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.
99 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
100 $data_path, and $header_path variables; fixed documentation bugs and
103 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
104 exipick to access non-standard spools, including the "frozen" queue
107 NM/01 Bugzilla 965: Support mysql stored procedures.
108 Patch from Alain Williams
110 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
112 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
113 Patch from Andreas Metzler
115 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
116 Patch from Kirill Miazine
118 NM/05 Bugzilla 671: Added umask to procmail example.
120 JJ/03 installed exipick 20100323.0, fixing doc bug
122 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
123 directory. Notification and patch from Dan Rosenberg.
125 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
127 TK/02 Improve log output when DKIM signing operation fails.
129 MH/01 Treat the transport option dkim_domain as a colon separated
130 list, not as a single string, and sign the message with each element,
131 omitting multiple occurences of the same signer.
133 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
134 Bugzilla 985, 986. Patch by Simon Arlott
136 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
137 Patch by Simon Arlott
139 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
140 MBX locking. Notification from Dan Rosenberg.
146 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
148 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
150 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
152 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
154 NM/04 Bugzilla 915: Flag broken perl installation during build.
160 TK/01 Added patch by Johannes Berg that expands the main option
161 "spamd_address" if it starts with a dollar sign.
163 TK/02 Write list of recipients to X-Envelope-Sender header when building
164 the mbox-format spool file for content scanning (suggested by Jakob
167 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
168 (http://www.dcc-servers.net/) support via dccifd. Activated by
169 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
171 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
172 by Mark Daniel Reidel <mr@df.eu>.
174 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
175 When building exim an external PCRE library is now needed -
176 PCRE is a system library on the majority of modern systems.
177 See entry on PCRE_LIBS in EDITME file.
179 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
180 conversation. Added nologin parameter to request.
181 Patch contributed by Kirill Miazine.
183 TF/01 Do not log submission mode rewrites if they do not change the address.
185 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
187 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
188 log files in place. Contributed by Roberto Lima.
190 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
192 TF/03 Bugzilla 615: When checking the local_parts router precondition
193 after a local_part_suffix or local_part_prefix option, Exim now
194 does not use the address's named list lookup cache, since this
195 contains cached lookups for the whole local part.
197 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
198 Robert Millan. Documentation is in experimental-spec.txt.
200 TF/04 Bugzilla 668: Fix parallel build (make -j).
202 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
204 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
205 Patch provided by Jan Srzednicki.
207 TF/05 Leading white space used to be stripped from $spam_report which
208 wrecked the formatting. Now it is preserved.
210 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
211 that they are available at delivery time.
213 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
215 TF/08 TLS error reporting now respects the incoming_interface and
216 incoming_port log selectors.
218 TF/09 Produce a more useful error message if an SMTP transport's hosts
219 setting expands to an empty string.
221 NM/06 Bugzilla 744: EXPN did not work under TLS.
222 Patch provided by Phil Pennock.
224 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
225 Patch provided by Richard Godbee.
227 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
228 acl_smtp_notquit, added index entry.
230 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
231 Patch provided by Eugene Bujak.
233 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
234 accept(). Patch provided by Maxim Dounin.
236 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
237 Patch provided by Phil Pennock.
239 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
241 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
242 Patch provided by Brad "anomie" Jorsch.
244 NM/14 Bugzilla 622: Added timeout setting to transport filter.
245 Patch provided by Dean Brooks.
247 TK/05 Add native DKIM support (does not depend on external libraries).
249 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
250 Patch provided by Graeme Fowler.
252 NM/16 Bugzilla 851: Documentation example syntax fix.
254 NM/17 Changed NOTICE file to remove references to embedded PCRE.
256 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
259 NM/19 Bugzilla 745: TLS version reporting.
260 Patch provided by Phil Pennock.
262 NM/20 Bugzilla 167: bool: condition support.
263 Patch provided by Phil Pennock.
265 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
266 clients. Patch provided by Phil Pennock.
268 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
269 Patch provided by Brad "anomie" Jorsch.
271 NM/23 Bugzilla 687: Fix misparses in eximstats.
272 Patch provided by Heiko Schlittermann.
274 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
275 Patch provided by Heiko Schlittermann.
277 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
278 plus update to original patch.
280 NM/26 Bugzilla 799: Documentation correction for ratelimit.
282 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
283 Patch provided by David Brownlee.
285 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
287 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
289 NM/30 Bugzilla 888: TLS documentation bugfixes.
291 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
293 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
294 Unlike the original bugzilla I have changed all shell scripts in src tree.
296 NM/33 Bugzilla 898: Transport filter timeout fix.
297 Patch by Todd Rinaldo.
299 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
300 Patch by Serge Demonchaux.
302 NM/35 Bugzilla 39: Base64 decode bug fixes.
303 Patch by Jakob Hirsch.
305 NM/36 Bugzilla 909: Correct connect() call in dcc code.
307 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
309 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
311 NM/39 Bugzilla 911: Fixed MakeLinks build script.
317 TK/01 Add preliminary DKIM support. Currently requires a forked version of
318 ALT-N's libdkim that I have put here:
319 http://duncanthrax.net/exim-experimental/
321 Note to Michael Haardt: I had to rename some vars in sieve.c. They
322 were called 'true' and it seems that C99 defines that as a reserved
323 keyword to be used with 'bool' variable types. That means you could
324 not include C99-style headers which use bools without triggering
325 build errors in sieve.c.
327 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
328 as mailq or other aliases. Changed the --help handling significantly
329 to do whats expected. exim_usage() emits usage/help information.
331 SC/01 Added the -bylocaldomain option to eximstats.
333 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
335 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
337 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
343 PH/01 Another patch from the Sieve maintainer.
345 PH/02 When an IPv6 address is converted to a string for single-key lookup
346 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
347 dots are used instead of colons so that keys in lsearch files need not
348 contain colons. This was done some time before quoting was made available
349 in lsearch files. However, iplsearch files do require colons in IPv6 keys
350 (notated using the quote facility) so as to distinguish them from IPv4
351 keys. This meant that lookups for IP addresses in host lists did not work
352 for iplsearch lookups.
354 This has been fixed by arranging for IPv6 addresses to be expressed with
355 colons if the lookup type is iplsearch. This is not incompatible, because
356 previously such lookups could never work.
358 The situation is now rather anomolous, since one *can* have colons in
359 ordinary lsearch keys. However, making the change in all cases is
360 incompatible and would probably break a number of configurations.
362 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
365 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
366 conversion specification without a maximum field width, thereby enabling
367 a rogue spamd server to cause a buffer overflow. While nobody in their
368 right mind would setup Exim to query an untrusted spamd server, an
369 attacker that gains access to a server running spamd could potentially
370 exploit this vulnerability to run arbitrary code as the Exim user.
372 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
373 $primary_hostname instead of what libspf2 thinks the hosts name is.
375 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
376 a directory entry by the name of the lookup key. Previously, if a
377 symlink pointed to a non-existing file or a file in a directory that
378 Exim lacked permissions to read, a lookup for a key matching that
379 symlink would fail. Now it is enough that a matching directory entry
380 exists, symlink or not. (Bugzilla 503.)
382 PH/03 The body_linecount and body_zerocount variables are now exported in the
385 PH/04 Added the $dnslist_matched variable.
387 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
388 This means they are set thereafter only if the connection becomes
391 PH/06 Added the client_condition to authenticators so that some can be skipped
392 by clients under certain conditions.
394 PH/07 The error message for a badly-placed control=no_multiline_responses left
395 "_responses" off the end of the name.
397 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
399 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
400 (without spaces) instead of just copying the configuration text.
402 PH/10 Added the /noupdate option to the ratelimit ACL condition.
404 PH/11 Added $max_received_linelength.
406 PH/12 Added +ignore_defer and +include_defer to host lists.
408 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
409 way in which PCRE > 7.0 is built.
411 PH/14 Implemented queue_only_load_latch.
413 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
414 MAIL command. The effect was to mangle the value on 64-bit systems.
416 PH/16 Another patch from the Sieve maintainer.
418 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
420 PH/18 If a system quota error occurred while trying to create the file for
421 a maildir delivery, the message "Mailbox is full" was not appended to the
422 bounce if the delivery eventually timed out. Change 4.67/27 below applied
423 only to a quota excession during the actual writing of the file.
425 PH/19 It seems that peer DN values may contain newlines (and other non-printing
426 characters?) which causes problems in log lines. The DN values are now
427 passed through string_printing() before being added to log lines.
429 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
430 and InterBase are left for another time.)
432 PH/21 Added message_body_newlines option.
434 PH/22 Guard against possible overflow in moan_check_errorcopy().
436 PH/23 POSIX allows open() to be a macro; guard against that.
438 PH/24 If the recipient of an error message contained an @ in the local part
439 (suitably quoted, of course), incorrect values were put in $domain and
440 $local_part during the evaluation of errors_copy.
446 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
447 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
450 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
451 issue a MAIL command.
453 PH/02 In an ACL statement such as
455 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
457 if a client was not listed at all, or was listed with a value other than
458 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
459 the condition was not true (as it should be), so access was not denied.
460 The bug was that the ! inversion was incorrectly passed on to the second
461 item. This has been fixed.
463 PH/03 Added additional dnslists conditions == and =& which are different from
464 = and & when the dns lookup returns more than one IP address.
466 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
467 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
469 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
470 FSYNC, which compiles an option called disable_fsync that allows for
471 bypassing fsync(). The documentation is heavily laced with warnings.
473 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
475 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
476 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
477 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
478 including adding "make clean"; (3) Added -fPIC when compiling the test
479 dynamically loaded module, to get rid of a warning.
481 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
482 message fails, move_frozen_messages = true and ignore_bounce_errors_after
483 = 0s. The bug is otherwise harmless.
485 PH/07 There was a bug in the dovecot authenticator such that the value of
486 $auth1 could be overwritten, and so not correctly preserved, after a
487 successful authentication. This usually meant that the value preserved by
488 the server_setid option was incorrect.
490 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
492 PH/09 Installed PCRE release 7.0.
494 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
495 run for batched SMTP input. It is now run at the start of every message
496 in the batch. While fixing this I discovered that the process information
497 (output by running exiwhat) was not always getting set for -bs and -bS
498 input. This is fixed, and it now also says "batched" for BSMTP.
500 PH/11 Added control=no_pipelining.
502 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
503 patch, slightly modified), and move the expansion of helo_data till after
504 the connection is made in the smtp transport (so it can use these
507 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
509 PH/14 Added log_selector = +pid.
511 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
513 PH/16 Add ${if forany and ${if forall.
515 PH/17 Added dsn_from option to vary the From: line in DSNs.
517 PH/18 Flush SMTP output before performing a callout, unless control =
518 no_callout_flush is set.
520 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
521 was true (the default) a successful delivery failed to delete the retry
522 item, thus causing premature timeout of the address. The bug is now
525 PH/20 Added hosts_avoid_pipelining to the smtp transport.
527 PH/21 Long custom messages for fakedefer and fakereject are now split up
528 into multiline reponses in the same way that messages for "deny" and
529 other ACL rejections are.
531 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
532 with slight modification.
534 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
535 draft, changing the syntax and factoring some duplicate code".
537 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
538 for deliveries of the second and subsequent messages over the same SMTP
541 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
542 ${reduce, with only minor "tidies".
544 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
546 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
547 expansion side effects.
549 PH/27 When a message times out after an over-quota error from an Exim-imposed
550 quota, the bounce message says "mailbox is full". This message was not
551 being given when it was a system quota that was exceeded. It now should
554 MH/03 Made $recipients available in local_scan(). local_scan() already has
555 better access to the recipient list through recipients_list[], but
556 $recipients can be useful in postmaster-provided expansion strings.
558 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
559 in the case of a MAIL command with additional options following the
560 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
561 were accidentally chopped off.
563 PH/29 SMTP synchronization checks are implemented when a command is read -
564 there is a check that no more input is waiting when there shouldn't be
565 any. However, for some commands, a delay in an ACL can mean that it is
566 some time before the response is written. In this time, more input might
567 arrive, invalidly. So now there are extra checks after an ACL has run for
568 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
569 pipelining has not been advertised.
571 PH/30 MH's patch to allow iscntrl() characters to be list separators.
573 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
574 returned in the SMTP response when smtp_return_error_details was false.
577 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
578 instead of the C I/O that was originally supplied, because problems were
581 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
582 Exim which did not show up earlier: it was assuming that a call to
583 SSL_CTX_set_info_callback() might give an error value. In fact, there is
584 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
585 was a macro that became an assignment, so it seemed to work. This has
586 changed to a proper function call with a void return, hence the compile
587 error. Exim's code has been fixed.
589 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
592 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
594 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
596 PH/37 If a message is not accepted after it has had an id assigned (e.g.
597 because it turns out to be too big or there is a timeout) there is no
598 "Completed" line in the log. When some messages of this type were
599 selected by exigrep, they were listed as "not completed". Others were
600 picked up by some special patterns. I have improved the selection
601 criteria to be more general.
603 PH/38 The host_find_failed option in the manualroute router can now be set
604 to "ignore", to completely ignore a host whose IP address cannot be
605 found. If all hosts are ignored, the behaviour is controlled by the new
606 host_all_ignored option.
608 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
609 homing or because of multiple MX records with /mx) generated more than
610 one IP address, and the following item turned out to be the local host,
611 all the secondary addresses of the first item were incorrectly removed
612 from the list, along with the local host and any following hosts (which
613 is what is supposed to happen).
615 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
616 whoever called Exim into the -H file. In the case of the daemon it was
617 behaving confusingly. When first started, it used values for whoever
618 started the daemon, but after a SIGHUP it used the Exim user (because it
619 calls itself on a restart). I have changed the code so that it now always
622 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
623 message are rejected with the same error (e.g. no authentication or bad
624 sender address), and a DATA command is nevertheless sent (as can happen
625 with PIPELINING or a stupid MUA), the error message that was given to the
626 RCPT commands is included in the rejection of the DATA command. This is
627 intended to be helpful for MUAs that show only the final error to their
630 PH/42 Another patch from the Sieve maintainer.
632 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
633 Eximstats - Fixed some broken HTML links and added missing column headers
635 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
636 columns for Rejects, Temp Rejects, Ham, and Spam rows.
638 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
640 PH/43 Yet another patch from the Sieve maintainer.
642 PH/44 I found a way to check for a TCP/IP connection going away before sending
643 the response to the final '.' that terminates a message, but only in the
644 case where the client has not sent further data following the '.'
645 (unfortunately, this is allowed). However, in many cases there won't be
646 any further data because there won't be any more messages to send. A call
647 to select() can be used: if it shows that the input is "ready", there is
648 either input waiting, or the socket has been closed. An attempt to read
649 the next input character can distinguish the two cases. Previously, Exim
650 would have sent an OK response which the client would never have see.
651 This could lead to message repetition. This fix should cure that, at
652 least in a lot of common cases.
654 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
655 advertised in response to EHLO.
661 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
662 fixed by 4.65/MH/01 (is this a record?) are fixed:
664 (i) An empty string was always treated as zero by the numeric comparison
665 operators. This behaviour has been restored.
667 (ii) It is documented that the numeric comparison operators always treat
668 their arguments as decimal numbers. This was broken in that numbers
669 starting with 0 were being interpreted as octal.
671 While fixing these problems I realized that there was another issue that
672 hadn't been noticed. Values of message_size_limit (both the global option
673 and the transport option) were treated as octal if they started with 0.
674 The documentation was vague. These values are now always treated as
675 decimal, and I will make that clear in the documentation.
681 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
682 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
685 MH/01 Don't check that the operands of numeric comparison operators are
686 integers when their expansion is in "skipping" mode (fixes bug
687 introduced by 4.64-PH/07).
689 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
690 child addresses, Exim now panics and dies. Previously, because the count
691 is held in a short int, deliveries were likely to be lost. As such a
692 large number of recipients for a single message is ridiculous
693 (performance will be very, very poor), I have chosen to impose a limit
694 rather than extend the field.
700 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
701 leftover -K file (the existence of which was triggered by #402).
702 While we were at it, introduced process PID as part of the -K
703 filename. This should rule out race conditions when creating
706 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
707 processing considerably. Previous code took too long for large mails,
708 triggering a timeout which in turn triggers #401.
710 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
711 in the DK code in transports.c. sendfile() is not really portable,
712 hence the _LINUX specificness.
714 TF/01 In the add_headers option to the mail command in an Exim filter,
715 there was a bug that Exim would claim a syntax error in any
716 header after the first one which had an odd number of characters
719 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
720 callout verification, Exim cached a "reject" for the entire domain. This
721 is correct for most verifications, but it is not correct for a recipient
722 verification with use_sender or use_postmaster set, because in that case
723 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
724 case of MAIL FROM:<> rejection from other early rejections (e.g.
725 rejection of HELO). When verifying a recipient using a non-null MAIL
726 address, the cache is ignored if it shows MAIL FROM:<> rejection.
727 Whatever the result of the callout, the value of the domain cache is
728 left unchanged (for any other kind of callout, getting as far as trying
729 RCPT means that the domain itself is ok).
731 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
734 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
735 manifest itself as EPIPE rather than ECONNECT. When tidying away a
736 session, the daemon ignores ECONNECT errors and logs others; it now
737 ignores EPIPE as well.
739 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
740 (quoted-printable decoding).
742 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
743 later the small subsequent patch to fix an introduced bug.
745 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
747 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
749 PH/08 An error is now given if message_size_limit is specified negative.
751 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
752 to be given (somewhat) arbitrary names.
754 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
757 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
758 miscellaneous code fixes
760 PH/10 Added the log_reject_target ACL modifier to specify where to log
763 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
764 hostname. This is wrong, because it relates to the incoming message (and
765 probably the interface on which it is arriving) and not to the outgoing
766 callout (which could be using a different interface). This has been
767 changed to use the value of the helo_data option from the smtp transport
768 instead - this is what is used when a message is actually being sent. If
769 there is no remote transport (possible with a router that sets up host
770 addresses), $smtp_active_hostname is used.
772 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
773 tweaks were necessary in order to get it to work (see also 21 below):
774 (a) The code assumed that strncpy() returns a negative number on buffer
775 overflow, which isn't the case. Replaced with Exim's string_format()
777 (b) There were several signed/unsigned issues. I just did the minimum
778 hacking in of casts. There is scope for a larger refactoring.
779 (c) The code used strcasecmp() which is not a standard C function.
780 Replaced with Exim's strcmpic() function.
781 (d) The code set only $1; it now sets $auth1 as well.
782 (e) A simple test gave the error "authentication client didn't specify
783 service in request". It would seem that Dovecot has changed its
784 interface. Fortunately there's a specification; I followed it and
785 changed what the client sends and it appears to be working now.
787 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
790 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
791 address A is aliased to B and C, where B exists and C does not. Without
792 -v the output is "A verified" because verification stops after a
793 successful redirection if more than one address is generated. However,
794 with -v the child addresses are also verified. Exim was outputting "A
795 failed to verify" and then showing the successful verification for C,
796 with its parentage. It now outputs "B failed to verify", showing B's
797 parentage before showing the successful verification of C.
799 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
800 look up a TXT record in a specific list after matching in a combined
803 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
804 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
805 they consult the DNS. I had assumed they would set it the way they
806 wanted; and indeed my experiments on Linux seem to show that in some
807 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
808 To be on the safe side, however, I have now made the interface to
809 host_find_byname() similar to host_find_bydns(), with an argument
810 containing the DNS resolver options. The host_find_byname() function now
811 sets these options at its start, just as host_find_bydns() does. The smtp
812 transport options dns_qualify_single and dns_search_parents are passed to
813 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
814 of host_find_byname() use the default settings of RES_DEFNAMES
815 (qualify_single) but not RES_DNSRCH (search_parents).
817 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
818 spool_read_header() do less string testing, by means of a preliminary
819 switch on the second character of optional "-foo" lines. (This is
820 overdue, caused by the large number of possibilities that now exist.
821 Originally there were few.) While I was there, I also converted the
822 str(n)cmp tests so they don't re-test the leading "-" and the first
823 character, in the hope this might squeeze out yet more improvement.
825 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
826 flag allowing group syntax was set by the header_syntax check but not
827 turned off, possible causing trouble later; (2) The flag was not being
828 set at all for the header_verify test, causing "group"-style headers to
829 be rejected. I have now set it in this case, and also caused header_
830 verify to ignore an empty address taken from a group. While doing this, I
831 came across some other cases where the code for allowing group syntax
832 while scanning a header line wasn't quite right (mostly, not resetting
833 the flag correctly in the right place). These bugs could have caused
834 trouble for malformed header lines. I hope it is now all correct.
836 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
837 with the "reply" argument non-NULL. The code, however (which originally
838 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
839 but it didn't always do it. This confused somebody who was copying the
840 code for some other use. I have removed all the tests.
842 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
843 feature that was used to support insecure browsers during the U.S. crypto
844 embargo. It requires special client support, and Exim is probably the
845 only MTA that supported it -- and would never use it because real RSA is
846 always available. This code has been removed, because it had the bad
847 effect of slowing Exim down by computing (never used) parameters for the
848 RSA_EXPORT functionality.
850 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
851 authenticator to fail if there's a tab character in the incoming data
852 (there should never be unless someone is messing about, as it's supposed
853 to be base64-encoded). Also added, on Timo's advice, the "secured" option
854 if the connection is using TLS or if the remote IP is the same as the
855 local IP, and the "valid-client-cert option" if a client certificate has
858 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
859 authenticators. This can be used for authorization after authentication
860 succeeds. (In the case of plaintext, it servers for both authentication
863 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
864 if any retry times were supplied.
866 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
867 connection, where there is no client IP address to check. In this
868 situation, the verify now always succeeds.
870 PH/25 Applied John Jetmore's -Mset patch.
872 PH/26 Added -bem to be like -Mset, but loading a message from a file.
874 PH/27 In a string expansion for a processed (not raw) header when multiple
875 headers of the same name were present, leading whitespace was being
876 removed from all of them, but trailing whitespace was being removed only
877 from the last one. Now trailing whitespace is removed from each header
878 before concatenation. Completely empty headers in a concatenation (as
881 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
882 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
884 PH/29 [Removed. This was a change that I later backed out, and forgot to
885 correct the ChangeLog entry (that I had efficiently created) before
886 committing the later change.]
888 PH/30 Exim was sometimes attempting to deliver messages that had suffered
889 address errors (4xx response to RCPT) over the same connection as other
890 messages routed to the same hosts. Such deliveries are always "forced",
891 so retry times are not inspected. This resulted in far too many retries
892 for the affected addresses. The effect occurred only when there were more
893 hosts than the hosts_max_try setting in the smtp transport when it had
894 the 4xx errors. Those hosts that it had tried were not added to the list
895 of hosts for which the message was waiting, so if all were tried, there
896 was no problem. Two fixes have been applied:
898 (i) If there are any address or message errors in an SMTP delivery, none
899 of the hosts (tried or untried) are now added to the list of hosts
900 for which the message is waiting, so the message should not be a
901 candidate for sending over the same connection that was used for a
902 successful delivery of some other message. This seems entirely
903 reasonable: after all the message is NOT "waiting for some host".
904 This is so "obvious" that I'm not sure why it wasn't done
905 previously. Hope I haven't missed anything, but it can't do any
906 harm, as the worst effect is to miss an optimization.
908 (ii) If, despite (i), such a delivery is accidentally attempted, the
909 routing retry time is respected, so at least it doesn't keep
910 hammering the server.
912 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
913 in ${readsocket because some servers need this prod.
915 PH/32 Added some extra debug output when updating a wait-xxx database.
917 PH/33 The hint "could be header name not terminated by colon", which has been
918 given for certain expansion errors for a long time, was not being given
919 for the ${if def:h_colon_omitted{... case.
921 PH/34 The spec says: "With one important exception, whenever a domain list is
922 being scanned, $domain contains the subject domain." There was at least
923 one case where this was not true.
925 PH/35 The error "getsockname() failed: connection reset by peer" was being
926 written to the panic log as well as the main log, but it isn't really
927 panic-worthy as it just means the connection died rather early on. I have
928 removed the panic log writing for the ECONNRESET error when getsockname()
931 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
932 runs only) independently of the message's sender address. This meant
933 that, if the 4xx error was in fact related to the sender, a different
934 message to the same recipient with a different sender could confuse
935 things. In particualar, this can happen when sending to a greylisting
936 server, but other circumstances could also provoke similar problems.
937 I have changed the default so that the retry time for these errors is now
938 based a combination of the sender and recipient addresses. This change
939 can be overridden by setting address_retry_include_sender=false in the
942 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
943 remote server are returned as part of bounce messages. This was not
944 happening for LMTP over a pipe (the lmtp transport), but now it is the
945 same for both kinds of LMTP.
947 PH/38 Despite being documented as not happening, Exim was rewriting addresses
948 in header lines that were in fact CNAMEs. This is no longer the case.
950 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
951 and queue runs started by the daemon processed all messages. This has
952 been fixed so that -R and -S can now usefully be given with -q<time>.
954 PH/40 Import PCRE release 6.7 (fixes some bugs).
956 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
958 PH/42 Give an error if -q is specified more than once.
960 PH/43 Renamed the variables $interface_address and $interface_port as
961 $received_ip_address and $received_port, to make it clear that these
962 values apply to message reception, and not to the outgoing interface when
963 a message is delivered. (The old names remain recognized, of course.)
965 PH/44 There was no timeout on the connect() call when using a Unix domain
966 socket in the ${readsocket expansion. There now is.
968 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
969 be meaningful with "accept".
971 SC/01 Eximstats V1.43
972 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
974 SC/02 Eximstats V1.44
975 Use a glob alias rather than an array ref in the generated
976 parser. This improves both readability and performance.
978 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
979 Collect SpamAssassin and rejection statistics.
980 Don't display local sender or destination tables unless
981 there is data to show.
982 Added average volumes into the top table text output.
984 SC/04 Eximstats V1.46
985 Collect data on the number of addresses (recipients)
986 as well as the number of messages.
988 SC/05 Eximstats V1.47
989 Added 'Message too big' to the list of mail rejection
990 reasons (thanks to Marco Gaiarin).
992 SC/06 Eximstats V1.48
993 Mainlog lines which have GMT offsets and are too short to
994 have a flag are now skipped.
996 SC/07 Eximstats V1.49 (Alain Williams)
997 Added the -emptyok flag.
999 SC/08 Eximstats V1.50
1000 Fixes for obtaining the IP address from reject messages.
1002 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1003 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1004 whitesspace changes from 4.64-PH/27
1006 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1009 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1010 are found, allow negative numbers in numeric criteria)
1012 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1014 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1015 to match changes made in 4.64-PH/43
1017 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1019 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1020 "domain" is actually an IP address, and force a failure. This locks out
1021 those revolvers/nameservers that support "A-for-A" lookups, in
1022 contravention of the specifications.
1024 PH/48 When a host name was looked up from an IP address, and the subsequent
1025 forward lookup of the name timed out, the host name was left in
1026 $sender_host_name, contrary to the specification.
1028 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1029 restricted to single-key lookups, Exim was not diagnosing an error if
1030 * or *@ was used with a query-style lookup.
1032 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1034 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1035 long ago, but noone interested enough thought of it. Let's just say that
1036 the "1.1" means that there are some new functions that weren't there at
1037 some point in the past.
1039 PH/51 Error processing for expansion failure of helo_data from an smtp
1040 transport during callout processing was broken.
1042 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1043 tested/used via the -bh/-bhc/-bs options.
1045 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1046 bug, fixed in subsequent PCRE releases).
1048 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1049 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1051 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1057 SC/01 Use a glob alias rather than an array ref in eximstats generated
1058 parser. This improves both readability and performance.
1060 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1061 Don't display local sender or destination tables in eximstats unless
1062 there is data to show.
1063 Added average volumes into the eximstats top table text output.
1065 SC/03 Collect data on the number of addresses (recipients) as well
1066 as the number of messages in eximstats.
1068 TF/01 Correct an error in the documentation for the redirect router. Exim
1069 does (usually) call initgroups() when daemonizing.
1071 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1072 with consistent privilege compared to when running as a daemon.
1074 TF/03 Note in the spec that $authenticated_id is not set for local
1075 submissions from trusted users.
1077 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1078 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1080 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1081 by adding some example configuration directives to the default
1082 configuration file. A little bit of work is required to uncomment the
1083 directives and define how usernames and passwords are checked, but
1084 there is now a framework to start from.
1086 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1087 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1088 without this. I don't know how relevant this is to other LDAP libraries.
1090 PH/02 Add the verb name to the "unknown ACL verb" error.
1092 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1094 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1096 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1097 directory not expanded when it should be if an expanded home directory
1098 was set for the address (which is overridden by the transport).
1100 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1103 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1104 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1105 because it is too late at that time, and has no effect.
1107 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1108 security issue with \' (bugzilla #107). I could not use the
1109 PQescapeStringConn() function, because it needs a PGconn value as one of
1112 PH/08 When testing addresses using -bt, indicate those final addresses that
1113 are duplicates that would not cause an additional delivery. At least one
1114 person was confused, thinking that -bt output corresponded to deliveries.
1115 (Suppressing duplicates isn't a good idea as you lose the information
1116 about possibly different redirections that led to the duplicates.)
1118 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1119 systems where poll() doesn't work, in particular OS X.
1121 PH/10 Added more information to debugging output for retry time not reached.
1123 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1124 operations in malware.c.
1126 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1129 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1130 syslog_duplication set false, log lines that would normally be written
1131 both the the main log and to the reject log were not written to syslog at
1134 PH/14 In the default configuration, change the use of "message" in ACL warn
1135 statements to "add_header".
1137 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1138 not followed by a command (e.g. "seen endif").
1140 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1141 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1144 PH/17 Added extra conditions to the default value of delay_warning_condition
1148 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1149 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1150 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1153 The Auto-Submitted: and various List- headers are standardised, whereas I
1154 don't think Precedence: ever was.
1156 PH/18 Refactored debugging code in route_finduser() to show more information,
1157 in particular, the error code if getpwnam() issues one.
1159 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1160 This is apparently needed in addition to the PH/07 change above to avoid
1161 any possible encoding problems.
1163 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1164 but not after initializing Perl.
1166 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1167 output them only if debugging. By default they are written stderr,
1168 apparently, which is not desirable.
1170 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1173 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1176 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1178 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1179 authenticated or an ident call has been made. Suppress the default
1180 values for $authenticated_id and $authenticated_sender (but permit -oMai
1181 and -oMas) when testing with -bh.
1183 PH/24 Re-jigged the order of the tests in the default configuration so that the
1184 tests for valid domains and recipients precede the DNS black list and CSA
1185 tests, on the grounds that those ones are more expensive.
1187 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1188 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1189 This bug exists in every version of Exim that I still have, right back to
1192 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1193 However, an attempt to turn on case-sensitivity in a regex key by
1194 including (?-i) didn't work because the subject string was already
1195 lowercased, and the effects were non-intuitive. It turns out that a
1196 one-line patch can be used to allow (?-i) to work as expected.
1202 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1203 other effects) broke the use of negated acl sub-conditions.
1205 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1208 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1209 "Deny" causes Exim to reject the incoming connection with a 554 error.
1210 Unfortunately, if there is a major crisis, such as a disk failure,
1211 tcp-wrappers gives "deny", whereas what one would like would be some
1212 kind of temporary error. A kludge has been added to help with this.
1213 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1214 554 error is used if errno is still zero or contains ENOENT (which occurs
1215 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1218 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1220 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1221 errors. Otherwise a message that provokes a temporary error (when other
1222 messages do not) can cause a whole host to time out.
1224 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1225 addresses were routed directly to files or pipes from a redirect router.
1226 File deliveries just didn't batch; pipe deliveries might have suffered
1229 PH/06 A failure to get a lock for a hints database would erroneously always say
1230 "Failed to get write lock", even when it was really a read lock.
1232 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1233 of 0600. This has been changed to use the value of the lockfile_mode
1234 option (which defaults to 0600).
1236 PH/08 Applied small patch from the Sieve maintainer.
1238 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1239 folder from quota calculations, a direct delivery into this folder messed
1240 up the contents of the maildirsize file. This was because the regex was
1241 used only to exclude .Trash (or whatever) when the size of the mailbox
1242 was calculated. There was no check that a delivery was happening into an
1243 excluded directory. This bug has been fixed by ignoring all quota
1244 processing for deliveries into excluded directories.
1246 PH/10 Added the maildirfolder_create_regex option to appendfile.
1252 PH/01 The code for finding all the local interface addresses on a FreeBSD
1253 system running IPv6 was broken. This may well have applied to all BSD
1254 systems, as well as to others that have similar system calls. The broken
1255 code found IPv4 interfaces correctly, but gave incorrect values for the
1256 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1257 that it would not match correctly against @[] and not recognize the IPv6
1260 PH/02 The ipliteral router was not recognizing addresses of the form user@
1261 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1263 PH/03 Added disable_ipv6.
1265 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1266 decoded form, because it is most often used to construct To: headers
1267 lines in autoreplies, and the decoded form may well be syntactically
1268 invalid. However, $reply_address has leading white space removed, and all
1269 newlines turned into spaces so that the autoreply transport does not
1272 PH/05 If group was specified without a user on a router, and no group or user
1273 was specified on a transport, the group from the router was ignored.
1275 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1276 for visible compile-time settings that can be used to change these
1277 numbers, for those that want even more. Backwards compatibility with old
1278 spool files has been maintained. However, going back to a previous Exim
1279 release will lost any variables that are in spool files.
1281 PH/07 Two small changes when running in the test harness: increase delay when
1282 passing a TCP/IP connection to a new process, in case the original
1283 process has to generate a bounce, and remove special handling of
1284 127.0.0.2 (sic), which is no longer necessary.
1286 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1287 be the same on different OS.
1289 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1292 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1293 whether --show-vars was specified or not
1295 JJ/02 exipick: Added support for new ACL variable spool format introduced
1298 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1299 syntactically invalid From: or Reply-to: line, and a filter used this to
1300 generate an autoreply, and therefore failed to obtain an address for the
1301 autoreply, Exim could try to deliver to a non-existent relative file
1302 name, causing unrelated and misleading errors. What now happens is that
1303 it logs this as a hard delivery error, but does not attempt to create a
1306 PH/11 The exinext utility has a -C option for testing purposes, but although
1307 the given file was scanned by exinext itself; it wasn't being passed on
1308 when Exim was called.
1310 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1311 an end-of-file indication when reading a command response.
1313 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1314 compiled. In many other places in Exim, IPv6 addresses are always
1315 recognized, so I have changed this. It also means that IPv4 domain
1316 literals of the form [IPV4:n.n.n.n] are now always recognized.
1318 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1319 used if the router is not running as root, for example, when verifying at
1320 ACL time, or when using -bh. The debugging output from this situation was
1321 non-existent - all you got was a failure to exec. I have made two
1324 (a) Failures to set uid/gid, the current directory, or a process leader
1325 in a subprocess such as that created by queryprogram now generate
1326 suitable debugging ouput when -d is set.
1328 (b) The queryprogram router detects when it is not running as root,
1329 outputs suitable debugging information if -d is set, and then runs
1330 the subprocess without attempting to change uid/gid.
1332 PH/15 Minor change to Makefile for building test_host (undocumented testing
1335 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1336 additional section of a DNS packet that returns MX or SRV records.
1337 Instead, it always explicitly searches for A/AAAA records. This avoids
1338 major problems that occur when a DNS server includes only records of one
1339 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1340 fixed another bug: if SRV records were looked up and the corresponding
1341 address records were *not* found in the additional section, the port
1342 values from the SRV records were lost.
1344 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1345 using the correct key (the original address) when searching the retry
1346 rules in order to find which one to use for generating the retry hint.
1348 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1349 adding the default one. Similarly, if it contains a Reply-To: header, the
1350 errors_reply_to option, if set, is not used.
1352 PH/19 When calculating a retry time, Exim used to measure the "time since
1353 failure" by looking at the "first failed" field in the retry record. Now
1354 it does not use this if it is later than than the arrival time of the
1355 message. Instead it uses the arrival time. This makes for better
1356 behaviour in cases where some deliveries succeed, thus re-setting the
1357 "first failed" field. An example is a quota failure for a huge message
1358 when small messages continue to be delivered. Without this change, the
1359 "time since failure" will always be short, possible causing more frequent
1360 delivery attempts for the huge message than are intended.
1361 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1363 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1364 $1, $2, $3) because the numerical variables can be reset during some
1365 expansion items (e.g. "match"), thereby losing the authentication data.
1367 PH/21 Make -bV show the size of off_t variables so that the test suite can
1368 decide whether to run tests for quotas > 2G.
1370 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1371 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1372 filecount value is greater than 2G or if a quota value is greater than 2G
1373 on a system where the size of off_t is not greater than 4, a panic error
1376 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1377 never match. The debug and -bh output now contains an explicit error
1378 message indicating a malformed IPv4 address or mask.
1380 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1381 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1382 PH/23 above applies.
1384 PH/25 Do not write to syslog when running in the test harness. The only
1385 occasion when this arises is a failure to open the main or panic logs
1386 (for which there is an explicit test).
1388 PH/26 Added the /no_tell option to "control=freeze".
1390 PH/27 If a host name lookup failed very early in a connection, for example, if
1391 the IP address matched host_lookup and the reverse lookup yielded a name
1392 that did not have a forward lookup, an error message of the form "no IP
1393 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1394 could be logged. Now it outputs the IP address instead of "NULL".
1396 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1397 allows the sender and the authenticated sender to be set when
1398 submitting a message from within Exim. Since child_open_exim() is
1399 documented for local_scan(), the new function should be too.
1401 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1402 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1403 results in an empty string is now treated as unset.
1405 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1407 PH/31 Added sender_verify_fail logging option.
1409 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1410 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1411 tidied the source and removed it altogether.
1413 PH/33 When a queue run was abandoned because the load average was too high, a
1414 log line was always written; now it is written only if the queue_run log
1415 selector is set. In addition, the log line for abandonment now contains
1416 information about the queue run such as the pid. This is always present
1417 in "start" and "stop" lines but was omitted from the "abandon" line.
1419 PH/34 Omit spaces between a header name and the colon in the error message that
1420 is given when verify = headers_syntax fails (if there are lots of them,
1421 the message gets confusing).
1423 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1424 names, as there are now some PTR records that contain slashes. This check
1425 is only to protect against broken name servers that fall over on strange
1426 characters, so the fact that it applies to all lookups doesn't matter.
1428 PH/36 Now that the new test suite is complete, we can remove some of the
1429 special code in Exim that was needed for the old test suite. For example,
1430 sorting DNS records because real resolvers return them in an arbitrary
1431 order. The new test suite's fake resolver always returns records in the
1434 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1435 bounces) except when queue_only is set, to avoid logging races between
1436 the different processes.
1438 PH/38 Panic-die if .include specifies a non-absolute path.
1440 PH/39 A tweak to the "H" retry rule from its user.
1442 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1443 a label. They prevented compilation on older perls.
1445 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1446 a warning to be raised on newish perls.
1448 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1449 on queue. Changes to match documented behaviour of showing count of
1450 messages matching specified criteria.
1452 PH/40 Changed the default ident timeout from 30s to 5s.
1454 PH/41 Added support for the use of login_cap features, on those BSD systems
1455 that have them, for controlling the resources used by pipe deliveries.
1457 PH/42 The content-scanning code uses fopen() to create files in which to put
1458 message data. Previously it was not paying any attention to the mode of
1459 the files. Exim runs with umask(0) because the rest of the code creates
1460 files with open(), and sets the required mode explicitly. Thus, these
1461 files were ending up world-writeable. This was not a big issue, because,
1462 being within the spool directory, they were not world-accessible. I have
1463 created a function called modefopen, which takes an additional mode
1464 argument. It sets umask(777), creates the file, chmods it to the required
1465 mode, then resets the umask. All the relevant calls to fopen() in the
1466 content scanning code have been changed to use this function.
1468 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1469 to 24 hours. This avoids potential overflow problems when processing G
1470 and H retry rules. I suspect nobody ever tinkers with this value.
1472 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1474 PH/45 When the plaintext authenticator is running as a client, the server's
1475 challenges are checked to ensure they are valid base64 strings. By
1476 default, the authentication attempt is cancelled if an invalid string is
1477 received. Setting client_ignore_invalid_base64 true ignores these errors.
1478 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1479 they are received. Thus, the responses can be made to depend on the
1480 challenges. If an invalid string is ignored, an empty string is placed in
1483 PH/46 Messages that are created by the autoreply transport now contains a
1484 References: header, in accordance with RFCs 2822 and 3834.
1486 PH/47 Added authenticated_sender_force to the smtp transport.
1488 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1490 PH/49 Installed latest patch from the Sieve maintainer.
1492 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1493 was also set, the appendfile transport was unnecessarily scanning a
1494 directory of message files (e.g. for maildir delivery) to find the count
1495 of files (along with the size), even though it did not need this
1496 information. It now does the scan only if it needs to find either the
1497 size of the count of files.
1499 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1501 PH/52 Two bugs concerned with error handling when the smtp transport is
1504 (i) Exim was not creating retry information for temporary errors given
1505 for individual recipients after the DATA command when the smtp transport
1506 was used in LMTP mode. This meant that they could be retried too
1507 frequently, and not timed out correctly.
1509 (ii) Exim was setting the flag that allows error details to be returned
1510 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1511 recipients that were returned after the DATA command.
1513 PH/53 This is related to PH/52, but is more general: for any failing address,
1514 when detailed error information was permitted to be returned to the
1515 sender, but the error was temporary, then after the final timeout, only
1516 "retry timeout exceeded" was returned. Now it returns the full error as
1517 well as "retry timeout exceeded".
1519 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1520 do this, and (what is worse) MTAs that accept it.
1522 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1523 will now be deprecated.
1525 PH/56 New os.c-cygwin from the Cygwin maintainer.
1527 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1528 formats (previously only available in exim formats via -bpr, -bpru,
1529 and -bpra. Now also available in native and exiqgrep formats)
1531 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1532 with very large, slow to parse queues
1534 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1536 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1538 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1539 responses to authentication challenges, though it was showing the
1540 challenges; (ii) I've removed the CR characters from the debug output for
1543 PH/58 Allow for the insertion of a newline as well as a space when a string
1544 is turned into more than one encoded-word during RFC 2047 encoding. The
1545 Sieve code now uses this.
1547 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1548 data_4xx, lost_connection, tls_required.
1550 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1551 message was being sent as an SMTP response.
1553 PH/61 Add -l and -k options to exicyclog.
1555 PH/62 When verifying, if an address was redirected to one new address, so that
1556 verification continued, and the new address failed or deferred after
1557 having set something in $address_data, the value of $address_data was not
1558 passed back to the ACL. This was different to the case when no
1559 redirection occurred. The value is now passed back in both cases.
1561 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1562 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1563 use login_cap.h, so on its own it isn't the distinguishing feature. The
1564 new name refers directly to the setclassresources() function.
1566 PH/65 Added configuration files for NetBSD3.
1568 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1570 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1571 is preferred over IPv4.
1573 PH/68 The bounce_return_message and bounce_return_body options were not being
1574 honoured for bounces generated during the reception of non-SMTP messages.
1575 In particular, this applied to messages rejected by the ACL. This bug has
1576 been fixed. However, if bounce_return_message is true and bounce_return_
1577 body is false, the headers that are returned for a non-SMTP message
1578 include only those that have been read before the error was detected.
1579 (In the case of an ACL rejection, they have all been read.)
1581 PH/69 The HTML version of the specification is now built in a directory called
1582 spec_html instead of spec.html, because the latter looks like a path with
1583 a MIME-type, and this confuses some software.
1585 PH/70 Catch two compiler warnings in sieve.c.
1587 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1588 function verify_get_ident() calls ip_connect() to connect a socket, but
1589 if the "connect()" function timed out, ip_connect() used to close the
1590 socket. However, verify_get_ident() also closes the socket later, and in
1591 between Exim writes to the log, which may get opened at this point. When
1592 the socket was closed in ip_connect(), the log could get the same file
1593 descriptor number as the socket. This naturally causes chaos. The fix is
1594 not to close the socket in ip_connect(); the socket should be closed by
1595 the function that creates it. There was only one place in the code where
1596 this was missing, in the iplookup router, which I don't think anybody now
1597 uses, but I've fixed it anyway.
1599 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1600 well as to direct DNS lookups. Otherwise the handling of names in host
1601 lists is inconsistent and therefore confusing.
1607 PH/01 Two changes to the default runtime configuration:
1609 (1) Move the checks for relay_from_hosts and authenticated clients from
1610 after to before the (commented out) DNS black list checks.
1612 (2) Add control=submission to the relay_from_hosts and authenticated
1613 clients checks, on the grounds that messages accepted by these
1614 statements are most likely to be submissions.
1616 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1618 (1) Generate an error if the third argument for the ${prvs expansion is
1621 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1624 (3) Reset the variables that are obtained from the first argument of
1625 ${prvscheck and used in the second argument before leaving the code,
1626 because their memory is reclaimed, so using them afterwards may do
1629 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1630 one (it's much easier than Tom thought :-).
1632 (5) Because of (4), we can now allow for the use of $prvscheck_result
1633 inside the third argument.
1635 PH/03 For some reason, the default setting of PATH when running a command from
1636 a pipe transport was just "/usr/bin". I have changed it to
1639 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1640 anything to be listed in the output from -bV.
1642 PH/05 When a filter generated an autoreply, the entire To: header line was
1643 quoted in the delivery log line, like this:
1645 => >A.N.Other <ano@some.domain> <original@ddress> ...
1647 This has been changed so that it extracts the operative address. There
1648 may be more than one such address. If so, they are comma-separated, like
1651 => >ano@some.domain,ona@other.domain <original@ddress> ...
1653 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1654 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1655 not being looked up in the rDNS to get a host name, Exim was showing the
1656 IP address twice in Received: lines, even though the IP addresses were
1657 identical. For example:
1659 Received: from [1.2.3.4] (helo=[1.2.3.4])
1661 However, if the real host name was known, it was omitting the HELO data
1662 if it matched the actual IP address. This has been tidied up so that it
1663 doesn't show the same IP address twice.
1665 PH/07 When both +timestamp and +memory debugging was on, the value given by
1666 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1667 called by the debug printing, thereby overwriting the timestamp buffer.
1668 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1670 PH/08 When the original message was included in an autoreply transport, it
1671 always said "this is a copy of the message, including all the headers",
1672 even if body_only or headers_only was set. It now gives an appropriate
1675 PH/09 Applied a patch from the Sieve maintainer which:
1677 o fixes some comments
1678 o adds the (disabled) notify extension core
1679 o adds some debug output for the result of if/elsif tests
1680 o points to the current vacation draft in the documentation
1681 and documents the missing references header update
1685 o fixes a bug in processing the envelope test (when testing
1686 multiple envelope elements, the last element determinted the
1689 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1690 Electronic Mail") by including:
1692 Auto-submitted: auto-generated
1694 in the messages that it generates (bounce messages and others, such as
1695 warnings). In the case of bounce messages for non-SMTP mesages, there was
1696 also a typo: it was using "Auto_submitted" (underscore instead of
1697 hyphen). Since every message generated by Exim is necessarily in response
1698 to another message, thes have all been changed to:
1700 Auto-Submitted: auto-replied
1702 in accordance with these statements in the RFC:
1704 The auto-replied keyword:
1706 - SHOULD be used on messages sent in direct response to another
1707 message by an automatic process,
1709 - MUST NOT be used on manually-generated messages,
1711 - MAY be used on Delivery Status Notifications (DSNs) and Message
1712 Disposition Notifications (MDNs),
1714 - MUST NOT be used on messages generated by automatic or periodic
1715 processes, except for messages which are automatic responses to
1718 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1719 to the default Received: header definition.
1721 PH/12 Added log selector acl_warn_skipped (default on).
1723 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1724 variables because (a) they are in the wrong storage pool and (b) even if
1725 they were copied, it wouldn't work properly because of the caching.
1727 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1728 checking when decoding. Apparently there are clients that generate
1729 overlong encoded strings. Why am I not surprised?
1731 PH/15 If the first argument of "${if match_address" was not empty, but did not
1732 contain an "@" character, Exim crashed. Now it writes a panic log message
1733 and treats the condition as false.
1735 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1737 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1738 extension "envelope-auth". The code is finished and in agreement with
1739 other implementations, but there is no documentation so far and in fact,
1740 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1741 not changing the active code.
1743 Print executed "if" and "elsif" statements when debugging is used. This
1744 helps a great deal to understand what a filter does.
1746 Document more things not specified clearly in RFC3028. I had all this
1747 sorted out, when out of a sudden new issues came to my mind. Oops."
1749 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1752 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1753 canonical form (as documented). However, after a host name lookup from
1754 the IP address, check_host() was doing a simple string comparison with
1755 addresses acquired from the DNS when checking that the found name did
1756 have the original IP as one of its addresses. Since any found IPv6
1757 addresses are likely to be in abbreviated form, the comparison could
1758 fail. Luckily, there already exists a function for doing the comparison
1759 by converting both addresses to binary, so now that is used instead of
1760 the text comparison.
1762 PH/20 There was another similar case to PH/19, when a complete host name was
1763 given in a host list; looking up its IP address could give an abbreviated
1764 form, whereas the current host's name might or might not be abbreviated.
1765 The same fix has been applied.
1771 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1772 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1775 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1776 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1778 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1780 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1781 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1782 Also, for "Auto-Submitted", treat anything other than "no" as
1783 non-personal, in accordance with RFC 3834. (Previously it treated
1784 anything starting "auto-" as non-personal.)
1786 TF/01 The control=submission/name=... option had a problem with syntax
1787 errors if the name included a slash character. The /name= option
1788 now slurps the rest of the string, so it can include any characters
1789 but it must come last in the list of options (after /sender_retain
1792 PH/05 Some modifications to the interface to the fake nameserver for the new
1800 TK/01 Added the "success_on_redirect" address verification option. See
1801 NewStuff for rationale and an example.
1803 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1805 PH/02 Patch to exigrep to allow it to work on syslog lines.
1807 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1808 fread() to skip over the body file's header line, because in Cygwin the
1809 header line is locked and is inaccessible.
1811 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1812 co-exist for some time) to make it clear that it is the Exim ID that is
1813 referenced, not the Message-ID: header line.
1815 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1816 string_format() function, because snprintf() does not exist on all
1819 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1820 ${stat: expansion item.
1822 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1823 protocol synchronization error", to keep the pedants happy.
1825 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1826 well as for IRIX systems, when gcc is being used. See the host.c source
1829 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1831 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1834 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1835 transport and to the smtp transport in LMTP mode.
1837 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1839 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1840 run a filter in a subprocess. This could lead to confusion in subsequent
1841 lookups in the parent process. There should also be a search_tidyup() at
1842 the end of the subprocess.
1844 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1845 only if the host matched helo_try_verify_hosts, which caused the
1846 verification to occur when the EHLO/HELO command was issued. The ACL just
1847 tested the remembered result. Now, if a previous verification attempt has
1848 not happened, "verify = helo" does it there and then.
1850 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1852 TK/03 Fix log output including CR from clamd.
1854 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1855 exist provoked a memory error which could cause a segfault.
1857 PH/15 Installed PCRE 6.2
1859 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1861 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1862 of the problem. Specifically, suggested +O2 rather than +O1 for the
1865 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1867 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1868 record, the port was not correctly logged when the outgoing_port log
1869 selector was set (it logged the transort's default port).
1871 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1872 fallback_hosts, and "hosts" in the smtp transport.
1874 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1875 host errors such as "Connection refused".
1877 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1878 authentication with radiusclient 0.4.9:
1880 - Error returned from rc_read_config was caught wrongly
1881 - Username/password not passed on to radius server due to wrong length.
1883 The presumption is that some radiusclient API changes for 4.51/PH/17
1884 were not taken care of correctly. The code is still untested by me (my
1885 Linux distribution still has 0.3.2 of radiusclient), but it was
1886 contributed by a Radius user.
1888 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1889 expanding the "port" option of the smtp transport.
1891 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1892 while reading a MIME header. Thanks to Tom Hughes for a patch.
1894 PH/24 Include config.h inside local_scan.h so that configuration settings are
1897 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1898 that in an ACL for RCPT (for example), you can examine exactly what was
1901 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1902 commands, but it was not correctly comparing the address with the actual
1903 client host address. Thus, it would show the EHLO address in Received:
1904 header lines when this was not necessary.
1906 PH/27 Added the % operator to ${eval:}.
1908 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1909 it should be ignoring failures (because with -C, for example, it has lost
1910 privilege). It wasn't ignoring creation failures other than "already
1913 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1916 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1917 because an input line was too long, either on its own, or by virtue of
1918 too many continuations, the temporary file was not being removed, and the
1919 return code was incorrect.
1921 PH/31 Missing "BOOL" in function definition in filtertest.c.
1923 PH/32 Applied Sieve patches from the maintainer.
1925 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1927 PH/33 Added "verify = not_blind".
1929 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1930 Local/Makefile (with some defaults set). These are used in built scripts
1931 such as exicyclog, but they have never been used in the exim_install
1932 script (though there are many overriding facilities there). I have
1933 arranged that the exim_install script now takes note of these two
1936 PH/35 Installed configuration files for Dragonfly.
1938 PH/36 When a locally submitted message by a trusted user did not contain a
1939 From: header, and the sender address was obtained from -f or from an SMTP
1940 MAIL command, and the trusted user did not use -F to supply a sender
1941 name, $originator_name was incorrectly used when constructing a From:
1942 header. Furthermore, $originator_name was used for submission mode
1943 messages from external hosts without From: headers in a similar way,
1944 which is clearly wrong.
1946 PH/37 Added control=suppress_local_fixups.
1948 PH/38 When log_selector = +received_sender was set, and the addition of the
1949 sender made the log line's construction buffer exactly full, or one byte
1950 less than full, an overflow happened when the terminating "\n" was
1953 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1954 when the result of a list match is failure because a DNS lookup failed.
1956 PH/40 RM_COMMAND is now used in the building process.
1958 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1959 the "build-* directories that it finds.
1961 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1962 address in a domain literal was a prefix of an interface address.
1964 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1965 when verifying a sender address, unless rewrite_headers is false.
1967 PH/44 Wrote a long comment about why errors_to addresses are verified as
1968 recipients, not senders.
1970 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1971 the ratelimit ACL was added.
1973 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1975 PH/46 Added extra information about PostgreSQL errors to the error string.
1977 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1978 avoiding the need to install special zones in a real server. This is
1979 backwards compatible; if it can't find the fake resolver, it drops back.
1980 Thus, both old and new test suites can be run.
1982 TF/02 Added util/ratelimit.pl
1984 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1985 clock is set back in time.
1987 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1988 Candler <B.Candler@pobox.com>.
1990 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1991 OK for addresses that are the result of redirections.
1993 PH/48 A number of further additions for the benefit of the new test suite,
1994 including a fake gethostbyname() that interfaces to the fake DNS resolver
1997 TF/06 The fix for widen_domains has also been applied to qualify_single and
1998 search_parents which are the other dnslookup options that can cause
2001 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2004 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2006 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2007 service" if it encountered an MX record with an empty target hostname.
2008 The message is now "an MX or SRV record indicated no SMTP service".
2010 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2011 if the DNS of the sending site is misconfigured. This is quite a
2012 common situation. This change restores the behaviour of treating a
2013 helo verification defer as a failure.
2015 PH/51 If self=fail was set on a router, the bounce message did not include the
2016 actual error message.
2022 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2024 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2025 command itself ended in error, the underlying message about the transport
2026 filter timeout was being overwritten with the pipe command error. Now the
2027 underlying error message should be appended to the second error message.
2029 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2031 PH/02 Reduce the amount of output that "make" produces by default. Full output
2032 can still be requested.
2034 PH/03 The warning log line about a condition test deferring for a "warn" verb
2035 was being output only once per connection, rather than after each
2036 occurrence (because it was using the same function as for successful
2037 "warn" verbs). This seems wrong, so I have changed it.
2039 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2040 it should not have, which might have caused a crash in the right
2041 circumstances, but probably never did.
2043 PH/04 Installed a modified version of Tony Finch's patch to make submission
2044 mode fix the return path as well as the Sender: header line, and to
2045 add a /name= option so that you can make the user's friendly name appear
2048 TF/03 Added the control = fakedefer ACL modifier.
2050 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2051 Mark Lowes for thorough testing.
2053 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2055 TK/03 Merged latest SRS patch from Miles Wilton.
2057 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2058 with the definition in sysexits.h (which is #included earlier).
2059 Fortunately, Exim does not actually use EX_OK. The code used to try to
2060 preserve the sysexits.h value, by assumimg that macro definitions were
2061 scanned for macro replacements. I have been disabused of this notion,
2062 so now the code just undefines EX_OK before #including unistd.h.
2064 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2065 in the smtp transport. When a block could not be written in a single
2066 write() function, the timeout was being re-applied to each part-write.
2067 This seems wrong - if the receiver was accepting one byte at a time it
2068 would take for ever. The timeout is now adjusted when this happens. It
2069 doesn't have to be particularly precise.
2071 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2072 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2074 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2075 without a domain if the check to <postmaster@domain> fails.
2077 SC/01 Eximstats: added -xls and the ability to specify output files
2078 (patch written by Frank Heydlauf).
2080 SC/02 Eximstats: use FileHandles for outputing results.
2082 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2084 SC/04 Eximstats: fixed display of large numbers with -nvr option
2086 SC/05 Eximstats: fixed merging of reports with empty tables.
2088 SC/06 Eximstats: added the -include_original_destination flag
2090 SC/07 Eximstats: removed tabs and trailing whitespace.
2092 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2094 TK/06 MBOX spool code: Add real "From " MBOX separator line
2095 so the .eml file is really in mbox format (even though
2096 most programs do not really care). Patch from Alex Miller.
2098 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2099 The latter is generated from $received_to and is only set if the
2100 message has one envelope recipient. SA can use these headers,
2101 obviously out-of-the-box. Patch from Alex Miller.
2103 PH/08 The ${def test on a variable was returning false if the variable's
2104 value was "0", contrary to what the specification has always said!
2105 The result should be true unless the variable is empty.
2107 PH/09 The syntax error of a character other than { following "${if
2108 def:variable_name" (after optional whitespace) was not being diagnosed.
2109 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2110 accidental colon was present, for example, could give incorrect results.
2112 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2113 result is used (not including appendfile, where other changes are about
2116 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2117 This involved changing a lot of size variables from int to off_t. It
2118 should work with maildirs and everything.
2120 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2121 spamd dying while we are connected to it.
2123 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2126 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2127 function for BDB 4.3.
2129 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2131 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2132 the sources for PCRE have been re-arranged and more files are now
2135 PH/15 The code I had for printing potentially long long variables in PH/11
2136 above was not the best (it lost precision). The length of off_t variables
2137 is now inspected at build time, and an appropriate printing format (%ld
2138 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2139 to be "long long int" or "long int". This is needed for the internal
2140 formatting function string_vformat().
2142 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2143 the configuration file to be ":syslog", then the script "guesses" where
2144 the logs files are, rather than using the compiled in default. In our
2145 case the guess is not the same as the compiled default, so the script
2146 suddenly stopped working when I started to use syslog. The patch checks
2147 to see if log_file_path is "". If so, it attempts to read it from exim
2148 with no configuration file to get the compiled in version, before it
2149 falls back to the previous guessing code."
2151 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2152 implementing BATV in an Exim configuration. See NewStuff for the gory
2155 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2156 Makefile that are specific to HP-UX.
2158 PH/18 If the "use_postmaster" option was set for a recipient callout together
2159 with the "random" option, the postmaster address was used as the MAIL
2160 FROM address for the random test, but not for the subsequent recipient
2161 test. It is now used for both.
2163 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2164 patch removes a few documentation additions to RFC 3028, because the
2165 latest draft now contains them. It adds the new en;ascii-case comparator
2166 and a new error check for 8bit text in MIME parts. Comparator and
2167 require names are now matched exactly. I enabled the subaddress
2168 extension, but it is not well tested yet (read: it works for me)."
2170 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2171 rework some of the code of TK/09 above to avoid the hardwired use of
2172 "%lld" and "long long". Replaced the call to snprintf() with a call to
2175 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2176 records point to non-existent hosts", "retry timeout exceeded", and
2177 "retry time not reached for any host after a long failure period".
2179 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2180 experimental DomainKeys support:
2182 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2183 (2) On an error such as an illegally used "control", the wrong name for
2184 the control was given.
2186 These problems did NOT occur unless DomainKeys support was compiled.
2188 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2190 PH/24 Added ${if match_ip condition.
2192 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2193 databases so that it will be absolutely obvious if a crash occurs in the
2194 DB library. This is a regular occurrence (often caused by mis-matched
2197 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2198 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2199 picked up on a user's system that detects such things. There doesn't seem
2200 to be a gcc warning option for this - only an attribute that has to be
2201 put on the function's prototype. It seems that in Fedora Core 4 they have
2202 set this on a number of new functions. No doubt there will be more in due
2205 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2206 specify a transport. However, if an address that was verified by such a
2207 router was the subject of a callout, Exim crashed because it tried to
2208 read the rcpt_include_affixes from the non-existent transport. Now it
2209 just assumes that the setting of that option is false. This bug was
2210 introduced by 4.51/PH/31.
2212 PH/28 Changed -d+all to exclude +memory, because that information is very
2213 rarely of interest, but it makes the output a lot bigger. People tend to
2214 do -d+all out of habit.
2216 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2217 code in os-type was giving problems when libc.so lives in lib64, like on
2220 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2221 aren't the modern standard, and it seems that some systems' include files
2222 don't always have them. Exim was already checking for some of the newer
2223 ones like T_AAAA, and defining it itself. I've added checks for all the
2224 record types that Exim uses.
2226 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2227 not automatically generating a new one, as it is supposed to. This
2228 prevented TLS from working. If the file did exist, but contained invalid
2229 data, a new version was generated, as expected. It was only the case of a
2230 non-existent file that was broken.
2232 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2233 with a change in libdomainkeys > 0.64.
2235 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2236 from DNS. If the selector record carries the flag, it now has
2237 precedence over the domain-wide flag.
2239 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2241 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2242 the use of a port name that isn't defined in /etc/services) occurred, the
2243 message was deferred as in a normal delivery, and thus remained on the
2244 spool, instead of being failed because of the mua_wrapper setting. This
2245 is now fixed, and I tidied up some of the mua_wrapper messages at the
2248 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2249 the messages in a hash of arrays rather than using individual hashes.
2250 This is a bit cleaner and results in dramatic memory savings, albeit
2251 at a slight CPU cost.
2253 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2254 as requested by Marc Sherman.
2256 SC/10 Eximstats: added histograms for user specified patterns as requested
2259 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2261 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2262 fopen() in the content-scanning modules that did not already have it.
2268 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2269 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2271 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2273 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2275 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2276 to test Sieve filters that use "vacation".
2278 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2279 that changes the way the GnuTLS parameters are stored in the cache file.
2280 The new format can be generated externally. For backward compatibility,
2281 if the data in the cache doesn't make sense, Exim assumes it has read an
2282 old-format file, and it generates new data and writes a new file. This
2283 means that you can't go back to an older release without removing the
2286 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2287 work if there are any delivery delays because "one_time" forces the
2288 parent to be marked "delivered", so its unseen clone is never tried
2289 again. For this reason, Exim now forbids the simultaneous setting of
2292 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2293 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2294 entry if you want to know the details. The fix, however, neglected to
2295 consider the case when local delivery batching is involved. The test for
2296 "previously delivered" was not happening when checking to see if an
2297 address could be batched with a previous (undelivered) one; under
2298 certain circumstances this could lead to multiple deliveries to the same
2301 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2302 in its include files, and this causes problems building Exim.
2304 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2305 header_syntax) but Exim was just ignoring anything given after a slash.
2306 In particular, this caused confusion with an attempt to use "verify =
2307 reverse_host_lookup/defer_ok". An error is now given when options are
2308 supplied for verify items that do not have them. (Maybe reverse_host_
2309 lookup should have a defer_ok option, but that's a different point.)
2311 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2312 defined by RFC 821) to 2048, because there were problems with some AUTH
2313 commands, and RFC 1869 says the size should be increased for extended
2314 SMTP commands that take arguments.
2316 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2319 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2320 "unknown" error; now it says that the functionality isn't in the binary.
2322 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2323 an address' error message when a string expansion fails (syntax or
2324 whatever). Otherwise the password may appear in the log. Following change
2325 PH/42 below, there is no longer a chance of it appearing in a bounce
2328 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2330 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2331 of its addresses was ever tried. (Bugzilla bug #2.)
2333 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2334 the result incorrectly in the debug output. (It correctly added a newline
2335 to what was transported.)
2337 TF/01 Added $received_time.
2339 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2340 commented out examples of how to interface to a virus scanner and to
2341 SpamAssassin. Also added commented examples of av_scanner and
2342 spamd_address settings.
2344 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2345 and controls are allowed in which ACLs. There were a couple of minor
2346 errors. Some of the entries in the conditions table (which is a table of
2347 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2348 negation of where the condition IS allowed.
2350 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2352 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2353 header file does not have a version number, so I've had to invent a new
2354 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2355 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2356 radiusclient), but it was contributed by a Radius user.
2358 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2359 files or directories, for OpenSSL.
2361 PH/19 When an Exim process that is running as root has to create an Exim log
2362 file, it does so in a subprocess that runs as exim:exim so as to get the
2363 ownership right at creation (otherwise, other Exim processes might see
2364 the file with the wrong ownership). There was no test for failure of this
2365 fork() call, which would lead to the process getting stuck as it waited
2366 for a non-existent subprocess. Forks do occasionally fail when resources
2367 run out. I reviewed all the other calls to fork(); they all seem to check
2370 PH/20 When checking for unexpected SMTP input at connect time (before writing
2371 the banner), Exim was not dealing correctly with a non-positive return
2372 from the read() function. If the client had disconnected by this time,
2373 the result was a log entry for a synchronization error with an empty
2374 string after "input=" when read() returned zero. If read() returned -1
2375 (an event I could not check), uninitialized data bytes were printed.
2376 There were reports of junk text (parts of files, etc) appearing after
2379 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2381 PH/22 Added support for macro redefinition, and (re)definition in between
2382 driver and ACL definitions.
2384 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2385 forgetting to use the resulting value; it was using the unexpanded value.
2387 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2388 hadn't been configured. The fix is from Juergen Kreileder, who
2389 understands it better than I do:
2391 "Here's what I see happening with three configured cyrus_sasl
2392 authenticators configured (plain, login, cram-md5):
2394 On startup auth_cyrus_sasl_init() gets called for each of these.
2395 This means three calls to sasl_listmech() without a specified mech_list.
2396 => SASL tests which mechs of all available mechs actually work
2397 => three warnings about OTP not working
2398 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2400 With the patch, sasl_listmech() also gets called three times. But now
2401 SASL's mech_list option is set to the server_mech specified in the the
2402 authenticator. Or in other words, the answer from sasl_listmech()
2403 gets limited to just the mech you're testing for (which is different
2405 => the return list contains just 'plain' or 'login', 'cram-md5' or
2406 nothing depending on the value of ob->server_mech.
2408 I've just tested the patch: Authentication still works fine,
2409 unavailable mechs specified in the exim configuration are still
2410 caught, and the auth.log warnings about OTP are gone."
2412 PH/25 When debugging is enabled, the contents of the command line are added
2413 to the debugging output, even when log_selector=+arguments is not
2416 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2417 answer is "GNU", and only if the return is "GNU/something" is the answer
2420 PH/27 $acl_verify_message is now set immediately after the failure of a
2421 verification in an ACL, and so is available in subsequent modifiers. In
2422 particular, the message can be preserved by coding like this:
2424 warn !verify = sender
2425 set acl_m0 = $acl_verify_message
2427 Previously, $acl_verify_message was set only while expanding "message"
2428 and "log_message" when a very denied access.
2430 PH/28 Modified OS/os.c-Linux with
2432 -#ifndef OS_LOAD_AVERAGE
2433 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2435 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2436 nomenclature these days.)
2438 PH/29 Installed patch from the Sieve maintainer that adds the options
2439 sieve_useraddress and sieve_subaddress to the redirect router.
2441 PH/30 In these circumstances:
2442 . Two addresses routed to the same list of hosts;
2443 . First host does not offer TLS;
2444 . First host accepts first address;
2445 . First host gives temporary error to second address;
2446 . Second host offers TLS and a TLS session is established;
2447 . Second host accepts second address.
2448 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2449 and peerdn, if requested) that were in fact used only for the second
2452 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2453 attention to any local part prefix or suffix that was matched by the
2454 router that accepted the address. It now behaves in the same way as it
2455 does for delivery: the affixes are removed from the local part unless
2456 rcpt_include_affixes is set on the transport.
2458 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2459 timeout during the DATA phase of an incoming message.
2461 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2462 applied a patch sanctioned by the Sieve maintainer.
2464 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2465 the uid or gid is negative. A case of a negative gid caused this to be
2466 noticed. The fix allows for either to be negative.
2468 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2469 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2472 PH/36 The change PH/12 above was broken. Fixed it.
2474 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2475 the grounds that routing the same address twice would always produce the
2476 same answer. This might have been true once, but it is certainly no
2477 longer true now. Routing a child address may depend on the previous
2478 routing that produced that child. Some complicated redirection strategies
2479 went wrong when messages had multiple recipients, and made Exim's
2480 behaviour dependent on the order in which the addresses were given.
2482 I have moved the duplicate checking until after the routing is complete.
2483 Exim scans the addresses that are assigned to local and remote
2484 transports, and removes any duplicates. This means that more work will be
2485 done, as duplicates will always all be routed, but duplicates are
2486 presumably rare, so I don't expect this is of any significance.
2488 For deliveries to pipes, files, and autoreplies, the duplicate checking
2489 still happens during the routing process, since they are not going to be
2492 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2493 It corrects a timeout issue with spamd. This is Ian's comment: "The
2494 background is that sometimes spamd either never reads data from a
2495 connection it has accepted, or it never writes response data. The exiscan
2496 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2497 blindly assumes that writes won't block so it may never time out."
2499 PH/39 Allow G after quota size as well as K and M.
2501 PH/40 The value set for $authenticated_id in an authenticator may not contain
2502 binary zeroes or newlines because the value is written to log lines and
2503 to spool files. There was no check on this. Now the value is run through
2504 the string_printing() function so that such characters are converted to
2505 printable escape sequences.
2507 PH/41 $message_linecount is a new variable that contains the total number of
2508 lines in the message. Compare $body_linecount, which is the count for the
2511 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2512 bounce and delay warning messages, except in certain special cases, which
2515 (a) An SMTP error message from a remote host;
2516 (b) A message specified in a :fail: redirection;
2517 (c) A message specified in a "fail" command in a system filter;
2518 (d) A message specified in a FAIL return from the queryprogram router;
2519 (e) A message specified by the cannot_route_message router option.
2521 In these cases only, Exim does include the error details in bounce and
2522 warning messages. There are also a few cases where bland messages such
2523 as "unrouteable address" or "local delivery error" are given.
2525 PH/43 $value is now also set for the "else" part of a ${run expansion.
2527 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2528 being worked on, but at least Exim now implements the latest version to
2531 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2532 process to complete was treated as a delivery failure, a timeout while
2533 writing the message to the pipe was logged, but erroneously treated as a
2534 successful delivery. Such timeouts include transport filter timeouts. For
2535 consistency with the overall process timeout, these timeouts are now
2536 treated as errors, giving rise to delivery failures by default. However,
2537 there is now a new Boolean option for the pipe transport called
2538 timeout_defer, which, if set TRUE, converts the failures into defers for
2539 both kinds of timeout. A transport filter timeout is now identified in
2542 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2543 systems where "make" and "gmake" are different, calling "gmake" at top
2544 level broke things. I've arranged for the value of $(MAKE) to be passed
2545 from the Makefile to this script so that it can call the same version of
2549 A note about Exim versions 4.44 and 4.50
2550 ----------------------------------------
2552 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2553 changes of various kinds. As a consequence, a big documentation update was
2554 needed. This delayed the release for rather longer than seemed good, especially
2555 in the light of a couple of (minor) security issues. Therefore, the changes
2556 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2557 release. So 4.44 and 4.50 are in effect two different branches that both start
2560 I have left the 4.50 change log unchanged; it contains all the changes since
2561 4.43. The change log for 4.44 is below; many of its items are identical to
2562 those for 4.50. This seems to be the most sensible way to preserve the
2563 historical information.
2569 1. Minor wording change to the doc/README.SIEVE file.
2571 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2572 computation of the current number of files was incorrect.
2574 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2575 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2578 4. Give more explanation in the error message when the command for a transport
2579 filter fails to execute.
2581 5. There are several places where Exim runs a non-Exim command in a
2582 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2583 was being done only for the command run by the queryprogram router. It is
2584 now done for all such subprocesses. The other cases are: ${run, transport
2585 filters, and the commands run by the lmtp and pipe transports.
2587 6. Added CONFIGURE_GROUP build-time option.
2589 7. Some older OS have a limit of 256 on the maximum number of file
2590 descriptors. Exim was using setrlimit() to set 1000 as a large value
2591 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2592 systems. I've change it so that if it can't get 1000, it tries for 256.
2594 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2595 was an oversight, and furthermore, ever since the addition of extra
2596 controls (e.g. 4.43/32), the checks on when to allow different forms of
2597 "control" were broken. There should now be diagnostics for all cases when a
2598 control that does not make sense is encountered.
2600 9. Added the /retain_sender option to "control=submission".
2602 10. $recipients is now available in the predata ACL (oversight).
2604 11. Tidy the search cache before the fork to do a delivery from a message
2605 received from the command line. Otherwise the child will trigger a lookup
2606 failure and thereby defer the delivery if it tries to use (for example) a
2607 cached ldap connection that the parent has called unbind on.
2609 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2610 of $address_data from the recipient verification was clobbered by the
2611 sender verification.
2613 13. The value of address_data from a sender verification is now available in
2614 $sender_address_data in subsequent conditions in the ACL statement.
2616 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2618 15. Added a new option "connect=<time>" to callout options, to set a different
2621 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2622 was its contents. (It was OK if the option was not defined at all.)
2624 17. A "Completed" log line is now written for messages that are removed from
2625 the spool by the -Mrm option.
2627 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2628 information about exactly what failed.
2630 19. Added -dd to debug only the daemon process.
2632 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2633 handles timeouts, both on the server side and network timeouts. Renamed the
2634 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2636 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2637 It is now set to "smtps".
2639 22. $host_address is now set to the target address during the checking of
2640 ignore_target_hosts.
2642 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2643 being passed; this would have caused $sender_host_name to have been used if
2644 matching the list had actually called for a host name (not very likely,
2645 since this list is usually IP addresses). A host name is now passed as
2648 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2649 code that specifies a non-restarting handler (typically sigaction() in
2650 modern systems) in an attempt to fix a rare and obscure crash bug.
2652 25. Narrowed the window for a race in the daemon that could cause it to ignore
2653 SIGCHLD signals. This is not a major problem, because they are used only to
2654 wake it up if nothing else does.
2656 26. A malformed maildirsize file could cause Exim to calculate negative values
2657 for the mailbox size or file count. Odd effects could occur as a result.
2658 The maildirsize information is now recalculated if the size or filecount
2661 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2662 support for a long time. Removed HAVE_SYS_VFS_H.
2664 28. Installed the latest version of exipick from John Jetmore.
2666 29. In an address list, if the pattern was not a regular expression, an empty
2667 subject address (from a bounce message) matched only if the pattern was an
2668 empty string. Non-empty patterns were not even tested. This was the wrong
2669 because it is perfectly reasonable to use an empty address as part of a
2670 database query. An empty address is now tested by patterns that are
2671 lookups. However, all the other forms of pattern expect the subject to
2672 contain a local part and a domain, and therefore, for them, an empty
2673 address still always fails if the pattern is not itself empty.
2675 30. Exim went into a mad DNS loop when attempting to do a callout where the
2676 host was specified on an smtp transport, and looking it up yielded more
2677 than one IP address.
2679 31. Re-factored the code for checking spool and log partition space into a
2680 function that finds that data and another that does the check. The former
2681 is then used to implement four new variables: $spool_space, $log_space,
2682 $spool_inodes, and $log_inodes.
2684 32. The RFC2047 encoding function was originally intended for short strings
2685 such as real names; it was not keeping to the 75-character limit for
2686 encoded words that the RFC imposes. It now respects the limit, and
2687 generates multiple encoded words if necessary. To be on the safe side, I
2688 have increased the buffer size for the ${rfc2047: expansion operator from
2691 33. It is now permitted to omit both strings after an "if" condition; if the
2692 condition is true, the result is "true". As before, when the second string
2693 is omitted, a false condition yields an empty string. This makes it less
2694 cumbersome to write custom ACL and router conditions.
2696 34. Failure to deliver a bounce message always caused it to be frozen, even if
2697 there was an errors_to setting on the router. The errors_to setting is now
2700 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2701 canonical form (fully expanded) before being placed in
2702 $sender_host_address.
2704 36. The table in the code that translates DNS record types into text (T_A to
2705 "A" for instance) was missing entries for NS and CNAME. It is just possible
2706 that this could have caused confusion if both these types were looked up
2707 for the same domain, because the text type is used as part of Exim's
2708 per-process caching. But the chance of anyone hitting this buglet seems
2711 37. The dnsdb lookup has been extended in a number of ways.
2713 (1) There is a new type, "zns", which walks up the domain tree until it
2714 finds some nameserver records. It should be used with care.
2716 (2) There is a new type, "mxh", which is like "mx" except that it returns
2717 just the host names, not the priorities.
2719 (3) It is now possible to give a list of domains (or IP addresses) to be
2720 looked up. The behaviour when one of the lookups defers can be
2721 controlled by a keyword.
2723 (4) It is now possible to specify the separator character for use when
2724 multiple records are returned.
2726 38. The dnslists ACL condition has been extended: it is now possible to supply
2727 a list of IP addresses and/or domains to be looked up in a particular DNS
2730 39. Added log_selector=+queue_time_overall.
2732 40. When running the queue in the test harness, wait just a tad after forking a
2733 delivery process, to get repeatability of debugging output.
2735 41. Include certificate and key file names in error message when GnuTLS fails
2736 to set them up, because the GnuTLS error message doesn't include the name
2737 of the failing file when there is a problem reading it.
2739 42. Allow both -bf and -bF in the same test run.
2741 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2743 44. The "Exiscan patch" is now merged into the mainline Exim source.
2745 45. Sometimes the final signoff response after QUIT could fail to get
2746 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2747 < 0 before doing a fflush(). This bug looks as though it goes back to the
2748 introduction of TLS in release 3.20, but "sometimes" must have been rare
2749 because the tests only now provoked it.
2751 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2752 (this can affect the format of dates).
2754 47. exim_tidydb, when checking for the continued existence of a message for
2755 which it has found a message-specific retry record, was not finding
2756 messages that were in split spool directories. Consequently, it was
2757 deleting retry records that should have stayed in existence.
2759 48. Steve fixed some bugs in eximstats.
2761 49. The SPA authentication driver was not abandoning authentication and moving
2762 on to the next authenticator when an expansion was forced to fail,
2763 contradicting the general specification for all authenticators. Instead it
2764 was generating a temporary error. It now behaves as specified.
2766 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2767 (the order specifies the preference for clients). The order is now AES256,
2768 AES128, 3DES, ARCFOUR128.
2770 51. Small patch to Sieve code - explicitly set From: when generating an
2773 52. Exim crashed if a remote delivery caused a very long error message to be
2774 recorded - for instance if somebody sent an entire SpamAssassin report back
2775 as a large number of 550 error lines. This bug was coincidentally fixed by
2776 increasing the size of one of Exim's internal buffers (big_buffer) that
2777 happened as part of the Exiscan merge. However, to be on the safe side, I
2778 have made the code more robust (and fixed the comments that describe what
2781 53. Now that there can be additional text after "Completed" in log lines (if
2782 the queue_time_overall log selector is set), a one-byte patch to exigrep
2783 was needed to allow it to recognize "Completed" as not the last thing in
2786 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2787 patch that reportedly fixes this has been added. I am not expert enough to
2788 create a test for it. This is what the patch creator wrote:
2790 "I found a little strange behaviour of ldap code when working with
2791 Windows 2003 AD Domain, where users was placed in more than one
2792 Organization Units. When I tried to give exim partial DN, the exit code
2793 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2794 But simultaneously result of request was absolutely normal ldap result,
2795 so I produce this patch..."
2797 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2798 REFERENCE, so I have modified the code to exclude the patch when that macro
2801 55. Some experimental protocols are using DNS PTR records for new purposes. The
2802 keys for these records are domain names, not reversed IP addresses. The
2803 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2804 leaves it alone. Component reversal etc. now happens only for IP addresses.
2807 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2809 57. Double the size of the debug message buffer (to 2048) so that more of very
2810 long debug lines gets shown.
2812 58. The exicyclog utility now does better if the number of log files to keep
2813 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2815 59. Two changes related to the smtp_active_hostname option:
2817 (1) $smtp_active_hostname is now available as a variable.
2818 (2) The default for smtp_banner uses $smtp_active_hostname instead
2819 of $primary_hostname.
2821 60. The host_aton() function is supposed to be passed a string that is known
2822 to be a valid IP address. However, in the case of IPv6 addresses, it was
2823 not checking this. This is a hostage to fortune. Exim now panics and dies
2824 if the condition is not met. A case was found where this could be provoked
2825 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2826 components; fortuitously, this particular loophole had already been fixed
2827 by change 4.50/55 above.
2829 If there are any other similar loopholes, the new check in host_aton()
2830 itself should stop them being exploited. The report I received stated that
2831 data on the command line could provoke the exploit when Exim was running as
2832 exim, but did not say which command line option was involved. All I could
2833 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2834 running as the user.
2837 61. There was a buffer overflow vulnerability in the SPA authentication code
2838 (which came originally from the Samba project). I have added a test to the
2839 spa_base64_to_bits() function which I hope fixes it.
2842 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2843 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2845 63. The daemon start-up calls getloadavg() while still root for those OS that
2846 need the first call to be done as root, but it missed one case: when
2847 deliver_queue_load_max is set with deliver_drop_privilege. This is
2848 necessary for the benefit of the queue runner, because there is no re-exec
2849 when deliver_drop_privilege is set.
2851 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2852 This has been fixed.
2854 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2855 was in use, was not putting the data itself into the right store pool;
2856 consequently, it could be overwritten for a subsequent message in the same
2857 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2860 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2862 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2863 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2864 the function were treating the return as a boolean value, which happened to
2865 work because 0=false and not-0=true, but is not correct code.
2867 68. The host_aton() function was not handling scoped IPv6 addresses (those
2868 with, for example, "%eth0" on the end) correctly.
2870 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2871 negated items (that is, ~something) in unsigned ints. Some compilers
2872 apparently mutter when there is no cast.
2874 70. If an address verification called from an ACL failed, and did not produce a
2875 user-specific message (i.e. there was only a "system" message), nothing was
2876 put in $acl_verify_message. In this situation, it now puts the system
2879 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2880 session; change 4.31/43 added the unwanted input to the log line - except
2881 that it did not do this in the start of session case. It now does.
2883 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2884 This is wrong and can cause the other end to generate a synchronization
2885 error if it is another Exim or anything else that does the synchronization
2886 check. A QUIT command is no longer sent after a timeout.
2888 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2889 during host lookups.
2891 74. The defer_ok option of callout verification was not working if it was used
2892 when verifying addresses in header lines, that is, for this case:
2894 verify = header_sender/callout=defer_ok
2896 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2897 those file descriptors could be used for SMTP connections. If anything
2898 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2899 could be sent to the SMTP client, causing chaos. The daemon now opens
2900 stdin, stdout, and stderr to /dev/null when it puts itself into the
2903 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2904 log by default. The user can override this with suitable Perl magic.
2906 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2907 the log message when discard triggers, was not working for the DATA ACL or
2908 for the non-SMTP ACL.
2910 78. Error message wording change in sieve.c.
2912 79. If smtp_accept_max_per_host was set, the number of connections could be
2913 restricted to fewer than expected, because the daemon was trying to set up
2914 a new connection before checking whether the processes handling previous
2915 connections had finished. The check for completed processes is now done
2916 earlier. On busy systems, this bug wouldn't be noticed because something
2917 else would have woken the daemon, and it would have reaped the completed
2920 80. If a message was submitted locally by a user whose login name contained one
2921 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2922 It caused a spool format error. I have fixed the spool reading code. A
2923 related problem was that the "from" clause in the Received: line became
2924 illegal because of the space(s). It is now covered by ${quote_local_part.
2926 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2929 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2931 83. Patch from Sieve maintainer to fix unterminated string problem in
2932 "vacation" handling.
2934 84. Some minor changes to the Linux configuration files to help with other
2935 OS variants using glibc.
2937 85. One more patch for Sieve to update vacation handling to latest spec.
2940 ----------------------------------------------------
2941 See the note above about the 4.44 and 4.50 releases.
2942 ----------------------------------------------------
2948 1. Change 4.43/35 introduced a bug that caused file counts to be
2949 incorrectly computed when quota_filecount was set in an appendfile
2952 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2953 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2956 3. Give more explanation in the error message when the command for a transport
2957 filter fails to execute.
2959 4. There are several places where Exim runs a non-Exim command in a
2960 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2961 was being done only for the command run by the queryprogram router. It is
2962 now done for all such subprocesses. The other cases are: ${run, transport
2963 filters, and the commands run by the lmtp and pipe transports.
2965 5. Some older OS have a limit of 256 on the maximum number of file
2966 descriptors. Exim was using setrlimit() to set 1000 as a large value
2967 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2968 systems. I've change it so that if it can't get 1000, it tries for 256.
2970 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2971 was an oversight, and furthermore, ever since the addition of extra
2972 controls (e.g. 4.43/32), the checks on when to allow different forms of
2973 "control" were broken. There should now be diagnostics for all cases when a
2974 control that does not make sense is encountered.
2976 7. $recipients is now available in the predata ACL (oversight).
2978 8. Tidy the search cache before the fork to do a delivery from a message
2979 received from the command line. Otherwise the child will trigger a lookup
2980 failure and thereby defer the delivery if it tries to use (for example) a
2981 cached ldap connection that the parent has called unbind on.
2983 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2984 of $address_data from the recipient verification was clobbered by the
2985 sender verification.
2987 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2988 was its contents. (It was OK if the option was not defined at all.)
2990 11. A "Completed" log line is now written for messages that are removed from
2991 the spool by the -Mrm option.
2993 12. $host_address is now set to the target address during the checking of
2994 ignore_target_hosts.
2996 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2997 being passed; this would have caused $sender_host_name to have been used if
2998 matching the list had actually called for a host name (not very likely,
2999 since this list is usually IP addresses). A host name is now passed as
3002 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3003 code that specifies a non-restarting handler (typically sigaction() in
3004 modern systems) in an attempt to fix a rare and obscure crash bug.
3006 15. Narrowed the window for a race in the daemon that could cause it to ignore
3007 SIGCHLD signals. This is not a major problem, because they are used only to
3008 wake it up if nothing else does.
3010 16. A malformed maildirsize file could cause Exim to calculate negative values
3011 for the mailbox size or file count. Odd effects could occur as a result.
3012 The maildirsize information is now recalculated if the size or filecount
3015 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3016 support for a long time. Removed HAVE_SYS_VFS_H.
3018 18. Updated exipick to current release from John Jetmore.
3020 19. Allow an empty sender to be matched against a lookup in an address list.
3021 Previously the only cases considered were a regular expression, or an
3024 20. Exim went into a mad DNS lookup loop when doing a callout where the
3025 host was specified on the transport, if the DNS lookup yielded more than
3028 21. The RFC2047 encoding function was originally intended for short strings
3029 such as real names; it was not keeping to the 75-character limit for
3030 encoded words that the RFC imposes. It now respects the limit, and
3031 generates multiple encoded words if necessary. To be on the safe side, I
3032 have increased the buffer size for the ${rfc2047: expansion operator from
3035 22. Failure to deliver a bounce message always caused it to be frozen, even if
3036 there was an errors_to setting on the router. The errors_to setting is now
3039 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3040 canonical form (fully expanded) before being placed in
3041 $sender_host_address.
3043 24. Updated eximstats to version 1.33
3045 25. Include certificate and key file names in error message when GnuTLS fails
3046 to set them up, because the GnuTLS error message doesn't include the name
3047 of the failing file when there is a problem reading it.
3049 26. Expand error message when OpenSSL has problems setting up cert/key files.
3052 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3053 (this can affect the format of dates).
3055 28. exim_tidydb, when checking for the continued existence of a message for
3056 which it has found a message-specific retry record, was not finding
3057 messages that were in split spool directories. Consequently, it was
3058 deleting retry records that should have stayed in existence.
3060 29. eximstats updated to version 1.35
3061 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3062 1.35 - bugfix such that pie charts by volume are generated correctly
3064 30. The SPA authentication driver was not abandoning authentication and moving
3065 on to the next authenticator when an expansion was forced to fail,
3066 contradicting the general specification for all authenticators. Instead it
3067 was generating a temporary error. It now behaves as specified.
3069 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3070 (the order specifies the preference for clients). The order is now AES256,
3071 AES128, 3DES, ARCFOUR128.
3073 31. Small patch to Sieve code - explicitly set From: when generating an
3076 32. Exim crashed if a remote delivery caused a very long error message to be
3077 recorded - for instance if somebody sent an entire SpamAssassin report back
3078 as a large number of 550 error lines. This bug was coincidentally fixed by
3079 increasing the size of one of Exim's internal buffers (big_buffer) that
3080 happened as part of the Exiscan merge. However, to be on the safe side, I
3081 have made the code more robust (and fixed the comments that describe what
3084 33. Some experimental protocols are using DNS PTR records for new purposes. The
3085 keys for these records are domain names, not reversed IP addresses. The
3086 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3087 leaves it alone. Component reversal etc. now happens only for IP addresses.
3090 34. The host_aton() function is supposed to be passed a string that is known
3091 to be a valid IP address. However, in the case of IPv6 addresses, it was
3092 not checking this. This is a hostage to fortune. Exim now panics and dies
3093 if the condition is not met. A case was found where this could be provoked
3094 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3095 components; fortuitously, this particular loophole had already been fixed
3096 by change 4.50/55 or 4.44/33 above.
3098 If there are any other similar loopholes, the new check in host_aton()
3099 itself should stop them being exploited. The report I received stated that
3100 data on the command line could provoke the exploit when Exim was running as
3101 exim, but did not say which command line option was involved. All I could
3102 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3103 running as the user.
3106 35. There was a buffer overflow vulnerability in the SPA authentication code
3107 (which came originally from the Samba project). I have added a test to the
3108 spa_base64_to_bits() function which I hope fixes it.
3111 36. The daemon start-up calls getloadavg() while still root for those OS that
3112 need the first call to be done as root, but it missed one case: when
3113 deliver_queue_load_max is set with deliver_drop_privilege. This is
3114 necessary for the benefit of the queue runner, because there is no re-exec
3115 when deliver_drop_privilege is set.
3117 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3118 was in use, was not putting the data itself into the right store pool;
3119 consequently, it could be overwritten for a subsequent message in the same
3120 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3123 38. Sometimes the final signoff response after QUIT could fail to get
3124 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3125 < 0 before doing a fflush(). This bug looks as though it goes back to the
3126 introduction of TLS in release 3.20, but "sometimes" must have been rare
3127 because the tests only now provoked it.
3133 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3134 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3135 Now it returns an integer. A number of calls were still expecting a T/F
3136 return. Fortuitously, in all cases, the tests worked in OK situations,
3137 which is the norm. However, things would have gone wrong on any write
3138 failures on the smtp file descriptor. This function is used when sending
3139 messages over SMTP and also when doing verify callouts.
3141 2. When Exim is called to do synchronous delivery of a locally submitted
3142 message (the -odf or -odi options), it no longer closes stderr before doing
3145 3. Implemented the mua_wrapper option.
3147 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3149 5. Implemented the functions header_remove(), header_testname(),
3150 header_add_at_position(), and receive_remove_recipient(), and exported them
3153 6. If an ACL "warn" statement specified the addition of headers, Exim already
3154 inserted X-ACL-Warn: at the start if there was no header name. However, it
3155 was not making this test for the second and subsequent header lines if
3156 there were newlines in the string. This meant that an invalid header could
3157 be inserted if Exim was badly configured.
3159 7. Allow an ACL "warn" statement to add header lines at the start or after all
3160 the Received: headers, as well as at the end.
3162 8. Added the rcpt_4xx retry error code.
3164 9. Added postmaster_mailfrom=xxx to callout verification option.
3166 10. Added mailfrom=xxxx to the callout verification option, for verify=
3169 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3170 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3172 12. Inserted some casts to stop certain compilers warning when using pointer
3173 differences as field lengths or precisions in printf-type calls (mostly
3174 affecting debugging statements).
3176 13. Added optional readline() support for -be (dynamically loaded).
3178 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3179 same clock tick as a message's arrival, so that its received time was the
3180 same as the "first fail" time on the retry record, and that message
3181 remained on the queue past the ultimate address timeout, every queue runner
3182 would try a delivery (because it was past the ultimate address timeout) but
3183 after another failure, the ultimate address timeout, which should have then
3184 bounced the address, did not kick in. This was a "< instead of <=" error;
3185 in most cases the first failure would have been in the next clock tick
3186 after the received time, and all would be well.
3188 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3189 being recognized when the domain list was tested by the match_domain
3190 condition in an expansion string.
3192 16. Added the ${str2b64: operator.
3194 17. Exim was always calling setrlimit() to set a large limit for the number of
3195 processes, without checking whether the existing limit was already
3196 adequate. (It did check for the limit on file descriptors.) Furthermore,
3197 errors from getrlimit() and setrlimit() were being ignored. Now they are
3198 logged to the main and panic logs, but Exim does carry on, to try to do its
3199 job under whatever limits there are.
3201 18. Imported PCRE 5.0.
3203 19. Trivial typo in log message " temporarily refused connection" (the leading
3206 20. If the log selector return_path_on_delivery was set and an address was
3207 redirected to /dev/null, the delivery process crashed because it assumed
3208 that a return path would always be set for a "successful" delivery. In this
3209 case, the whole delivery is bypassed as an optimization, and therefore no
3212 21. Internal re-arrangement: the function for sending a challenge and reading
3213 a response while authentication was assuming a zero-terminated challenge
3214 string. It's now changed to take a pointer and a length, to allow for
3215 binary data in such strings.
3217 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3219 23. Exim was not respecting finduser_retries when seeking the login of the
3220 uid under which it was called; it was always trying 10 times. (The default
3221 setting of finduser_retries is zero.) Also, it was sleeping after the final
3222 failure, which is pointless.
3224 24. Implemented tls_on_connect_ports.
3226 25. Implemented acl_smtp_predata.
3228 26. If the domain in control=submission is set empty, Exim assumes that the
3229 authenticated id is a complete email address when it generates From: or
3230 Sender: header lines.
3232 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3233 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3234 chown and chgrp in /bin and hostname in /usr/bin.
3236 28. Exim was keeping the "process log" file open after each use, just as it
3237 does for the main log. This opens the possibility of it remaining open for
3238 long periods when the USR1 signal hits a daemon. Occasional processlog
3239 errors were reported, that could have been caused by this. Anyway, it seems
3240 much more sensible not to leave this file open at all, so that is what now
3243 29. The long-running daemon process does not normally write to the log once it
3244 has entered its main loop, and it closes the log before doing so. This is
3245 so that log files can straightforwardly be renamed and moved. However,
3246 there are a couple of unusual error situations where the daemon does write
3247 log entries, and I had neglected to close the log afterwards.
3249 30. The text of an SMTP error response that was received during a remote
3250 delivery was being truncated at 512 bytes. This is too short for some of
3251 the long messages that one sometimes sees. I've increased the limit to
3254 31. It is now possible to make retry rules that apply only when a message has a
3255 specific sender, in particular, an empty sender.
3257 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3258 it possible to be selective about when SMTP synchronization is enforced.
3260 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3262 32. Implemented hosts_connection_nolog.
3264 33. Added an ACL for QUIT.
3266 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3269 35. Added mailbox_size and mailbox_filecount to appendfile.
3271 36. Added control = no_multiline_responses to ACLs.
3273 37. There was a bug in the logic of the code that waits for the clock to tick
3274 in the case where the clock went backwards by a substantial amount such
3275 that the microsecond fraction of "now" was more than the microsecond
3276 fraction of "then" (but the whole seconds number was less).
3278 38. Added support for the libradius Radius client library this is found on
3279 FreeBSD (previously only the radiusclient library was supported).
3285 1. When certain lookups returned multiple values in the form name=value, the
3286 quoting of the values was not always being done properly. Specifically:
3287 (a) If the value started with a double quote, but contained no whitespace,
3289 (b) If the value contained whitespace other than a space character (i.e.
3290 tabs or newlines or carriage returns) it was not quoted.
3291 This fix has been applied to the mysql and pgsql lookups by writing a
3292 separate quoting function and calling it from the lookup code. The fix
3293 should probably also be applied to nisplus, ibase and oracle lookups, but
3294 since I cannot test any of those, I have not disturbed their existing code.
3296 2. A hit in the callout cache for a specific address caused a log line with no
3297 reason for rejecting RCPT. Now it says "Previous (cached) callout
3298 verification failure".
3300 3. There was an off-by-one bug in the queryprogram router. An over-long
3301 return line was truncated at 256 instead of 255 characters, thereby
3302 overflowing its buffer with the terminating zero. As well as fixing this, I
3303 have increased the buffer size to 1024 (and made a note to document this).
3305 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3306 when Exim is waiting for an SMTP response from a remote server, Exim
3307 restarts its select() call on the socket, thereby resetting its timeout.
3308 This is not a problem when such interrupts are rare. Somebody set up a cron
3309 job to run exiwhat every 2 minutes, which is less than the normal select()
3310 timeout (5 or 10 minutes). This meant that the select() timeout never
3311 kicked in because it was always reset. I have fixed this by comparing the
3312 time when an interrupt arrives with the time at the start of the first call
3313 to select(). If more time than the timeout has elapsed, the interrupt is
3314 treated as a timeout.
3316 5. Some internal re-factoring in preparation for the addition of Sieve
3317 extensions (by MH). In particular, the "personal" test is moved to a
3318 separate function, and given an option for scanning Cc: and Bcc: (which is
3319 not set for Exim filters).
3321 6. When Exim created an email address using the login of the caller as the
3322 local part (e.g. when creating a From: or Sender: header line), it was not
3323 quoting the local part when it contained special characters such as @.
3325 7. Installed new OpenBSD configuration files.
3327 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3328 try to make them clearer.
3330 9. Callout options, other than the timeout value, were being ignored when
3331 verifying sender addresses in header lines. For example, when using
3333 verify = header_sender/callout=no_cache
3335 the cache was (incorrectly) being used.
3337 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3338 only the Cygwin environment.
3340 11. When return_path_on_delivery was set as a log selector, if different remote
3341 addresses in the same message used different return paths and parallel
3342 remote delivery occurred, the wrong values would sometimes be logged.
3343 (Whenever a remote delivery process finished, the return path value from
3344 the most recently started remote delivery process was logged.)
3346 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3347 lines when AUTH and/or TLS are in use. This is the "received protocol"
3348 field. Exim used to use "asmtp" for authenticated SMTP, without any
3349 indication (in the protocol name) for TLS use. Now it follows the RFC and
3350 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3351 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3352 names appear in log lines as well as in Received: header lines.
3354 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3355 extensions, and comparison tests, and to fix some bugs.
3357 14. Changes to the "personal" filter test:
3359 (1) The test was buggy in that it was just doing the equivalent of
3360 "contains" tests on header lines. For example, if a user's address was
3361 anne@some.where, the "personal" test would incorrectly be true for
3363 To: susanne@some.where
3365 This test is now done by extracting each address from the header in turn,
3366 and checking the entire address. Other tests that are part of "personal"
3367 are now done using regular expressions (for example, to check local parts
3368 of addresses in From: header lines).
3370 (2) The list of non-personal local parts in From: addresses has been
3371 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3372 taken from the Sieve specification recommendations.
3374 (3) If the message contains any header line starting with "List-" it is
3375 treated as non-personal.
3377 (4) The test for "circular" in the Subject: header line has been removed
3378 because it now seems ill-conceived.
3380 15. Minor typos in src/EDITME comments corrected.
3382 16. Installed latest exipick from John Jetmore.
3384 17. If headers_add on a router specified a text string that was too long for
3385 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3386 of string_sprintf() is now avoided.
3388 18. $message_body_size was not set (it was always zero) when running the DATA
3389 ACL and the local_scan() function.
3391 19. For the "mail" command in an Exim filter, no default was being set for
3392 the once_repeat time, causing a random time value to be used if "once" was
3393 specified. (If the value happened to be <= 0, no repeat happened.) The
3394 default is now 0s, meaning "never repeat". The "vacation" command was OK
3395 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3396 (I found it when inspecting the code).
3398 20. There is now an overall timeout for performing a callout verification. It
3399 defaults to 4 times the callout timeout, which applies to individual SMTP
3400 commands during the callout. The overall timeout applies when there is more
3401 than one host that can be tried. The timeout is checked before trying the
3402 next host. This prevents very long delays if there are a large number of
3403 hosts and all are timing out (e.g. when the network connections are timing
3404 out). The value of the overall timeout can be changed by specifying an
3405 additional sub-option for "callout", called "maxwait". For example:
3407 verify = sender/callout=5s,maxwait=20s
3409 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3410 to the end before writing, but this should make it even safer).
3412 22. Exim was forgetting that it had advertised PIPELINING for the second and
3413 subsequent messages on an SMTP connection. It was also not resetting its
3414 memory on STARTTLS and an internal HELO.
3416 23. When Exim logs an SMTP synchronization error within a session, it now
3417 records whether PIPELINING has been advertised or not.
3419 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3420 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3421 rather than long int.
3423 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3425 26. Added the never_mail option to autoreply.
3431 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3432 crash if the getsockname() call failed; this can happen if a connection is
3433 closed very soon after it is established. The problem was simply in the
3434 order in which certain operations were done, causing Exim to try to write
3435 to the SMTP stream before it had set up the file descriptor. The bug has
3436 been fixed by making things happen in the correct order.
3442 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3443 before the connection was closed, thus losing the rejection response.
3445 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3446 some early Solaris releases, but causes trouble in current releases where
3447 socklen_t is defined.
3449 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3452 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3455 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3456 that releasing the top of it at the end releases what was used for sub-
3457 expansions (unless the block got too big). However, discard this block if
3458 the first thing is a variable or header, so that we can use its block when
3459 it is dynamic (useful for very large $message_headers, for example).
3461 6. Lookups now cache *every* query, not just the most recent. A new, separate
3462 store pool is used for this. It can be recovered when all lookup caches are
3463 flushed. Lookups now release memory at the end of their result strings.
3464 This has involved some general refactoring of the lookup sources.
3466 7. Some code has been added to the store_xxx() functions to reduce the amount
3467 of flapping under certain conditions.
3469 8. log_incoming_interface used to affect only the <= reception log lines. Now
3470 it causes the local interface and port to be added to several more SMTP log
3471 lines, for example "SMTP connection from", and rejection lines.
3473 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3475 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3477 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3478 could be overwritten at the end of the current message (or the start of a
3479 new message if it was set in a HELO ACL). The value is now preserved for
3480 the duration of the SMTP connection.
3482 12. If a transport had a headers_rewrite setting, and a matching header line
3483 contained an unqualified address, that address was qualified, even if it
3484 did not match any rewriting rules. The underlying bug was that the values
3485 of the flags that permit the existence of unqualified sender and recipient
3486 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3487 non-local messages, and by -bnq for local messages) were not being
3488 preserved with the message after it was received.
3490 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3491 "next input=" as part of the text comprising the host identity instead of
3492 the correct text. The code was using the same buffer for two different
3493 strings. However, depending on which order the printing function evaluated
3494 its arguments, the bug did not always show up. Under Linux, for example, my
3495 test suite worked just fine.
3497 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3498 doesn't work with some older versions of Perl. It has been changed to "my",
3499 which in any case is probably the better facility to use.
3501 15. A really picky compiler found some instances of statements for creating
3502 error messages that either had too many or two few arguments for the format
3505 16. The size of the buffer for calls to the DNS resolver has been increased
3506 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3507 for addresses that have a lot of PTR records. This alleviates a problem; it
3508 does not fully solve it.
3510 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3511 buffer now truncates the list and logs the incident, which is the same
3512 action as happens when Exim is looking up a host name and its aliases.
3513 Previously in this situation something unpredictable would happen;
3514 sometimes it was "internal error: store_reset failed".
3516 18. If a server dropped the connection unexpectedly when an Exim client was
3517 using GnuTLS and trying to read a response, the client delivery process
3518 crashed while trying to generate an error log message.
3520 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3521 string, for example:
3523 warn message = H1: something\nH2: something
3525 the text was added as a single header line from Exim's point of view
3526 though it ended up OK in the delivered message. However, searching for the
3527 second and subsequent header lines using $h_h2: did not work. This has been
3528 fixed. Similarly, if a system filter added multiple headers in this way,
3529 the routers could not see them.
3531 20. Expanded the error message when iplsearch is called with an invalid key to
3532 suggest using net-iplsearch in a host list.
3534 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3535 ACLs are no longer actually imposed (and a message to that effect is
3538 22. If a "gecos" field in a passwd entry contained escaped characters, in
3539 particular, if it contained a \" sequence, Exim got it wrong when building
3540 a From: or a Sender: header from that name. A second bug also caused
3541 incorrect handling when an unquoted " was present following a character
3542 that needed quoting.
3544 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3545 was not being matched caselessly.
3547 24. Arranged for all hyphens in the exim.8 source to be escaped with
3550 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3551 properly. Recipient callout cache records were still being keyed to include
3552 the sender, even when use_sender was set false. This led to far more
3553 callouts that were necessary. The sender is no longer included in the key
3554 when use_sender is false.
3556 26. Added "control = submission" modifier to ACLs.
3558 27. Added the ${base62d: operator to decode base 62 numbers.
3560 28. dnsdb lookups can now access SRV records.
3562 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3563 the configuration file.
3565 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3566 (-v) mode. This makes the output for a verbose queue run more intelligible.
3568 31. Added a use_postmaster feature to recipient callouts.
3570 32. Added the $body_zerocount variable, containing the number of binary zero
3571 bytes in the message body.
3573 33. The time of last modification of the "new" subdirectory is now used as the
3574 "mailbox time last read" when there is a quota error for a maildir
3577 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3579 35. Added +ignore_unknown as a special item in host lists.
3581 36. Code for decoding IPv6 addresses in host lists is now included, even if
3582 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3583 address was recognized as an IP address, but was then not correctly decoded
3584 into binary, causing unexpected and incorrect effects when compared with
3591 1. Very minor rewording of debugging text in manualroute to say "list of
3592 hosts" instead of "hostlist".
3594 2. If verify=header_syntax was set, and a header line with an unqualified
3595 address (no domain) and a large number of spaces between the end of the
3596 name and the colon was received, the reception process suffered a buffer
3597 overflow, and (when I tested it) crashed. This was caused by some obsolete
3598 code that should have been removed. The fix is to remove it!
3600 3. When running in the test harness, delay a bit after writing a bounce
3601 message to get a bit more predictability in the log output.
3603 4. Added a call to search_tidyup() just before forking a reception process. In
3604 theory, someone could use a lookup in the expansion of smtp_accept_max_
3605 per_host which, without the tidyup, could leave open a database connection.
3607 5. Added the variables $recipient_data and $sender_data which get set from a
3608 lookup success in an ACL "recipients" or "senders" condition, or a router
3609 "senders" option, similar to $domain_data and $local_part_data.
3611 6. Moved the writing of debug_print from before to after the "senders" test
3614 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3615 problems for message scanning, either using a data ACL, or using
3616 local_scan() because the Received: header was not generated till after they
3617 were called (in order to set the time as the time of reception completion).
3618 I have revised the way this works. The header is now generated after the
3619 body is received, but before the ACL or local_scan() are called. After they
3620 are run, the timestamp in the header is updated.
3626 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3627 before starting a queue runner without re-exec. This happened only when
3628 deliver_drop_privilege was set or when the Exim user was set to root. The
3629 effect of the bug was that timeouts during subsequent deliveries caused
3630 crashes instead of being properly handled. The handler is now left at its
3631 default (and expected) setting.
3633 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3634 message, again when deliver_drop_privilege is set or Exim is run as root.
3635 The bug described in (1) was not present in this case, but the tidying up
3636 of the other signals was missing. I have made the two cases consistent.
3638 3. The ignore_target_hosts setting on a manualroute router was being ignored
3639 for hosts that were looked up using the /MX notation.
3641 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3644 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3645 operated on the sender address. After changing the $sender_address to <>
3646 for the sender address verify, Exim was re-instated it as the original
3647 (before rewriting) address, but remembering that it had rewritten it, so it
3648 wasn't rewriting it again. This bug also had the effect of breaking the
3649 sender address verification caching when the sender address was rewritten.
3651 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3652 This has been changed so that if the ip literal address matches
3653 ignore_target_hosts, the router declines.
3655 7. Added expansion conditions match_domain, match_address, and match_local_
3656 part (NOT match_host).
3658 8. The placeholder for the Received: header didn't have a length field set.
3660 9. Added code to Exim itself and to exim_lock to test for a specific race
3661 condition that could lead to file corruption when using MBX delivery. The
3662 issue is with the lockfile that is created in /tmp. If this file is removed
3663 after a process has opened it but before that process has acquired a lock,
3664 there is the potential for a second process to recreate the file and also
3665 acquire a lock. This could lead to two Exim processes writing to the file
3666 at the same time. The added code performs the same test as UW imapd; it
3667 checks after acquiring the lock that its file descriptor still refers to
3668 the same named file.
3670 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3671 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3674 11. Added the smtp_active_hostname option. If used, this will typically be made
3675 to depend on the incoming interface address. Because $interface_address is
3676 not set up until the daemon has forked a reception process, error responses
3677 that can happen earlier (such as "too many connections") no longer contain
3680 12. If an expansion in a condition on a "warn" statement fails because a lookup
3681 defers, the "warn" statement is abandoned, and the next ACL statement is
3682 processed. Previously this caused the whole ACL to be aborted.
3684 13. Added the iplsearch lookup type.
3686 14. Added ident_timeout as a log selector.
3688 15. Added tls_certificate_verified as a log selector.
3690 16. Added a global option tls_require_ciphers (compare the smtp transport
3691 option of the same name). This controls incoming TLS connections.
3693 17. I finally figured out how to make tls_require_ciphers do a similar thing
3694 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3695 before starting the TLS session.
3697 18. Tabs are now shown as \t in -bP output.
3699 19. If the log selector return_path_on_delivery was set, Exim crashed when
3700 bouncing a message because it had too many Received: header lines.
3702 20. If two routers both had headers_remove settings, and the first one included
3703 a superfluous trailing colon, the final name in the first list and the
3704 first name in the second list were incorrectly joined into one item (with a
3705 colon in the middle).
3711 1. Added -C and -D options to the exinext utility, mainly to make it easier
3712 to include in the automated testing, but these could be helpful when
3713 multiple configurations are in use.
3715 2. The exinext utility was not formatting the output nicely when there was
3716 an alternate port involved in the retry record key, nor when there was a
3717 message id as well (for retries that were specific to a specific message
3718 and a specific host). It was also confused by IPv6 addresses, because of
3719 the additional colons they contain. I have fixed the IPv4 problem, and
3720 patched it up to do a reasonable job for IPv6.
3722 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3723 delivery, the log line now contains "pipelined" if PIPELINING was used.
3725 4. An SMTP transport process used to panic and die if the bind() call to set
3726 an explicit outgoing interface failed. This has been changed; it is now
3727 treated in the same way as a connect() failure.
3729 5. A reference to $sender_host_name in the part of a conditional expansion
3730 that was being skipped was still causing a DNS lookup. This no longer
3733 6. The def: expansion condition was not recognizing references to header lines
3734 that used bh_ and bheader_.
3736 7. Added the _cache feature to named lists.
3738 8. The code for checking quota_filecount in the appendfile transport was
3739 allowing one more file than it should have been.
3741 9. For compatibility with Sendmail, the command line option
3749 and sets the incoming protocol and host name (for trusted callers). The
3750 host name and its colon can be omitted when only the protocol is to be set.
3751 Note the Exim already has two private options, -pd and -ps, that refer to
3752 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3753 "s", but I don't think that's a major issue.
3755 10. A number of refactoring changes to the code, none of which should affect
3758 (a) The number of logging options was getting close to filling up the
3759 32-bit word that was used as a bit map. I have split them into two classes:
3760 those that are passed in the argument to log_write(), and those that are
3761 only ever tested independently outside of that function. These are now in
3762 separate 32-bit words, so there is plenty of room for expansion again.
3763 There is no change in the user interface or the logging behaviour.
3765 (b) When building, for example, log lines, the code previously used a
3766 macro that called string_cat() twice, in order to add two strings. This is
3767 not really sufficiently general. Furthermore, there was one instance where
3768 it was actually wrong because one of the argument was used twice, and in
3769 one call a function was used. (As it happened, calling the function twice
3770 did not affect the overall behaviour.) The macro has been replaced by a
3771 function that can join an arbitrary number of extra strings onto a growing
3774 (c) The code for expansion conditions now uses a table and a binary chop
3775 instead of a serial search (which was left over from when there were very
3776 few conditions). Also, it now recognizes conditions like "pam" even when
3777 the relevant support is not compiled in: a suitably worded error message is
3778 given if an attempt is made to use such a condition.
3780 11. Added ${time_interval:xxxxx}.
3782 12. A bug was causing one of the ddress fields not to be passed back correctly
3783 from remote delivery subprocesses. The field in question was not being
3784 subsequently used, so this caused to problems in practice.
3786 13. Added new log selectors queue_time and deliver_time.
3788 14. Might have fixed a bug in maildirsizefile handling that threw up
3789 "unexpected character" debug warnings, and recalculated the data
3790 unnecessarily. In any case, I expanded the warning message to give more
3793 15. Added the message "Restricted characters in address" to the statements in
3794 the default ACL that block characters like @ and % in local parts.
3796 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3797 Three changes have been made:
3799 (a) There was a serious bug; a negative response to MAIL caused the whole
3800 recipient domain to be cached as invalid, thereby blocking all messages
3801 to all local parts at the same domain, from all senders. This bug has
3802 been fixed. The domain is no longer cached after a negative response to
3803 MAIL if the sender used is not empty.
3805 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3808 (c) A new callout option, "use_sender" has been added for people who want
3809 the modified behaviour.
3815 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3818 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3819 indeed breaks things for older releases.
3821 3. Added additional logging to the case where there is a problem reading data
3822 from a filter that is running in a subprocess using a pipe, in order to
3823 try to track down a specific problem.
3825 4. Testing facility fudge: when running in the test harness and attempting
3826 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3827 getting "No route to host". Convert this to a timeout.
3829 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3832 6. Some OS don't have socklen_t but use size_t instead. This affects the
3833 fifth argument of getsockopt() amongst other things. This is now
3834 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3835 can be set for individual OS. I have set it for SunOS5, OSF1, and
3836 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3837 some earlier ones do not.
3839 7. Change 4.30/15 was not doing the test caselessly.
3841 8. The standard form for an IPv6 address literal was being rejected by address
3842 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3843 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3844 this, as well as the form without the "IPv6" on the front (but only when
3845 address literals are enabled, of course).
3847 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3849 10. Exim crashed if a message with an empty sender address specified by -f
3850 encountered a router with an errors_to setting. This could be provoked only
3851 by a command such as
3855 where an empty string was supplied; "<>" did not hit this bug.
3857 11. Installed PCRE release 4.5.
3859 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3860 remained set. It is now erased.
3862 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3863 times from message ids (which are base 36 rather than the normal 62).
3865 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3866 were being counted as actual protocol errors, and logged if the log
3867 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3868 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3869 and DATA following a set of rejected RCPTs do not count as protocol errors.
3870 In other words, Exim assumes they were pipelined, though this may not
3871 actually be the case. Of course, in all cases the client gets an
3872 appropriate error code.
3874 15. If a lookup fails in an ACL condition, a message about the failure may
3875 be available; it is used if testing the ACL cannot continue, because most
3876 such messages specify what the cause of the deferral is. However, some
3877 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3878 that caused an old message to be retained and used if a later statement
3879 caused a defer, replacing the real cause of the deferral.
3881 16. If an IP address had so many PTR records that the DNS lookup buffer
3882 was not large enough to hold them, Exim could crash while trying to process
3883 the truncated data. It now detects and logs this case.
3885 17. Further to 4.21/58, another change has been made: if (and only if) the
3886 first line of a message (the first header line) ends with CRLF, a bare LF
3887 in a subsequent header line has a space inserted after it, so as not to
3888 terminate the header.
3890 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3891 unnecessarily, used atoi() instead of strtol(), and didn't check the
3892 termination when getting file sizes from file names by regex.
3894 19. Completely re-implemented the support for maildirsize files, in the light
3895 of a number of problems with the previous contributed implementation
3896 (4.30/29). In particular:
3898 . If the quota is zero, the maildirsize file is maintained, but no quota is
3901 . If the maildir directory does not exist, it is created before any attempt
3902 to write a maildirsize file.
3904 . The quota value in the file is just a cache; if the quota is changed in
3905 the transport, the new value overrides.
3907 . A regular expression is available for excluding directories from the
3910 20. The autoreply transport checks the characters in options that define the
3911 message's headers; it allows continued headers, but it was checking with
3912 isspace() after an embedded newline instead of explicitly looking for a
3915 21. If all the "regular" hosts to which an address was routed had passed their
3916 expiry times, and had not reached their retry times, the address was
3917 bounced, even if fallback hosts were defined. Now Exim should go on to try
3920 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3921 equivalent code in the SMTP transport. Some hosts send humungous responses
3922 to HELO/EHLO, more than 1024 it seems.
3924 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3925 gives compiler warnings in some environments. I've now done it "properly",
3928 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3929 (because of problems with the built-in one) was declared to return uschar *
3930 instead of char *, causing compiler failure.
3932 25. Fixed a file descriptor leak when processing alias/forward files.
3934 26. Fixed a minor format string issue in dbfn.c.
3936 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3938 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3939 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3942 29. When a sender address is verified, it is cached, to save repeating the test
3943 when there is more than one recipient in a message. However, when the
3944 verification involves a callout, it is possible for different callout
3945 options to be set for different recipients. It is too complicated to keep
3946 track of this in the cache, so now Exim always runs a verification when a
3947 callout is required, relying on the callout cache for the optimization.
3948 The overhead is duplication of the address routing, but this should not be
3951 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3952 to be verified with callout=postmaster, and the main callout worked but the
3953 postmaster check failed, the verification correctly failed. However, if a
3954 subsequent RCPT command asked for sender verification *without* the
3955 postmaster check, incorrect caching caused this verification also to fail,
3958 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3959 it was not caching the DNS options (qualify_single, search_parents) that
3960 were used when the lookup failed. A subsequent lookup with different
3961 options therefore always gave the same answer, though there were cases
3962 where it should not have. (Example: a "domains = !$mx_any" option on a
3963 dnslookup router: the "domains" option is always processed without any
3964 widening, but the router might have qualify_single set.) Now Exim uses the
3965 cached value only when the same options are set.
3967 32. Added John Jetmore's "exipick" utility to the distribution.
3969 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3970 than a timeout (e.g. a certificate is required, and is not provided), an
3971 Exim server now closes the connection immediately. Previously it waited for
3972 the client to close - but if the client is SSL, it seems that they each
3973 wait for each other, leading to a delay before one of them times out.
3975 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3976 maintained 0.8.x compatibility because I don't think many are using it, and
3977 it is clearly obsolete.
3979 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3982 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3983 client certificate was expired. A simple patch fixes this, though I don't
3984 understand the full logic of why the verify callback is called multiple
3987 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3988 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3989 which causes problems with some clients (such as the Certicom SSL Plus
3990 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3991 disables the coutermeasure allowing Eudora to connect."
3993 38. Exim was not checking that a write() to a log file succeeded. This could
3994 lead to Bad Things if a log got too big, in particular if it hit a file
3995 size limit. Exim now panics and dies if it cannot write to a log file, just
3996 as it does if it cannot open a log file.
3998 39. Modified OS/Makefile-Linux so that it now contains
4000 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4002 The two -D definitions ensure that Exim is compiled with large file
4003 support, which makes it possible to handle log files that are bigger than
4006 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4007 instance) a domain was checked against a named list that involved a lookup,
4008 causing $domain_data to be set, then another domain was checked against the
4009 same list, then the first domain was re-checked, the value of $domain_data
4010 after the final check could be wrong. In particular, if the second check
4011 failed, it could be set empty. This bug probably also applied to
4014 41. The strip_trailing_dot option was not being applied to the address given
4015 with the -f command-line option.
4017 42. The code for reading a message's header from the spool was incrementing
4018 $received_count, but never initializing it. This meant that the value was
4019 incorrect (doubled) while delivering a message in the same process in which
4020 it was received. In the most common configuration of Exim, this never
4021 happens - a fresh exec is done - but it can happen when
4022 deliver_drop_privilege is set.
4024 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4025 it now includes up to 150 characters of the unexpected data in the log
4028 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4029 and building data strings. The size of both of these buffers was 10 000
4030 bytes - far larger than anybody would *ever* want, thought I. Needless to
4031 say, somebody hit the limit. I have increased the maximum line length to
4032 20 000 and the maximum data length of concatenated lines to 100 000. I have
4033 also fixed two bugs, because there was no checking on these buffers. Tsk,
4034 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4035 buffer is too small.
4037 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4038 lsearch lookups. Now it does.
4040 46. When parsing a route_list item in a manualroute router, a fixed-length
4041 buffer was used for the list of hosts. I made this 1024 bytes long,
4042 thinking that nobody would ever have a list of hosts that long. Wrong.
4043 Somebody had a whole pile of complicated expansion conditions, and the
4044 string was silently truncated, leading to an expansion error. It turns out
4045 that it is easier to change to an unlimited length (owing to other changes
4046 that have happened since this code was originally written) than to build
4047 structure for giving a limitation error. The length of the item that
4048 expands into the list of hosts is now unlimited.
4050 47. The lsearch lookup could not handle data where the length of text line was
4051 more than 4095 characters. Such lines were truncated, leading to shortened
4052 data being returned. It should now handle lines of any length.
4054 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4055 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4058 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4059 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4060 no longer happens in comment lines. A list of replacements is now placed
4061 at the head of all of the source files, except those whose only change is
4062 to replace PERL_COMMAND in the very first #! line.
4064 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4065 messages on the queue, with a bottom-up merge sort, using code contributed
4066 by Michael Haardt. This should make operations like -bp somewhat faster on
4067 large queues. It won't affect queue runners, except when queue_run_in_order
4070 51. Installed eximstats 1.31 in the distribution.
4072 52. Added support for SRV lookups to the dnslookup router.
4074 53. If an ACL referred to $message_body or $message_body_end, the value was not
4075 reset for any messages that followed in the same SMTP session.
4077 54. The store-handling optimization for building very long strings was not
4078 differentiating between the different store pools. I don't think this
4079 actually made any difference in practice, but I've tidied it.
4081 55. While running the routers to verify a sender address, $sender_address
4082 was still set to the sender address. This is wrong, because when routing to
4083 send a bounce to the sender, it would be empty. Therefore, I have changed
4084 it so that, while verifying a sender address, $sender_address is set to <>.
4085 (There is no change to what happens when verifying a recipient address.)
4087 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4088 target A or AAAA records (if not already returned) without resetting the
4089 qualify_single or search_parents options of the DNS resolver. These are
4090 inappropriate in this case because the targets of MX and SRV records must
4091 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4092 target that, when qualified, matched something in the local domain. These
4093 two options are now turned off when doing these lookups.
4095 57. It seems that at least some releases of Reiserfs (which does not have the
4096 concept of a fixed number of inodes) returns zero and not -1 for the
4097 number of available inodes. This interacted badly with check_spool_inodes,
4098 which assumed that -1 was the "no such thing" setting. What I have done is
4099 to check that the total number of inodes is greater than zero before doing
4100 the test of how many are available.
4102 58. When a "warn" ACL statement has a log_message modifier, the message is
4103 remembered, and not repeated. This is to avoid a lot of repetition when a
4104 message has many recipients that cause the same warning to be written.
4105 Howewer, Exim was preserving the list of already written lines for an
4106 entire SMTP session, which doesn't seem right. The memory is now reset if a
4107 new message is started.
4109 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4110 debugging output unless log_rewrite was also set.
4112 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4113 of (int)(handle) when we know that handle contains (void *)(-1).
4115 61. The Exim daemon panic-logs an error return when it closes the incoming
4116 connection. However "connection reset by peer" seems to be common, and
4117 isn't really an error worthy of noting specially, so that particular error
4120 62. When Exim is trying to find all the local interfaces, it used to panic and
4121 die if the ioctl to get the interface flags failed. However, it seems that
4122 on at least one OS (Solaris 9) it is possible to have an interface that is
4123 included in the list of interfaces, but for which you get a failure error
4124 for this call. This happens when the interface is not "plumbed" into a
4125 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4126 failure of the "get flags" call assumes that the interface is down.
4128 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4129 makes life easier for people who are doing arithmetic on fields extracted
4130 from dates, where you often get leading zeros that should not be
4131 interpreted as octal.
4133 64. Added qualify_domain to the redirect router, to override the global
4136 65. If a pathologically long header line contained very many addresses (the
4137 report of this problem mentioned 10 000) and each of them was rewritten,
4138 Exim could use up a very large amount of memory. (It kept on making new
4139 copies of the header line as it rewrote, and never released the old ones.)
4140 At the expense of a bit more processing, the header rewriting function has
4141 been changed so that it no longer eats memory in this way.
4143 66. The generation of the Received: header has been moved from the time that a
4144 message starts to be received, to the time that it finishes. The timestamp
4145 in the Received: header should now be very close to that of the <= log
4146 line. There are two side-effects of this change:
4148 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4149 logged header lines no longer include the local Received: line, because
4150 it has not yet been created. The same applies to a copy of the message
4151 that is returned to a non-SMTP sender when a message is rejected.
4153 (b) When a filter file is tested using -bf, no additional Received: header
4154 is added to the test message. After some thought, I decided that this
4157 This change does not affect the value of $received_for. It is still set
4158 after address rewriting, but before local_scan() is called.
4160 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4162 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4163 gave an unhelpful panic error message, and a defer error. I have managed to
4164 change this behaviour so that it now rejects any supplied certificate,
4165 which seems right, as the list of acceptable certificates is empty.
4167 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4168 gave an unhelpful defer error. I have not managed to make this reject any
4169 supplied certificates, but the error message it gives is "no certificate
4170 supplied", which is not helpful.
4172 70. exigrep's output now also includes lines that are not associated with any
4173 message, but which match the given pattern. Implemented by a patch from
4174 Martin Sluka, which also tidied up the Perl a bit.
4176 71. Recipient callout verification, like sender verification, was using <> in
4177 the MAIL FROM command. This isn't really the right thing, since the actual
4178 sender may affect whether the remote host accepts the recipient or not. I
4179 have changed it to use the actual sender in the callout; this means that
4180 the cache record is now keyed on a recipient/sender pair, not just the
4181 recipient address. There doesn't seem to be a real danger of callout loops,
4182 since a callout by the remote host to check the sender would use <>.
4183 [SEE ABOVE: changed after hitting problems.]
4185 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4186 temporary errors. However, in the case of such a code being given after
4187 the end of a data transmission (i.e. after ".") Exim was failing to write
4188 a retry record for the message. (Yes, there was some broken host that was
4189 actually sending 8xx at this point.)
4191 73. An unknown lookup type in a host list could cause Exim to panic-die when
4192 the list was checked. (An example that provoked this was putting <; in the
4193 middle of a list instead of at the start.) If this happened during a DATA
4194 ACL check, a -D file could be left lying around. This kind of configuration
4195 error no longer causes Exim to die; instead it causes a defer errror. The
4196 incident is still logged to the main and panic logs.
4198 74. Buglet left over from Exim 3 conversion. The message "too many messages
4199 in one connection" was written to the rejectlog but not the mainlog, except
4200 when address rewriting (yes!) was being logged.
4202 75. Added write_rejectlog option.
4204 76. When a system filter was run not as root (that is, when system_filter_user
4205 was set), the values of the $n variables were not being returned to the
4206 main process; thus, they were not subsequently available in the $sn
4209 77. Added +return_path_on_delivery log selector.
4211 78. A connection timeout was being treated differently from recipients deferred
4212 when testing hosts_max_try with a message that was older than the host's
4213 retry timeout. (The host should not be counted, thus allowing all hosts to
4214 be tried at least once before bouncing.) This may have been the cause of an
4215 occasionally reported bug whereby a message would remain on the queue
4216 longer than the retry timeout, but would be bounced if a delivery was
4217 forced. I say "may" because I never totally pinned down the problem;
4218 setting up timeout/retry tests is difficult. See also the next item.
4220 79. The ultimate address timeout was not being applied to errors that involved
4221 a combination of host plus message (for example, a timeout on a MAIL
4222 command). When an address resolved to a number of possible hosts, and they
4223 were not all tried for each delivery (e.g. because of hosts_max_try), a
4224 message could remain on the queue longer than the retry timeout.
4226 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4229 81. Fixed an obscure SMTP outgoing bug which required at least the following
4230 conditions: (a) there was another message waiting for the same server;
4231 (b) the server returned 5xx to all RCPT commands in the first message so
4232 that the message was not completed; (c) the server dropped the connection
4233 or gave a negative response to the RSET that Exim sends to abort the
4234 transaction. The observed case was a dropped connection after DATA that had
4235 been sent in pipelining mode. That is, the server had advertised PIPELINING
4236 but was not implementing it correctly. The effect of the bug was incorrect
4237 behaviour, such as trying another host, and this could lead to a crash.
4243 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4244 and daemon.c were passed as pointers to ints; they should have been
4245 pointers to socklen_t variables (which are typically unsigned ints).
4247 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4250 3. Fixed a really odd bug that affected only the testing scheme; patching a
4251 certain fixed string in the binary changed the value of another string that
4252 happened to be identical to the end of the original first string.
4254 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4255 name", it returns that address as the IP address. On some operating
4256 systems (e.g. Solaris), it also passes back the IP address string as the
4257 "host name". However, on others (e.g. Linux), it passes back an empty
4258 string. Exim wasn't checking for this, and was changing the host name to an
4259 empty string, assuming it had been canonicized.
4261 5. Although rare, it is permitted to have more than one PTR record for a given
4262 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4263 all the names associated with an address, because they do in Solaris.
4264 However, it seems that they do not in Linux for data that comes from the
4265 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4266 I found this out when I moved to a new Linux workstation and tried to run
4267 the Exim test suite.
4269 To get round this problem I have changed the code so that it now does its
4270 own call to the DNS to look up PTR records when searching for a host name.
4271 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4272 addresses that are only in /etc/hosts are still found.
4274 This behaviour is, however, controlled by an option called host_lookup_
4275 order, which defaults to "bydns:byaddr". If people want to use the other
4276 order, or indeed, just use one or the other means of lookup, they can
4277 specify it in this variable.
4279 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4280 some operating systems, this comes back from gethostbyaddr() as an empty
4281 string, and this is what Exim used to test for. However, it seems that in
4282 other systems, "." is yielded. Exim now tests for this case too.
4284 7. The values of check_spool_space and check_log_space are now held internally
4285 as a number of kilobytes instead of an absolute number of bytes. If a
4286 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4287 kilobyte. This means that much larger values can be stored.
4289 8. Exim monitor: an attempt to get the action menu when not actually pointing
4290 at a message produces an empty menu entitled "No message selected". This
4291 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4292 no entries in it ("Shell widget menu has zero width and/or height"). So I
4293 have added a single, blank menu entry in this case.
4295 9. Added ${quote_local_part.
4297 10. MIME decoding is now applied to the contents of Subject: header lines when
4300 11. Now that a reference to $sender_host_address automatically causes a reverse
4301 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4302 host lookup before query-style lookups in lists that might use this
4303 variable. This has therefore been abolished, and the "net-" prefix is no
4304 longer necessary for query-style lookups.
4306 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4307 have been a typo for LFLAGS, so it has been changed.
4309 13. The install script calls Exim with "-C /dev/null" in order to find the
4310 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4311 to be output. Howeve, since Exim outputs its version number before the
4312 error, it didn't break the script. It just looked ugly. I fixed this by
4313 always allowing "-C /dev/null" if the caller is root.
4315 14. Ignore overlarge ACL variable number when reading spool file - insurance
4316 against a later release with more variables having written the file.
4318 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4319 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4320 this, as well as the form without the "IPv6" on the front.
4322 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4323 OS/Makefile-Darwin file.
4325 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4326 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4328 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4329 variable. However, some OS use ints for inodes. I've added cast to long int
4330 to get rid of the compiler warning.
4332 19. I had forgotten to lock out "/../" in configuration file names when
4333 ALT_CONFIG_PREFIX was set.
4335 20. Routers used for verification do not need to specify transports. However,
4336 if such a router generated a host list, and callout was configured, Exim
4337 crashed, because it could not find a port number from the (non-existent)
4338 transport. It now assumes port 25 in this circumstance.
4340 21. Added the -t option to exigrep.
4342 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4343 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4344 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4345 them is set, LOOKUP_LSEARCH is forced.
4347 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4349 24. Added sender and host information to the "rejected by local_scan()" log
4350 line; previously there was no indication of these.
4352 25. Added .include_if_exists.
4354 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4355 Linux. It turns out that not all file systems support this. Apparently some
4356 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4357 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4358 not supported on the file descriptor, is now ignored when Exim is trying to
4359 sync a directory. This applies only to Linux.
4361 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4363 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4364 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4367 29. Added support for maildirsize files from supplied patch (modified a bit).
4369 30. The use of :fail: followed by an empty string could lead Exim to respond to
4370 sender verification failures with (e.g.):
4372 550 Verification failed for <xxx>
4373 550 Sender verify failed
4375 where the first response line was missing the '-' that indicates it is not
4376 the final line of the response.
4378 31. The loop for finding the name of the user that called Exim had a hardwired
4379 limit of 10; it now uses the value of finduser_retries, which is used for
4380 all other user lookups.
4382 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4385 33. Exim was neglecting to zero errno before one call of strtol() when
4386 expanding a string and expecting an integer value. On some systems this
4387 resulted in spurious "integer overflow" errors. Also, it was casting the
4388 result into an int without checking.
4390 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4391 did not work. The code looks as if it has *never* worked, though it appears
4392 to have been documented since at least releast 1.62. I have made it work.
4394 35. The "timeout_DNS" error in retry rules, also documented since at least
4395 1.62, also never worked. As it isn't clear exactly what this means, and
4396 clearly it isn't a major issue, I have abolished the feature by treating it
4397 as "timeout", and writing a warning to the main and panic logs.
4399 36. The display of retry rules for -brt wasn't always showing the error code
4402 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4403 timeout_connect_A, timeout_connect_MX.
4405 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4406 to the empty sender.
4408 39. The daemon was not analysing the content of -oX till after it had closed
4409 stderr and disconnected from the controlling terminal. This meant that any
4410 syntax errors were only noted on the panic log, and the return code from
4411 the command was 0. By re-arranging the code a little, I've made the
4412 decoding happen first, so such errors now appear on stderr, and the return
4413 code is 1. However, the actual setting up of the sockets still happens in
4414 the disconnected process, so errors there are still only recorded on the
4417 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4418 connections (as happens on some IP stacks) was logged at start up time as
4419 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4420 it from "IPv6 and IPv4", which means that two separate sockets are being
4423 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4424 says whether AF_INET or AF_INET6 was passed as an argument.
4426 42. Exiwhat output was messed up when time zones were included in log
4429 43. Exiwhat now gives more information about the daemon's listening ports,
4430 and whether -tls-on-connect was used.
4432 44. The "port" option of the smtp transport is now expanded.
4434 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4435 silently ignored. Now an error message is written to the main and panic
4438 46. There's a new ACL modifier called "logwrite" which writes to a log file
4439 as soon as it is encountered.
4441 47. Added $local_user_uid and $local_user_gid at routing time.
4443 48. Exim crashed when trying to verify a sender address that was being
4446 49. Exim was recognizing only a space character after ".include". It now also
4447 recognizes a tab character.
4449 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4450 extracting the relevant information from the specification. The man page no
4451 longer contains scrambled data for the -d option, and I've added a section
4452 at the front about calling Exim under different names.
4454 51. Added "extra_headers" argument to the "mail" command in filter files.
4456 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4459 53. Installed eximstats 1.29.
4461 54. Added transport_filter_timeout as a generic transport option.
4463 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4464 Cc: header lines. This was required by RFC 822, but it not required by RFC
4467 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4468 incoming messages that did not have them. Now it does so only if the
4469 message originates locally, that is, if there is no associated remote host
4470 address. When Resent- header lines are present, this applies to the Resent-
4471 lines rather than the non-Resent- lines.
4473 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4474 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4476 58. Messages for configuration errors now include the name of the main
4477 configuration file - useful now that there may be more than one file in a
4478 list (.included file names were always shown).
4480 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4481 for those rare installations that do not start the daemon as root or run it
4482 setuid root. I've cut out the call to initgroups() if the daemon is not
4485 60. The Exim user and group can now be bound into the binary as text strings
4486 that are looked up at the start of Exim's processing.
4488 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4490 62. Added $mailstore_basename variable.
4492 63. Installed patch to sieve.c from Michael Haardt.
4494 64. When Exim failed to open the panic log after failing to open the main log,
4495 the original message it was trying to log was written to stderr and debug
4496 output, but if they were not available (the usual case in production), it
4497 was lost. Now it is written to syslog before the two lines that record the
4498 failures to open the logs.
4500 65. Users' Exim filters run in subprocesses under the user's uid. It is
4501 possible for a "deliver" command or an alias in a "personal" command to
4502 provoke an address rewrite. If logging of address rewriting is configured,
4503 this fails because the process is not running as root or exim. There may be
4504 a better way of dealing with this, but for the moment (because 4.30 needs
4505 to be released), I have disabled address rewrite logging when running a
4506 filter in a non-root, non-exim process.
4512 1. The buildconfig auxiliary program wasn't quoting the value set for
4513 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4514 not defined. This bug was masked in 4.22 by the effect that was fixed in
4517 2. Some messages that were rejected after a message id was allocated were
4518 shown as "incomplete" by exigrep. It no longer does this for messages that
4519 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4521 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4522 have allow_domain_literals set, the ID did not get logged in the <= line.
4523 Domain literals are now always recognized in Message-ID: header lines.
4525 4. The first argument for a ${extract expansion item is the key name or field
4526 number. Leading and trailing spaces in this item were not being ignored,
4527 causing some misleading effects.
4529 5. When deliver_drop_privilege was set, single queue runner processes started
4530 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4531 same command in the process it spins off) were not dropping privilege.
4533 6. When the daemon running as "exim" started a queue runner, it always
4534 re-executed Exim in the spun-off process. This is a waste of effort when
4535 deliver_drop_privilege is set. The new process now just calls the
4536 queue-runner function directly.
4542 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4545 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4546 ignored. Though the use of -oP was forcing the writing of a pid file, it
4547 was always written to the default place.
4549 3. If the message "no IP address found for host xxxx" is generated during
4550 incoming verification, it is now followed by identification of the incoming
4551 connection (so you can more easily find what provoked it).
4553 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4555 5. Added some features to "harden" Exim a bit more against certain attacks:
4557 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4558 be put in Local/Makefile. This is like the never_users runtime option,
4559 but it cannot be overridden. The default setting is "root".
4561 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4562 prefix string with which any file named in a -C command line option
4565 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4566 is retained for -C and -D only if the caller of Exim is root. Without
4567 it, the exim user may also use -C and -D and retain privilege.
4569 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4570 command line option is disabled.
4572 6. Macro names set by the -D option must start with an upper case letter, just
4573 like macro names defined in the configuration file.
4575 7. Added "dereference=" facility to LDAP.
4577 8. Two instances of the typo "uknown" in the source files are fixed.
4579 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4580 the Configure-Makefile script screwed up while processing it.
4582 10. Incorporated PCRE 4.4.
4584 11. The SMTP synchronization check was not operating right at the start of an
4585 SMTP session. For example, it could not catch a HELO sent before the client
4586 waited for the greeting. There is now a check for outstanding input at the
4587 point when the greeting is written. Because of the duplex, asynchronous
4588 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4589 way, but not yet received, when the check is performed.
4591 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4592 on TCP/IP sockets, because this apparently causes some broken clients to
4595 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4596 unchanged) from the Cygwin maintainer.
4598 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4599 was supported instead of testing for mbx. Effectively a typo.
4601 15. The spa authenticator server code was not checking that the input it
4602 received was valid base64.
4604 16. The debug output line for the "set" modifier in ACLs was not showing the
4605 name of the variable that was being set.
4607 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4609 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4610 lookup on its own. Something else was needed to trigger the lookup. For
4611 example, a match in host_lookup or the need for a host name in a host list.
4612 Now, if $sender_host_name is referenced and the host name has not yet been
4613 looked up, a lookup is performed. If the lookup fails, the variable remains
4614 empty, and $host_lookup_failed is set to "1".
4616 19. Added "eqi" as a case-independent comparison operator.
4618 20. The saslauthd authentication condition could segfault if neither service
4619 nor realm was specified.
4621 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4622 error message that was logged was misleading, and incoming SMTP
4623 connections were dropped. The message is now more accurate, and temporary
4624 errors are given to SMTP connections.
4626 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4627 (except RSET) with a 421 error, until QUIT is received. However, it was
4628 failing to send a response to QUIT.
4630 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4631 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4632 ACL is now run after the helo_try_verify_hosts code.
4634 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4635 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4636 case-independent, so other case variants are also recognized). Apparently
4637 some systems use these upper case variants.
4639 25. If more than two messages were waiting for the same host, and a transport
4640 filter was specified for the transport, Exim sent two messages over the
4641 same TCP/IP connection, and then failed with "socket operation on non-
4642 socket" when it tried to send the third.
4644 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4646 27. The extern definition of crypt16() in expand.c was not being excluded when
4647 the OS had its own crypt16() function.
4649 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4650 as a preferred synonym for return_size_limit, both as an option and as an
4653 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4655 30. Changed the default configuration ACL to relax the local part checking rule
4656 for addresses that are not in any local domains. For these addresses,
4657 slashes and pipe symbols are allowed within local parts, but the sequence
4658 /../ is explicitly forbidden.
4660 31. SPA server authentication was not clearing the challenge buffer before
4663 32. log_message in a "warn" ACL statement was writing to the reject log as
4664 well as to the main log, which contradicts the documentation and doesn't
4665 seem right (because no rejection is happening). So I have stopped it.
4667 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4668 However, I am unable to do any testing of this.
4670 34. Fixed an infelicity in the appendfile transport. When checking directories
4671 for a mailbox, to see if any needed to be created, it was accidentally
4672 using path names with one or more superfluous leading slashes; tracing
4673 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4675 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4676 contents are added to the log line that is written for every discarded
4677 recipient. (Previously a log_message setting was ignored.)
4679 36. The ${quote: operator now quotes the string if it is empty.
4681 37. The install script runs exim in order to find its version number. If for
4682 some reason other than non-existence or emptiness, which it checks, it
4683 could not run './exim', it was installing it with an empty version number,
4684 i.e. as "exim-". This error state is now caught, and the installation is
4687 38. An argument was missing from the function that creates an error message
4688 when Exim fails to connect to the socket for saslauthd authentication.
4689 This could cause Exim to crash, or give a corrupted message.
4691 39. Added isip, isip4, and isip6 to ${if conditions.
4693 40. The ACL variables $acl_xx are now saved with the message, and can be
4694 accessed later in routers, transports, and filters.
4696 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4697 strings in the file are not string-expanded.
4699 42. If a MAIL command specified a SIZE value that was too large to fit into an
4700 int variable, the check against message_size_limit failed. Such values are
4701 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4702 day this will have to be increased, but I don't think I want to be around
4703 when emails are that large.
4710 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4711 iconv() is not standard in FreeBSD.
4713 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4714 IPv6 enabled. The observed symptom was a segmentation fault on return from
4715 the function os_common_find_running_interfaces() in src/os.c.
4717 3. In the check_special_case() function in daemon.c I had used "errno" as an
4718 argument name, which causes warnings on some systems. This was basically a
4719 typo, since it was named "eno" in the comments!
4721 4. The code that waits for the clock to tick (at a resolution of some fraction
4722 of a second) so as to ensure message-id uniqueness was always waiting for
4723 at least one whole tick, when it could have waited for less. [This is
4724 almost certainly not relevant at current processor speeds, where it is
4725 unlikely to ever wait at all. But we try to future-proof.]
4727 5. The function that sleeps for a time interval that includes fractions of a
4728 second contained a race. It did not block SIGALRM between setting the
4729 timer, and suspending (a couple of lines later). If the interval was short
4730 and the sigsuspend() was delayed until after it had expired, the suspension
4731 never ended. On busy systems this could lead to processes getting stuck for
4734 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4735 process, before it forks any delivery processes. The open lookup caching
4736 mechanism meant that the open file or database connection was passed into
4737 the delivery process. The problem was that delivery processes always tidy
4738 up cached lookup data. This could cause a problem for the next delivery
4739 process started by the queue runner, because the external queue runner
4740 process does not know about the closure. So the next delivery process
4741 still has data in the lookup cache. In the case of a file lookup, there was
4742 no problem because closing a file descriptor in a subprocess doesn't affect
4743 the parent. However, if the lookup was caching a connection to a database,
4744 the connection was closed, and the second delivery process was likely to
4745 see errors such as "PGSQL: query failed: server closed the connection
4746 unexpectedly". The problem has been fixed by closing all cached lookups
4747 in a queue runner before running a delivery process.
4749 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4750 seem to have the "const" qualifier which it has on other OS. I've
4753 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4754 *of the same type* (client or server) with the same public name that an
4755 error should be diagnosed.
4757 9. When Exim looked up a host name for an IP address, but failed to find the
4758 original IP address when looking up the host name (a safety check), it
4759 output the message "<ip address> does not match any IP for NULL", which was
4760 confusing, to say the least. The bug was that the host name should have
4761 appeared instead of "NULL".
4763 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4764 user that is built into the binary, and the -C or -D options is used, root
4765 privilege is dropped before the configuration file is read. In addition,
4766 logging is switched to stderr instead of the normal log files. If the
4767 configuration then re-defines the Exim user, the unprivileged environment
4768 is probably not what is expected, so Exim logs a panic warning message (but
4771 However, if deliver_drop_privilege is set, the unprivileged state may well
4772 be exactly what is intended, so the warning has been cut out in that case,
4773 and Exim is allowed to try to write to its normal log files.
4779 1. smtp_return_error_details was not giving details for temporary sender
4780 or receiver verification errors.
4782 2. Diagnose a configuration error if two authenticators have the same public
4785 3. Exim used not to create the message log file for a message until the first
4786 delivery attempt. This could be confusing when incoming messages were held
4787 for policy or load reasons. The message log file is now created at the time
4788 the message is received, and an initial "Received" line is written to it.
4790 4. The automatically generated man page for command line options had a minor
4791 bug that caused no ill effects; however, a more serious problem was that
4792 the procedure for building the man page automatically didn't always
4793 operate. Consequently, release 4.20 contains an out-of-date version. This
4794 shouldn't happen again.
4796 5. When building Exim with embedded Perl support, the script that builds the
4797 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4798 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4800 6. The freeze_tell option was not being used for messages that were frozen on
4801 arrival, either by an ACL or by local_scan().
4803 7. Added the smtp_incomplete_transaction log selector.
4805 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4806 was accepting AUTH without a new EHLO.
4808 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4809 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4812 10. Logging of TCP/IP connections (when configured) now happens in the main
4813 daemon process instead of the child process, so that the TCP/IP connection
4814 count is more accurate (but it can never be perfect).
4816 11. The use of "drop" in a nested ACL was not being handled correctly in the
4817 outer ACL. Now, if condition failure induced by the nested "drop" causes
4818 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4819 or "require"), the connection is dropped.
4821 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4822 that yield "discard" can now be used with an "accept" or a "discard" verb,
4823 but an error is generated for any others (because I can't see a useful way
4824 to define what should happen).
4826 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4827 are now processed in the same way as lines in the Exim configuration file.
4828 In particular, continuation lines are supported.
4830 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4832 15. Added -ti meaning -t -i.
4834 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4835 domains, and warn by logging if others are found.
4837 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4838 returned from GIFCONF when Exim is trying to find the list of interfaces on
4839 a host. The code in os.c has been modified to copy each ifreq to an aligned
4840 structure in all cases.
4842 Also, in some cases, the returned ifreq's were being copied to a 'struct
4843 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4844 means the last couple of bytes of an IPv6 address could be chopped if the
4845 ifreq contained only a normal sockaddr (14 bytes storage).
4847 18. Named domain lists were not supported in the hosts_treat_as_local option.
4848 An entry such as +xxxx was not recognized, and was treated as a literal
4851 19. Ensure that header lines added by a DATA ACL are included in the reject log
4852 if the ACL subsequently rejects the message.
4854 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4855 MD5 (which is deprecated).
4857 21. When testing a filter file using -bf, Exim was writing a message when it
4858 took the sender from a "From " line in the message, but it was not doing so
4859 when it took $return_path from a Return-Path: header line. It now does.
4861 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4862 with a valid header line field name (a series of printing characters
4863 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4865 23. Changed "disc" in the source to "disk" to conform to the documentation and
4866 the book and for uniformity.
4868 24. Ignore Sendmail's -Ooption=value command line item.
4870 25. When execve() failed while trying to run a command in a pipe transport,
4871 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4872 could be confused with a return value of 69 from the command itself. This
4873 has been changed to 127, the value the shell returns if it is asked to run
4874 a non-existent command. The wording for the related log line suggests a
4875 non-existent command as the problem.
4877 26. If received_header_text expands to an empty string, do not add a Received:
4878 header line to the message. (Well, it adds a token one on the spool, but
4879 marks it "old" so that it doesn't get used or transmitted.)
4881 27. Installed eximstats 1.28 (addition of -nt option).
4883 28. There was no check for failure on the call to getsockname() in the daemon
4884 code. This can fail if there is a shortage of resources on the system, with
4885 ENOMEM, for example. A temporary error is now given on failure.
4887 29. Contrary to the C standard, it seems that in some environments, the
4888 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4889 program. Exim now does this explicitly; it affects the formatting of
4890 timestamps using strftime().
4892 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4893 complaints. I've now initialized all the variables, to avoid this.
4895 32. Header lines added by a system filter were not being "seen" during
4896 transport-time rewrites.
4898 33. The info_callback() function passed to OpenSSL is set up with type void
4899 (*)(SSL *, int, int), as described somewhere. However, when calling the
4900 function (actually a macro) that sets it up, the type void(*)() is
4901 expected. I've put in a cast to prevent warnings from picky compilers.
4903 34. If a DNS black list lookup found a CNAME record, but there were no A
4904 records associated with the domain it pointed at, Exim crashed.
4906 35. If a DNS black list lookup returned more than one A record, Exim ignored
4907 all but the first. It now scans all returned addresses if a particular IP
4908 value is being sought. In this situation, the contents of the
4909 $dnslist_value variable are a list of all the addresses, separated by a
4912 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4913 to accept a host name and all its aliases if the forward lookup for any of
4914 them yielded the IP address of the incoming connection. Now it accepts only
4915 those names whose forward lookup yields the correct IP address. Any other
4916 names are discarded. This closes a loophole whereby a rogue DNS
4917 administrator could create reverse DNS records to break through a
4918 wildcarded host restriction in an ACL.
4920 37. If a user filter or a system filter that ran in a subprocess used any of
4921 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4922 the wrong values were passed to the pipe command ($thisaddress had the
4923 value of $0, $0 had the value of $1, etc). This bug was introduced by
4924 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4926 38. Improved the line breaking for long SMTP error messages from ACLs.
4927 Previously, if there was no break point between 40 and 75 characters, Exim
4928 left the rest of the message alone. Two changes have been made: (a) I've
4929 reduced the minimum length to 35 characters; (b) if it can't find a break
4930 point between 35 and 75 characters, it looks ahead and uses the first one
4931 that it finds. This may give the occasional overlong line, but at least the
4932 remaining text gets split now.
4934 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4935 file descriptors might be low, and that setting 1000 would always raise it.
4936 It turns out that in some environments, the limit is already over 1000 and
4937 that lowering it causes trouble. So now Exim takes care not to decrease it.
4939 40. When delivering a message, the value of $return_path is set to $sender_
4940 address at the start of routing (routers may change the value). By an
4941 oversight, this default was not being set up when an address was tested by
4942 -bt or -bv, which affected the outcome if any router or filter referred to
4945 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4946 log only when "message" or "log_message" are set. However, if one of the
4947 conditions was an address verification, or a call to a nested ACL, the
4948 messages generated by the underlying test were being passed through. This
4949 no longer happens. The underlying message is available in $acl_verify_
4950 message for both "message" and "log_message" expansions, so it can be
4951 passed through if needed.
4953 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4954 new expansion $bh_ to give the encoded byte string without charset
4955 translation. Translation happens only if iconv() is available; HAVE_ICONV
4956 indicates this at build time. HEADERS_CHARSET gives the charset to
4957 translate to; headers_charset can change it in the configuration, and
4958 "headers charset" can change it in an individual filter file.
4960 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4961 that creates RFC 2047 encoded "words" labels them as that charset instead
4962 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4963 expansion operator; (ii) when Exim creates a From: line for a local
4964 message; (iii) when a header line is rewritten to include a "phrase" part.
4966 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4967 buggy, causing it to skip the first lines of messages whose message ID
4968 ended in 'D'. This would not have bitten before Exim release 4.14, because
4969 message IDs were unlikely to end in 'D' before then. The effect was to have
4970 incorrect size information for certain domains.
4972 45. #include "config.h" was missing at the start of the crypt16.c module. This
4973 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4976 46. If there was a timeout during a "random" callout check, Exim treated it as
4977 a failure of the random address, and carried on sending RSET and the real
4978 address. If the delay was just some slowness somewhere, the response to the
4979 original RCPT would be taken as a response to RSET and so on, causing
4980 mayhem of various kinds.
4982 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4983 when I implemented it. It didn't allow for the fact that some option values
4984 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4985 the right test for positive values.
4987 48. Domain names in DNS records are case-independent. Exim always looks them up
4988 in lower case. Some resolvers return domain names in exactly the case they
4989 appear in the zone file, that is, they may contain uppercase letters. Not
4990 all resolvers do this - some return always lower case. Exim was treating a
4991 change of case by a resolver as a change of domain, similar to a widening
4992 of a domain abbreviation. This triggered its re-routing code and so it was
4993 trying to route what was effectively the same domain again. This normally
4994 caused routing to fail (because the router wouldn't handle the domain
4995 twice). Now Exim checks for this case specially, and just changes the
4996 casing of the domain that it ultimately uses when it transmits the message
4999 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5002 50. If a filter generated a file delivery with a non-absolute name (possible if
5003 no home directory exists for the router), the forbid_file option was not
5006 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5007 the existing equality matching.
5009 52. Exim was using ints instead of ino_t variables in some places where it was
5010 dealing with inode numbers.
5012 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5013 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5014 called TMPDIR, and if it finds it is different, it changes its value.
5016 54. The smtp_printf() function is now made available to local_scan() so
5017 additional output lines can be written before returning. There is also an
5018 smtp_fflush() function to enable the detection of a dropped connection.
5019 The variables smtp_input and smtp_batched_input are exported to
5022 55. Changed the default runtime configuration: the message "Unknown user"
5023 has been removed from the ACL, and instead placed on the localuser router,
5024 using the cannot_route_message feature. This means that any verification
5025 failures that generate their own messages won't get overridden. Similarly,
5026 the "Unrouteable address" message that was in the ACL for unverifiable
5027 relay addresses has also been removed.
5029 56. Added hosts_avoid_esmtp to the smtp transport.
5031 57. The exicyclog script was not checking for the esoteric option
5032 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5033 will work only if exicyclog is run under the appropriate euid.
5035 58. Following a discussion on the list, the rules by which Exim recognises line
5036 endings on incoming messages have been changed. The -dropcr and drop_cr
5037 options are now no-ops, retained only for backwards compatibility. The
5038 following line terminators are recognized: LF CRLF CR. However, special
5039 processing applies to CR:
5041 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5042 nor a local message in the state where . is a terminator.
5044 (ii) If a bare CR is encountered in a header line, an extra space is added
5045 after the line terminator so as not to end the header. The reasoning
5046 behind this is that bare CRs in header lines are most likely either
5047 to be mistakes, or people trying to play silly games.
5049 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5050 was being incorrectly given as 18 bytes larger than it should have been.
5051 This is a VOB (very old bug).
5053 60. This may never have affected anything current, but just in case it has:
5054 When the local host is found other than at the start of a list of hosts,
5055 the local host, those with the same MX, and any that follow, are discarded.
5056 When the list in question was part of a longer list of hosts, the following
5057 hosts (not currently being processed) were also being discarded. This no
5058 longer happens. I'm not sure if this situation could ever has previously
5061 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5064 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5065 header. This is something that is recommended in a new Internet Draft, and
5066 is something that is documented as being done by Sendmail. There are two
5067 possible values. For messages generated by the autoreply transport, Exim
5070 Auto-Submitted: auto-replied
5072 whereas for all other generated messages (e.g. bounces) it adds
5074 Auto-Submitted: auto-generated
5076 63. The "personal" condition in filters now includes a test for the
5077 Auto-Submitted: header. If it contains the string "auto-" the message it
5078 not considered personal.
5080 64. Added rcpt_include_affixes as a generic transport option.
5082 65. Added queue_only_override (default true).
5084 66. Added the syslog_duplication option.
5086 67. If what should have been the first header line of a message consisted of
5087 a space followed by a colon, Exim was mis-interpreting it as a header line.
5088 It isn't of course - it is syntactically invalid and should therefore be
5089 treated as the start of the message body. The misbehaviour could have
5090 caused a number of strange effects, including loss of data in subsequent
5091 header lines, and spool format errors.
5093 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5094 client host had authenticated. This control can now be exercised by an ACL
5095 for more flexibility.
5097 69. By default, callouts do not happen when testing with -bh. There is now a
5098 variant, -bhc, which does actually run the callout code, including
5099 consulting and updating the callout cache.
5101 70. Added support for saslauthd authentication, courtesy of Alexander
5104 71. If statvfs() failed on the spool or log directories while checking their
5105 size for availability, Exim confusingly gave the error "space shortage".
5106 Furthermore, in debugging mode it crashed with a floating point exception.
5107 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5108 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5109 serious problem, Exim now writes to the main and panic logs when this
5110 happens, with details of the failure. It then refuses to accept the
5111 incoming message, giving the message "spool directory problem" or "log
5112 directory problem" with a 421 code for SMTP messages.
5114 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5115 0, 1, and 2 exist, because some OS complain for execs without them (see
5116 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5117 descriptors. However, the code omitted to check that the open succeeded,
5118 causing mysterious errors if for some reason the permissions on /dev/null
5119 got screwed. Now Exim writes a message to the main and panic logs, and
5120 bombs out if it can't open /dev/null.
5122 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5123 interact so that it is all more flexible. It is supposed to remain
5124 backwards compatible. Also added extra_local_interfaces.
5126 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5127 to bomb out with an assertion failure - to the client this appears as a
5128 connection drop. This problem occurs in the part of the code that was taken
5129 from the Samba project. Fortunately, the assertion is in a very simple
5130 function, so I have fixed this by reproducing the function inline in the
5131 one place where it is called, and arranging for authentication to fail
5132 instead of killing the process with assert().
5134 75. The SPA client code was not working when the server requested OEM rather
5135 than Unicode encoding.
5137 76. Added code to make require_files with a specific uid setting more usable in
5138 the case where statting the file as root fails - usually a non-root-mounted
5139 NFS file system. When this happens and the failure is EACCES, Exim now
5140 forks a subprocess and does the per-uid checking as the relevant uid.
5142 77. Added process_log_path.
5144 78. If log_file_path was not explicitly set, a setting of check_log_space or
5145 check_log_inodes was ignored.
5147 79. If a space check for the spool or log partitions fails, the incident is now
5148 logged. Of course, in the latter case the data may get lost...
5150 80. Added the %p formatting code to string_format() so that it can be used to
5151 print addresses in debug_print(). Adjusted all the address printing in the
5152 debugging in store.c to use %p rather than %d.
5154 81. There was a concern that a line of code in smtp_in.c could overflow a
5155 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5156 initially expressed, the concern was not well-founded, because trailing
5157 spaces are removed early. However, if the trailing spaces were followed by
5158 a NULL, they did not get removed, so the overflow was possible. Two fixes
5161 (a) I re-wrote the offending code in a cleaner fashion.
5162 (b) If an incoming SMTP command contains a NULL character, it is rejected
5165 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5166 runs initgroups(), so that if the Exim user is in any additional groups,
5167 they will be used during message reception.
5173 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob