it
2082 would take for ever. The timeout is now adjusted when this happens. It
2083 doesn't have to be particularly precise.
2085 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2086 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2088 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2089 without a domain if the check to <postmaster@domain> fails.
2091 SC/01 Eximstats: added -xls and the ability to specify output files
2092 (patch written by Frank Heydlauf).
2094 SC/02 Eximstats: use FileHandles for outputing results.
2096 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2098 SC/04 Eximstats: fixed display of large numbers with -nvr option
2100 SC/05 Eximstats: fixed merging of reports with empty tables.
2102 SC/06 Eximstats: added the -include_original_destination flag
2104 SC/07 Eximstats: removed tabs and trailing whitespace.
2106 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2108 TK/06 MBOX spool code: Add real "From " MBOX separator line
2109 so the .eml file is really in mbox format (even though
2110 most programs do not really care). Patch from Alex Miller.
2112 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2113 The latter is generated from $received_to and is only set if the
2114 message has one envelope recipient. SA can use these headers,
2115 obviously out-of-the-box. Patch from Alex Miller.
2117 PH/08 The ${def test on a variable was returning false if the variable's
2118 value was "0", contrary to what the specification has always said!
2119 The result should be true unless the variable is empty.
2121 PH/09 The syntax error of a character other than { following "${if
2122 def:variable_name" (after optional whitespace) was not being diagnosed.
2123 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2124 accidental colon was present, for example, could give incorrect results.
2126 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2127 result is used (not including appendfile, where other changes are about
2130 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2131 This involved changing a lot of size variables from int to off_t. It
2132 should work with maildirs and everything.
2134 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2135 spamd dying while we are connected to it.
2137 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2140 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2141 function for BDB 4.3.
2143 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2145 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2146 the sources for PCRE have been re-arranged and more files are now
2149 PH/15 The code I had for printing potentially long long variables in PH/11
2150 above was not the best (it lost precision). The length of off_t variables
2151 is now inspected at build time, and an appropriate printing format (%ld
2152 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2153 to be "long long int" or "long int". This is needed for the internal
2154 formatting function string_vformat().
2156 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2157 the configuration file to be ":syslog", then the script "guesses" where
2158 the logs files are, rather than using the compiled in default. In our
2159 case the guess is not the same as the compiled default, so the script
2160 suddenly stopped working when I started to use syslog. The patch checks
2161 to see if log_file_path is "". If so, it attempts to read it from exim
2162 with no configuration file to get the compiled in version, before it
2163 falls back to the previous guessing code."
2165 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2166 implementing BATV in an Exim configuration. See NewStuff for the gory
2169 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2170 Makefile that are specific to HP-UX.
2172 PH/18 If the "use_postmaster" option was set for a recipient callout together
2173 with the "random" option, the postmaster address was used as the MAIL
2174 FROM address for the random test, but not for the subsequent recipient
2175 test. It is now used for both.
2177 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2178 patch removes a few documentation additions to RFC 3028, because the
2179 latest draft now contains them. It adds the new en;ascii-case comparator
2180 and a new error check for 8bit text in MIME parts. Comparator and
2181 require names are now matched exactly. I enabled the subaddress
2182 extension, but it is not well tested yet (read: it works for me)."
2184 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2185 rework some of the code of TK/09 above to avoid the hardwired use of
2186 "%lld" and "long long". Replaced the call to snprintf() with a call to
2189 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2190 records point to non-existent hosts", "retry timeout exceeded", and
2191 "retry time not reached for any host after a long failure period".
2193 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2194 experimental DomainKeys support:
2196 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2197 (2) On an error such as an illegally used "control", the wrong name for
2198 the control was given.
2200 These problems did NOT occur unless DomainKeys support was compiled.
2202 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2204 PH/24 Added ${if match_ip condition.
2206 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2207 databases so that it will be absolutely obvious if a crash occurs in the
2208 DB library. This is a regular occurrence (often caused by mis-matched
2211 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2212 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2213 picked up on a user's system that detects such things. There doesn't seem
2214 to be a gcc warning option for this - only an attribute that has to be
2215 put on the function's prototype. It seems that in Fedora Core 4 they have
2216 set this on a number of new functions. No doubt there will be more in due
2219 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2220 specify a transport. However, if an address that was verified by such a
2221 router was the subject of a callout, Exim crashed because it tried to
2222 read the rcpt_include_affixes from the non-existent transport. Now it
2223 just assumes that the setting of that option is false. This bug was
2224 introduced by 4.51/PH/31.
2226 PH/28 Changed -d+all to exclude +memory, because that information is very
2227 rarely of interest, but it makes the output a lot bigger. People tend to
2228 do -d+all out of habit.
2230 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2231 code in os-type was giving problems when libc.so lives in lib64, like on
2234 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2235 aren't the modern standard, and it seems that some systems' include files
2236 don't always have them. Exim was already checking for some of the newer
2237 ones like T_AAAA, and defining it itself. I've added checks for all the
2238 record types that Exim uses.
2240 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2241 not automatically generating a new one, as it is supposed to. This
2242 prevented TLS from working. If the file did exist, but contained invalid
2243 data, a new version was generated, as expected. It was only the case of a
2244 non-existent file that was broken.
2246 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2247 with a change in libdomainkeys > 0.64.
2249 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2250 from DNS. If the selector record carries the flag, it now has
2251 precedence over the domain-wide flag.
2253 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2255 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2256 the use of a port name that isn't defined in /etc/services) occurred, the
2257 message was deferred as in a normal delivery, and thus remained on the
2258 spool, instead of being failed because of the mua_wrapper setting. This
2259 is now fixed, and I tidied up some of the mua_wrapper messages at the
2262 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2263 the messages in a hash of arrays rather than using individual hashes.
2264 This is a bit cleaner and results in dramatic memory savings, albeit
2265 at a slight CPU cost.
2267 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2268 as requested by Marc Sherman.
2270 SC/10 Eximstats: added histograms for user specified patterns as requested
2273 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2275 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2276 fopen() in the content-scanning modules that did not already have it.
2282 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2283 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2285 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2287 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2289 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2290 to test Sieve filters that use "vacation".
2292 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2293 that changes the way the GnuTLS parameters are stored in the cache file.
2294 The new format can be generated externally. For backward compatibility,
2295 if the data in the cache doesn't make sense, Exim assumes it has read an
2296 old-format file, and it generates new data and writes a new file. This
2297 means that you can't go back to an older release without removing the
2300 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2301 work if there are any delivery delays because "one_time" forces the
2302 parent to be marked "delivered", so its unseen clone is never tried
2303 again. For this reason, Exim now forbids the simultaneous setting of
2306 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2307 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2308 entry if you want to know the details. The fix, however, neglected to
2309 consider the case when local delivery batching is involved. The test for
2310 "previously delivered" was not happening when checking to see if an
2311 address could be batched with a previous (undelivered) one; under
2312 certain circumstances this could lead to multiple deliveries to the same
2315 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2316 in its include files, and this causes problems building Exim.
2318 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2319 header_syntax) but Exim was just ignoring anything given after a slash.
2320 In particular, this caused confusion with an attempt to use "verify =
2321 reverse_host_lookup/defer_ok". An error is now given when options are
2322 supplied for verify items that do not have them. (Maybe reverse_host_
2323 lookup should have a defer_ok option, but that's a different point.)
2325 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2326 defined by RFC 821) to 2048, because there were problems with some AUTH
2327 commands, and RFC 1869 says the size should be increased for extended
2328 SMTP commands that take arguments.
2330 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2333 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2334 "unknown" error; now it says that the functionality isn't in the binary.
2336 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2337 an address' error message when a string expansion fails (syntax or
2338 whatever). Otherwise the password may appear in the log. Following change
2339 PH/42 below, there is no longer a chance of it appearing in a bounce
2342 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2344 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2345 of its addresses was ever tried. (Bugzilla bug #2.)
2347 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2348 the result incorrectly in the debug output. (It correctly added a newline
2349 to what was transported.)
2351 TF/01 Added $received_time.
2353 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2354 commented out examples of how to interface to a virus scanner and to
2355 SpamAssassin. Also added commented examples of av_scanner and
2356 spamd_address settings.
2358 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2359 and controls are allowed in which ACLs. There were a couple of minor
2360 errors. Some of the entries in the conditions table (which is a table of
2361 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2362 negation of where the condition IS allowed.
2364 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2366 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2367 header file does not have a version number, so I've had to invent a new
2368 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2369 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2370 radiusclient), but it was contributed by a Radius user.
2372 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2373 files or directories, for OpenSSL.
2375 PH/19 When an Exim process that is running as root has to create an Exim log
2376 file, it does so in a subprocess that runs as exim:exim so as to get the
2377 ownership right at creation (otherwise, other Exim processes might see
2378 the file with the wrong ownership). There was no test for failure of this
2379 fork() call, which would lead to the process getting stuck as it waited
2380 for a non-existent subprocess. Forks do occasionally fail when resources
2381 run out. I reviewed all the other calls to fork(); they all seem to check
2384 PH/20 When checking for unexpected SMTP input at connect time (before writing
2385 the banner), Exim was not dealing correctly with a non-positive return
2386 from the read() function. If the client had disconnected by this time,
2387 the result was a log entry for a synchronization error with an empty
2388 string after "input=" when read() returned zero. If read() returned -1
2389 (an event I could not check), uninitialized data bytes were printed.
2390 There were reports of junk text (parts of files, etc) appearing after
2393 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2395 PH/22 Added support for macro redefinition, and (re)definition in between
2396 driver and ACL definitions.
2398 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2399 forgetting to use the resulting value; it was using the unexpanded value.
2401 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2402 hadn't been configured. The fix is from Juergen Kreileder, who
2403 understands it better than I do:
2405 "Here's what I see happening with three configured cyrus_sasl
2406 authenticators configured (plain, login, cram-md5):
2408 On startup auth_cyrus_sasl_init() gets called for each of these.
2409 This means three calls to sasl_listmech() without a specified mech_list.
2410 => SASL tests which mechs of all available mechs actually work
2411 => three warnings about OTP not working
2412 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2414 With the patch, sasl_listmech() also gets called three times. But now
2415 SASL's mech_list option is set to the server_mech specified in the the
2416 authenticator. Or in other words, the answer from sasl_listmech()
2417 gets limited to just the mech you're testing for (which is different
2419 => the return list contains just 'plain' or 'login', 'cram-md5' or
2420 nothing depending on the value of ob->server_mech.
2422 I've just tested the patch: Authentication still works fine,
2423 unavailable mechs specified in the exim configuration are still
2424 caught, and the auth.log warnings about OTP are gone."
2426 PH/25 When debugging is enabled, the contents of the command line are added
2427 to the debugging output, even when log_selector=+arguments is not
2430 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2431 answer is "GNU", and only if the return is "GNU/something" is the answer
2434 PH/27 $acl_verify_message is now set immediately after the failure of a
2435 verification in an ACL, and so is available in subsequent modifiers. In
2436 particular, the message can be preserved by coding like this:
2438 warn !verify = sender
2439 set acl_m0 = $acl_verify_message
2441 Previously, $acl_verify_message was set only while expanding "message"
2442 and "log_message" when a very denied access.
2444 PH/28 Modified OS/os.c-Linux with
2446 -#ifndef OS_LOAD_AVERAGE
2447 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2449 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2450 nomenclature these days.)
2452 PH/29 Installed patch from the Sieve maintainer that adds the options
2453 sieve_useraddress and sieve_subaddress to the redirect router.
2455 PH/30 In these circumstances:
2456 . Two addresses routed to the same list of hosts;
2457 . First host does not offer TLS;
2458 . First host accepts first address;
2459 . First host gives temporary error to second address;
2460 . Second host offers TLS and a TLS session is established;
2461 . Second host accepts second address.
2462 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2463 and peerdn, if requested) that were in fact used only for the second
2466 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2467 attention to any local part prefix or suffix that was matched by the
2468 router that accepted the address. It now behaves in the same way as it
2469 does for delivery: the affixes are removed from the local part unless
2470 rcpt_include_affixes is set on the transport.
2472 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2473 timeout during the DATA phase of an incoming message.
2475 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2476 applied a patch sanctioned by the Sieve maintainer.
2478 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2479 the uid or gid is negative. A case of a negative gid caused this to be
2480 noticed. The fix allows for either to be negative.
2482 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2483 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2486 PH/36 The change PH/12 above was broken. Fixed it.
2488 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2489 the grounds that routing the same address twice would always produce the
2490 same answer. This might have been true once, but it is certainly no
2491 longer true now. Routing a child address may depend on the previous
2492 routing that produced that child. Some complicated redirection strategies
2493 went wrong when messages had multiple recipients, and made Exim's
2494 behaviour dependent on the order in which the addresses were given.
2496 I have moved the duplicate checking until after the routing is complete.
2497 Exim scans the addresses that are assigned to local and remote
2498 transports, and removes any duplicates. This means that more work will be
2499 done, as duplicates will always all be routed, but duplicates are
2500 presumably rare, so I don't expect this is of any significance.
2502 For deliveries to pipes, files, and autoreplies, the duplicate checking
2503 still happens during the routing process, since they are not going to be
2506 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2507 It corrects a timeout issue with spamd. This is Ian's comment: "The
2508 background is that sometimes spamd either never reads data from a
2509 connection it has accepted, or it never writes response data. The exiscan
2510 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2511 blindly assumes that writes won't block so it may never time out."
2513 PH/39 Allow G after quota size as well as K and M.
2515 PH/40 The value set for $authenticated_id in an authenticator may not contain
2516 binary zeroes or newlines because the value is written to log lines and
2517 to spool files. There was no check on this. Now the value is run through
2518 the string_printing() function so that such characters are converted to
2519 printable escape sequences.
2521 PH/41 $message_linecount is a new variable that contains the total number of
2522 lines in the message. Compare $body_linecount, which is the count for the
2525 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2526 bounce and delay warning messages, except in certain special cases, which
2529 (a) An SMTP error message from a remote host;
2530 (b) A message specified in a :fail: redirection;
2531 (c) A message specified in a "fail" command in a system filter;
2532 (d) A message specified in a FAIL return from the queryprogram router;
2533 (e) A message specified by the cannot_route_message router option.
2535 In these cases only, Exim does include the error details in bounce and
2536 warning messages. There are also a few cases where bland messages such
2537 as "unrouteable address" or "local delivery error" are given.
2539 PH/43 $value is now also set for the "else" part of a ${run expansion.
2541 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2542 being worked on, but at least Exim now implements the latest version to
2545 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2546 process to complete was treated as a delivery failure, a timeout while
2547 writing the message to the pipe was logged, but erroneously treated as a
2548 successful delivery. Such timeouts include transport filter timeouts. For
2549 consistency with the overall process timeout, these timeouts are now
2550 treated as errors, giving rise to delivery failures by default. However,
2551 there is now a new Boolean option for the pipe transport called
2552 timeout_defer, which, if set TRUE, converts the failures into defers for
2553 both kinds of timeout. A transport filter timeout is now identified in
2556 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2557 systems where "make" and "gmake" are different, calling "gmake" at top
2558 level broke things. I've arranged for the value of $(MAKE) to be passed
2559 from the Makefile to this script so that it can call the same version of
2563 A note about Exim versions 4.44 and 4.50
2564 ----------------------------------------
2566 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2567 changes of various kinds. As a consequence, a big documentation update was
2568 needed. This delayed the release for rather longer than seemed good, especially
2569 in the light of a couple of (minor) security issues. Therefore, the changes
2570 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2571 release. So 4.44 and 4.50 are in effect two different branches that both start
2574 I have left the 4.50 change log unchanged; it contains all the changes since
2575 4.43. The change log for 4.44 is below; many of its items are identical to
2576 those for 4.50. This seems to be the most sensible way to preserve the
2577 historical information.
2583 1. Minor wording change to the doc/README.SIEVE file.
2585 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2586 computation of the current number of files was incorrect.
2588 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2589 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2592 4. Give more explanation in the error message when the command for a transport
2593 filter fails to execute.
2595 5. There are several places where Exim runs a non-Exim command in a
2596 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2597 was being done only for the command run by the queryprogram router. It is
2598 now done for all such subprocesses. The other cases are: ${run, transport
2599 filters, and the commands run by the lmtp and pipe transports.
2601 6. Added CONFIGURE_GROUP build-time option.
2603 7. Some older OS have a limit of 256 on the maximum number of file
2604 descriptors. Exim was using setrlimit() to set 1000 as a large value
2605 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2606 systems. I've change it so that if it can't get 1000, it tries for 256.
2608 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2609 was an oversight, and furthermore, ever since the addition of extra
2610 controls (e.g. 4.43/32), the checks on when to allow different forms of
2611 "control" were broken. There should now be diagnostics for all cases when a
2612 control that does not make sense is encountered.
2614 9. Added the /retain_sender option to "control=submission".
2616 10. $recipients is now available in the predata ACL (oversight).
2618 11. Tidy the search cache before the fork to do a delivery from a message
2619 received from the command line. Otherwise the child will trigger a lookup
2620 failure and thereby defer the delivery if it tries to use (for example) a
2621 cached ldap connection that the parent has called unbind on.
2623 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2624 of $address_data from the recipient verification was clobbered by the
2625 sender verification.
2627 13. The value of address_data from a sender verification is now available in
2628 $sender_address_data in subsequent conditions in the ACL statement.
2630 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2632 15. Added a new option "connect=<time>" to callout options, to set a different
2635 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2636 was its contents. (It was OK if the option was not defined at all.)
2638 17. A "Completed" log line is now written for messages that are removed from
2639 the spool by the -Mrm option.
2641 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2642 information about exactly what failed.
2644 19. Added -dd to debug only the daemon process.
2646 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2647 handles timeouts, both on the server side and network timeouts. Renamed the
2648 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2650 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2651 It is now set to "smtps".
2653 22. $host_address is now set to the target address during the checking of
2654 ignore_target_hosts.
2656 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2657 being passed; this would have caused $sender_host_name to have been used if
2658 matching the list had actually called for a host name (not very likely,
2659 since this list is usually IP addresses). A host name is now passed as
2662 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2663 code that specifies a non-restarting handler (typically sigaction() in
2664 modern systems) in an attempt to fix a rare and obscure crash bug.
2666 25. Narrowed the window for a race in the daemon that could cause it to ignore
2667 SIGCHLD signals. This is not a major problem, because they are used only to
2668 wake it up if nothing else does.
2670 26. A malformed maildirsize file could cause Exim to calculate negative values
2671 for the mailbox size or file count. Odd effects could occur as a result.
2672 The maildirsize information is now recalculated if the size or filecount
2675 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2676 support for a long time. Removed HAVE_SYS_VFS_H.
2678 28. Installed the latest version of exipick from John Jetmore.
2680 29. In an address list, if the pattern was not a regular expression, an empty
2681 subject address (from a bounce message) matched only if the pattern was an
2682 empty string. Non-empty patterns were not even tested. This was the wrong
2683 because it is perfectly reasonable to use an empty address as part of a
2684 database query. An empty address is now tested by patterns that are
2685 lookups. However, all the other forms of pattern expect the subject to
2686 contain a local part and a domain, and therefore, for them, an empty
2687 address still always fails if the pattern is not itself empty.
2689 30. Exim went into a mad DNS loop when attempting to do a callout where the
2690 host was specified on an smtp transport, and looking it up yielded more
2691 than one IP address.
2693 31. Re-factored the code for checking spool and log partition space into a
2694 function that finds that data and another that does the check. The former
2695 is then used to implement four new variables: $spool_space, $log_space,
2696 $spool_inodes, and $log_inodes.
2698 32. The RFC2047 encoding function was originally intended for short strings
2699 such as real names; it was not keeping to the 75-character limit for
2700 encoded words that the RFC imposes. It now respects the limit, and
2701 generates multiple encoded words if necessary. To be on the safe side, I
2702 have increased the buffer size for the ${rfc2047: expansion operator from
2705 33. It is now permitted to omit both strings after an "if" condition; if the
2706 condition is true, the result is "true". As before, when the second string
2707 is omitted, a false condition yields an empty string. This makes it less
2708 cumbersome to write custom ACL and router conditions.
2710 34. Failure to deliver a bounce message always caused it to be frozen, even if
2711 there was an errors_to setting on the router. The errors_to setting is now
2714 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2715 canonical form (fully expanded) before being placed in
2716 $sender_host_address.
2718 36. The table in the code that translates DNS record types into text (T_A to
2719 "A" for instance) was missing entries for NS and CNAME. It is just possible
2720 that this could have caused confusion if both these types were looked up
2721 for the same domain, because the text type is used as part of Exim's
2722 per-process caching. But the chance of anyone hitting this buglet seems
2725 37. The dnsdb lookup has been extended in a number of ways.
2727 (1) There is a new type, "zns", which walks up the domain tree until it
2728 finds some nameserver records. It should be used with care.
2730 (2) There is a new type, "mxh", which is like "mx" except that it returns
2731 just the host names, not the priorities.
2733 (3) It is now possible to give a list of domains (or IP addresses) to be
2734 looked up. The behaviour when one of the lookups defers can be
2735 controlled by a keyword.
2737 (4) It is now possible to specify the separator character for use when
2738 multiple records are returned.
2740 38. The dnslists ACL condition has been extended: it is now possible to supply
2741 a list of IP addresses and/or domains to be looked up in a particular DNS
2744 39. Added log_selector=+queue_time_overall.
2746 40. When running the queue in the test harness, wait just a tad after forking a
2747 delivery process, to get repeatability of debugging output.
2749 41. Include certificate and key file names in error message when GnuTLS fails
2750 to set them up, because the GnuTLS error message doesn't include the name
2751 of the failing file when there is a problem reading it.
2753 42. Allow both -bf and -bF in the same test run.
2755 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2757 44. The "Exiscan patch" is now merged into the mainline Exim source.
2759 45. Sometimes the final signoff response after QUIT could fail to get
2760 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2761 < 0 before doing a fflush(). This bug looks as though it goes back to the
2762 introduction of TLS in release 3.20, but "sometimes" must have been rare
2763 because the tests only now provoked it.
2765 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2766 (this can affect the format of dates).
2768 47. exim_tidydb, when checking for the continued existence of a message for
2769 which it has found a message-specific retry record, was not finding
2770 messages that were in split spool directories. Consequently, it was
2771 deleting retry records that should have stayed in existence.
2773 48. Steve fixed some bugs in eximstats.
2775 49. The SPA authentication driver was not abandoning authentication and moving
2776 on to the next authenticator when an expansion was forced to fail,
2777 contradicting the general specification for all authenticators. Instead it
2778 was generating a temporary error. It now behaves as specified.
2780 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2781 (the order specifies the preference for clients). The order is now AES256,
2782 AES128, 3DES, ARCFOUR128.
2784 51. Small patch to Sieve code - explicitly set From: when generating an
2787 52. Exim crashed if a remote delivery caused a very long error message to be
2788 recorded - for instance if somebody sent an entire SpamAssassin report back
2789 as a large number of 550 error lines. This bug was coincidentally fixed by
2790 increasing the size of one of Exim's internal buffers (big_buffer) that
2791 happened as part of the Exiscan merge. However, to be on the safe side, I
2792 have made the code more robust (and fixed the comments that describe what
2795 53. Now that there can be additional text after "Completed" in log lines (if
2796 the queue_time_overall log selector is set), a one-byte patch to exigrep
2797 was needed to allow it to recognize "Completed" as not the last thing in
2800 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2801 patch that reportedly fixes this has been added. I am not expert enough to
2802 create a test for it. This is what the patch creator wrote:
2804 "I found a little strange behaviour of ldap code when working with
2805 Windows 2003 AD Domain, where users was placed in more than one
2806 Organization Units. When I tried to give exim partial DN, the exit code
2807 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2808 But simultaneously result of request was absolutely normal ldap result,
2809 so I produce this patch..."
2811 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2812 REFERENCE, so I have modified the code to exclude the patch when that macro
2815 55. Some experimental protocols are using DNS PTR records for new purposes. The
2816 keys for these records are domain names, not reversed IP addresses. The
2817 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2818 leaves it alone. Component reversal etc. now happens only for IP addresses.
2821 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2823 57. Double the size of the debug message buffer (to 2048) so that more of very
2824 long debug lines gets shown.
2826 58. The exicyclog utility now does better if the number of log files to keep
2827 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2829 59. Two changes related to the smtp_active_hostname option:
2831 (1) $smtp_active_hostname is now available as a variable.
2832 (2) The default for smtp_banner uses $smtp_active_hostname instead
2833 of $primary_hostname.
2835 60. The host_aton() function is supposed to be passed a string that is known
2836 to be a valid IP address. However, in the case of IPv6 addresses, it was
2837 not checking this. This is a hostage to fortune. Exim now panics and dies
2838 if the condition is not met. A case was found where this could be provoked
2839 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2840 components; fortuitously, this particular loophole had already been fixed
2841 by change 4.50/55 above.
2843 If there are any other similar loopholes, the new check in host_aton()
2844 itself should stop them being exploited. The report I received stated that
2845 data on the command line could provoke the exploit when Exim was running as
2846 exim, but did not say which command line option was involved. All I could
2847 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2848 running as the user.
2851 61. There was a buffer overflow vulnerability in the SPA authentication code
2852 (which came originally from the Samba project). I have added a test to the
2853 spa_base64_to_bits() function which I hope fixes it.
2856 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2857 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2859 63. The daemon start-up calls getloadavg() while still root for those OS that
2860 need the first call to be done as root, but it missed one case: when
2861 deliver_queue_load_max is set with deliver_drop_privilege. This is
2862 necessary for the benefit of the queue runner, because there is no re-exec
2863 when deliver_drop_privilege is set.
2865 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2866 This has been fixed.
2868 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2869 was in use, was not putting the data itself into the right store pool;
2870 consequently, it could be overwritten for a subsequent message in the same
2871 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2874 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2876 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2877 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2878 the function were treating the return as a boolean value, which happened to
2879 work because 0=false and not-0=true, but is not correct code.
2881 68. The host_aton() function was not handling scoped IPv6 addresses (those
2882 with, for example, "%eth0" on the end) correctly.
2884 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2885 negated items (that is, ~something) in unsigned ints. Some compilers
2886 apparently mutter when there is no cast.
2888 70. If an address verification called from an ACL failed, and did not produce a
2889 user-specific message (i.e. there was only a "system" message), nothing was
2890 put in $acl_verify_message. In this situation, it now puts the system
2893 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2894 session; change 4.31/43 added the unwanted input to the log line - except
2895 that it did not do this in the start of session case. It now does.
2897 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2898 This is wrong and can cause the other end to generate a synchronization
2899 error if it is another Exim or anything else that does the synchronization
2900 check. A QUIT command is no longer sent after a timeout.
2902 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2903 during host lookups.
2905 74. The defer_ok option of callout verification was not working if it was used
2906 when verifying addresses in header lines, that is, for this case:
2908 verify = header_sender/callout=defer_ok
2910 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2911 those file descriptors could be used for SMTP connections. If anything
2912 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2913 could be sent to the SMTP client, causing chaos. The daemon now opens
2914 stdin, stdout, and stderr to /dev/null when it puts itself into the
2917 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2918 log by default. The user can override this with suitable Perl magic.
2920 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2921 the log message when discard triggers, was not working for the DATA ACL or
2922 for the non-SMTP ACL.
2924 78. Error message wording change in sieve.c.
2926 79. If smtp_accept_max_per_host was set, the number of connections could be
2927 restricted to fewer than expected, because the daemon was trying to set up
2928 a new connection before checking whether the processes handling previous
2929 connections had finished. The check for completed processes is now done
2930 earlier. On busy systems, this bug wouldn't be noticed because something
2931 else would have woken the daemon, and it would have reaped the completed
2934 80. If a message was submitted locally by a user whose login name contained one
2935 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2936 It caused a spool format error. I have fixed the spool reading code. A
2937 related problem was that the "from" clause in the Received: line became
2938 illegal because of the space(s). It is now covered by ${quote_local_part.
2940 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2943 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2945 83. Patch from Sieve maintainer to fix unterminated string problem in
2946 "vacation" handling.
2948 84. Some minor changes to the Linux configuration files to help with other
2949 OS variants using glibc.
2951 85. One more patch for Sieve to update vacation handling to latest spec.
2954 ----------------------------------------------------
2955 See the note above about the 4.44 and 4.50 releases.
2956 ----------------------------------------------------
2962 1. Change 4.43/35 introduced a bug that caused file counts to be
2963 incorrectly computed when quota_filecount was set in an appendfile
2966 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2967 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2970 3. Give more explanation in the error message when the command for a transport
2971 filter fails to execute.
2973 4. There are several places where Exim runs a non-Exim command in a
2974 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2975 was being done only for the command run by the queryprogram router. It is
2976 now done for all such subprocesses. The other cases are: ${run, transport
2977 filters, and the commands run by the lmtp and pipe transports.
2979 5. Some older OS have a limit of 256 on the maximum number of file
2980 descriptors. Exim was using setrlimit() to set 1000 as a large value
2981 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2982 systems. I've change it so that if it can't get 1000, it tries for 256.
2984 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2985 was an oversight, and furthermore, ever since the addition of extra
2986 controls (e.g. 4.43/32), the checks on when to allow different forms of
2987 "control" were broken. There should now be diagnostics for all cases when a
2988 control that does not make sense is encountered.
2990 7. $recipients is now available in the predata ACL (oversight).
2992 8. Tidy the search cache before the fork to do a delivery from a message
2993 received from the command line. Otherwise the child will trigger a lookup
2994 failure and thereby defer the delivery if it tries to use (for example) a
2995 cached ldap connection that the parent has called unbind on.
2997 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2998 of $address_data from the recipient verification was clobbered by the
2999 sender verification.
3001 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3002 was its contents. (It was OK if the option was not defined at all.)
3004 11. A "Completed" log line is now written for messages that are removed from
3005 the spool by the -Mrm option.
3007 12. $host_address is now set to the target address during the checking of
3008 ignore_target_hosts.
3010 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3011 being passed; this would have caused $sender_host_name to have been used if
3012 matching the list had actually called for a host name (not very likely,
3013 since this list is usually IP addresses). A host name is now passed as
3016 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3017 code that specifies a non-restarting handler (typically sigaction() in
3018 modern systems) in an attempt to fix a rare and obscure crash bug.
3020 15. Narrowed the window for a race in the daemon that could cause it to ignore
3021 SIGCHLD signals. This is not a major problem, because they are used only to
3022 wake it up if nothing else does.
3024 16. A malformed maildirsize file could cause Exim to calculate negative values
3025 for the mailbox size or file count. Odd effects could occur as a result.
3026 The maildirsize information is now recalculated if the size or filecount
3029 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3030 support for a long time. Removed HAVE_SYS_VFS_H.
3032 18. Updated exipick to current release from John Jetmore.
3034 19. Allow an empty sender to be matched against a lookup in an address list.
3035 Previously the only cases considered were a regular expression, or an
3038 20. Exim went into a mad DNS lookup loop when doing a callout where the
3039 host was specified on the transport, if the DNS lookup yielded more than
3042 21. The RFC2047 encoding function was originally intended for short strings
3043 such as real names; it was not keeping to the 75-character limit for
3044 encoded words that the RFC imposes. It now respects the limit, and
3045 generates multiple encoded words if necessary. To be on the safe side, I
3046 have increased the buffer size for the ${rfc2047: expansion operator from
3049 22. Failure to deliver a bounce message always caused it to be frozen, even if
3050 there was an errors_to setting on the router. The errors_to setting is now
3053 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3054 canonical form (fully expanded) before being placed in
3055 $sender_host_address.
3057 24. Updated eximstats to version 1.33
3059 25. Include certificate and key file names in error message when GnuTLS fails
3060 to set them up, because the GnuTLS error message doesn't include the name
3061 of the failing file when there is a problem reading it.
3063 26. Expand error message when OpenSSL has problems setting up cert/key files.
3066 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3067 (this can affect the format of dates).
3069 28. exim_tidydb, when checking for the continued existence of a message for
3070 which it has found a message-specific retry record, was not finding
3071 messages that were in split spool directories. Consequently, it was
3072 deleting retry records that should have stayed in existence.
3074 29. eximstats updated to version 1.35
3075 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3076 1.35 - bugfix such that pie charts by volume are generated correctly
3078 30. The SPA authentication driver was not abandoning authentication and moving
3079 on to the next authenticator when an expansion was forced to fail,
3080 contradicting the general specification for all authenticators. Instead it
3081 was generating a temporary error. It now behaves as specified.
3083 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3084 (the order specifies the preference for clients). The order is now AES256,
3085 AES128, 3DES, ARCFOUR128.
3087 31. Small patch to Sieve code - explicitly set From: when generating an
3090 32. Exim crashed if a remote delivery caused a very long error message to be
3091 recorded - for instance if somebody sent an entire SpamAssassin report back
3092 as a large number of 550 error lines. This bug was coincidentally fixed by
3093 increasing the size of one of Exim's internal buffers (big_buffer) that
3094 happened as part of the Exiscan merge. However, to be on the safe side, I
3095 have made the code more robust (and fixed the comments that describe what
3098 33. Some experimental protocols are using DNS PTR records for new purposes. The
3099 keys for these records are domain names, not reversed IP addresses. The
3100 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3101 leaves it alone. Component reversal etc. now happens only for IP addresses.
3104 34. The host_aton() function is supposed to be passed a string that is known
3105 to be a valid IP address. However, in the case of IPv6 addresses, it was
3106 not checking this. This is a hostage to fortune. Exim now panics and dies
3107 if the condition is not met. A case was found where this could be provoked
3108 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3109 components; fortuitously, this particular loophole had already been fixed
3110 by change 4.50/55 or 4.44/33 above.
3112 If there are any other similar loopholes, the new check in host_aton()
3113 itself should stop them being exploited. The report I received stated that
3114 data on the command line could provoke the exploit when Exim was running as
3115 exim, but did not say which command line option was involved. All I could
3116 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3117 running as the user.
3120 35. There was a buffer overflow vulnerability in the SPA authentication code
3121 (which came originally from the Samba project). I have added a test to the
3122 spa_base64_to_bits() function which I hope fixes it.
3125 36. The daemon start-up calls getloadavg() while still root for those OS that
3126 need the first call to be done as root, but it missed one case: when
3127 deliver_queue_load_max is set with deliver_drop_privilege. This is
3128 necessary for the benefit of the queue runner, because there is no re-exec
3129 when deliver_drop_privilege is set.
3131 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3132 was in use, was not putting the data itself into the right store pool;
3133 consequently, it could be overwritten for a subsequent message in the same
3134 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3137 38. Sometimes the final signoff response after QUIT could fail to get
3138 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3139 < 0 before doing a fflush(). This bug looks as though it goes back to the
3140 introduction of TLS in release 3.20, but "sometimes" must have been rare
3141 because the tests only now provoked it.
3147 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3148 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3149 Now it returns an integer. A number of calls were still expecting a T/F
3150 return. Fortuitously, in all cases, the tests worked in OK situations,
3151 which is the norm. However, things would have gone wrong on any write
3152 failures on the smtp file descriptor. This function is used when sending
3153 messages over SMTP and also when doing verify callouts.
3155 2. When Exim is called to do synchronous delivery of a locally submitted
3156 message (the -odf or -odi options), it no longer closes stderr before doing
3159 3. Implemented the mua_wrapper option.
3161 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3163 5. Implemented the functions header_remove(), header_testname(),
3164 header_add_at_position(), and receive_remove_recipient(), and exported them
3167 6. If an ACL "warn" statement specified the addition of headers, Exim already
3168 inserted X-ACL-Warn: at the start if there was no header name. However, it
3169 was not making this test for the second and subsequent header lines if
3170 there were newlines in the string. This meant that an invalid header could
3171 be inserted if Exim was badly configured.
3173 7. Allow an ACL "warn" statement to add header lines at the start or after all
3174 the Received: headers, as well as at the end.
3176 8. Added the rcpt_4xx retry error code.
3178 9. Added postmaster_mailfrom=xxx to callout verification option.
3180 10. Added mailfrom=xxxx to the callout verification option, for verify=
3183 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3184 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3186 12. Inserted some casts to stop certain compilers warning when using pointer
3187 differences as field lengths or precisions in printf-type calls (mostly
3188 affecting debugging statements).
3190 13. Added optional readline() support for -be (dynamically loaded).
3192 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3193 same clock tick as a message's arrival, so that its received time was the
3194 same as the "first fail" time on the retry record, and that message
3195 remained on the queue past the ultimate address timeout, every queue runner
3196 would try a delivery (because it was past the ultimate address timeout) but
3197 after another failure, the ultimate address timeout, which should have then
3198 bounced the address, did not kick in. This was a "< instead of <=" error;
3199 in most cases the first failure would have been in the next clock tick
3200 after the received time, and all would be well.
3202 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3203 being recognized when the domain list was tested by the match_domain
3204 condition in an expansion string.
3206 16. Added the ${str2b64: operator.
3208 17. Exim was always calling setrlimit() to set a large limit for the number of
3209 processes, without checking whether the existing limit was already
3210 adequate. (It did check for the limit on file descriptors.) Furthermore,
3211 errors from getrlimit() and setrlimit() were being ignored. Now they are
3212 logged to the main and panic logs, but Exim does carry on, to try to do its
3213 job under whatever limits there are.
3215 18. Imported PCRE 5.0.
3217 19. Trivial typo in log message " temporarily refused connection" (the leading
3220 20. If the log selector return_path_on_delivery was set and an address was
3221 redirected to /dev/null, the delivery process crashed because it assumed
3222 that a return path would always be set for a "successful" delivery. In this
3223 case, the whole delivery is bypassed as an optimization, and therefore no
3226 21. Internal re-arrangement: the function for sending a challenge and reading
3227 a response while authentication was assuming a zero-terminated challenge
3228 string. It's now changed to take a pointer and a length, to allow for
3229 binary data in such strings.
3231 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3233 23. Exim was not respecting finduser_retries when seeking the login of the
3234 uid under which it was called; it was always trying 10 times. (The default
3235 setting of finduser_retries is zero.) Also, it was sleeping after the final
3236 failure, which is pointless.
3238 24. Implemented tls_on_connect_ports.
3240 25. Implemented acl_smtp_predata.
3242 26. If the domain in control=submission is set empty, Exim assumes that the
3243 authenticated id is a complete email address when it generates From: or
3244 Sender: header lines.
3246 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3247 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3248 chown and chgrp in /bin and hostname in /usr/bin.
3250 28. Exim was keeping the "process log" file open after each use, just as it
3251 does for the main log. This opens the possibility of it remaining open for
3252 long periods when the USR1 signal hits a daemon. Occasional processlog
3253 errors were reported, that could have been caused by this. Anyway, it seems
3254 much more sensible not to leave this file open at all, so that is what now
3257 29. The long-running daemon process does not normally write to the log once it
3258 has entered its main loop, and it closes the log before doing so. This is
3259 so that log files can straightforwardly be renamed and moved. However,
3260 there are a couple of unusual error situations where the daemon does write
3261 log entries, and I had neglected to close the log afterwards.
3263 30. The text of an SMTP error response that was received during a remote
3264 delivery was being truncated at 512 bytes. This is too short for some of
3265 the long messages that one sometimes sees. I've increased the limit to
3268 31. It is now possible to make retry rules that apply only when a message has a
3269 specific sender, in particular, an empty sender.
3271 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3272 it possible to be selective about when SMTP synchronization is enforced.
3274 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3276 32. Implemented hosts_connection_nolog.
3278 33. Added an ACL for QUIT.
3280 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3283 35. Added mailbox_size and mailbox_filecount to appendfile.
3285 36. Added control = no_multiline_responses to ACLs.
3287 37. There was a bug in the logic of the code that waits for the clock to tick
3288 in the case where the clock went backwards by a substantial amount such
3289 that the microsecond fraction of "now" was more than the microsecond
3290 fraction of "then" (but the whole seconds number was less).
3292 38. Added support for the libradius Radius client library this is found on
3293 FreeBSD (previously only the radiusclient library was supported).
3299 1. When certain lookups returned multiple values in the form name=value, the
3300 quoting of the values was not always being done properly. Specifically:
3301 (a) If the value started with a double quote, but contained no whitespace,
3303 (b) If the value contained whitespace other than a space character (i.e.
3304 tabs or newlines or carriage returns) it was not quoted.
3305 This fix has been applied to the mysql and pgsql lookups by writing a
3306 separate quoting function and calling it from the lookup code. The fix
3307 should probably also be applied to nisplus, ibase and oracle lookups, but
3308 since I cannot test any of those, I have not disturbed their existing code.
3310 2. A hit in the callout cache for a specific address caused a log line with no
3311 reason for rejecting RCPT. Now it says "Previous (cached) callout
3312 verification failure".
3314 3. There was an off-by-one bug in the queryprogram router. An over-long
3315 return line was truncated at 256 instead of 255 characters, thereby
3316 overflowing its buffer with the terminating zero. As well as fixing this, I
3317 have increased the buffer size to 1024 (and made a note to document this).
3319 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3320 when Exim is waiting for an SMTP response from a remote server, Exim
3321 restarts its select() call on the socket, thereby resetting its timeout.
3322 This is not a problem when such interrupts are rare. Somebody set up a cron
3323 job to run exiwhat every 2 minutes, which is less than the normal select()
3324 timeout (5 or 10 minutes). This meant that the select() timeout never
3325 kicked in because it was always reset. I have fixed this by comparing the
3326 time when an interrupt arrives with the time at the start of the first call
3327 to select(). If more time than the timeout has elapsed, the interrupt is
3328 treated as a timeout.
3330 5. Some internal re-factoring in preparation for the addition of Sieve
3331 extensions (by MH). In particular, the "personal" test is moved to a
3332 separate function, and given an option for scanning Cc: and Bcc: (which is
3333 not set for Exim filters).
3335 6. When Exim created an email address using the login of the caller as the
3336 local part (e.g. when creating a From: or Sender: header line), it was not
3337 quoting the local part when it contained special characters such as @.
3339 7. Installed new OpenBSD configuration files.
3341 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3342 try to make them clearer.
3344 9. Callout options, other than the timeout value, were being ignored when
3345 verifying sender addresses in header lines. For example, when using
3347 verify = header_sender/callout=no_cache
3349 the cache was (incorrectly) being used.
3351 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3352 only the Cygwin environment.
3354 11. When return_path_on_delivery was set as a log selector, if different remote
3355 addresses in the same message used different return paths and parallel
3356 remote delivery occurred, the wrong values would sometimes be logged.
3357 (Whenever a remote delivery process finished, the return path value from
3358 the most recently started remote delivery process was logged.)
3360 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3361 lines when AUTH and/or TLS are in use. This is the "received protocol"
3362 field. Exim used to use "asmtp" for authenticated SMTP, without any
3363 indication (in the protocol name) for TLS use. Now it follows the RFC and
3364 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3365 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3366 names appear in log lines as well as in Received: header lines.
3368 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3369 extensions, and comparison tests, and to fix some bugs.
3371 14. Changes to the "personal" filter test:
3373 (1) The test was buggy in that it was just doing the equivalent of
3374 "contains" tests on header lines. For example, if a user's address was
3375 anne@some.where, the "personal" test would incorrectly be true for
3377 To: susanne@some.where
3379 This test is now done by extracting each address from the header in turn,
3380 and checking the entire address. Other tests that are part of "personal"
3381 are now done using regular expressions (for example, to check local parts
3382 of addresses in From: header lines).
3384 (2) The list of non-personal local parts in From: addresses has been
3385 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3386 taken from the Sieve specification recommendations.
3388 (3) If the message contains any header line starting with "List-" it is
3389 treated as non-personal.
3391 (4) The test for "circular" in the Subject: header line has been removed
3392 because it now seems ill-conceived.
3394 15. Minor typos in src/EDITME comments corrected.
3396 16. Installed latest exipick from John Jetmore.
3398 17. If headers_add on a router specified a text string that was too long for
3399 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3400 of string_sprintf() is now avoided.
3402 18. $message_body_size was not set (it was always zero) when running the DATA
3403 ACL and the local_scan() function.
3405 19. For the "mail" command in an Exim filter, no default was being set for
3406 the once_repeat time, causing a random time value to be used if "once" was
3407 specified. (If the value happened to be <= 0, no repeat happened.) The
3408 default is now 0s, meaning "never repeat". The "vacation" command was OK
3409 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3410 (I found it when inspecting the code).
3412 20. There is now an overall timeout for performing a callout verification. It
3413 defaults to 4 times the callout timeout, which applies to individual SMTP
3414 commands during the callout. The overall timeout applies when there is more
3415 than one host that can be tried. The timeout is checked before trying the
3416 next host. This prevents very long delays if there are a large number of
3417 hosts and all are timing out (e.g. when the network connections are timing
3418 out). The value of the overall timeout can be changed by specifying an
3419 additional sub-option for "callout", called "maxwait". For example:
3421 verify = sender/callout=5s,maxwait=20s
3423 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3424 to the end before writing, but this should make it even safer).
3426 22. Exim was forgetting that it had advertised PIPELINING for the second and
3427 subsequent messages on an SMTP connection. It was also not resetting its
3428 memory on STARTTLS and an internal HELO.
3430 23. When Exim logs an SMTP synchronization error within a session, it now
3431 records whether PIPELINING has been advertised or not.
3433 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3434 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3435 rather than long int.
3437 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3439 26. Added the never_mail option to autoreply.
3445 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3446 crash if the getsockname() call failed; this can happen if a connection is
3447 closed very soon after it is established. The problem was simply in the
3448 order in which certain operations were done, causing Exim to try to write
3449 to the SMTP stream before it had set up the file descriptor. The bug has
3450 been fixed by making things happen in the correct order.
3456 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3457 before the connection was closed, thus losing the rejection response.
3459 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3460 some early Solaris releases, but causes trouble in current releases where
3461 socklen_t is defined.
3463 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3466 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3469 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3470 that releasing the top of it at the end releases what was used for sub-
3471 expansions (unless the block got too big). However, discard this block if
3472 the first thing is a variable or header, so that we can use its block when
3473 it is dynamic (useful for very large $message_headers, for example).
3475 6. Lookups now cache *every* query, not just the most recent. A new, separate
3476 store pool is used for this. It can be recovered when all lookup caches are
3477 flushed. Lookups now release memory at the end of their result strings.
3478 This has involved some general refactoring of the lookup sources.
3480 7. Some code has been added to the store_xxx() functions to reduce the amount
3481 of flapping under certain conditions.
3483 8. log_incoming_interface used to affect only the <= reception log lines. Now
3484 it causes the local interface and port to be added to several more SMTP log
3485 lines, for example "SMTP connection from", and rejection lines.
3487 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3489 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3491 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3492 could be overwritten at the end of the current message (or the start of a
3493 new message if it was set in a HELO ACL). The value is now preserved for
3494 the duration of the SMTP connection.
3496 12. If a transport had a headers_rewrite setting, and a matching header line
3497 contained an unqualified address, that address was qualified, even if it
3498 did not match any rewriting rules. The underlying bug was that the values
3499 of the flags that permit the existence of unqualified sender and recipient
3500 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3501 non-local messages, and by -bnq for local messages) were not being
3502 preserved with the message after it was received.
3504 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3505 "next input=" as part of the text comprising the host identity instead of
3506 the correct text. The code was using the same buffer for two different
3507 strings. However, depending on which order the printing function evaluated
3508 its arguments, the bug did not always show up. Under Linux, for example, my
3509 test suite worked just fine.
3511 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3512 doesn't work with some older versions of Perl. It has been changed to "my",
3513 which in any case is probably the better facility to use.
3515 15. A really picky compiler found some instances of statements for creating
3516 error messages that either had too many or two few arguments for the format
3519 16. The size of the buffer for calls to the DNS resolver has been increased
3520 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3521 for addresses that have a lot of PTR records. This alleviates a problem; it
3522 does not fully solve it.
3524 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3525 buffer now truncates the list and logs the incident, which is the same
3526 action as happens when Exim is looking up a host name and its aliases.
3527 Previously in this situation something unpredictable would happen;
3528 sometimes it was "internal error: store_reset failed".
3530 18. If a server dropped the connection unexpectedly when an Exim client was
3531 using GnuTLS and trying to read a response, the client delivery process
3532 crashed while trying to generate an error log message.
3534 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3535 string, for example:
3537 warn message = H1: something\nH2: something
3539 the text was added as a single header line from Exim's point of view
3540 though it ended up OK in the delivered message. However, searching for the
3541 second and subsequent header lines using $h_h2: did not work. This has been
3542 fixed. Similarly, if a system filter added multiple headers in this way,
3543 the routers could not see them.
3545 20. Expanded the error message when iplsearch is called with an invalid key to
3546 suggest using net-iplsearch in a host list.
3548 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3549 ACLs are no longer actually imposed (and a message to that effect is
3552 22. If a "gecos" field in a passwd entry contained escaped characters, in
3553 particular, if it contained a \" sequence, Exim got it wrong when building
3554 a From: or a Sender: header from that name. A second bug also caused
3555 incorrect handling when an unquoted " was present following a character
3556 that needed quoting.
3558 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3559 was not being matched caselessly.
3561 24. Arranged for all hyphens in the exim.8 source to be escaped with
3564 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3565 properly. Recipient callout cache records were still being keyed to include
3566 the sender, even when use_sender was set false. This led to far more
3567 callouts that were necessary. The sender is no longer included in the key
3568 when use_sender is false.
3570 26. Added "control = submission" modifier to ACLs.
3572 27. Added the ${base62d: operator to decode base 62 numbers.
3574 28. dnsdb lookups can now access SRV records.
3576 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3577 the configuration file.
3579 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3580 (-v) mode. This makes the output for a verbose queue run more intelligible.
3582 31. Added a use_postmaster feature to recipient callouts.
3584 32. Added the $body_zerocount variable, containing the number of binary zero
3585 bytes in the message body.
3587 33. The time of last modification of the "new" subdirectory is now used as the
3588 "mailbox time last read" when there is a quota error for a maildir
3591 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3593 35. Added +ignore_unknown as a special item in host lists.
3595 36. Code for decoding IPv6 addresses in host lists is now included, even if
3596 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3597 address was recognized as an IP address, but was then not correctly decoded
3598 into binary, causing unexpected and incorrect effects when compared with
3605 1. Very minor rewording of debugging text in manualroute to say "list of
3606 hosts" instead of "hostlist".
3608 2. If verify=header_syntax was set, and a header line with an unqualified
3609 address (no domain) and a large number of spaces between the end of the
3610 name and the colon was received, the reception process suffered a buffer
3611 overflow, and (when I tested it) crashed. This was caused by some obsolete
3612 code that should have been removed. The fix is to remove it!
3614 3. When running in the test harness, delay a bit after writing a bounce
3615 message to get a bit more predictability in the log output.
3617 4. Added a call to search_tidyup() just before forking a reception process. In
3618 theory, someone could use a lookup in the expansion of smtp_accept_max_
3619 per_host which, without the tidyup, could leave open a database connection.
3621 5. Added the variables $recipient_data and $sender_data which get set from a
3622 lookup success in an ACL "recipients" or "senders" condition, or a router
3623 "senders" option, similar to $domain_data and $local_part_data.
3625 6. Moved the writing of debug_print from before to after the "senders" test
3628 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3629 problems for message scanning, either using a data ACL, or using
3630 local_scan() because the Received: header was not generated till after they
3631 were called (in order to set the time as the time of reception completion).
3632 I have revised the way this works. The header is now generated after the
3633 body is received, but before the ACL or local_scan() are called. After they
3634 are run, the timestamp in the header is updated.
3640 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3641 before starting a queue runner without re-exec. This happened only when
3642 deliver_drop_privilege was set or when the Exim user was set to root. The
3643 effect of the bug was that timeouts during subsequent deliveries caused
3644 crashes instead of being properly handled. The handler is now left at its
3645 default (and expected) setting.
3647 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3648 message, again when deliver_drop_privilege is set or Exim is run as root.
3649 The bug described in (1) was not present in this case, but the tidying up
3650 of the other signals was missing. I have made the two cases consistent.
3652 3. The ignore_target_hosts setting on a manualroute router was being ignored
3653 for hosts that were looked up using the /MX notation.
3655 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3658 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3659 operated on the sender address. After changing the $sender_address to <>
3660 for the sender address verify, Exim was re-instated it as the original
3661 (before rewriting) address, but remembering that it had rewritten it, so it
3662 wasn't rewriting it again. This bug also had the effect of breaking the
3663 sender address verification caching when the sender address was rewritten.
3665 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3666 This has been changed so that if the ip literal address matches
3667 ignore_target_hosts, the router declines.
3669 7. Added expansion conditions match_domain, match_address, and match_local_
3670 part (NOT match_host).
3672 8. The placeholder for the Received: header didn't have a length field set.
3674 9. Added code to Exim itself and to exim_lock to test for a specific race
3675 condition that could lead to file corruption when using MBX delivery. The
3676 issue is with the lockfile that is created in /tmp. If this file is removed
3677 after a process has opened it but before that process has acquired a lock,
3678 there is the potential for a second process to recreate the file and also
3679 acquire a lock. This could lead to two Exim processes writing to the file
3680 at the same time. The added code performs the same test as UW imapd; it
3681 checks after acquiring the lock that its file descriptor still refers to
3682 the same named file.
3684 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3685 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3688 11. Added the smtp_active_hostname option. If used, this will typically be made
3689 to depend on the incoming interface address. Because $interface_address is
3690 not set up until the daemon has forked a reception process, error responses
3691 that can happen earlier (such as "too many connections") no longer contain
3694 12. If an expansion in a condition on a "warn" statement fails because a lookup
3695 defers, the "warn" statement is abandoned, and the next ACL statement is
3696 processed. Previously this caused the whole ACL to be aborted.
3698 13. Added the iplsearch lookup type.
3700 14. Added ident_timeout as a log selector.
3702 15. Added tls_certificate_verified as a log selector.
3704 16. Added a global option tls_require_ciphers (compare the smtp transport
3705 option of the same name). This controls incoming TLS connections.
3707 17. I finally figured out how to make tls_require_ciphers do a similar thing
3708 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3709 before starting the TLS session.
3711 18. Tabs are now shown as \t in -bP output.
3713 19. If the log selector return_path_on_delivery was set, Exim crashed when
3714 bouncing a message because it had too many Received: header lines.
3716 20. If two routers both had headers_remove settings, and the first one included
3717 a superfluous trailing colon, the final name in the first list and the
3718 first name in the second list were incorrectly joined into one item (with a
3719 colon in the middle).
3725 1. Added -C and -D options to the exinext utility, mainly to make it easier
3726 to include in the automated testing, but these could be helpful when
3727 multiple configurations are in use.
3729 2. The exinext utility was not formatting the output nicely when there was
3730 an alternate port involved in the retry record key, nor when there was a
3731 message id as well (for retries that were specific to a specific message
3732 and a specific host). It was also confused by IPv6 addresses, because of
3733 the additional colons they contain. I have fixed the IPv4 problem, and
3734 patched it up to do a reasonable job for IPv6.
3736 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3737 delivery, the log line now contains "pipelined" if PIPELINING was used.
3739 4. An SMTP transport process used to panic and die if the bind() call to set
3740 an explicit outgoing interface failed. This has been changed; it is now
3741 treated in the same way as a connect() failure.
3743 5. A reference to $sender_host_name in the part of a conditional expansion
3744 that was being skipped was still causing a DNS lookup. This no longer
3747 6. The def: expansion condition was not recognizing references to header lines
3748 that used bh_ and bheader_.
3750 7. Added the _cache feature to named lists.
3752 8. The code for checking quota_filecount in the appendfile transport was
3753 allowing one more file than it should have been.
3755 9. For compatibility with Sendmail, the command line option
3763 and sets the incoming protocol and host name (for trusted callers). The
3764 host name and its colon can be omitted when only the protocol is to be set.
3765 Note the Exim already has two private options, -pd and -ps, that refer to
3766 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3767 "s", but I don't think that's a major issue.
3769 10. A number of refactoring changes to the code, none of which should affect
3772 (a) The number of logging options was getting close to filling up the
3773 32-bit word that was used as a bit map. I have split them into two classes:
3774 those that are passed in the argument to log_write(), and those that are
3775 only ever tested independently outside of that function. These are now in
3776 separate 32-bit words, so there is plenty of room for expansion again.
3777 There is no change in the user interface or the logging behaviour.
3779 (b) When building, for example, log lines, the code previously used a
3780 macro that called string_cat() twice, in order to add two strings. This is
3781 not really sufficiently general. Furthermore, there was one instance where
3782 it was actually wrong because one of the argument was used twice, and in
3783 one call a function was used. (As it happened, calling the function twice
3784 did not affect the overall behaviour.) The macro has been replaced by a
3785 function that can join an arbitrary number of extra strings onto a growing
3788 (c) The code for expansion conditions now uses a table and a binary chop
3789 instead of a serial search (which was left over from when there were very
3790 few conditions). Also, it now recognizes conditions like "pam" even when
3791 the relevant support is not compiled in: a suitably worded error message is
3792 given if an attempt is made to use such a condition.
3794 11. Added ${time_interval:xxxxx}.
3796 12. A bug was causing one of the ddress fields not to be passed back correctly
3797 from remote delivery subprocesses. The field in question was not being
3798 subsequently used, so this caused to problems in practice.
3800 13. Added new log selectors queue_time and deliver_time.
3802 14. Might have fixed a bug in maildirsizefile handling that threw up
3803 "unexpected character" debug warnings, and recalculated the data
3804 unnecessarily. In any case, I expanded the warning message to give more
3807 15. Added the message "Restricted characters in address" to the statements in
3808 the default ACL that block characters like @ and % in local parts.
3810 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3811 Three changes have been made:
3813 (a) There was a serious bug; a negative response to MAIL caused the whole
3814 recipient domain to be cached as invalid, thereby blocking all messages
3815 to all local parts at the same domain, from all senders. This bug has
3816 been fixed. The domain is no longer cached after a negative response to
3817 MAIL if the sender used is not empty.
3819 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3822 (c) A new callout option, "use_sender" has been added for people who want
3823 the modified behaviour.
3829 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3832 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3833 indeed breaks things for older releases.
3835 3. Added additional logging to the case where there is a problem reading data
3836 from a filter that is running in a subprocess using a pipe, in order to
3837 try to track down a specific problem.
3839 4. Testing facility fudge: when running in the test harness and attempting
3840 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3841 getting "No route to host". Convert this to a timeout.
3843 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3846 6. Some OS don't have socklen_t but use size_t instead. This affects the
3847 fifth argument of getsockopt() amongst other things. This is now
3848 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3849 can be set for individual OS. I have set it for SunOS5, OSF1, and
3850 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3851 some earlier ones do not.
3853 7. Change 4.30/15 was not doing the test caselessly.
3855 8. The standard form for an IPv6 address literal was being rejected by address
3856 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3857 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3858 this, as well as the form without the "IPv6" on the front (but only when
3859 address literals are enabled, of course).
3861 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3863 10. Exim crashed if a message with an empty sender address specified by -f
3864 encountered a router with an errors_to setting. This could be provoked only
3865 by a command such as
3869 where an empty string was supplied; "<>" did not hit this bug.
3871 11. Installed PCRE release 4.5.
3873 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3874 remained set. It is now erased.
3876 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3877 times from message ids (which are base 36 rather than the normal 62).
3879 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3880 were being counted as actual protocol errors, and logged if the log
3881 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3882 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3883 and DATA following a set of rejected RCPTs do not count as protocol errors.
3884 In other words, Exim assumes they were pipelined, though this may not
3885 actually be the case. Of course, in all cases the client gets an
3886 appropriate error code.
3888 15. If a lookup fails in an ACL condition, a message about the failure may
3889 be available; it is used if testing the ACL cannot continue, because most
3890 such messages specify what the cause of the deferral is. However, some
3891 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3892 that caused an old message to be retained and used if a later statement
3893 caused a defer, replacing the real cause of the deferral.
3895 16. If an IP address had so many PTR records that the DNS lookup buffer
3896 was not large enough to hold them, Exim could crash while trying to process
3897 the truncated data. It now detects and logs this case.
3899 17. Further to 4.21/58, another change has been made: if (and only if) the
3900 first line of a message (the first header line) ends with CRLF, a bare LF
3901 in a subsequent header line has a space inserted after it, so as not to
3902 terminate the header.
3904 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3905 unnecessarily, used atoi() instead of strtol(), and didn't check the
3906 termination when getting file sizes from file names by regex.
3908 19. Completely re-implemented the support for maildirsize files, in the light
3909 of a number of problems with the previous contributed implementation
3910 (4.30/29). In particular:
3912 . If the quota is zero, the maildirsize file is maintained, but no quota is
3915 . If the maildir directory does not exist, it is created before any attempt
3916 to write a maildirsize file.
3918 . The quota value in the file is just a cache; if the quota is changed in
3919 the transport, the new value overrides.
3921 . A regular expression is available for excluding directories from the
3924 20. The autoreply transport checks the characters in options that define the
3925 message's headers; it allows continued headers, but it was checking with
3926 isspace() after an embedded newline instead of explicitly looking for a
3929 21. If all the "regular" hosts to which an address was routed had passed their
3930 expiry times, and had not reached their retry times, the address was
3931 bounced, even if fallback hosts were defined. Now Exim should go on to try
3934 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3935 equivalent code in the SMTP transport. Some hosts send humungous responses
3936 to HELO/EHLO, more than 1024 it seems.
3938 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3939 gives compiler warnings in some environments. I've now done it "properly",
3942 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3943 (because of problems with the built-in one) was declared to return uschar *
3944 instead of char *, causing compiler failure.
3946 25. Fixed a file descriptor leak when processing alias/forward files.
3948 26. Fixed a minor format string issue in dbfn.c.
3950 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3952 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3953 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3956 29. When a sender address is verified, it is cached, to save repeating the test
3957 when there is more than one recipient in a message. However, when the
3958 verification involves a callout, it is possible for different callout
3959 options to be set for different recipients. It is too complicated to keep
3960 track of this in the cache, so now Exim always runs a verification when a
3961 callout is required, relying on the callout cache for the optimization.
3962 The overhead is duplication of the address routing, but this should not be
3965 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3966 to be verified with callout=postmaster, and the main callout worked but the
3967 postmaster check failed, the verification correctly failed. However, if a
3968 subsequent RCPT command asked for sender verification *without* the
3969 postmaster check, incorrect caching caused this verification also to fail,
3972 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3973 it was not caching the DNS options (qualify_single, search_parents) that
3974 were used when the lookup failed. A subsequent lookup with different
3975 options therefore always gave the same answer, though there were cases
3976 where it should not have. (Example: a "domains = !$mx_any" option on a
3977 dnslookup router: the "domains" option is always processed without any
3978 widening, but the router might have qualify_single set.) Now Exim uses the
3979 cached value only when the same options are set.
3981 32. Added John Jetmore's "exipick" utility to the distribution.
3983 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3984 than a timeout (e.g. a certificate is required, and is not provided), an
3985 Exim server now closes the connection immediately. Previously it waited for
3986 the client to close - but if the client is SSL, it seems that they each
3987 wait for each other, leading to a delay before one of them times out.
3989 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3990 maintained 0.8.x compatibility because I don't think many are using it, and
3991 it is clearly obsolete.
3993 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3996 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3997 client certificate was expired. A simple patch fixes this, though I don't
3998 understand the full logic of why the verify callback is called multiple
4001 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4002 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4003 which causes problems with some clients (such as the Certicom SSL Plus
4004 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4005 disables the coutermeasure allowing Eudora to connect."
4007 38. Exim was not checking that a write() to a log file succeeded. This could
4008 lead to Bad Things if a log got too big, in particular if it hit a file
4009 size limit. Exim now panics and dies if it cannot write to a log file, just
4010 as it does if it cannot open a log file.
4012 39. Modified OS/Makefile-Linux so that it now contains
4014 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4016 The two -D definitions ensure that Exim is compiled with large file
4017 support, which makes it possible to handle log files that are bigger than
4020 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4021 instance) a domain was checked against a named list that involved a lookup,
4022 causing $domain_data to be set, then another domain was checked against the
4023 same list, then the first domain was re-checked, the value of $domain_data
4024 after the final check could be wrong. In particular, if the second check
4025 failed, it could be set empty. This bug probably also applied to
4028 41. The strip_trailing_dot option was not being applied to the address given
4029 with the -f command-line option.
4031 42. The code for reading a message's header from the spool was incrementing
4032 $received_count, but never initializing it. This meant that the value was
4033 incorrect (doubled) while delivering a message in the same process in which
4034 it was received. In the most common configuration of Exim, this never
4035 happens - a fresh exec is done - but it can happen when
4036 deliver_drop_privilege is set.
4038 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4039 it now includes up to 150 characters of the unexpected data in the log
4042 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4043 and building data strings. The size of both of these buffers was 10 000
4044 bytes - far larger than anybody would *ever* want, thought I. Needless to
4045 say, somebody hit the limit. I have increased the maximum line length to
4046 20 000 and the maximum data length of concatenated lines to 100 000. I have
4047 also fixed two bugs, because there was no checking on these buffers. Tsk,
4048 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4049 buffer is too small.
4051 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4052 lsearch lookups. Now it does.
4054 46. When parsing a route_list item in a manualroute router, a fixed-length
4055 buffer was used for the list of hosts. I made this 1024 bytes long,
4056 thinking that nobody would ever have a list of hosts that long. Wrong.
4057 Somebody had a whole pile of complicated expansion conditions, and the
4058 string was silently truncated, leading to an expansion error. It turns out
4059 that it is easier to change to an unlimited length (owing to other changes
4060 that have happened since this code was originally written) than to build
4061 structure for giving a limitation error. The length of the item that
4062 expands into the list of hosts is now unlimited.
4064 47. The lsearch lookup could not handle data where the length of text line was
4065 more than 4095 characters. Such lines were truncated, leading to shortened
4066 data being returned. It should now handle lines of any length.
4068 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4069 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4072 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4073 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4074 no longer happens in comment lines. A list of replacements is now placed
4075 at the head of all of the source files, except those whose only change is
4076 to replace PERL_COMMAND in the very first #! line.
4078 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4079 messages on the queue, with a bottom-up merge sort, using code contributed
4080 by Michael Haardt. This should make operations like -bp somewhat faster on
4081 large queues. It won't affect queue runners, except when queue_run_in_order
4084 51. Installed eximstats 1.31 in the distribution.
4086 52. Added support for SRV lookups to the dnslookup router.
4088 53. If an ACL referred to $message_body or $message_body_end, the value was not
4089 reset for any messages that followed in the same SMTP session.
4091 54. The store-handling optimization for building very long strings was not
4092 differentiating between the different store pools. I don't think this
4093 actually made any difference in practice, but I've tidied it.
4095 55. While running the routers to verify a sender address, $sender_address
4096 was still set to the sender address. This is wrong, because when routing to
4097 send a bounce to the sender, it would be empty. Therefore, I have changed
4098 it so that, while verifying a sender address, $sender_address is set to <>.
4099 (There is no change to what happens when verifying a recipient address.)
4101 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4102 target A or AAAA records (if not already returned) without resetting the
4103 qualify_single or search_parents options of the DNS resolver. These are
4104 inappropriate in this case because the targets of MX and SRV records must
4105 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4106 target that, when qualified, matched something in the local domain. These
4107 two options are now turned off when doing these lookups.
4109 57. It seems that at least some releases of Reiserfs (which does not have the
4110 concept of a fixed number of inodes) returns zero and not -1 for the
4111 number of available inodes. This interacted badly with check_spool_inodes,
4112 which assumed that -1 was the "no such thing" setting. What I have done is
4113 to check that the total number of inodes is greater than zero before doing
4114 the test of how many are available.
4116 58. When a "warn" ACL statement has a log_message modifier, the message is
4117 remembered, and not repeated. This is to avoid a lot of repetition when a
4118 message has many recipients that cause the same warning to be written.
4119 Howewer, Exim was preserving the list of already written lines for an
4120 entire SMTP session, which doesn't seem right. The memory is now reset if a
4121 new message is started.
4123 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4124 debugging output unless log_rewrite was also set.
4126 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4127 of (int)(handle) when we know that handle contains (void *)(-1).
4129 61. The Exim daemon panic-logs an error return when it closes the incoming
4130 connection. However "connection reset by peer" seems to be common, and
4131 isn't really an error worthy of noting specially, so that particular error
4134 62. When Exim is trying to find all the local interfaces, it used to panic and
4135 die if the ioctl to get the interface flags failed. However, it seems that
4136 on at least one OS (Solaris 9) it is possible to have an interface that is
4137 included in the list of interfaces, but for which you get a failure error
4138 for this call. This happens when the interface is not "plumbed" into a
4139 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4140 failure of the "get flags" call assumes that the interface is down.
4142 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4143 makes life easier for people who are doing arithmetic on fields extracted
4144 from dates, where you often get leading zeros that should not be
4145 interpreted as octal.
4147 64. Added qualify_domain to the redirect router, to override the global
4150 65. If a pathologically long header line contained very many addresses (the
4151 report of this problem mentioned 10 000) and each of them was rewritten,
4152 Exim could use up a very large amount of memory. (It kept on making new
4153 copies of the header line as it rewrote, and never released the old ones.)
4154 At the expense of a bit more processing, the header rewriting function has
4155 been changed so that it no longer eats memory in this way.
4157 66. The generation of the Received: header has been moved from the time that a
4158 message starts to be received, to the time that it finishes. The timestamp
4159 in the Received: header should now be very close to that of the <= log
4160 line. There are two side-effects of this change:
4162 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4163 logged header lines no longer include the local Received: line, because
4164 it has not yet been created. The same applies to a copy of the message
4165 that is returned to a non-SMTP sender when a message is rejected.
4167 (b) When a filter file is tested using -bf, no additional Received: header
4168 is added to the test message. After some thought, I decided that this
4171 This change does not affect the value of $received_for. It is still set
4172 after address rewriting, but before local_scan() is called.
4174 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4176 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4177 gave an unhelpful panic error message, and a defer error. I have managed to
4178 change this behaviour so that it now rejects any supplied certificate,
4179 which seems right, as the list of acceptable certificates is empty.
4181 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4182 gave an unhelpful defer error. I have not managed to make this reject any
4183 supplied certificates, but the error message it gives is "no certificate
4184 supplied", which is not helpful.
4186 70. exigrep's output now also includes lines that are not associated with any
4187 message, but which match the given pattern. Implemented by a patch from
4188 Martin Sluka, which also tidied up the Perl a bit.
4190 71. Recipient callout verification, like sender verification, was using <> in
4191 the MAIL FROM command. This isn't really the right thing, since the actual
4192 sender may affect whether the remote host accepts the recipient or not. I
4193 have changed it to use the actual sender in the callout; this means that
4194 the cache record is now keyed on a recipient/sender pair, not just the
4195 recipient address. There doesn't seem to be a real danger of callout loops,
4196 since a callout by the remote host to check the sender would use <>.
4197 [SEE ABOVE: changed after hitting problems.]
4199 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4200 temporary errors. However, in the case of such a code being given after
4201 the end of a data transmission (i.e. after ".") Exim was failing to write
4202 a retry record for the message. (Yes, there was some broken host that was
4203 actually sending 8xx at this point.)
4205 73. An unknown lookup type in a host list could cause Exim to panic-die when
4206 the list was checked. (An example that provoked this was putting <; in the
4207 middle of a list instead of at the start.) If this happened during a DATA
4208 ACL check, a -D file could be left lying around. This kind of configuration
4209 error no longer causes Exim to die; instead it causes a defer errror. The
4210 incident is still logged to the main and panic logs.
4212 74. Buglet left over from Exim 3 conversion. The message "too many messages
4213 in one connection" was written to the rejectlog but not the mainlog, except
4214 when address rewriting (yes!) was being logged.
4216 75. Added write_rejectlog option.
4218 76. When a system filter was run not as root (that is, when system_filter_user
4219 was set), the values of the $n variables were not being returned to the
4220 main process; thus, they were not subsequently available in the $sn
4223 77. Added +return_path_on_delivery log selector.
4225 78. A connection timeout was being treated differently from recipients deferred
4226 when testing hosts_max_try with a message that was older than the host's
4227 retry timeout. (The host should not be counted, thus allowing all hosts to
4228 be tried at least once before bouncing.) This may have been the cause of an
4229 occasionally reported bug whereby a message would remain on the queue
4230 longer than the retry timeout, but would be bounced if a delivery was
4231 forced. I say "may" because I never totally pinned down the problem;
4232 setting up timeout/retry tests is difficult. See also the next item.
4234 79. The ultimate address timeout was not being applied to errors that involved
4235 a combination of host plus message (for example, a timeout on a MAIL
4236 command). When an address resolved to a number of possible hosts, and they
4237 were not all tried for each delivery (e.g. because of hosts_max_try), a
4238 message could remain on the queue longer than the retry timeout.
4240 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4243 81. Fixed an obscure SMTP outgoing bug which required at least the following
4244 conditions: (a) there was another message waiting for the same server;
4245 (b) the server returned 5xx to all RCPT commands in the first message so
4246 that the message was not completed; (c) the server dropped the connection
4247 or gave a negative response to the RSET that Exim sends to abort the
4248 transaction. The observed case was a dropped connection after DATA that had
4249 been sent in pipelining mode. That is, the server had advertised PIPELINING
4250 but was not implementing it correctly. The effect of the bug was incorrect
4251 behaviour, such as trying another host, and this could lead to a crash.
4257 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4258 and daemon.c were passed as pointers to ints; they should have been
4259 pointers to socklen_t variables (which are typically unsigned ints).
4261 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4264 3. Fixed a really odd bug that affected only the testing scheme; patching a
4265 certain fixed string in the binary changed the value of another string that
4266 happened to be identical to the end of the original first string.
4268 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4269 name", it returns that address as the IP address. On some operating
4270 systems (e.g. Solaris), it also passes back the IP address string as the
4271 "host name". However, on others (e.g. Linux), it passes back an empty
4272 string. Exim wasn't checking for this, and was changing the host name to an
4273 empty string, assuming it had been canonicized.
4275 5. Although rare, it is permitted to have more than one PTR record for a given
4276 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4277 all the names associated with an address, because they do in Solaris.
4278 However, it seems that they do not in Linux for data that comes from the
4279 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4280 I found this out when I moved to a new Linux workstation and tried to run
4281 the Exim test suite.
4283 To get round this problem I have changed the code so that it now does its
4284 own call to the DNS to look up PTR records when searching for a host name.
4285 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4286 addresses that are only in /etc/hosts are still found.
4288 This behaviour is, however, controlled by an option called host_lookup_
4289 order, which defaults to "bydns:byaddr". If people want to use the other
4290 order, or indeed, just use one or the other means of lookup, they can
4291 specify it in this variable.
4293 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4294 some operating systems, this comes back from gethostbyaddr() as an empty
4295 string, and this is what Exim used to test for. However, it seems that in
4296 other systems, "." is yielded. Exim now tests for this case too.
4298 7. The values of check_spool_space and check_log_space are now held internally
4299 as a number of kilobytes instead of an absolute number of bytes. If a
4300 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4301 kilobyte. This means that much larger values can be stored.
4303 8. Exim monitor: an attempt to get the action menu when not actually pointing
4304 at a message produces an empty menu entitled "No message selected". This
4305 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4306 no entries in it ("Shell widget menu has zero width and/or height"). So I
4307 have added a single, blank menu entry in this case.
4309 9. Added ${quote_local_part.
4311 10. MIME decoding is now applied to the contents of Subject: header lines when
4314 11. Now that a reference to $sender_host_address automatically causes a reverse
4315 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4316 host lookup before query-style lookups in lists that might use this
4317 variable. This has therefore been abolished, and the "net-" prefix is no
4318 longer necessary for query-style lookups.
4320 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4321 have been a typo for LFLAGS, so it has been changed.
4323 13. The install script calls Exim with "-C /dev/null" in order to find the
4324 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4325 to be output. Howeve, since Exim outputs its version number before the
4326 error, it didn't break the script. It just looked ugly. I fixed this by
4327 always allowing "-C /dev/null" if the caller is root.
4329 14. Ignore overlarge ACL variable number when reading spool file - insurance
4330 against a later release with more variables having written the file.
4332 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4333 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4334 this, as well as the form without the "IPv6" on the front.
4336 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4337 OS/Makefile-Darwin file.
4339 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4340 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4342 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4343 variable. However, some OS use ints for inodes. I've added cast to long int
4344 to get rid of the compiler warning.
4346 19. I had forgotten to lock out "/../" in configuration file names when
4347 ALT_CONFIG_PREFIX was set.
4349 20. Routers used for verification do not need to specify transports. However,
4350 if such a router generated a host list, and callout was configured, Exim
4351 crashed, because it could not find a port number from the (non-existent)
4352 transport. It now assumes port 25 in this circumstance.
4354 21. Added the -t option to exigrep.
4356 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4357 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4358 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4359 them is set, LOOKUP_LSEARCH is forced.
4361 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4363 24. Added sender and host information to the "rejected by local_scan()" log
4364 line; previously there was no indication of these.
4366 25. Added .include_if_exists.
4368 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4369 Linux. It turns out that not all file systems support this. Apparently some
4370 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4371 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4372 not supported on the file descriptor, is now ignored when Exim is trying to
4373 sync a directory. This applies only to Linux.
4375 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4377 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4378 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4381 29. Added support for maildirsize files from supplied patch (modified a bit).
4383 30. The use of :fail: followed by an empty string could lead Exim to respond to
4384 sender verification failures with (e.g.):
4386 550 Verification failed for <xxx>
4387 550 Sender verify failed
4389 where the first response line was missing the '-' that indicates it is not
4390 the final line of the response.
4392 31. The loop for finding the name of the user that called Exim had a hardwired
4393 limit of 10; it now uses the value of finduser_retries, which is used for
4394 all other user lookups.
4396 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4399 33. Exim was neglecting to zero errno before one call of strtol() when
4400 expanding a string and expecting an integer value. On some systems this
4401 resulted in spurious "integer overflow" errors. Also, it was casting the
4402 result into an int without checking.
4404 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4405 did not work. The code looks as if it has *never* worked, though it appears
4406 to have been documented since at least releast 1.62. I have made it work.
4408 35. The "timeout_DNS" error in retry rules, also documented since at least
4409 1.62, also never worked. As it isn't clear exactly what this means, and
4410 clearly it isn't a major issue, I have abolished the feature by treating it
4411 as "timeout", and writing a warning to the main and panic logs.
4413 36. The display of retry rules for -brt wasn't always showing the error code
4416 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4417 timeout_connect_A, timeout_connect_MX.
4419 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4420 to the empty sender.
4422 39. The daemon was not analysing the content of -oX till after it had closed
4423 stderr and disconnected from the controlling terminal. This meant that any
4424 syntax errors were only noted on the panic log, and the return code from
4425 the command was 0. By re-arranging the code a little, I've made the
4426 decoding happen first, so such errors now appear on stderr, and the return
4427 code is 1. However, the actual setting up of the sockets still happens in
4428 the disconnected process, so errors there are still only recorded on the
4431 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4432 connections (as happens on some IP stacks) was logged at start up time as
4433 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4434 it from "IPv6 and IPv4", which means that two separate sockets are being
4437 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4438 says whether AF_INET or AF_INET6 was passed as an argument.
4440 42. Exiwhat output was messed up when time zones were included in log
4443 43. Exiwhat now gives more information about the daemon's listening ports,
4444 and whether -tls-on-connect was used.
4446 44. The "port" option of the smtp transport is now expanded.
4448 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4449 silently ignored. Now an error message is written to the main and panic
4452 46. There's a new ACL modifier called "logwrite" which writes to a log file
4453 as soon as it is encountered.
4455 47. Added $local_user_uid and $local_user_gid at routing time.
4457 48. Exim crashed when trying to verify a sender address that was being
4460 49. Exim was recognizing only a space character after ".include". It now also
4461 recognizes a tab character.
4463 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4464 extracting the relevant information from the specification. The man page no
4465 longer contains scrambled data for the -d option, and I've added a section
4466 at the front about calling Exim under different names.
4468 51. Added "extra_headers" argument to the "mail" command in filter files.
4470 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4473 53. Installed eximstats 1.29.
4475 54. Added transport_filter_timeout as a generic transport option.
4477 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4478 Cc: header lines. This was required by RFC 822, but it not required by RFC
4481 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4482 incoming messages that did not have them. Now it does so only if the
4483 message originates locally, that is, if there is no associated remote host
4484 address. When Resent- header lines are present, this applies to the Resent-
4485 lines rather than the non-Resent- lines.
4487 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4488 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4490 58. Messages for configuration errors now include the name of the main
4491 configuration file - useful now that there may be more than one file in a
4492 list (.included file names were always shown).
4494 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4495 for those rare installations that do not start the daemon as root or run it
4496 setuid root. I've cut out the call to initgroups() if the daemon is not
4499 60. The Exim user and group can now be bound into the binary as text strings
4500 that are looked up at the start of Exim's processing.
4502 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4504 62. Added $mailstore_basename variable.
4506 63. Installed patch to sieve.c from Michael Haardt.
4508 64. When Exim failed to open the panic log after failing to open the main log,
4509 the original message it was trying to log was written to stderr and debug
4510 output, but if they were not available (the usual case in production), it
4511 was lost. Now it is written to syslog before the two lines that record the
4512 failures to open the logs.
4514 65. Users' Exim filters run in subprocesses under the user's uid. It is
4515 possible for a "deliver" command or an alias in a "personal" command to
4516 provoke an address rewrite. If logging of address rewriting is configured,
4517 this fails because the process is not running as root or exim. There may be
4518 a better way of dealing with this, but for the moment (because 4.30 needs
4519 to be released), I have disabled address rewrite logging when running a
4520 filter in a non-root, non-exim process.
4526 1. The buildconfig auxiliary program wasn't quoting the value set for
4527 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4528 not defined. This bug was masked in 4.22 by the effect that was fixed in
4531 2. Some messages that were rejected after a message id was allocated were
4532 shown as "incomplete" by exigrep. It no longer does this for messages that
4533 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4535 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4536 have allow_domain_literals set, the ID did not get logged in the <= line.
4537 Domain literals are now always recognized in Message-ID: header lines.
4539 4. The first argument for a ${extract expansion item is the key name or field
4540 number. Leading and trailing spaces in this item were not being ignored,
4541 causing some misleading effects.
4543 5. When deliver_drop_privilege was set, single queue runner processes started
4544 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4545 same command in the process it spins off) were not dropping privilege.
4547 6. When the daemon running as "exim" started a queue runner, it always
4548 re-executed Exim in the spun-off process. This is a waste of effort when
4549 deliver_drop_privilege is set. The new process now just calls the
4550 queue-runner function directly.
4556 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4559 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4560 ignored. Though the use of -oP was forcing the writing of a pid file, it
4561 was always written to the default place.
4563 3. If the message "no IP address found for host xxxx" is generated during
4564 incoming verification, it is now followed by identification of the incoming
4565 connection (so you can more easily find what provoked it).
4567 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4569 5. Added some features to "harden" Exim a bit more against certain attacks:
4571 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4572 be put in Local/Makefile. This is like the never_users runtime option,
4573 but it cannot be overridden. The default setting is "root".
4575 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4576 prefix string with which any file named in a -C command line option
4579 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4580 is retained for -C and -D only if the caller of Exim is root. Without
4581 it, the exim user may also use -C and -D and retain privilege.
4583 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4584 command line option is disabled.
4586 6. Macro names set by the -D option must start with an upper case letter, just
4587 like macro names defined in the configuration file.
4589 7. Added "dereference=" facility to LDAP.
4591 8. Two instances of the typo "uknown" in the source files are fixed.
4593 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4594 the Configure-Makefile script screwed up while processing it.
4596 10. Incorporated PCRE 4.4.
4598 11. The SMTP synchronization check was not operating right at the start of an
4599 SMTP session. For example, it could not catch a HELO sent before the client
4600 waited for the greeting. There is now a check for outstanding input at the
4601 point when the greeting is written. Because of the duplex, asynchronous
4602 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4603 way, but not yet received, when the check is performed.
4605 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4606 on TCP/IP sockets, because this apparently causes some broken clients to
4609 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4610 unchanged) from the Cygwin maintainer.
4612 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4613 was supported instead of testing for mbx. Effectively a typo.
4615 15. The spa authenticator server code was not checking that the input it
4616 received was valid base64.
4618 16. The debug output line for the "set" modifier in ACLs was not showing the
4619 name of the variable that was being set.
4621 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4623 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4624 lookup on its own. Something else was needed to trigger the lookup. For
4625 example, a match in host_lookup or the need for a host name in a host list.
4626 Now, if $sender_host_name is referenced and the host name has not yet been
4627 looked up, a lookup is performed. If the lookup fails, the variable remains
4628 empty, and $host_lookup_failed is set to "1".
4630 19. Added "eqi" as a case-independent comparison operator.
4632 20. The saslauthd authentication condition could segfault if neither service
4633 nor realm was specified.
4635 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4636 error message that was logged was misleading, and incoming SMTP
4637 connections were dropped. The message is now more accurate, and temporary
4638 errors are given to SMTP connections.
4640 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4641 (except RSET) with a 421 error, until QUIT is received. However, it was
4642 failing to send a response to QUIT.
4644 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4645 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4646 ACL is now run after the helo_try_verify_hosts code.
4648 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4649 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4650 case-independent, so other case variants are also recognized). Apparently
4651 some systems use these upper case variants.
4653 25. If more than two messages were waiting for the same host, and a transport
4654 filter was specified for the transport, Exim sent two messages over the
4655 same TCP/IP connection, and then failed with "socket operation on non-
4656 socket" when it tried to send the third.
4658 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4660 27. The extern definition of crypt16() in expand.c was not being excluded when
4661 the OS had its own crypt16() function.
4663 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4664 as a preferred synonym for return_size_limit, both as an option and as an
4667 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4669 30. Changed the default configuration ACL to relax the local part checking rule
4670 for addresses that are not in any local domains. For these addresses,
4671 slashes and pipe symbols are allowed within local parts, but the sequence
4672 /../ is explicitly forbidden.
4674 31. SPA server authentication was not clearing the challenge buffer before
4677 32. log_message in a "warn" ACL statement was writing to the reject log as
4678 well as to the main log, which contradicts the documentation and doesn't
4679 seem right (because no rejection is happening). So I have stopped it.
4681 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4682 However, I am unable to do any testing of this.
4684 34. Fixed an infelicity in the appendfile transport. When checking directories
4685 for a mailbox, to see if any needed to be created, it was accidentally
4686 using path names with one or more superfluous leading slashes; tracing
4687 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4689 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4690 contents are added to the log line that is written for every discarded
4691 recipient. (Previously a log_message setting was ignored.)
4693 36. The ${quote: operator now quotes the string if it is empty.
4695 37. The install script runs exim in order to find its version number. If for
4696 some reason other than non-existence or emptiness, which it checks, it
4697 could not run './exim', it was installing it with an empty version number,
4698 i.e. as "exim-". This error state is now caught, and the installation is
4701 38. An argument was missing from the function that creates an error message
4702 when Exim fails to connect to the socket for saslauthd authentication.
4703 This could cause Exim to crash, or give a corrupted message.
4705 39. Added isip, isip4, and isip6 to ${if conditions.
4707 40. The ACL variables $acl_xx are now saved with the message, and can be
4708 accessed later in routers, transports, and filters.
4710 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4711 strings in the file are not string-expanded.
4713 42. If a MAIL command specified a SIZE value that was too large to fit into an
4714 int variable, the check against message_size_limit failed. Such values are
4715 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4716 day this will have to be increased, but I don't think I want to be around
4717 when emails are that large.
4724 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4725 iconv() is not standard in FreeBSD.
4727 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4728 IPv6 enabled. The observed symptom was a segmentation fault on return from
4729 the function os_common_find_running_interfaces() in src/os.c.
4731 3. In the check_special_case() function in daemon.c I had used "errno" as an
4732 argument name, which causes warnings on some systems. This was basically a
4733 typo, since it was named "eno" in the comments!
4735 4. The code that waits for the clock to tick (at a resolution of some fraction
4736 of a second) so as to ensure message-id uniqueness was always waiting for
4737 at least one whole tick, when it could have waited for less. [This is
4738 almost certainly not relevant at current processor speeds, where it is
4739 unlikely to ever wait at all. But we try to future-proof.]
4741 5. The function that sleeps for a time interval that includes fractions of a
4742 second contained a race. It did not block SIGALRM between setting the
4743 timer, and suspending (a couple of lines later). If the interval was short
4744 and the sigsuspend() was delayed until after it had expired, the suspension
4745 never ended. On busy systems this could lead to processes getting stuck for
4748 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4749 process, before it forks any delivery processes. The open lookup caching
4750 mechanism meant that the open file or database connection was passed into
4751 the delivery process. The problem was that delivery processes always tidy
4752 up cached lookup data. This could cause a problem for the next delivery
4753 process started by the queue runner, because the external queue runner
4754 process does not know about the closure. So the next delivery process
4755 still has data in the lookup cache. In the case of a file lookup, there was
4756 no problem because closing a file descriptor in a subprocess doesn't affect
4757 the parent. However, if the lookup was caching a connection to a database,
4758 the connection was closed, and the second delivery process was likely to
4759 see errors such as "PGSQL: query failed: server closed the connection
4760 unexpectedly". The problem has been fixed by closing all cached lookups
4761 in a queue runner before running a delivery process.
4763 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4764 seem to have the "const" qualifier which it has on other OS. I've
4767 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4768 *of the same type* (client or server) with the same public name that an
4769 error should be diagnosed.
4771 9. When Exim looked up a host name for an IP address, but failed to find the
4772 original IP address when looking up the host name (a safety check), it
4773 output the message "<ip address> does not match any IP for NULL", which was
4774 confusing, to say the least. The bug was that the host name should have
4775 appeared instead of "NULL".
4777 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4778 user that is built into the binary, and the -C or -D options is used, root
4779 privilege is dropped before the configuration file is read. In addition,
4780 logging is switched to stderr instead of the normal log files. If the
4781 configuration then re-defines the Exim user, the unprivileged environment
4782 is probably not what is expected, so Exim logs a panic warning message (but
4785 However, if deliver_drop_privilege is set, the unprivileged state may well
4786 be exactly what is intended, so the warning has been cut out in that case,
4787 and Exim is allowed to try to write to its normal log files.
4793 1. smtp_return_error_details was not giving details for temporary sender
4794 or receiver verification errors.
4796 2. Diagnose a configuration error if two authenticators have the same public
4799 3. Exim used not to create the message log file for a message until the first
4800 delivery attempt. This could be confusing when incoming messages were held
4801 for policy or load reasons. The message log file is now created at the time
4802 the message is received, and an initial "Received" line is written to it.
4804 4. The automatically generated man page for command line options had a minor
4805 bug that caused no ill effects; however, a more serious problem was that
4806 the procedure for building the man page automatically didn't always
4807 operate. Consequently, release 4.20 contains an out-of-date version. This
4808 shouldn't happen again.
4810 5. When building Exim with embedded Perl support, the script that builds the
4811 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4812 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4814 6. The freeze_tell option was not being used for messages that were frozen on
4815 arrival, either by an ACL or by local_scan().
4817 7. Added the smtp_incomplete_transaction log selector.
4819 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4820 was accepting AUTH without a new EHLO.
4822 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4823 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4826 10. Logging of TCP/IP connections (when configured) now happens in the main
4827 daemon process instead of the child process, so that the TCP/IP connection
4828 count is more accurate (but it can never be perfect).
4830 11. The use of "drop" in a nested ACL was not being handled correctly in the
4831 outer ACL. Now, if condition failure induced by the nested "drop" causes
4832 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4833 or "require"), the connection is dropped.
4835 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4836 that yield "discard" can now be used with an "accept" or a "discard" verb,
4837 but an error is generated for any others (because I can't see a useful way
4838 to define what should happen).
4840 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4841 are now processed in the same way as lines in the Exim configuration file.
4842 In particular, continuation lines are supported.
4844 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4846 15. Added -ti meaning -t -i.
4848 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4849 domains, and warn by logging if others are found.
4851 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4852 returned from GIFCONF when Exim is trying to find the list of interfaces on
4853 a host. The code in os.c has been modified to copy each ifreq to an aligned
4854 structure in all cases.
4856 Also, in some cases, the returned ifreq's were being copied to a 'struct
4857 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4858 means the last couple of bytes of an IPv6 address could be chopped if the
4859 ifreq contained only a normal sockaddr (14 bytes storage).
4861 18. Named domain lists were not supported in the hosts_treat_as_local option.
4862 An entry such as +xxxx was not recognized, and was treated as a literal
4865 19. Ensure that header lines added by a DATA ACL are included in the reject log
4866 if the ACL subsequently rejects the message.
4868 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4869 MD5 (which is deprecated).
4871 21. When testing a filter file using -bf, Exim was writing a message when it
4872 took the sender from a "From " line in the message, but it was not doing so
4873 when it took $return_path from a Return-Path: header line. It now does.
4875 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4876 with a valid header line field name (a series of printing characters
4877 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4879 23. Changed "disc" in the source to "disk" to conform to the documentation and
4880 the book and for uniformity.
4882 24. Ignore Sendmail's -Ooption=value command line item.
4884 25. When execve() failed while trying to run a command in a pipe transport,
4885 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4886 could be confused with a return value of 69 from the command itself. This
4887 has been changed to 127, the value the shell returns if it is asked to run
4888 a non-existent command. The wording for the related log line suggests a
4889 non-existent command as the problem.
4891 26. If received_header_text expands to an empty string, do not add a Received:
4892 header line to the message. (Well, it adds a token one on the spool, but
4893 marks it "old" so that it doesn't get used or transmitted.)
4895 27. Installed eximstats 1.28 (addition of -nt option).
4897 28. There was no check for failure on the call to getsockname() in the daemon
4898 code. This can fail if there is a shortage of resources on the system, with
4899 ENOMEM, for example. A temporary error is now given on failure.
4901 29. Contrary to the C standard, it seems that in some environments, the
4902 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4903 program. Exim now does this explicitly; it affects the formatting of
4904 timestamps using strftime().
4906 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4907 complaints. I've now initialized all the variables, to avoid this.
4909 32. Header lines added by a system filter were not being "seen" during
4910 transport-time rewrites.
4912 33. The info_callback() function passed to OpenSSL is set up with type void
4913 (*)(SSL *, int, int), as described somewhere. However, when calling the
4914 function (actually a macro) that sets it up, the type void(*)() is
4915 expected. I've put in a cast to prevent warnings from picky compilers.
4917 34. If a DNS black list lookup found a CNAME record, but there were no A
4918 records associated with the domain it pointed at, Exim crashed.
4920 35. If a DNS black list lookup returned more than one A record, Exim ignored
4921 all but the first. It now scans all returned addresses if a particular IP
4922 value is being sought. In this situation, the contents of the
4923 $dnslist_value variable are a list of all the addresses, separated by a
4926 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4927 to accept a host name and all its aliases if the forward lookup for any of
4928 them yielded the IP address of the incoming connection. Now it accepts only
4929 those names whose forward lookup yields the correct IP address. Any other
4930 names are discarded. This closes a loophole whereby a rogue DNS
4931 administrator could create reverse DNS records to break through a
4932 wildcarded host restriction in an ACL.
4934 37. If a user filter or a system filter that ran in a subprocess used any of
4935 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4936 the wrong values were passed to the pipe command ($thisaddress had the
4937 value of $0, $0 had the value of $1, etc). This bug was introduced by
4938 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4940 38. Improved the line breaking for long SMTP error messages from ACLs.
4941 Previously, if there was no break point between 40 and 75 characters, Exim
4942 left the rest of the message alone. Two changes have been made: (a) I've
4943 reduced the minimum length to 35 characters; (b) if it can't find a break
4944 point between 35 and 75 characters, it looks ahead and uses the first one
4945 that it finds. This may give the occasional overlong line, but at least the
4946 remaining text gets split now.
4948 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4949 file descriptors might be low, and that setting 1000 would always raise it.
4950 It turns out that in some environments, the limit is already over 1000 and
4951 that lowering it causes trouble. So now Exim takes care not to decrease it.
4953 40. When delivering a message, the value of $return_path is set to $sender_
4954 address at the start of routing (routers may change the value). By an
4955 oversight, this default was not being set up when an address was tested by
4956 -bt or -bv, which affected the outcome if any router or filter referred to
4959 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4960 log only when "message" or "log_message" are set. However, if one of the
4961 conditions was an address verification, or a call to a nested ACL, the
4962 messages generated by the underlying test were being passed through. This
4963 no longer happens. The underlying message is available in $acl_verify_
4964 message for both "message" and "log_message" expansions, so it can be
4965 passed through if needed.
4967 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4968 new expansion $bh_ to give the encoded byte string without charset
4969 translation. Translation happens only if iconv() is available; HAVE_ICONV
4970 indicates this at build time. HEADERS_CHARSET gives the charset to
4971 translate to; headers_charset can change it in the configuration, and
4972 "headers charset" can change it in an individual filter file.
4974 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4975 that creates RFC 2047 encoded "words" labels them as that charset instead
4976 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4977 expansion operator; (ii) when Exim creates a From: line for a local
4978 message; (iii) when a header line is rewritten to include a "phrase" part.
4980 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4981 buggy, causing it to skip the first lines of messages whose message ID
4982 ended in 'D'. This would not have bitten before Exim release 4.14, because
4983 message IDs were unlikely to end in 'D' before then. The effect was to have
4984 incorrect size information for certain domains.
4986 45. #include "config.h" was missing at the start of the crypt16.c module. This
4987 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4990 46. If there was a timeout during a "random" callout check, Exim treated it as
4991 a failure of the random address, and carried on sending RSET and the real
4992 address. If the delay was just some slowness somewhere, the response to the
4993 original RCPT would be taken as a response to RSET and so on, causing
4994 mayhem of various kinds.
4996 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4997 when I implemented it. It didn't allow for the fact that some option values
4998 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4999 the right test for positive values.
5001 48. Domain names in DNS records are case-independent. Exim always looks them up
5002 in lower case. Some resolvers return domain names in exactly the case they
5003 appear in the zone file, that is, they may contain uppercase letters. Not
5004 all resolvers do this - some return always lower case. Exim was treating a
5005 change of case by a resolver as a change of domain, similar to a widening
5006 of a domain abbreviation. This triggered its re-routing code and so it was
5007 trying to route what was effectively the same domain again. This normally
5008 caused routing to fail (because the router wouldn't handle the domain
5009 twice). Now Exim checks for this case specially, and just changes the
5010 casing of the domain that it ultimately uses when it transmits the message
5013 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5016 50. If a filter generated a file delivery with a non-absolute name (possible if
5017 no home directory exists for the router), the forbid_file option was not
5020 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5021 the existing equality matching.
5023 52. Exim was using ints instead of ino_t variables in some places where it was
5024 dealing with inode numbers.
5026 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5027 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5028 called TMPDIR, and if it finds it is different, it changes its value.
5030 54. The smtp_printf() function is now made available to local_scan() so
5031 additional output lines can be written before returning. There is also an
5032 smtp_fflush() function to enable the detection of a dropped connection.
5033 The variables smtp_input and smtp_batched_input are exported to
5036 55. Changed the default runtime configuration: the message "Unknown user"
5037 has been removed from the ACL, and instead placed on the localuser router,
5038 using the cannot_route_message feature. This means that any verification
5039 failures that generate their own messages won't get overridden. Similarly,
5040 the "Unrouteable address" message that was in the ACL for unverifiable
5041 relay addresses has also been removed.
5043 56. Added hosts_avoid_esmtp to the smtp transport.
5045 57. The exicyclog script was not checking for the esoteric option
5046 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5047 will work only if exicyclog is run under the appropriate euid.
5049 58. Following a discussion on the list, the rules by which Exim recognises line
5050 endings on incoming messages have been changed. The -dropcr and drop_cr
5051 options are now no-ops, retained only for backwards compatibility. The
5052 following line terminators are recognized: LF CRLF CR. However, special
5053 processing applies to CR:
5055 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5056 nor a local message in the state where . is a terminator.
5058 (ii) If a bare CR is encountered in a header line, an extra space is added
5059 after the line terminator so as not to end the header. The reasoning
5060 behind this is that bare CRs in header lines are most likely either
5061 to be mistakes, or people trying to play silly games.
5063 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5064 was being incorrectly given as 18 bytes larger than it should have been.
5065 This is a VOB (very old bug).
5067 60. This may never have affected anything current, but just in case it has:
5068 When the local host is found other than at the start of a list of hosts,
5069 the local host, those with the same MX, and any that follow, are discarded.
5070 When the list in question was part of a longer list of hosts, the following
5071 hosts (not currently being processed) were also being discarded. This no
5072 longer happens. I'm not sure if this situation could ever has previously
5075 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5078 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5079 header. This is something that is recommended in a new Internet Draft, and
5080 is something that is documented as being done by Sendmail. There are two
5081 possible values. For messages generated by the autoreply transport, Exim
5084 Auto-Submitted: auto-replied
5086 whereas for all other generated messages (e.g. bounces) it adds
5088 Auto-Submitted: auto-generated
5090 63. The "personal" condition in filters now includes a test for the
5091 Auto-Submitted: header. If it contains the string "auto-" the message it
5092 not considered personal.
5094 64. Added rcpt_include_affixes as a generic transport option.
5096 65. Added queue_only_override (default true).
5098 66. Added the syslog_duplication option.
5100 67. If what should have been the first header line of a message consisted of
5101 a space followed by a colon, Exim was mis-interpreting it as a header line.
5102 It isn't of course - it is syntactically invalid and should therefore be
5103 treated as the start of the message body. The misbehaviour could have
5104 caused a number of strange effects, including loss of data in subsequent
5105 header lines, and spool format errors.
5107 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5108 client host had authenticated. This control can now be exercised by an ACL
5109 for more flexibility.
5111 69. By default, callouts do not happen when testing with -bh. There is now a
5112 variant, -bhc, which does actually run the callout code, including
5113 consulting and updating the callout cache.
5115 70. Added support for saslauthd authentication, courtesy of Alexander
5118 71. If statvfs() failed on the spool or log directories while checking their
5119 size for availability, Exim confusingly gave the error "space shortage".
5120 Furthermore, in debugging mode it crashed with a floating point exception.
5121 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5122 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5123 serious problem, Exim now writes to the main and panic logs when this
5124 happens, with details of the failure. It then refuses to accept the
5125 incoming message, giving the message "spool directory problem" or "log
5126 directory problem" with a 421 code for SMTP messages.
5128 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5129 0, 1, and 2 exist, because some OS complain for execs without them (see
5130 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5131 descriptors. However, the code omitted to check that the open succeeded,
5132 causing mysterious errors if for some reason the permissions on /dev/null
5133 got screwed. Now Exim writes a message to the main and panic logs, and
5134 bombs out if it can't open /dev/null.
5136 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5137 interact so that it is all more flexible. It is supposed to remain
5138 backwards compatible. Also added extra_local_interfaces.
5140 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5141 to bomb out with an assertion failure - to the client this appears as a
5142 connection drop. This problem occurs in the part of the code that was taken
5143 from the Samba project. Fortunately, the assertion is in a very simple
5144 function, so I have fixed this by reproducing the function inline in the
5145 one place where it is called, and arranging for authentication to fail
5146 instead of killing the process with assert().
5148 75. The SPA client code was not working when the server requested OEM rather
5149 than Unicode encoding.
5151 76. Added code to make require_files with a specific uid setting more usable in
5152 the case where statting the file as root fails - usually a non-root-mounted
5153 NFS file system. When this happens and the failure is EACCES, Exim now
5154 forks a subprocess and does the per-uid checking as the relevant uid.
5156 77. Added process_log_path.
5158 78. If log_file_path was not explicitly set, a setting of check_log_space or
5159 check_log_inodes was ignored.
5161 79. If a space check for the spool or log partitions fails, the incident is now
5162 logged. Of course, in the latter case the data may get lost...
5164 80. Added the %p formatting code to string_format() so that it can be used to
5165 print addresses in debug_print(). Adjusted all the address printing in the
5166 debugging in store.c to use %p rather than %d.
5168 81. There was a concern that a line of code in smtp_in.c could overflow a
5169 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5170 initially expressed, the concern was not well-founded, because trailing
5171 spaces are removed early. However, if the trailing spaces were followed by
5172 a NULL, they did not get removed, so the overflow was possible. Two fixes
5175 (a) I re-wrote the offending code in a cleaner fashion.
5176 (b) If an incoming SMTP command contains a NULL character, it is rejected
5179 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5180 runs initgroups(), so that if the Exim user is in any additional groups,
5181 they will be used during message reception.
5187 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob