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 The new ldap_require_cert option would segfault if used. Fixed.
11 PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
18 NM/01 Workround for PCRE version dependency in version reporting
21 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
22 This fixes portability to compilers other than gcc, notably
23 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
25 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
26 makefiles for portability to HP-UX and POSIX correctness.
28 PP/01 Permit LOOKUP_foo enabling on the make command-line.
29 Also via indented variable definition in the Makefile.
30 (Debugging by Oliver Heesakkers).
32 PP/02 Restore caching of spamd results with expanded spamd_address.
33 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
35 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
36 Improves build reliability. Fix from: Frank Elsner
38 NM/02 Fix wide character breakage in the rfc2047 coding
39 Fixes bug 1064. Patch from Andrey N. Oktyabrski
41 NM/03 Allow underscore in dnslist lookups
42 Fixes bug 1026. Patch from Graeme Fowler
44 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
45 Code patches from Adam Ciarcinski of NetBSD.
47 NM/04 Fixed exiqgrep to cope with mailq missing size issue
50 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
51 is logged, to avoid truncation. Patch from John Horne.
53 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
54 Patch from Jakob Hirsch.
56 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
57 SQL string expansion failure details.
58 Patch from Andrey Oktyabrski.
60 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
61 Patch from Simon Arlott.
63 PP/09 New lookups functionality failed to compile on old gcc which rejects
64 extern declarations in function scope.
65 Patch from Oliver Fleischmann
67 PP/10 Use sig_atomic_t for flags set from signal handlers.
68 Check getgroups() return and improve debugging.
69 Fixed developed for diagnosis in bug 927 (which turned out to be
72 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
73 Patch from Mark Zealey.
75 PP/12 Bugzilla 1056: Improved spamd server selection.
76 Patch from Mark Zealey.
78 PP/13 Bugzilla 1086: Deal with maildir quota file races.
79 Based on patch from Heiko Schlittermann.
81 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
82 Patch from Uwe Doering, sign-off by Michael Haardt.
84 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
85 variable declaration deep within a block. Bug and patch from
88 PP/15 lookups-Makefile IRIX compatibilty coercion.
90 PP/16 Make DISABLE_DKIM build knob functional.
92 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
95 TF/03 Fix valgrind.h portability to C89 compilers that do not support
96 variable argument macros. Our copy now differs from upstream.
102 TF/01 Failure to get a lock on a hints database can have serious
103 consequences so log it to the panic log.
105 TF/02 Log LMTP confirmation messages in the same way as SMTP,
106 controlled using the smtp_confirmation log selector.
108 TF/03 Include the error message when we fail to unlink a spool file.
110 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
111 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
112 for maintaining out-of-tree patches for some time.
114 PP/01 Bugzilla 139: Documentation and portability issues.
115 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
116 Handle per-OS dynamic-module compilation flags.
118 PP/02 Let /dev/null have normal permissions.
119 The 4.73 fixes were a little too stringent and complained about the
120 permissions on /dev/null. Exempt it from some checks.
121 Reported by Andreas M. Kirchwitz.
123 PP/03 Report version information for many libraries, including
124 Exim version information for dynamically loaded libraries. Created
125 version.h, now support a version extension string for distributors
126 who patch heavily. Dynamic module ABI change.
128 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
129 privilege escalation vulnerability whereby the Exim run-time user
130 can cause root to append content of the attacker's choosing to
133 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
136 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
137 If dropping privileges for untrusted macros, we disabled normal logging
138 on the basis that it would fail; for the Exim run-time user, this is not
139 the case, and it resulted in successful deliveries going unlogged.
140 Fixed. Reported by Andreas Metzler.
146 PP/01 Date: & Message-Id: revert to normally being appended to a message,
147 only prepend for the Resent-* case. Fixes regression introduced in
148 Exim 4.70 by NM/22 for Bugzilla 607.
150 PP/02 Include check_rfc2047_length in configure.default because we're seeing
151 increasing numbers of administrators be bitten by this.
153 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
155 PP/03 Bugzilla 994: added openssl_options main configuration option.
157 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
159 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
161 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
163 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
164 without a peer certificate, leading to a segfault because of an
165 assumption that peers always have certificates. Be a little more
166 paranoid. Problem reported by Martin Tscholak.
168 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
169 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
170 NB: ClamAV planning to remove STREAM in "middle of 2010".
171 CL also introduces -bmalware, various -d+acl logging additions and
172 more caution in buffer sizes.
174 PP/09 Implemented reverse_ip expansion operator.
176 PP/10 Bugzilla 937: provide a "debug" ACL control.
178 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
180 PP/12 Bugzilla 973: Implement --version.
182 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
184 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
186 PP/15 Bugzilla 816: support multiple condition rules on Routers.
188 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
189 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
190 ignore trailing whitespace.
192 JJ/02 prevent non-panic DKIM error from being sent to paniclog
194 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
197 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
198 Notification from Dr Andrew Aitchison.
200 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
201 ExtendedDetectionInfo response format.
202 Notification from John Horne.
204 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
207 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
208 XSL and documented dependency on system catalogs, with examples of how
211 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
214 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
215 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
216 configuration file which is writeable by the Exim user or group.
218 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
219 of configuration files to cover files specified with the -C option if
220 they are going to be used with root privileges, not just the default
223 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
224 option (effectively making it always true).
226 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
227 files to be used while preserving root privileges.
229 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
230 that rogue child processes cannot use them.
232 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
233 run-time user, instead of root.
235 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
236 Exim run-time user without dropping privileges.
238 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
239 result string, instead of calling string_vformat() twice with the same
242 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
243 for other users. Others should always drop root privileges if they use
244 -C on the command line, even for a whitelisted configure file.
246 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
248 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
254 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
255 $data_path, and $header_path variables; fixed documentation bugs and
258 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
259 exipick to access non-standard spools, including the "frozen" queue
262 NM/01 Bugzilla 965: Support mysql stored procedures.
263 Patch from Alain Williams
265 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
267 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
268 Patch from Andreas Metzler
270 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
271 Patch from Kirill Miazine
273 NM/05 Bugzilla 671: Added umask to procmail example.
275 JJ/03 installed exipick 20100323.0, fixing doc bug
277 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
278 directory. Notification and patch from Dan Rosenberg.
280 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
282 TK/02 Improve log output when DKIM signing operation fails.
284 MH/01 Treat the transport option dkim_domain as a colon separated
285 list, not as a single string, and sign the message with each element,
286 omitting multiple occurences of the same signer.
288 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
289 Bugzilla 985, 986. Patch by Simon Arlott
291 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
292 Patch by Simon Arlott
294 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
295 MBX locking. Notification from Dan Rosenberg.
301 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
303 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
305 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
307 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
309 NM/04 Bugzilla 915: Flag broken perl installation during build.
315 TK/01 Added patch by Johannes Berg that expands the main option
316 "spamd_address" if it starts with a dollar sign.
318 TK/02 Write list of recipients to X-Envelope-Sender header when building
319 the mbox-format spool file for content scanning (suggested by Jakob
322 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
323 (http://www.dcc-servers.net/) support via dccifd. Activated by
324 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
326 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
327 by Mark Daniel Reidel <mr@df.eu>.
329 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
330 When building exim an external PCRE library is now needed -
331 PCRE is a system library on the majority of modern systems.
332 See entry on PCRE_LIBS in EDITME file.
334 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
335 conversation. Added nologin parameter to request.
336 Patch contributed by Kirill Miazine.
338 TF/01 Do not log submission mode rewrites if they do not change the address.
340 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
342 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
343 log files in place. Contributed by Roberto Lima.
345 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
347 TF/03 Bugzilla 615: When checking the local_parts router precondition
348 after a local_part_suffix or local_part_prefix option, Exim now
349 does not use the address's named list lookup cache, since this
350 contains cached lookups for the whole local part.
352 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
353 Robert Millan. Documentation is in experimental-spec.txt.
355 TF/04 Bugzilla 668: Fix parallel build (make -j).
357 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
359 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
360 Patch provided by Jan Srzednicki.
362 TF/05 Leading white space used to be stripped from $spam_report which
363 wrecked the formatting. Now it is preserved.
365 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
366 that they are available at delivery time.
368 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
370 TF/08 TLS error reporting now respects the incoming_interface and
371 incoming_port log selectors.
373 TF/09 Produce a more useful error message if an SMTP transport's hosts
374 setting expands to an empty string.
376 NM/06 Bugzilla 744: EXPN did not work under TLS.
377 Patch provided by Phil Pennock.
379 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
380 Patch provided by Richard Godbee.
382 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
383 acl_smtp_notquit, added index entry.
385 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
386 Patch provided by Eugene Bujak.
388 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
389 accept(). Patch provided by Maxim Dounin.
391 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
392 Patch provided by Phil Pennock.
394 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
396 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
397 Patch provided by Brad "anomie" Jorsch.
399 NM/14 Bugzilla 622: Added timeout setting to transport filter.
400 Patch provided by Dean Brooks.
402 TK/05 Add native DKIM support (does not depend on external libraries).
404 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
405 Patch provided by Graeme Fowler.
407 NM/16 Bugzilla 851: Documentation example syntax fix.
409 NM/17 Changed NOTICE file to remove references to embedded PCRE.
411 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
414 NM/19 Bugzilla 745: TLS version reporting.
415 Patch provided by Phil Pennock.
417 NM/20 Bugzilla 167: bool: condition support.
418 Patch provided by Phil Pennock.
420 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
421 clients. Patch provided by Phil Pennock.
423 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
424 Patch provided by Brad "anomie" Jorsch.
426 NM/23 Bugzilla 687: Fix misparses in eximstats.
427 Patch provided by Heiko Schlittermann.
429 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
430 Patch provided by Heiko Schlittermann.
432 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
433 plus update to original patch.
435 NM/26 Bugzilla 799: Documentation correction for ratelimit.
437 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
438 Patch provided by David Brownlee.
440 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
442 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
444 NM/30 Bugzilla 888: TLS documentation bugfixes.
446 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
448 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
449 Unlike the original bugzilla I have changed all shell scripts in src tree.
451 NM/33 Bugzilla 898: Transport filter timeout fix.
452 Patch by Todd Rinaldo.
454 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
455 Patch by Serge Demonchaux.
457 NM/35 Bugzilla 39: Base64 decode bug fixes.
458 Patch by Jakob Hirsch.
460 NM/36 Bugzilla 909: Correct connect() call in dcc code.
462 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
464 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
466 NM/39 Bugzilla 911: Fixed MakeLinks build script.
472 TK/01 Add preliminary DKIM support. Currently requires a forked version of
473 ALT-N's libdkim that I have put here:
474 http://duncanthrax.net/exim-experimental/
476 Note to Michael Haardt: I had to rename some vars in sieve.c. They
477 were called 'true' and it seems that C99 defines that as a reserved
478 keyword to be used with 'bool' variable types. That means you could
479 not include C99-style headers which use bools without triggering
480 build errors in sieve.c.
482 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
483 as mailq or other aliases. Changed the --help handling significantly
484 to do whats expected. exim_usage() emits usage/help information.
486 SC/01 Added the -bylocaldomain option to eximstats.
488 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
490 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
492 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
498 PH/01 Another patch from the Sieve maintainer.
500 PH/02 When an IPv6 address is converted to a string for single-key lookup
501 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
502 dots are used instead of colons so that keys in lsearch files need not
503 contain colons. This was done some time before quoting was made available
504 in lsearch files. However, iplsearch files do require colons in IPv6 keys
505 (notated using the quote facility) so as to distinguish them from IPv4
506 keys. This meant that lookups for IP addresses in host lists did not work
507 for iplsearch lookups.
509 This has been fixed by arranging for IPv6 addresses to be expressed with
510 colons if the lookup type is iplsearch. This is not incompatible, because
511 previously such lookups could never work.
513 The situation is now rather anomolous, since one *can* have colons in
514 ordinary lsearch keys. However, making the change in all cases is
515 incompatible and would probably break a number of configurations.
517 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
520 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
521 conversion specification without a maximum field width, thereby enabling
522 a rogue spamd server to cause a buffer overflow. While nobody in their
523 right mind would setup Exim to query an untrusted spamd server, an
524 attacker that gains access to a server running spamd could potentially
525 exploit this vulnerability to run arbitrary code as the Exim user.
527 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
528 $primary_hostname instead of what libspf2 thinks the hosts name is.
530 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
531 a directory entry by the name of the lookup key. Previously, if a
532 symlink pointed to a non-existing file or a file in a directory that
533 Exim lacked permissions to read, a lookup for a key matching that
534 symlink would fail. Now it is enough that a matching directory entry
535 exists, symlink or not. (Bugzilla 503.)
537 PH/03 The body_linecount and body_zerocount variables are now exported in the
540 PH/04 Added the $dnslist_matched variable.
542 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
543 This means they are set thereafter only if the connection becomes
546 PH/06 Added the client_condition to authenticators so that some can be skipped
547 by clients under certain conditions.
549 PH/07 The error message for a badly-placed control=no_multiline_responses left
550 "_responses" off the end of the name.
552 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
554 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
555 (without spaces) instead of just copying the configuration text.
557 PH/10 Added the /noupdate option to the ratelimit ACL condition.
559 PH/11 Added $max_received_linelength.
561 PH/12 Added +ignore_defer and +include_defer to host lists.
563 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
564 way in which PCRE > 7.0 is built.
566 PH/14 Implemented queue_only_load_latch.
568 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
569 MAIL command. The effect was to mangle the value on 64-bit systems.
571 PH/16 Another patch from the Sieve maintainer.
573 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
575 PH/18 If a system quota error occurred while trying to create the file for
576 a maildir delivery, the message "Mailbox is full" was not appended to the
577 bounce if the delivery eventually timed out. Change 4.67/27 below applied
578 only to a quota excession during the actual writing of the file.
580 PH/19 It seems that peer DN values may contain newlines (and other non-printing
581 characters?) which causes problems in log lines. The DN values are now
582 passed through string_printing() before being added to log lines.
584 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
585 and InterBase are left for another time.)
587 PH/21 Added message_body_newlines option.
589 PH/22 Guard against possible overflow in moan_check_errorcopy().
591 PH/23 POSIX allows open() to be a macro; guard against that.
593 PH/24 If the recipient of an error message contained an @ in the local part
594 (suitably quoted, of course), incorrect values were put in $domain and
595 $local_part during the evaluation of errors_copy.
601 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
602 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
605 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
606 issue a MAIL command.
608 PH/02 In an ACL statement such as
610 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
612 if a client was not listed at all, or was listed with a value other than
613 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
614 the condition was not true (as it should be), so access was not denied.
615 The bug was that the ! inversion was incorrectly passed on to the second
616 item. This has been fixed.
618 PH/03 Added additional dnslists conditions == and =& which are different from
619 = and & when the dns lookup returns more than one IP address.
621 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
622 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
624 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
625 FSYNC, which compiles an option called disable_fsync that allows for
626 bypassing fsync(). The documentation is heavily laced with warnings.
628 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
630 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
631 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
632 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
633 including adding "make clean"; (3) Added -fPIC when compiling the test
634 dynamically loaded module, to get rid of a warning.
636 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
637 message fails, move_frozen_messages = true and ignore_bounce_errors_after
638 = 0s. The bug is otherwise harmless.
640 PH/07 There was a bug in the dovecot authenticator such that the value of
641 $auth1 could be overwritten, and so not correctly preserved, after a
642 successful authentication. This usually meant that the value preserved by
643 the server_setid option was incorrect.
645 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
647 PH/09 Installed PCRE release 7.0.
649 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
650 run for batched SMTP input. It is now run at the start of every message
651 in the batch. While fixing this I discovered that the process information
652 (output by running exiwhat) was not always getting set for -bs and -bS
653 input. This is fixed, and it now also says "batched" for BSMTP.
655 PH/11 Added control=no_pipelining.
657 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
658 patch, slightly modified), and move the expansion of helo_data till after
659 the connection is made in the smtp transport (so it can use these
662 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
664 PH/14 Added log_selector = +pid.
666 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
668 PH/16 Add ${if forany and ${if forall.
670 PH/17 Added dsn_from option to vary the From: line in DSNs.
672 PH/18 Flush SMTP output before performing a callout, unless control =
673 no_callout_flush is set.
675 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
676 was true (the default) a successful delivery failed to delete the retry
677 item, thus causing premature timeout of the address. The bug is now
680 PH/20 Added hosts_avoid_pipelining to the smtp transport.
682 PH/21 Long custom messages for fakedefer and fakereject are now split up
683 into multiline reponses in the same way that messages for "deny" and
684 other ACL rejections are.
686 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
687 with slight modification.
689 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
690 draft, changing the syntax and factoring some duplicate code".
692 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
693 for deliveries of the second and subsequent messages over the same SMTP
696 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
697 ${reduce, with only minor "tidies".
699 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
701 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
702 expansion side effects.
704 PH/27 When a message times out after an over-quota error from an Exim-imposed
705 quota, the bounce message says "mailbox is full". This message was not
706 being given when it was a system quota that was exceeded. It now should
709 MH/03 Made $recipients available in local_scan(). local_scan() already has
710 better access to the recipient list through recipients_list[], but
711 $recipients can be useful in postmaster-provided expansion strings.
713 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
714 in the case of a MAIL command with additional options following the
715 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
716 were accidentally chopped off.
718 PH/29 SMTP synchronization checks are implemented when a command is read -
719 there is a check that no more input is waiting when there shouldn't be
720 any. However, for some commands, a delay in an ACL can mean that it is
721 some time before the response is written. In this time, more input might
722 arrive, invalidly. So now there are extra checks after an ACL has run for
723 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
724 pipelining has not been advertised.
726 PH/30 MH's patch to allow iscntrl() characters to be list separators.
728 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
729 returned in the SMTP response when smtp_return_error_details was false.
732 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
733 instead of the C I/O that was originally supplied, because problems were
736 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
737 Exim which did not show up earlier: it was assuming that a call to
738 SSL_CTX_set_info_callback() might give an error value. In fact, there is
739 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
740 was a macro that became an assignment, so it seemed to work. This has
741 changed to a proper function call with a void return, hence the compile
742 error. Exim's code has been fixed.
744 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
747 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
749 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
751 PH/37 If a message is not accepted after it has had an id assigned (e.g.
752 because it turns out to be too big or there is a timeout) there is no
753 "Completed" line in the log. When some messages of this type were
754 selected by exigrep, they were listed as "not completed". Others were
755 picked up by some special patterns. I have improved the selection
756 criteria to be more general.
758 PH/38 The host_find_failed option in the manualroute router can now be set
759 to "ignore", to completely ignore a host whose IP address cannot be
760 found. If all hosts are ignored, the behaviour is controlled by the new
761 host_all_ignored option.
763 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
764 homing or because of multiple MX records with /mx) generated more than
765 one IP address, and the following item turned out to be the local host,
766 all the secondary addresses of the first item were incorrectly removed
767 from the list, along with the local host and any following hosts (which
768 is what is supposed to happen).
770 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
771 whoever called Exim into the -H file. In the case of the daemon it was
772 behaving confusingly. When first started, it used values for whoever
773 started the daemon, but after a SIGHUP it used the Exim user (because it
774 calls itself on a restart). I have changed the code so that it now always
777 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
778 message are rejected with the same error (e.g. no authentication or bad
779 sender address), and a DATA command is nevertheless sent (as can happen
780 with PIPELINING or a stupid MUA), the error message that was given to the
781 RCPT commands is included in the rejection of the DATA command. This is
782 intended to be helpful for MUAs that show only the final error to their
785 PH/42 Another patch from the Sieve maintainer.
787 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
788 Eximstats - Fixed some broken HTML links and added missing column headers
790 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
791 columns for Rejects, Temp Rejects, Ham, and Spam rows.
793 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
795 PH/43 Yet another patch from the Sieve maintainer.
797 PH/44 I found a way to check for a TCP/IP connection going away before sending
798 the response to the final '.' that terminates a message, but only in the
799 case where the client has not sent further data following the '.'
800 (unfortunately, this is allowed). However, in many cases there won't be
801 any further data because there won't be any more messages to send. A call
802 to select() can be used: if it shows that the input is "ready", there is
803 either input waiting, or the socket has been closed. An attempt to read
804 the next input character can distinguish the two cases. Previously, Exim
805 would have sent an OK response which the client would never have see.
806 This could lead to message repetition. This fix should cure that, at
807 least in a lot of common cases.
809 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
810 advertised in response to EHLO.
816 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
817 fixed by 4.65/MH/01 (is this a record?) are fixed:
819 (i) An empty string was always treated as zero by the numeric comparison
820 operators. This behaviour has been restored.
822 (ii) It is documented that the numeric comparison operators always treat
823 their arguments as decimal numbers. This was broken in that numbers
824 starting with 0 were being interpreted as octal.
826 While fixing these problems I realized that there was another issue that
827 hadn't been noticed. Values of message_size_limit (both the global option
828 and the transport option) were treated as octal if they started with 0.
829 The documentation was vague. These values are now always treated as
830 decimal, and I will make that clear in the documentation.
836 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
837 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
840 MH/01 Don't check that the operands of numeric comparison operators are
841 integers when their expansion is in "skipping" mode (fixes bug
842 introduced by 4.64-PH/07).
844 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
845 child addresses, Exim now panics and dies. Previously, because the count
846 is held in a short int, deliveries were likely to be lost. As such a
847 large number of recipients for a single message is ridiculous
848 (performance will be very, very poor), I have chosen to impose a limit
849 rather than extend the field.
855 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
856 leftover -K file (the existence of which was triggered by #402).
857 While we were at it, introduced process PID as part of the -K
858 filename. This should rule out race conditions when creating
861 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
862 processing considerably. Previous code took too long for large mails,
863 triggering a timeout which in turn triggers #401.
865 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
866 in the DK code in transports.c. sendfile() is not really portable,
867 hence the _LINUX specificness.
869 TF/01 In the add_headers option to the mail command in an Exim filter,
870 there was a bug that Exim would claim a syntax error in any
871 header after the first one which had an odd number of characters
874 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
875 callout verification, Exim cached a "reject" for the entire domain. This
876 is correct for most verifications, but it is not correct for a recipient
877 verification with use_sender or use_postmaster set, because in that case
878 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
879 case of MAIL FROM:<> rejection from other early rejections (e.g.
880 rejection of HELO). When verifying a recipient using a non-null MAIL
881 address, the cache is ignored if it shows MAIL FROM:<> rejection.
882 Whatever the result of the callout, the value of the domain cache is
883 left unchanged (for any other kind of callout, getting as far as trying
884 RCPT means that the domain itself is ok).
886 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
889 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
890 manifest itself as EPIPE rather than ECONNECT. When tidying away a
891 session, the daemon ignores ECONNECT errors and logs others; it now
892 ignores EPIPE as well.
894 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
895 (quoted-printable decoding).
897 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
898 later the small subsequent patch to fix an introduced bug.
900 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
902 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
904 PH/08 An error is now given if message_size_limit is specified negative.
906 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
907 to be given (somewhat) arbitrary names.
909 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
912 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
913 miscellaneous code fixes
915 PH/10 Added the log_reject_target ACL modifier to specify where to log
918 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
919 hostname. This is wrong, because it relates to the incoming message (and
920 probably the interface on which it is arriving) and not to the outgoing
921 callout (which could be using a different interface). This has been
922 changed to use the value of the helo_data option from the smtp transport
923 instead - this is what is used when a message is actually being sent. If
924 there is no remote transport (possible with a router that sets up host
925 addresses), $smtp_active_hostname is used.
927 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
928 tweaks were necessary in order to get it to work (see also 21 below):
929 (a) The code assumed that strncpy() returns a negative number on buffer
930 overflow, which isn't the case. Replaced with Exim's string_format()
932 (b) There were several signed/unsigned issues. I just did the minimum
933 hacking in of casts. There is scope for a larger refactoring.
934 (c) The code used strcasecmp() which is not a standard C function.
935 Replaced with Exim's strcmpic() function.
936 (d) The code set only $1; it now sets $auth1 as well.
937 (e) A simple test gave the error "authentication client didn't specify
938 service in request". It would seem that Dovecot has changed its
939 interface. Fortunately there's a specification; I followed it and
940 changed what the client sends and it appears to be working now.
942 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
945 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
946 address A is aliased to B and C, where B exists and C does not. Without
947 -v the output is "A verified" because verification stops after a
948 successful redirection if more than one address is generated. However,
949 with -v the child addresses are also verified. Exim was outputting "A
950 failed to verify" and then showing the successful verification for C,
951 with its parentage. It now outputs "B failed to verify", showing B's
952 parentage before showing the successful verification of C.
954 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
955 look up a TXT record in a specific list after matching in a combined
958 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
959 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
960 they consult the DNS. I had assumed they would set it the way they
961 wanted; and indeed my experiments on Linux seem to show that in some
962 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
963 To be on the safe side, however, I have now made the interface to
964 host_find_byname() similar to host_find_bydns(), with an argument
965 containing the DNS resolver options. The host_find_byname() function now
966 sets these options at its start, just as host_find_bydns() does. The smtp
967 transport options dns_qualify_single and dns_search_parents are passed to
968 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
969 of host_find_byname() use the default settings of RES_DEFNAMES
970 (qualify_single) but not RES_DNSRCH (search_parents).
972 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
973 spool_read_header() do less string testing, by means of a preliminary
974 switch on the second character of optional "-foo" lines. (This is
975 overdue, caused by the large number of possibilities that now exist.
976 Originally there were few.) While I was there, I also converted the
977 str(n)cmp tests so they don't re-test the leading "-" and the first
978 character, in the hope this might squeeze out yet more improvement.
980 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
981 flag allowing group syntax was set by the header_syntax check but not
982 turned off, possible causing trouble later; (2) The flag was not being
983 set at all for the header_verify test, causing "group"-style headers to
984 be rejected. I have now set it in this case, and also caused header_
985 verify to ignore an empty address taken from a group. While doing this, I
986 came across some other cases where the code for allowing group syntax
987 while scanning a header line wasn't quite right (mostly, not resetting
988 the flag correctly in the right place). These bugs could have caused
989 trouble for malformed header lines. I hope it is now all correct.
991 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
992 with the "reply" argument non-NULL. The code, however (which originally
993 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
994 but it didn't always do it. This confused somebody who was copying the
995 code for some other use. I have removed all the tests.
997 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
998 feature that was used to support insecure browsers during the U.S. crypto
999 embargo. It requires special client support, and Exim is probably the
1000 only MTA that supported it -- and would never use it because real RSA is
1001 always available. This code has been removed, because it had the bad
1002 effect of slowing Exim down by computing (never used) parameters for the
1003 RSA_EXPORT functionality.
1005 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1006 authenticator to fail if there's a tab character in the incoming data
1007 (there should never be unless someone is messing about, as it's supposed
1008 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1009 if the connection is using TLS or if the remote IP is the same as the
1010 local IP, and the "valid-client-cert option" if a client certificate has
1013 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1014 authenticators. This can be used for authorization after authentication
1015 succeeds. (In the case of plaintext, it servers for both authentication
1018 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1019 if any retry times were supplied.
1021 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1022 connection, where there is no client IP address to check. In this
1023 situation, the verify now always succeeds.
1025 PH/25 Applied John Jetmore's -Mset patch.
1027 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1029 PH/27 In a string expansion for a processed (not raw) header when multiple
1030 headers of the same name were present, leading whitespace was being
1031 removed from all of them, but trailing whitespace was being removed only
1032 from the last one. Now trailing whitespace is removed from each header
1033 before concatenation. Completely empty headers in a concatenation (as
1034 before) are ignored.
1036 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1037 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1039 PH/29 [Removed. This was a change that I later backed out, and forgot to
1040 correct the ChangeLog entry (that I had efficiently created) before
1041 committing the later change.]
1043 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1044 address errors (4xx response to RCPT) over the same connection as other
1045 messages routed to the same hosts. Such deliveries are always "forced",
1046 so retry times are not inspected. This resulted in far too many retries
1047 for the affected addresses. The effect occurred only when there were more
1048 hosts than the hosts_max_try setting in the smtp transport when it had
1049 the 4xx errors. Those hosts that it had tried were not added to the list
1050 of hosts for which the message was waiting, so if all were tried, there
1051 was no problem. Two fixes have been applied:
1053 (i) If there are any address or message errors in an SMTP delivery, none
1054 of the hosts (tried or untried) are now added to the list of hosts
1055 for which the message is waiting, so the message should not be a
1056 candidate for sending over the same connection that was used for a
1057 successful delivery of some other message. This seems entirely
1058 reasonable: after all the message is NOT "waiting for some host".
1059 This is so "obvious" that I'm not sure why it wasn't done
1060 previously. Hope I haven't missed anything, but it can't do any
1061 harm, as the worst effect is to miss an optimization.
1063 (ii) If, despite (i), such a delivery is accidentally attempted, the
1064 routing retry time is respected, so at least it doesn't keep
1065 hammering the server.
1067 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1068 in ${readsocket because some servers need this prod.
1070 PH/32 Added some extra debug output when updating a wait-xxx database.
1072 PH/33 The hint "could be header name not terminated by colon", which has been
1073 given for certain expansion errors for a long time, was not being given
1074 for the ${if def:h_colon_omitted{... case.
1076 PH/34 The spec says: "With one important exception, whenever a domain list is
1077 being scanned, $domain contains the subject domain." There was at least
1078 one case where this was not true.
1080 PH/35 The error "getsockname() failed: connection reset by peer" was being
1081 written to the panic log as well as the main log, but it isn't really
1082 panic-worthy as it just means the connection died rather early on. I have
1083 removed the panic log writing for the ECONNRESET error when getsockname()
1086 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1087 runs only) independently of the message's sender address. This meant
1088 that, if the 4xx error was in fact related to the sender, a different
1089 message to the same recipient with a different sender could confuse
1090 things. In particualar, this can happen when sending to a greylisting
1091 server, but other circumstances could also provoke similar problems.
1092 I have changed the default so that the retry time for these errors is now
1093 based a combination of the sender and recipient addresses. This change
1094 can be overridden by setting address_retry_include_sender=false in the
1097 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1098 remote server are returned as part of bounce messages. This was not
1099 happening for LMTP over a pipe (the lmtp transport), but now it is the
1100 same for both kinds of LMTP.
1102 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1103 in header lines that were in fact CNAMEs. This is no longer the case.
1105 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1106 and queue runs started by the daemon processed all messages. This has
1107 been fixed so that -R and -S can now usefully be given with -q<time>.
1109 PH/40 Import PCRE release 6.7 (fixes some bugs).
1111 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1113 PH/42 Give an error if -q is specified more than once.
1115 PH/43 Renamed the variables $interface_address and $interface_port as
1116 $received_ip_address and $received_port, to make it clear that these
1117 values apply to message reception, and not to the outgoing interface when
1118 a message is delivered. (The old names remain recognized, of course.)
1120 PH/44 There was no timeout on the connect() call when using a Unix domain
1121 socket in the ${readsocket expansion. There now is.
1123 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1124 be meaningful with "accept".
1126 SC/01 Eximstats V1.43
1127 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1129 SC/02 Eximstats V1.44
1130 Use a glob alias rather than an array ref in the generated
1131 parser. This improves both readability and performance.
1133 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1134 Collect SpamAssassin and rejection statistics.
1135 Don't display local sender or destination tables unless
1136 there is data to show.
1137 Added average volumes into the top table text output.
1139 SC/04 Eximstats V1.46
1140 Collect data on the number of addresses (recipients)
1141 as well as the number of messages.
1143 SC/05 Eximstats V1.47
1144 Added 'Message too big' to the list of mail rejection
1145 reasons (thanks to Marco Gaiarin).
1147 SC/06 Eximstats V1.48
1148 Mainlog lines which have GMT offsets and are too short to
1149 have a flag are now skipped.
1151 SC/07 Eximstats V1.49 (Alain Williams)
1152 Added the -emptyok flag.
1154 SC/08 Eximstats V1.50
1155 Fixes for obtaining the IP address from reject messages.
1157 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1158 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1159 whitesspace changes from 4.64-PH/27
1161 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1164 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1165 are found, allow negative numbers in numeric criteria)
1167 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1169 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1170 to match changes made in 4.64-PH/43
1172 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1174 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1175 "domain" is actually an IP address, and force a failure. This locks out
1176 those revolvers/nameservers that support "A-for-A" lookups, in
1177 contravention of the specifications.
1179 PH/48 When a host name was looked up from an IP address, and the subsequent
1180 forward lookup of the name timed out, the host name was left in
1181 $sender_host_name, contrary to the specification.
1183 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1184 restricted to single-key lookups, Exim was not diagnosing an error if
1185 * or *@ was used with a query-style lookup.
1187 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1189 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1190 long ago, but noone interested enough thought of it. Let's just say that
1191 the "1.1" means that there are some new functions that weren't there at
1192 some point in the past.
1194 PH/51 Error processing for expansion failure of helo_data from an smtp
1195 transport during callout processing was broken.
1197 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1198 tested/used via the -bh/-bhc/-bs options.
1200 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1201 bug, fixed in subsequent PCRE releases).
1203 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1204 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1206 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1212 SC/01 Use a glob alias rather than an array ref in eximstats generated
1213 parser. This improves both readability and performance.
1215 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1216 Don't display local sender or destination tables in eximstats unless
1217 there is data to show.
1218 Added average volumes into the eximstats top table text output.
1220 SC/03 Collect data on the number of addresses (recipients) as well
1221 as the number of messages in eximstats.
1223 TF/01 Correct an error in the documentation for the redirect router. Exim
1224 does (usually) call initgroups() when daemonizing.
1226 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1227 with consistent privilege compared to when running as a daemon.
1229 TF/03 Note in the spec that $authenticated_id is not set for local
1230 submissions from trusted users.
1232 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1233 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1235 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1236 by adding some example configuration directives to the default
1237 configuration file. A little bit of work is required to uncomment the
1238 directives and define how usernames and passwords are checked, but
1239 there is now a framework to start from.
1241 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1242 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1243 without this. I don't know how relevant this is to other LDAP libraries.
1245 PH/02 Add the verb name to the "unknown ACL verb" error.
1247 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1249 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1251 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1252 directory not expanded when it should be if an expanded home directory
1253 was set for the address (which is overridden by the transport).
1255 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1258 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1259 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1260 because it is too late at that time, and has no effect.
1262 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1263 security issue with \' (bugzilla #107). I could not use the
1264 PQescapeStringConn() function, because it needs a PGconn value as one of
1267 PH/08 When testing addresses using -bt, indicate those final addresses that
1268 are duplicates that would not cause an additional delivery. At least one
1269 person was confused, thinking that -bt output corresponded to deliveries.
1270 (Suppressing duplicates isn't a good idea as you lose the information
1271 about possibly different redirections that led to the duplicates.)
1273 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1274 systems where poll() doesn't work, in particular OS X.
1276 PH/10 Added more information to debugging output for retry time not reached.
1278 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1279 operations in malware.c.
1281 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1284 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1285 syslog_duplication set false, log lines that would normally be written
1286 both the the main log and to the reject log were not written to syslog at
1289 PH/14 In the default configuration, change the use of "message" in ACL warn
1290 statements to "add_header".
1292 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1293 not followed by a command (e.g. "seen endif").
1295 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1296 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1299 PH/17 Added extra conditions to the default value of delay_warning_condition
1303 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1304 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1305 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1308 The Auto-Submitted: and various List- headers are standardised, whereas I
1309 don't think Precedence: ever was.
1311 PH/18 Refactored debugging code in route_finduser() to show more information,
1312 in particular, the error code if getpwnam() issues one.
1314 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1315 This is apparently needed in addition to the PH/07 change above to avoid
1316 any possible encoding problems.
1318 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1319 but not after initializing Perl.
1321 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1322 output them only if debugging. By default they are written stderr,
1323 apparently, which is not desirable.
1325 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1328 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1331 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1333 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1334 authenticated or an ident call has been made. Suppress the default
1335 values for $authenticated_id and $authenticated_sender (but permit -oMai
1336 and -oMas) when testing with -bh.
1338 PH/24 Re-jigged the order of the tests in the default configuration so that the
1339 tests for valid domains and recipients precede the DNS black list and CSA
1340 tests, on the grounds that those ones are more expensive.
1342 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1343 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1344 This bug exists in every version of Exim that I still have, right back to
1347 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1348 However, an attempt to turn on case-sensitivity in a regex key by
1349 including (?-i) didn't work because the subject string was already
1350 lowercased, and the effects were non-intuitive. It turns out that a
1351 one-line patch can be used to allow (?-i) to work as expected.
1357 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1358 other effects) broke the use of negated acl sub-conditions.
1360 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1363 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1364 "Deny" causes Exim to reject the incoming connection with a 554 error.
1365 Unfortunately, if there is a major crisis, such as a disk failure,
1366 tcp-wrappers gives "deny", whereas what one would like would be some
1367 kind of temporary error. A kludge has been added to help with this.
1368 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1369 554 error is used if errno is still zero or contains ENOENT (which occurs
1370 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1373 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1375 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1376 errors. Otherwise a message that provokes a temporary error (when other
1377 messages do not) can cause a whole host to time out.
1379 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1380 addresses were routed directly to files or pipes from a redirect router.
1381 File deliveries just didn't batch; pipe deliveries might have suffered
1384 PH/06 A failure to get a lock for a hints database would erroneously always say
1385 "Failed to get write lock", even when it was really a read lock.
1387 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1388 of 0600. This has been changed to use the value of the lockfile_mode
1389 option (which defaults to 0600).
1391 PH/08 Applied small patch from the Sieve maintainer.
1393 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1394 folder from quota calculations, a direct delivery into this folder messed
1395 up the contents of the maildirsize file. This was because the regex was
1396 used only to exclude .Trash (or whatever) when the size of the mailbox
1397 was calculated. There was no check that a delivery was happening into an
1398 excluded directory. This bug has been fixed by ignoring all quota
1399 processing for deliveries into excluded directories.
1401 PH/10 Added the maildirfolder_create_regex option to appendfile.
1407 PH/01 The code for finding all the local interface addresses on a FreeBSD
1408 system running IPv6 was broken. This may well have applied to all BSD
1409 systems, as well as to others that have similar system calls. The broken
1410 code found IPv4 interfaces correctly, but gave incorrect values for the
1411 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1412 that it would not match correctly against @[] and not recognize the IPv6
1415 PH/02 The ipliteral router was not recognizing addresses of the form user@
1416 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1418 PH/03 Added disable_ipv6.
1420 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1421 decoded form, because it is most often used to construct To: headers
1422 lines in autoreplies, and the decoded form may well be syntactically
1423 invalid. However, $reply_address has leading white space removed, and all
1424 newlines turned into spaces so that the autoreply transport does not
1427 PH/05 If group was specified without a user on a router, and no group or user
1428 was specified on a transport, the group from the router was ignored.
1430 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1431 for visible compile-time settings that can be used to change these
1432 numbers, for those that want even more. Backwards compatibility with old
1433 spool files has been maintained. However, going back to a previous Exim
1434 release will lost any variables that are in spool files.
1436 PH/07 Two small changes when running in the test harness: increase delay when
1437 passing a TCP/IP connection to a new process, in case the original
1438 process has to generate a bounce, and remove special handling of
1439 127.0.0.2 (sic), which is no longer necessary.
1441 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1442 be the same on different OS.
1444 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1447 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1448 whether --show-vars was specified or not
1450 JJ/02 exipick: Added support for new ACL variable spool format introduced
1453 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1454 syntactically invalid From: or Reply-to: line, and a filter used this to
1455 generate an autoreply, and therefore failed to obtain an address for the
1456 autoreply, Exim could try to deliver to a non-existent relative file
1457 name, causing unrelated and misleading errors. What now happens is that
1458 it logs this as a hard delivery error, but does not attempt to create a
1461 PH/11 The exinext utility has a -C option for testing purposes, but although
1462 the given file was scanned by exinext itself; it wasn't being passed on
1463 when Exim was called.
1465 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1466 an end-of-file indication when reading a command response.
1468 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1469 compiled. In many other places in Exim, IPv6 addresses are always
1470 recognized, so I have changed this. It also means that IPv4 domain
1471 literals of the form [IPV4:n.n.n.n] are now always recognized.
1473 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1474 used if the router is not running as root, for example, when verifying at
1475 ACL time, or when using -bh. The debugging output from this situation was
1476 non-existent - all you got was a failure to exec. I have made two
1479 (a) Failures to set uid/gid, the current directory, or a process leader
1480 in a subprocess such as that created by queryprogram now generate
1481 suitable debugging ouput when -d is set.
1483 (b) The queryprogram router detects when it is not running as root,
1484 outputs suitable debugging information if -d is set, and then runs
1485 the subprocess without attempting to change uid/gid.
1487 PH/15 Minor change to Makefile for building test_host (undocumented testing
1490 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1491 additional section of a DNS packet that returns MX or SRV records.
1492 Instead, it always explicitly searches for A/AAAA records. This avoids
1493 major problems that occur when a DNS server includes only records of one
1494 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1495 fixed another bug: if SRV records were looked up and the corresponding
1496 address records were *not* found in the additional section, the port
1497 values from the SRV records were lost.
1499 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1500 using the correct key (the original address) when searching the retry
1501 rules in order to find which one to use for generating the retry hint.
1503 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1504 adding the default one. Similarly, if it contains a Reply-To: header, the
1505 errors_reply_to option, if set, is not used.
1507 PH/19 When calculating a retry time, Exim used to measure the "time since
1508 failure" by looking at the "first failed" field in the retry record. Now
1509 it does not use this if it is later than than the arrival time of the
1510 message. Instead it uses the arrival time. This makes for better
1511 behaviour in cases where some deliveries succeed, thus re-setting the
1512 "first failed" field. An example is a quota failure for a huge message
1513 when small messages continue to be delivered. Without this change, the
1514 "time since failure" will always be short, possible causing more frequent
1515 delivery attempts for the huge message than are intended.
1516 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1518 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1519 $1, $2, $3) because the numerical variables can be reset during some
1520 expansion items (e.g. "match"), thereby losing the authentication data.
1522 PH/21 Make -bV show the size of off_t variables so that the test suite can
1523 decide whether to run tests for quotas > 2G.
1525 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1526 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1527 filecount value is greater than 2G or if a quota value is greater than 2G
1528 on a system where the size of off_t is not greater than 4, a panic error
1531 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1532 never match. The debug and -bh output now contains an explicit error
1533 message indicating a malformed IPv4 address or mask.
1535 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1536 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1537 PH/23 above applies.
1539 PH/25 Do not write to syslog when running in the test harness. The only
1540 occasion when this arises is a failure to open the main or panic logs
1541 (for which there is an explicit test).
1543 PH/26 Added the /no_tell option to "control=freeze".
1545 PH/27 If a host name lookup failed very early in a connection, for example, if
1546 the IP address matched host_lookup and the reverse lookup yielded a name
1547 that did not have a forward lookup, an error message of the form "no IP
1548 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1549 could be logged. Now it outputs the IP address instead of "NULL".
1551 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1552 allows the sender and the authenticated sender to be set when
1553 submitting a message from within Exim. Since child_open_exim() is
1554 documented for local_scan(), the new function should be too.
1556 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1557 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1558 results in an empty string is now treated as unset.
1560 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1562 PH/31 Added sender_verify_fail logging option.
1564 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1565 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1566 tidied the source and removed it altogether.
1568 PH/33 When a queue run was abandoned because the load average was too high, a
1569 log line was always written; now it is written only if the queue_run log
1570 selector is set. In addition, the log line for abandonment now contains
1571 information about the queue run such as the pid. This is always present
1572 in "start" and "stop" lines but was omitted from the "abandon" line.
1574 PH/34 Omit spaces between a header name and the colon in the error message that
1575 is given when verify = headers_syntax fails (if there are lots of them,
1576 the message gets confusing).
1578 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1579 names, as there are now some PTR records that contain slashes. This check
1580 is only to protect against broken name servers that fall over on strange
1581 characters, so the fact that it applies to all lookups doesn't matter.
1583 PH/36 Now that the new test suite is complete, we can remove some of the
1584 special code in Exim that was needed for the old test suite. For example,
1585 sorting DNS records because real resolvers return them in an arbitrary
1586 order. The new test suite's fake resolver always returns records in the
1589 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1590 bounces) except when queue_only is set, to avoid logging races between
1591 the different processes.
1593 PH/38 Panic-die if .include specifies a non-absolute path.
1595 PH/39 A tweak to the "H" retry rule from its user.
1597 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1598 a label. They prevented compilation on older perls.
1600 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1601 a warning to be raised on newish perls.
1603 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1604 on queue. Changes to match documented behaviour of showing count of
1605 messages matching specified criteria.
1607 PH/40 Changed the default ident timeout from 30s to 5s.
1609 PH/41 Added support for the use of login_cap features, on those BSD systems
1610 that have them, for controlling the resources used by pipe deliveries.
1612 PH/42 The content-scanning code uses fopen() to create files in which to put
1613 message data. Previously it was not paying any attention to the mode of
1614 the files. Exim runs with umask(0) because the rest of the code creates
1615 files with open(), and sets the required mode explicitly. Thus, these
1616 files were ending up world-writeable. This was not a big issue, because,
1617 being within the spool directory, they were not world-accessible. I have
1618 created a function called modefopen, which takes an additional mode
1619 argument. It sets umask(777), creates the file, chmods it to the required
1620 mode, then resets the umask. All the relevant calls to fopen() in the
1621 content scanning code have been changed to use this function.
1623 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1624 to 24 hours. This avoids potential overflow problems when processing G
1625 and H retry rules. I suspect nobody ever tinkers with this value.
1627 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1629 PH/45 When the plaintext authenticator is running as a client, the server's
1630 challenges are checked to ensure they are valid base64 strings. By
1631 default, the authentication attempt is cancelled if an invalid string is
1632 received. Setting client_ignore_invalid_base64 true ignores these errors.
1633 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1634 they are received. Thus, the responses can be made to depend on the
1635 challenges. If an invalid string is ignored, an empty string is placed in
1638 PH/46 Messages that are created by the autoreply transport now contains a
1639 References: header, in accordance with RFCs 2822 and 3834.
1641 PH/47 Added authenticated_sender_force to the smtp transport.
1643 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1645 PH/49 Installed latest patch from the Sieve maintainer.
1647 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1648 was also set, the appendfile transport was unnecessarily scanning a
1649 directory of message files (e.g. for maildir delivery) to find the count
1650 of files (along with the size), even though it did not need this
1651 information. It now does the scan only if it needs to find either the
1652 size of the count of files.
1654 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1656 PH/52 Two bugs concerned with error handling when the smtp transport is
1659 (i) Exim was not creating retry information for temporary errors given
1660 for individual recipients after the DATA command when the smtp transport
1661 was used in LMTP mode. This meant that they could be retried too
1662 frequently, and not timed out correctly.
1664 (ii) Exim was setting the flag that allows error details to be returned
1665 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1666 recipients that were returned after the DATA command.
1668 PH/53 This is related to PH/52, but is more general: for any failing address,
1669 when detailed error information was permitted to be returned to the
1670 sender, but the error was temporary, then after the final timeout, only
1671 "retry timeout exceeded" was returned. Now it returns the full error as
1672 well as "retry timeout exceeded".
1674 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1675 do this, and (what is worse) MTAs that accept it.
1677 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1678 will now be deprecated.
1680 PH/56 New os.c-cygwin from the Cygwin maintainer.
1682 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1683 formats (previously only available in exim formats via -bpr, -bpru,
1684 and -bpra. Now also available in native and exiqgrep formats)
1686 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1687 with very large, slow to parse queues
1689 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1691 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1693 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1694 responses to authentication challenges, though it was showing the
1695 challenges; (ii) I've removed the CR characters from the debug output for
1698 PH/58 Allow for the insertion of a newline as well as a space when a string
1699 is turned into more than one encoded-word during RFC 2047 encoding. The
1700 Sieve code now uses this.
1702 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1703 data_4xx, lost_connection, tls_required.
1705 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1706 message was being sent as an SMTP response.
1708 PH/61 Add -l and -k options to exicyclog.
1710 PH/62 When verifying, if an address was redirected to one new address, so that
1711 verification continued, and the new address failed or deferred after
1712 having set something in $address_data, the value of $address_data was not
1713 passed back to the ACL. This was different to the case when no
1714 redirection occurred. The value is now passed back in both cases.
1716 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1717 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1718 use login_cap.h, so on its own it isn't the distinguishing feature. The
1719 new name refers directly to the setclassresources() function.
1721 PH/65 Added configuration files for NetBSD3.
1723 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1725 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1726 is preferred over IPv4.
1728 PH/68 The bounce_return_message and bounce_return_body options were not being
1729 honoured for bounces generated during the reception of non-SMTP messages.
1730 In particular, this applied to messages rejected by the ACL. This bug has
1731 been fixed. However, if bounce_return_message is true and bounce_return_
1732 body is false, the headers that are returned for a non-SMTP message
1733 include only those that have been read before the error was detected.
1734 (In the case of an ACL rejection, they have all been read.)
1736 PH/69 The HTML version of the specification is now built in a directory called
1737 spec_html instead of spec.html, because the latter looks like a path with
1738 a MIME-type, and this confuses some software.
1740 PH/70 Catch two compiler warnings in sieve.c.
1742 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1743 function verify_get_ident() calls ip_connect() to connect a socket, but
1744 if the "connect()" function timed out, ip_connect() used to close the
1745 socket. However, verify_get_ident() also closes the socket later, and in
1746 between Exim writes to the log, which may get opened at this point. When
1747 the socket was closed in ip_connect(), the log could get the same file
1748 descriptor number as the socket. This naturally causes chaos. The fix is
1749 not to close the socket in ip_connect(); the socket should be closed by
1750 the function that creates it. There was only one place in the code where
1751 this was missing, in the iplookup router, which I don't think anybody now
1752 uses, but I've fixed it anyway.
1754 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1755 well as to direct DNS lookups. Otherwise the handling of names in host
1756 lists is inconsistent and therefore confusing.
1762 PH/01 Two changes to the default runtime configuration:
1764 (1) Move the checks for relay_from_hosts and authenticated clients from
1765 after to before the (commented out) DNS black list checks.
1767 (2) Add control=submission to the relay_from_hosts and authenticated
1768 clients checks, on the grounds that messages accepted by these
1769 statements are most likely to be submissions.
1771 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1773 (1) Generate an error if the third argument for the ${prvs expansion is
1776 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1779 (3) Reset the variables that are obtained from the first argument of
1780 ${prvscheck and used in the second argument before leaving the code,
1781 because their memory is reclaimed, so using them afterwards may do
1784 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1785 one (it's much easier than Tom thought :-).
1787 (5) Because of (4), we can now allow for the use of $prvscheck_result
1788 inside the third argument.
1790 PH/03 For some reason, the default setting of PATH when running a command from
1791 a pipe transport was just "/usr/bin". I have changed it to
1794 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1795 anything to be listed in the output from -bV.
1797 PH/05 When a filter generated an autoreply, the entire To: header line was
1798 quoted in the delivery log line, like this:
1800 => >A.N.Other <ano@some.domain> <original@ddress> ...
1802 This has been changed so that it extracts the operative address. There
1803 may be more than one such address. If so, they are comma-separated, like
1806 => >ano@some.domain,ona@other.domain <original@ddress> ...
1808 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1809 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1810 not being looked up in the rDNS to get a host name, Exim was showing the
1811 IP address twice in Received: lines, even though the IP addresses were
1812 identical. For example:
1814 Received: from [1.2.3.4] (helo=[1.2.3.4])
1816 However, if the real host name was known, it was omitting the HELO data
1817 if it matched the actual IP address. This has been tidied up so that it
1818 doesn't show the same IP address twice.
1820 PH/07 When both +timestamp and +memory debugging was on, the value given by
1821 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1822 called by the debug printing, thereby overwriting the timestamp buffer.
1823 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1825 PH/08 When the original message was included in an autoreply transport, it
1826 always said "this is a copy of the message, including all the headers",
1827 even if body_only or headers_only was set. It now gives an appropriate
1830 PH/09 Applied a patch from the Sieve maintainer which:
1832 o fixes some comments
1833 o adds the (disabled) notify extension core
1834 o adds some debug output for the result of if/elsif tests
1835 o points to the current vacation draft in the documentation
1836 and documents the missing references header update
1840 o fixes a bug in processing the envelope test (when testing
1841 multiple envelope elements, the last element determinted the
1844 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1845 Electronic Mail") by including:
1847 Auto-submitted: auto-generated
1849 in the messages that it generates (bounce messages and others, such as
1850 warnings). In the case of bounce messages for non-SMTP mesages, there was
1851 also a typo: it was using "Auto_submitted" (underscore instead of
1852 hyphen). Since every message generated by Exim is necessarily in response
1853 to another message, thes have all been changed to:
1855 Auto-Submitted: auto-replied
1857 in accordance with these statements in the RFC:
1859 The auto-replied keyword:
1861 - SHOULD be used on messages sent in direct response to another
1862 message by an automatic process,
1864 - MUST NOT be used on manually-generated messages,
1866 - MAY be used on Delivery Status Notifications (DSNs) and Message
1867 Disposition Notifications (MDNs),
1869 - MUST NOT be used on messages generated by automatic or periodic
1870 processes, except for messages which are automatic responses to
1873 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1874 to the default Received: header definition.
1876 PH/12 Added log selector acl_warn_skipped (default on).
1878 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1879 variables because (a) they are in the wrong storage pool and (b) even if
1880 they were copied, it wouldn't work properly because of the caching.
1882 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1883 checking when decoding. Apparently there are clients that generate
1884 overlong encoded strings. Why am I not surprised?
1886 PH/15 If the first argument of "${if match_address" was not empty, but did not
1887 contain an "@" character, Exim crashed. Now it writes a panic log message
1888 and treats the condition as false.
1890 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1892 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1893 extension "envelope-auth". The code is finished and in agreement with
1894 other implementations, but there is no documentation so far and in fact,
1895 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1896 not changing the active code.
1898 Print executed "if" and "elsif" statements when debugging is used. This
1899 helps a great deal to understand what a filter does.
1901 Document more things not specified clearly in RFC3028. I had all this
1902 sorted out, when out of a sudden new issues came to my mind. Oops."
1904 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1907 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1908 canonical form (as documented). However, after a host name lookup from
1909 the IP address, check_host() was doing a simple string comparison with
1910 addresses acquired from the DNS when checking that the found name did
1911 have the original IP as one of its addresses. Since any found IPv6
1912 addresses are likely to be in abbreviated form, the comparison could
1913 fail. Luckily, there already exists a function for doing the comparison
1914 by converting both addresses to binary, so now that is used instead of
1915 the text comparison.
1917 PH/20 There was another similar case to PH/19, when a complete host name was
1918 given in a host list; looking up its IP address could give an abbreviated
1919 form, whereas the current host's name might or might not be abbreviated.
1920 The same fix has been applied.
1926 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1927 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1930 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1931 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1933 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1935 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1936 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1937 Also, for "Auto-Submitted", treat anything other than "no" as
1938 non-personal, in accordance with RFC 3834. (Previously it treated
1939 anything starting "auto-" as non-personal.)
1941 TF/01 The control=submission/name=... option had a problem with syntax
1942 errors if the name included a slash character. The /name= option
1943 now slurps the rest of the string, so it can include any characters
1944 but it must come last in the list of options (after /sender_retain
1947 PH/05 Some modifications to the interface to the fake nameserver for the new
1955 TK/01 Added the "success_on_redirect" address verification option. See
1956 NewStuff for rationale and an example.
1958 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1960 PH/02 Patch to exigrep to allow it to work on syslog lines.
1962 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1963 fread() to skip over the body file's header line, because in Cygwin the
1964 header line is locked and is inaccessible.
1966 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1967 co-exist for some time) to make it clear that it is the Exim ID that is
1968 referenced, not the Message-ID: header line.
1970 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1971 string_format() function, because snprintf() does not exist on all
1974 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1975 ${stat: expansion item.
1977 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1978 protocol synchronization error", to keep the pedants happy.
1980 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1981 well as for IRIX systems, when gcc is being used. See the host.c source
1984 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1986 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1989 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1990 transport and to the smtp transport in LMTP mode.
1992 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1994 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1995 run a filter in a subprocess. This could lead to confusion in subsequent
1996 lookups in the parent process. There should also be a search_tidyup() at
1997 the end of the subprocess.
1999 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2000 only if the host matched helo_try_verify_hosts, which caused the
2001 verification to occur when the EHLO/HELO command was issued. The ACL just
2002 tested the remembered result. Now, if a previous verification attempt has
2003 not happened, "verify = helo" does it there and then.
2005 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2007 TK/03 Fix log output including CR from clamd.
2009 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2010 exist provoked a memory error which could cause a segfault.
2012 PH/15 Installed PCRE 6.2
2014 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2016 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2017 of the problem. Specifically, suggested +O2 rather than +O1 for the
2020 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2022 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2023 record, the port was not correctly logged when the outgoing_port log
2024 selector was set (it logged the transort's default port).
2026 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2027 fallback_hosts, and "hosts" in the smtp transport.
2029 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2030 host errors such as "Connection refused".
2032 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2033 authentication with radiusclient 0.4.9:
2035 - Error returned from rc_read_config was caught wrongly
2036 - Username/password not passed on to radius server due to wrong length.
2038 The presumption is that some radiusclient API changes for 4.51/PH/17
2039 were not taken care of correctly. The code is still untested by me (my
2040 Linux distribution still has 0.3.2 of radiusclient), but it was
2041 contributed by a Radius user.
2043 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2044 expanding the "port" option of the smtp transport.
2046 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2047 while reading a MIME header. Thanks to Tom Hughes for a patch.
2049 PH/24 Include config.h inside local_scan.h so that configuration settings are
2052 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2053 that in an ACL for RCPT (for example), you can examine exactly what was
2056 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2057 commands, but it was not correctly comparing the address with the actual
2058 client host address. Thus, it would show the EHLO address in Received:
2059 header lines when this was not necessary.
2061 PH/27 Added the % operator to ${eval:}.
2063 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2064 it should be ignoring failures (because with -C, for example, it has lost
2065 privilege). It wasn't ignoring creation failures other than "already
2068 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2071 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2072 because an input line was too long, either on its own, or by virtue of
2073 too many continuations, the temporary file was not being removed, and the
2074 return code was incorrect.
2076 PH/31 Missing "BOOL" in function definition in filtertest.c.
2078 PH/32 Applied Sieve patches from the maintainer.
2080 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2082 PH/33 Added "verify = not_blind".
2084 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2085 Local/Makefile (with some defaults set). These are used in built scripts
2086 such as exicyclog, but they have never been used in the exim_install
2087 script (though there are many overriding facilities there). I have
2088 arranged that the exim_install script now takes note of these two
2091 PH/35 Installed configuration files for Dragonfly.
2093 PH/36 When a locally submitted message by a trusted user did not contain a
2094 From: header, and the sender address was obtained from -f or from an SMTP
2095 MAIL command, and the trusted user did not use -F to supply a sender
2096 name, $originator_name was incorrectly used when constructing a From:
2097 header. Furthermore, $originator_name was used for submission mode
2098 messages from external hosts without From: headers in a similar way,
2099 which is clearly wrong.
2101 PH/37 Added control=suppress_local_fixups.
2103 PH/38 When log_selector = +received_sender was set, and the addition of the
2104 sender made the log line's construction buffer exactly full, or one byte
2105 less than full, an overflow happened when the terminating "\n" was
2108 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2109 when the result of a list match is failure because a DNS lookup failed.
2111 PH/40 RM_COMMAND is now used in the building process.
2113 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2114 the "build-* directories that it finds.
2116 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2117 address in a domain literal was a prefix of an interface address.
2119 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2120 when verifying a sender address, unless rewrite_headers is false.
2122 PH/44 Wrote a long comment about why errors_to addresses are verified as
2123 recipients, not senders.
2125 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2126 the ratelimit ACL was added.
2128 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2130 PH/46 Added extra information about PostgreSQL errors to the error string.
2132 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2133 avoiding the need to install special zones in a real server. This is
2134 backwards compatible; if it can't find the fake resolver, it drops back.
2135 Thus, both old and new test suites can be run.
2137 TF/02 Added util/ratelimit.pl
2139 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2140 clock is set back in time.
2142 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2143 Candler <B.Candler@pobox.com>.
2145 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2146 OK for addresses that are the result of redirections.
2148 PH/48 A number of further additions for the benefit of the new test suite,
2149 including a fake gethostbyname() that interfaces to the fake DNS resolver
2152 TF/06 The fix for widen_domains has also been applied to qualify_single and
2153 search_parents which are the other dnslookup options that can cause
2156 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2159 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2161 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2162 service" if it encountered an MX record with an empty target hostname.
2163 The message is now "an MX or SRV record indicated no SMTP service".
2165 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2166 if the DNS of the sending site is misconfigured. This is quite a
2167 common situation. This change restores the behaviour of treating a
2168 helo verification defer as a failure.
2170 PH/51 If self=fail was set on a router, the bounce message did not include the
2171 actual error message.
2177 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2179 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2180 command itself ended in error, the underlying message about the transport
2181 filter timeout was being overwritten with the pipe command error. Now the
2182 underlying error message should be appended to the second error message.
2184 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2186 PH/02 Reduce the amount of output that "make" produces by default. Full output
2187 can still be requested.
2189 PH/03 The warning log line about a condition test deferring for a "warn" verb
2190 was being output only once per connection, rather than after each
2191 occurrence (because it was using the same function as for successful
2192 "warn" verbs). This seems wrong, so I have changed it.
2194 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2195 it should not have, which might have caused a crash in the right
2196 circumstances, but probably never did.
2198 PH/04 Installed a modified version of Tony Finch's patch to make submission
2199 mode fix the return path as well as the Sender: header line, and to
2200 add a /name= option so that you can make the user's friendly name appear
2203 TF/03 Added the control = fakedefer ACL modifier.
2205 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2206 Mark Lowes for thorough testing.
2208 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2210 TK/03 Merged latest SRS patch from Miles Wilton.
2212 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2213 with the definition in sysexits.h (which is #included earlier).
2214 Fortunately, Exim does not actually use EX_OK. The code used to try to
2215 preserve the sysexits.h value, by assumimg that macro definitions were
2216 scanned for macro replacements. I have been disabused of this notion,
2217 so now the code just undefines EX_OK before #including unistd.h.
2219 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2220 in the smtp transport. When a block could not be written in a single
2221 write() function, the timeout was being re-applied to each part-write.
2222 This seems wrong - if the receiver was accepting one byte at a time it
2223 would take for ever. The timeout is now adjusted when this happens. It
2224 doesn't have to be particularly precise.
2226 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2227 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2229 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2230 without a domain if the check to <postmaster@domain> fails.
2232 SC/01 Eximstats: added -xls and the ability to specify output files
2233 (patch written by Frank Heydlauf).
2235 SC/02 Eximstats: use FileHandles for outputing results.
2237 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2239 SC/04 Eximstats: fixed display of large numbers with -nvr option
2241 SC/05 Eximstats: fixed merging of reports with empty tables.
2243 SC/06 Eximstats: added the -include_original_destination flag
2245 SC/07 Eximstats: removed tabs and trailing whitespace.
2247 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2249 TK/06 MBOX spool code: Add real "From " MBOX separator line
2250 so the .eml file is really in mbox format (even though
2251 most programs do not really care). Patch from Alex Miller.
2253 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2254 The latter is generated from $received_to and is only set if the
2255 message has one envelope recipient. SA can use these headers,
2256 obviously out-of-the-box. Patch from Alex Miller.
2258 PH/08 The ${def test on a variable was returning false if the variable's
2259 value was "0", contrary to what the specification has always said!
2260 The result should be true unless the variable is empty.
2262 PH/09 The syntax error of a character other than { following "${if
2263 def:variable_name" (after optional whitespace) was not being diagnosed.
2264 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2265 accidental colon was present, for example, could give incorrect results.
2267 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2268 result is used (not including appendfile, where other changes are about
2271 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2272 This involved changing a lot of size variables from int to off_t. It
2273 should work with maildirs and everything.
2275 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2276 spamd dying while we are connected to it.
2278 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2281 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2282 function for BDB 4.3.
2284 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2286 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2287 the sources for PCRE have been re-arranged and more files are now
2290 PH/15 The code I had for printing potentially long long variables in PH/11
2291 above was not the best (it lost precision). The length of off_t variables
2292 is now inspected at build time, and an appropriate printing format (%ld
2293 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2294 to be "long long int" or "long int". This is needed for the internal
2295 formatting function string_vformat().
2297 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2298 the configuration file to be ":syslog", then the script "guesses" where
2299 the logs files are, rather than using the compiled in default. In our
2300 case the guess is not the same as the compiled default, so the script
2301 suddenly stopped working when I started to use syslog. The patch checks
2302 to see if log_file_path is "". If so, it attempts to read it from exim
2303 with no configuration file to get the compiled in version, before it
2304 falls back to the previous guessing code."
2306 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2307 implementing BATV in an Exim configuration. See NewStuff for the gory
2310 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2311 Makefile that are specific to HP-UX.
2313 PH/18 If the "use_postmaster" option was set for a recipient callout together
2314 with the "random" option, the postmaster address was used as the MAIL
2315 FROM address for the random test, but not for the subsequent recipient
2316 test. It is now used for both.
2318 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2319 patch removes a few documentation additions to RFC 3028, because the
2320 latest draft now contains them. It adds the new en;ascii-case comparator
2321 and a new error check for 8bit text in MIME parts. Comparator and
2322 require names are now matched exactly. I enabled the subaddress
2323 extension, but it is not well tested yet (read: it works for me)."
2325 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2326 rework some of the code of TK/09 above to avoid the hardwired use of
2327 "%lld" and "long long". Replaced the call to snprintf() with a call to
2330 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2331 records point to non-existent hosts", "retry timeout exceeded", and
2332 "retry time not reached for any host after a long failure period".
2334 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2335 experimental DomainKeys support:
2337 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2338 (2) On an error such as an illegally used "control", the wrong name for
2339 the control was given.
2341 These problems did NOT occur unless DomainKeys support was compiled.
2343 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2345 PH/24 Added ${if match_ip condition.
2347 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2348 databases so that it will be absolutely obvious if a crash occurs in the
2349 DB library. This is a regular occurrence (often caused by mis-matched
2352 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2353 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2354 picked up on a user's system that detects such things. There doesn't seem
2355 to be a gcc warning option for this - only an attribute that has to be
2356 put on the function's prototype. It seems that in Fedora Core 4 they have
2357 set this on a number of new functions. No doubt there will be more in due
2360 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2361 specify a transport. However, if an address that was verified by such a
2362 router was the subject of a callout, Exim crashed because it tried to
2363 read the rcpt_include_affixes from the non-existent transport. Now it
2364 just assumes that the setting of that option is false. This bug was
2365 introduced by 4.51/PH/31.
2367 PH/28 Changed -d+all to exclude +memory, because that information is very
2368 rarely of interest, but it makes the output a lot bigger. People tend to
2369 do -d+all out of habit.
2371 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2372 code in os-type was giving problems when libc.so lives in lib64, like on
2375 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2376 aren't the modern standard, and it seems that some systems' include files
2377 don't always have them. Exim was already checking for some of the newer
2378 ones like T_AAAA, and defining it itself. I've added checks for all the
2379 record types that Exim uses.
2381 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2382 not automatically generating a new one, as it is supposed to. This
2383 prevented TLS from working. If the file did exist, but contained invalid
2384 data, a new version was generated, as expected. It was only the case of a
2385 non-existent file that was broken.
2387 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2388 with a change in libdomainkeys > 0.64.
2390 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2391 from DNS. If the selector record carries the flag, it now has
2392 precedence over the domain-wide flag.
2394 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2396 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2397 the use of a port name that isn't defined in /etc/services) occurred, the
2398 message was deferred as in a normal delivery, and thus remained on the
2399 spool, instead of being failed because of the mua_wrapper setting. This
2400 is now fixed, and I tidied up some of the mua_wrapper messages at the
2403 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2404 the messages in a hash of arrays rather than using individual hashes.
2405 This is a bit cleaner and results in dramatic memory savings, albeit
2406 at a slight CPU cost.
2408 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2409 as requested by Marc Sherman.
2411 SC/10 Eximstats: added histograms for user specified patterns as requested
2414 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2416 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2417 fopen() in the content-scanning modules that did not already have it.
2423 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2424 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2426 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2428 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2430 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2431 to test Sieve filters that use "vacation".
2433 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2434 that changes the way the GnuTLS parameters are stored in the cache file.
2435 The new format can be generated externally. For backward compatibility,
2436 if the data in the cache doesn't make sense, Exim assumes it has read an
2437 old-format file, and it generates new data and writes a new file. This
2438 means that you can't go back to an older release without removing the
2441 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2442 work if there are any delivery delays because "one_time" forces the
2443 parent to be marked "delivered", so its unseen clone is never tried
2444 again. For this reason, Exim now forbids the simultaneous setting of
2447 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2448 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2449 entry if you want to know the details. The fix, however, neglected to
2450 consider the case when local delivery batching is involved. The test for
2451 "previously delivered" was not happening when checking to see if an
2452 address could be batched with a previous (undelivered) one; under
2453 certain circumstances this could lead to multiple deliveries to the same
2456 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2457 in its include files, and this causes problems building Exim.
2459 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2460 header_syntax) but Exim was just ignoring anything given after a slash.
2461 In particular, this caused confusion with an attempt to use "verify =
2462 reverse_host_lookup/defer_ok". An error is now given when options are
2463 supplied for verify items that do not have them. (Maybe reverse_host_
2464 lookup should have a defer_ok option, but that's a different point.)
2466 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2467 defined by RFC 821) to 2048, because there were problems with some AUTH
2468 commands, and RFC 1869 says the size should be increased for extended
2469 SMTP commands that take arguments.
2471 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2474 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2475 "unknown" error; now it says that the functionality isn't in the binary.
2477 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2478 an address' error message when a string expansion fails (syntax or
2479 whatever). Otherwise the password may appear in the log. Following change
2480 PH/42 below, there is no longer a chance of it appearing in a bounce
2483 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2485 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2486 of its addresses was ever tried. (Bugzilla bug #2.)
2488 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2489 the result incorrectly in the debug output. (It correctly added a newline
2490 to what was transported.)
2492 TF/01 Added $received_time.
2494 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2495 commented out examples of how to interface to a virus scanner and to
2496 SpamAssassin. Also added commented examples of av_scanner and
2497 spamd_address settings.
2499 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2500 and controls are allowed in which ACLs. There were a couple of minor
2501 errors. Some of the entries in the conditions table (which is a table of
2502 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2503 negation of where the condition IS allowed.
2505 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2507 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2508 header file does not have a version number, so I've had to invent a new
2509 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2510 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2511 radiusclient), but it was contributed by a Radius user.
2513 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2514 files or directories, for OpenSSL.
2516 PH/19 When an Exim process that is running as root has to create an Exim log
2517 file, it does so in a subprocess that runs as exim:exim so as to get the
2518 ownership right at creation (otherwise, other Exim processes might see
2519 the file with the wrong ownership). There was no test for failure of this
2520 fork() call, which would lead to the process getting stuck as it waited
2521 for a non-existent subprocess. Forks do occasionally fail when resources
2522 run out. I reviewed all the other calls to fork(); they all seem to check
2525 PH/20 When checking for unexpected SMTP input at connect time (before writing
2526 the banner), Exim was not dealing correctly with a non-positive return
2527 from the read() function. If the client had disconnected by this time,
2528 the result was a log entry for a synchronization error with an empty
2529 string after "input=" when read() returned zero. If read() returned -1
2530 (an event I could not check), uninitialized data bytes were printed.
2531 There were reports of junk text (parts of files, etc) appearing after
2534 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2536 PH/22 Added support for macro redefinition, and (re)definition in between
2537 driver and ACL definitions.
2539 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2540 forgetting to use the resulting value; it was using the unexpanded value.
2542 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2543 hadn't been configured. The fix is from Juergen Kreileder, who
2544 understands it better than I do:
2546 "Here's what I see happening with three configured cyrus_sasl
2547 authenticators configured (plain, login, cram-md5):
2549 On startup auth_cyrus_sasl_init() gets called for each of these.
2550 This means three calls to sasl_listmech() without a specified mech_list.
2551 => SASL tests which mechs of all available mechs actually work
2552 => three warnings about OTP not working
2553 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2555 With the patch, sasl_listmech() also gets called three times. But now
2556 SASL's mech_list option is set to the server_mech specified in the the
2557 authenticator. Or in other words, the answer from sasl_listmech()
2558 gets limited to just the mech you're testing for (which is different
2560 => the return list contains just 'plain' or 'login', 'cram-md5' or
2561 nothing depending on the value of ob->server_mech.
2563 I've just tested the patch: Authentication still works fine,
2564 unavailable mechs specified in the exim configuration are still
2565 caught, and the auth.log warnings about OTP are gone."
2567 PH/25 When debugging is enabled, the contents of the command line are added
2568 to the debugging output, even when log_selector=+arguments is not
2571 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2572 answer is "GNU", and only if the return is "GNU/something" is the answer
2575 PH/27 $acl_verify_message is now set immediately after the failure of a
2576 verification in an ACL, and so is available in subsequent modifiers. In
2577 particular, the message can be preserved by coding like this:
2579 warn !verify = sender
2580 set acl_m0 = $acl_verify_message
2582 Previously, $acl_verify_message was set only while expanding "message"
2583 and "log_message" when a very denied access.
2585 PH/28 Modified OS/os.c-Linux with
2587 -#ifndef OS_LOAD_AVERAGE
2588 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2590 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2591 nomenclature these days.)
2593 PH/29 Installed patch from the Sieve maintainer that adds the options
2594 sieve_useraddress and sieve_subaddress to the redirect router.
2596 PH/30 In these circumstances:
2597 . Two addresses routed to the same list of hosts;
2598 . First host does not offer TLS;
2599 . First host accepts first address;
2600 . First host gives temporary error to second address;
2601 . Second host offers TLS and a TLS session is established;
2602 . Second host accepts second address.
2603 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2604 and peerdn, if requested) that were in fact used only for the second
2607 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2608 attention to any local part prefix or suffix that was matched by the
2609 router that accepted the address. It now behaves in the same way as it
2610 does for delivery: the affixes are removed from the local part unless
2611 rcpt_include_affixes is set on the transport.
2613 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2614 timeout during the DATA phase of an incoming message.
2616 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2617 applied a patch sanctioned by the Sieve maintainer.
2619 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2620 the uid or gid is negative. A case of a negative gid caused this to be
2621 noticed. The fix allows for either to be negative.
2623 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2624 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2627 PH/36 The change PH/12 above was broken. Fixed it.
2629 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2630 the grounds that routing the same address twice would always produce the
2631 same answer. This might have been true once, but it is certainly no
2632 longer true now. Routing a child address may depend on the previous
2633 routing that produced that child. Some complicated redirection strategies
2634 went wrong when messages had multiple recipients, and made Exim's
2635 behaviour dependent on the order in which the addresses were given.
2637 I have moved the duplicate checking until after the routing is complete.
2638 Exim scans the addresses that are assigned to local and remote
2639 transports, and removes any duplicates. This means that more work will be
2640 done, as duplicates will always all be routed, but duplicates are
2641 presumably rare, so I don't expect this is of any significance.
2643 For deliveries to pipes, files, and autoreplies, the duplicate checking
2644 still happens during the routing process, since they are not going to be
2647 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2648 It corrects a timeout issue with spamd. This is Ian's comment: "The
2649 background is that sometimes spamd either never reads data from a
2650 connection it has accepted, or it never writes response data. The exiscan
2651 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2652 blindly assumes that writes won't block so it may never time out."
2654 PH/39 Allow G after quota size as well as K and M.
2656 PH/40 The value set for $authenticated_id in an authenticator may not contain
2657 binary zeroes or newlines because the value is written to log lines and
2658 to spool files. There was no check on this. Now the value is run through
2659 the string_printing() function so that such characters are converted to
2660 printable escape sequences.
2662 PH/41 $message_linecount is a new variable that contains the total number of
2663 lines in the message. Compare $body_linecount, which is the count for the
2666 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2667 bounce and delay warning messages, except in certain special cases, which
2670 (a) An SMTP error message from a remote host;
2671 (b) A message specified in a :fail: redirection;
2672 (c) A message specified in a "fail" command in a system filter;
2673 (d) A message specified in a FAIL return from the queryprogram router;
2674 (e) A message specified by the cannot_route_message router option.
2676 In these cases only, Exim does include the error details in bounce and
2677 warning messages. There are also a few cases where bland messages such
2678 as "unrouteable address" or "local delivery error" are given.
2680 PH/43 $value is now also set for the "else" part of a ${run expansion.
2682 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2683 being worked on, but at least Exim now implements the latest version to
2686 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2687 process to complete was treated as a delivery failure, a timeout while
2688 writing the message to the pipe was logged, but erroneously treated as a
2689 successful delivery. Such timeouts include transport filter timeouts. For
2690 consistency with the overall process timeout, these timeouts are now
2691 treated as errors, giving rise to delivery failures by default. However,
2692 there is now a new Boolean option for the pipe transport called
2693 timeout_defer, which, if set TRUE, converts the failures into defers for
2694 both kinds of timeout. A transport filter timeout is now identified in
2697 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2698 systems where "make" and "gmake" are different, calling "gmake" at top
2699 level broke things. I've arranged for the value of $(MAKE) to be passed
2700 from the Makefile to this script so that it can call the same version of
2704 A note about Exim versions 4.44 and 4.50
2705 ----------------------------------------
2707 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2708 changes of various kinds. As a consequence, a big documentation update was
2709 needed. This delayed the release for rather longer than seemed good, especially
2710 in the light of a couple of (minor) security issues. Therefore, the changes
2711 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2712 release. So 4.44 and 4.50 are in effect two different branches that both start
2715 I have left the 4.50 change log unchanged; it contains all the changes since
2716 4.43. The change log for 4.44 is below; many of its items are identical to
2717 those for 4.50. This seems to be the most sensible way to preserve the
2718 historical information.
2724 1. Minor wording change to the doc/README.SIEVE file.
2726 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2727 computation of the current number of files was incorrect.
2729 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2730 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2733 4. Give more explanation in the error message when the command for a transport
2734 filter fails to execute.
2736 5. There are several places where Exim runs a non-Exim command in a
2737 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2738 was being done only for the command run by the queryprogram router. It is
2739 now done for all such subprocesses. The other cases are: ${run, transport
2740 filters, and the commands run by the lmtp and pipe transports.
2742 6. Added CONFIGURE_GROUP build-time option.
2744 7. Some older OS have a limit of 256 on the maximum number of file
2745 descriptors. Exim was using setrlimit() to set 1000 as a large value
2746 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2747 systems. I've change it so that if it can't get 1000, it tries for 256.
2749 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2750 was an oversight, and furthermore, ever since the addition of extra
2751 controls (e.g. 4.43/32), the checks on when to allow different forms of
2752 "control" were broken. There should now be diagnostics for all cases when a
2753 control that does not make sense is encountered.
2755 9. Added the /retain_sender option to "control=submission".
2757 10. $recipients is now available in the predata ACL (oversight).
2759 11. Tidy the search cache before the fork to do a delivery from a message
2760 received from the command line. Otherwise the child will trigger a lookup
2761 failure and thereby defer the delivery if it tries to use (for example) a
2762 cached ldap connection that the parent has called unbind on.
2764 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2765 of $address_data from the recipient verification was clobbered by the
2766 sender verification.
2768 13. The value of address_data from a sender verification is now available in
2769 $sender_address_data in subsequent conditions in the ACL statement.
2771 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2773 15. Added a new option "connect=<time>" to callout options, to set a different
2776 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2777 was its contents. (It was OK if the option was not defined at all.)
2779 17. A "Completed" log line is now written for messages that are removed from
2780 the spool by the -Mrm option.
2782 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2783 information about exactly what failed.
2785 19. Added -dd to debug only the daemon process.
2787 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2788 handles timeouts, both on the server side and network timeouts. Renamed the
2789 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2791 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2792 It is now set to "smtps".
2794 22. $host_address is now set to the target address during the checking of
2795 ignore_target_hosts.
2797 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2798 being passed; this would have caused $sender_host_name to have been used if
2799 matching the list had actually called for a host name (not very likely,
2800 since this list is usually IP addresses). A host name is now passed as
2803 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2804 code that specifies a non-restarting handler (typically sigaction() in
2805 modern systems) in an attempt to fix a rare and obscure crash bug.
2807 25. Narrowed the window for a race in the daemon that could cause it to ignore
2808 SIGCHLD signals. This is not a major problem, because they are used only to
2809 wake it up if nothing else does.
2811 26. A malformed maildirsize file could cause Exim to calculate negative values
2812 for the mailbox size or file count. Odd effects could occur as a result.
2813 The maildirsize information is now recalculated if the size or filecount
2816 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2817 support for a long time. Removed HAVE_SYS_VFS_H.
2819 28. Installed the latest version of exipick from John Jetmore.
2821 29. In an address list, if the pattern was not a regular expression, an empty
2822 subject address (from a bounce message) matched only if the pattern was an
2823 empty string. Non-empty patterns were not even tested. This was the wrong
2824 because it is perfectly reasonable to use an empty address as part of a
2825 database query. An empty address is now tested by patterns that are
2826 lookups. However, all the other forms of pattern expect the subject to
2827 contain a local part and a domain, and therefore, for them, an empty
2828 address still always fails if the pattern is not itself empty.
2830 30. Exim went into a mad DNS loop when attempting to do a callout where the
2831 host was specified on an smtp transport, and looking it up yielded more
2832 than one IP address.
2834 31. Re-factored the code for checking spool and log partition space into a
2835 function that finds that data and another that does the check. The former
2836 is then used to implement four new variables: $spool_space, $log_space,
2837 $spool_inodes, and $log_inodes.
2839 32. The RFC2047 encoding function was originally intended for short strings
2840 such as real names; it was not keeping to the 75-character limit for
2841 encoded words that the RFC imposes. It now respects the limit, and
2842 generates multiple encoded words if necessary. To be on the safe side, I
2843 have increased the buffer size for the ${rfc2047: expansion operator from
2846 33. It is now permitted to omit both strings after an "if" condition; if the
2847 condition is true, the result is "true". As before, when the second string
2848 is omitted, a false condition yields an empty string. This makes it less
2849 cumbersome to write custom ACL and router conditions.
2851 34. Failure to deliver a bounce message always caused it to be frozen, even if
2852 there was an errors_to setting on the router. The errors_to setting is now
2855 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2856 canonical form (fully expanded) before being placed in
2857 $sender_host_address.
2859 36. The table in the code that translates DNS record types into text (T_A to
2860 "A" for instance) was missing entries for NS and CNAME. It is just possible
2861 that this could have caused confusion if both these types were looked up
2862 for the same domain, because the text type is used as part of Exim's
2863 per-process caching. But the chance of anyone hitting this buglet seems
2866 37. The dnsdb lookup has been extended in a number of ways.
2868 (1) There is a new type, "zns", which walks up the domain tree until it
2869 finds some nameserver records. It should be used with care.
2871 (2) There is a new type, "mxh", which is like "mx" except that it returns
2872 just the host names, not the priorities.
2874 (3) It is now possible to give a list of domains (or IP addresses) to be
2875 looked up. The behaviour when one of the lookups defers can be
2876 controlled by a keyword.
2878 (4) It is now possible to specify the separator character for use when
2879 multiple records are returned.
2881 38. The dnslists ACL condition has been extended: it is now possible to supply
2882 a list of IP addresses and/or domains to be looked up in a particular DNS
2885 39. Added log_selector=+queue_time_overall.
2887 40. When running the queue in the test harness, wait just a tad after forking a
2888 delivery process, to get repeatability of debugging output.
2890 41. Include certificate and key file names in error message when GnuTLS fails
2891 to set them up, because the GnuTLS error message doesn't include the name
2892 of the failing file when there is a problem reading it.
2894 42. Allow both -bf and -bF in the same test run.
2896 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2898 44. The "Exiscan patch" is now merged into the mainline Exim source.
2900 45. Sometimes the final signoff response after QUIT could fail to get
2901 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2902 < 0 before doing a fflush(). This bug looks as though it goes back to the
2903 introduction of TLS in release 3.20, but "sometimes" must have been rare
2904 because the tests only now provoked it.
2906 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2907 (this can affect the format of dates).
2909 47. exim_tidydb, when checking for the continued existence of a message for
2910 which it has found a message-specific retry record, was not finding
2911 messages that were in split spool directories. Consequently, it was
2912 deleting retry records that should have stayed in existence.
2914 48. Steve fixed some bugs in eximstats.
2916 49. The SPA authentication driver was not abandoning authentication and moving
2917 on to the next authenticator when an expansion was forced to fail,
2918 contradicting the general specification for all authenticators. Instead it
2919 was generating a temporary error. It now behaves as specified.
2921 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2922 (the order specifies the preference for clients). The order is now AES256,
2923 AES128, 3DES, ARCFOUR128.
2925 51. Small patch to Sieve code - explicitly set From: when generating an
2928 52. Exim crashed if a remote delivery caused a very long error message to be
2929 recorded - for instance if somebody sent an entire SpamAssassin report back
2930 as a large number of 550 error lines. This bug was coincidentally fixed by
2931 increasing the size of one of Exim's internal buffers (big_buffer) that
2932 happened as part of the Exiscan merge. However, to be on the safe side, I
2933 have made the code more robust (and fixed the comments that describe what
2936 53. Now that there can be additional text after "Completed" in log lines (if
2937 the queue_time_overall log selector is set), a one-byte patch to exigrep
2938 was needed to allow it to recognize "Completed" as not the last thing in
2941 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2942 patch that reportedly fixes this has been added. I am not expert enough to
2943 create a test for it. This is what the patch creator wrote:
2945 "I found a little strange behaviour of ldap code when working with
2946 Windows 2003 AD Domain, where users was placed in more than one
2947 Organization Units. When I tried to give exim partial DN, the exit code
2948 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2949 But simultaneously result of request was absolutely normal ldap result,
2950 so I produce this patch..."
2952 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2953 REFERENCE, so I have modified the code to exclude the patch when that macro
2956 55. Some experimental protocols are using DNS PTR records for new purposes. The
2957 keys for these records are domain names, not reversed IP addresses. The
2958 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2959 leaves it alone. Component reversal etc. now happens only for IP addresses.
2962 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2964 57. Double the size of the debug message buffer (to 2048) so that more of very
2965 long debug lines gets shown.
2967 58. The exicyclog utility now does better if the number of log files to keep
2968 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2970 59. Two changes related to the smtp_active_hostname option:
2972 (1) $smtp_active_hostname is now available as a variable.
2973 (2) The default for smtp_banner uses $smtp_active_hostname instead
2974 of $primary_hostname.
2976 60. The host_aton() function is supposed to be passed a string that is known
2977 to be a valid IP address. However, in the case of IPv6 addresses, it was
2978 not checking this. This is a hostage to fortune. Exim now panics and dies
2979 if the condition is not met. A case was found where this could be provoked
2980 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2981 components; fortuitously, this particular loophole had already been fixed
2982 by change 4.50/55 above.
2984 If there are any other similar loopholes, the new check in host_aton()
2985 itself should stop them being exploited. The report I received stated that
2986 data on the command line could provoke the exploit when Exim was running as
2987 exim, but did not say which command line option was involved. All I could
2988 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2989 running as the user.
2992 61. There was a buffer overflow vulnerability in the SPA authentication code
2993 (which came originally from the Samba project). I have added a test to the
2994 spa_base64_to_bits() function which I hope fixes it.
2997 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2998 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3000 63. The daemon start-up calls getloadavg() while still root for those OS that
3001 need the first call to be done as root, but it missed one case: when
3002 deliver_queue_load_max is set with deliver_drop_privilege. This is
3003 necessary for the benefit of the queue runner, because there is no re-exec
3004 when deliver_drop_privilege is set.
3006 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3007 This has been fixed.
3009 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3010 was in use, was not putting the data itself into the right store pool;
3011 consequently, it could be overwritten for a subsequent message in the same
3012 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3015 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3017 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3018 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3019 the function were treating the return as a boolean value, which happened to
3020 work because 0=false and not-0=true, but is not correct code.
3022 68. The host_aton() function was not handling scoped IPv6 addresses (those
3023 with, for example, "%eth0" on the end) correctly.
3025 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3026 negated items (that is, ~something) in unsigned ints. Some compilers
3027 apparently mutter when there is no cast.
3029 70. If an address verification called from an ACL failed, and did not produce a
3030 user-specific message (i.e. there was only a "system" message), nothing was
3031 put in $acl_verify_message. In this situation, it now puts the system
3034 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3035 session; change 4.31/43 added the unwanted input to the log line - except
3036 that it did not do this in the start of session case. It now does.
3038 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3039 This is wrong and can cause the other end to generate a synchronization
3040 error if it is another Exim or anything else that does the synchronization
3041 check. A QUIT command is no longer sent after a timeout.
3043 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3044 during host lookups.
3046 74. The defer_ok option of callout verification was not working if it was used
3047 when verifying addresses in header lines, that is, for this case:
3049 verify = header_sender/callout=defer_ok
3051 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3052 those file descriptors could be used for SMTP connections. If anything
3053 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3054 could be sent to the SMTP client, causing chaos. The daemon now opens
3055 stdin, stdout, and stderr to /dev/null when it puts itself into the
3058 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3059 log by default. The user can override this with suitable Perl magic.
3061 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3062 the log message when discard triggers, was not working for the DATA ACL or
3063 for the non-SMTP ACL.
3065 78. Error message wording change in sieve.c.
3067 79. If smtp_accept_max_per_host was set, the number of connections could be
3068 restricted to fewer than expected, because the daemon was trying to set up
3069 a new connection before checking whether the processes handling previous
3070 connections had finished. The check for completed processes is now done
3071 earlier. On busy systems, this bug wouldn't be noticed because something
3072 else would have woken the daemon, and it would have reaped the completed
3075 80. If a message was submitted locally by a user whose login name contained one
3076 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3077 It caused a spool format error. I have fixed the spool reading code. A
3078 related problem was that the "from" clause in the Received: line became
3079 illegal because of the space(s). It is now covered by ${quote_local_part.
3081 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3084 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3086 83. Patch from Sieve maintainer to fix unterminated string problem in
3087 "vacation" handling.
3089 84. Some minor changes to the Linux configuration files to help with other
3090 OS variants using glibc.
3092 85. One more patch for Sieve to update vacation handling to latest spec.
3095 ----------------------------------------------------
3096 See the note above about the 4.44 and 4.50 releases.
3097 ----------------------------------------------------
3103 1. Change 4.43/35 introduced a bug that caused file counts to be
3104 incorrectly computed when quota_filecount was set in an appendfile
3107 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3108 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3111 3. Give more explanation in the error message when the command for a transport
3112 filter fails to execute.
3114 4. There are several places where Exim runs a non-Exim command in a
3115 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3116 was being done only for the command run by the queryprogram router. It is
3117 now done for all such subprocesses. The other cases are: ${run, transport
3118 filters, and the commands run by the lmtp and pipe transports.
3120 5. Some older OS have a limit of 256 on the maximum number of file
3121 descriptors. Exim was using setrlimit() to set 1000 as a large value
3122 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3123 systems. I've change it so that if it can't get 1000, it tries for 256.
3125 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3126 was an oversight, and furthermore, ever since the addition of extra
3127 controls (e.g. 4.43/32), the checks on when to allow different forms of
3128 "control" were broken. There should now be diagnostics for all cases when a
3129 control that does not make sense is encountered.
3131 7. $recipients is now available in the predata ACL (oversight).
3133 8. Tidy the search cache before the fork to do a delivery from a message
3134 received from the command line. Otherwise the child will trigger a lookup
3135 failure and thereby defer the delivery if it tries to use (for example) a
3136 cached ldap connection that the parent has called unbind on.
3138 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3139 of $address_data from the recipient verification was clobbered by the
3140 sender verification.
3142 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3143 was its contents. (It was OK if the option was not defined at all.)
3145 11. A "Completed" log line is now written for messages that are removed from
3146 the spool by the -Mrm option.
3148 12. $host_address is now set to the target address during the checking of
3149 ignore_target_hosts.
3151 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3152 being passed; this would have caused $sender_host_name to have been used if
3153 matching the list had actually called for a host name (not very likely,
3154 since this list is usually IP addresses). A host name is now passed as
3157 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3158 code that specifies a non-restarting handler (typically sigaction() in
3159 modern systems) in an attempt to fix a rare and obscure crash bug.
3161 15. Narrowed the window for a race in the daemon that could cause it to ignore
3162 SIGCHLD signals. This is not a major problem, because they are used only to
3163 wake it up if nothing else does.
3165 16. A malformed maildirsize file could cause Exim to calculate negative values
3166 for the mailbox size or file count. Odd effects could occur as a result.
3167 The maildirsize information is now recalculated if the size or filecount
3170 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3171 support for a long time. Removed HAVE_SYS_VFS_H.
3173 18. Updated exipick to current release from John Jetmore.
3175 19. Allow an empty sender to be matched against a lookup in an address list.
3176 Previously the only cases considered were a regular expression, or an
3179 20. Exim went into a mad DNS lookup loop when doing a callout where the
3180 host was specified on the transport, if the DNS lookup yielded more than
3183 21. The RFC2047 encoding function was originally intended for short strings
3184 such as real names; it was not keeping to the 75-character limit for
3185 encoded words that the RFC imposes. It now respects the limit, and
3186 generates multiple encoded words if necessary. To be on the safe side, I
3187 have increased the buffer size for the ${rfc2047: expansion operator from
3190 22. Failure to deliver a bounce message always caused it to be frozen, even if
3191 there was an errors_to setting on the router. The errors_to setting is now
3194 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3195 canonical form (fully expanded) before being placed in
3196 $sender_host_address.
3198 24. Updated eximstats to version 1.33
3200 25. Include certificate and key file names in error message when GnuTLS fails
3201 to set them up, because the GnuTLS error message doesn't include the name
3202 of the failing file when there is a problem reading it.
3204 26. Expand error message when OpenSSL has problems setting up cert/key files.
3207 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3208 (this can affect the format of dates).
3210 28. exim_tidydb, when checking for the continued existence of a message for
3211 which it has found a message-specific retry record, was not finding
3212 messages that were in split spool directories. Consequently, it was
3213 deleting retry records that should have stayed in existence.
3215 29. eximstats updated to version 1.35
3216 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3217 1.35 - bugfix such that pie charts by volume are generated correctly
3219 30. The SPA authentication driver was not abandoning authentication and moving
3220 on to the next authenticator when an expansion was forced to fail,
3221 contradicting the general specification for all authenticators. Instead it
3222 was generating a temporary error. It now behaves as specified.
3224 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3225 (the order specifies the preference for clients). The order is now AES256,
3226 AES128, 3DES, ARCFOUR128.
3228 31. Small patch to Sieve code - explicitly set From: when generating an
3231 32. Exim crashed if a remote delivery caused a very long error message to be
3232 recorded - for instance if somebody sent an entire SpamAssassin report back
3233 as a large number of 550 error lines. This bug was coincidentally fixed by
3234 increasing the size of one of Exim's internal buffers (big_buffer) that
3235 happened as part of the Exiscan merge. However, to be on the safe side, I
3236 have made the code more robust (and fixed the comments that describe what
3239 33. Some experimental protocols are using DNS PTR records for new purposes. The
3240 keys for these records are domain names, not reversed IP addresses. The
3241 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3242 leaves it alone. Component reversal etc. now happens only for IP addresses.
3245 34. The host_aton() function is supposed to be passed a string that is known
3246 to be a valid IP address. However, in the case of IPv6 addresses, it was
3247 not checking this. This is a hostage to fortune. Exim now panics and dies
3248 if the condition is not met. A case was found where this could be provoked
3249 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3250 components; fortuitously, this particular loophole had already been fixed
3251 by change 4.50/55 or 4.44/33 above.
3253 If there are any other similar loopholes, the new check in host_aton()
3254 itself should stop them being exploited. The report I received stated that
3255 data on the command line could provoke the exploit when Exim was running as
3256 exim, but did not say which command line option was involved. All I could
3257 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3258 running as the user.
3261 35. There was a buffer overflow vulnerability in the SPA authentication code
3262 (which came originally from the Samba project). I have added a test to the
3263 spa_base64_to_bits() function which I hope fixes it.
3266 36. The daemon start-up calls getloadavg() while still root for those OS that
3267 need the first call to be done as root, but it missed one case: when
3268 deliver_queue_load_max is set with deliver_drop_privilege. This is
3269 necessary for the benefit of the queue runner, because there is no re-exec
3270 when deliver_drop_privilege is set.
3272 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3273 was in use, was not putting the data itself into the right store pool;
3274 consequently, it could be overwritten for a subsequent message in the same
3275 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3278 38. Sometimes the final signoff response after QUIT could fail to get
3279 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3280 < 0 before doing a fflush(). This bug looks as though it goes back to the
3281 introduction of TLS in release 3.20, but "sometimes" must have been rare
3282 because the tests only now provoked it.
3288 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3289 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3290 Now it returns an integer. A number of calls were still expecting a T/F
3291 return. Fortuitously, in all cases, the tests worked in OK situations,
3292 which is the norm. However, things would have gone wrong on any write
3293 failures on the smtp file descriptor. This function is used when sending
3294 messages over SMTP and also when doing verify callouts.
3296 2. When Exim is called to do synchronous delivery of a locally submitted
3297 message (the -odf or -odi options), it no longer closes stderr before doing
3300 3. Implemented the mua_wrapper option.
3302 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3304 5. Implemented the functions header_remove(), header_testname(),
3305 header_add_at_position(), and receive_remove_recipient(), and exported them
3308 6. If an ACL "warn" statement specified the addition of headers, Exim already
3309 inserted X-ACL-Warn: at the start if there was no header name. However, it
3310 was not making this test for the second and subsequent header lines if
3311 there were newlines in the string. This meant that an invalid header could
3312 be inserted if Exim was badly configured.
3314 7. Allow an ACL "warn" statement to add header lines at the start or after all
3315 the Received: headers, as well as at the end.
3317 8. Added the rcpt_4xx retry error code.
3319 9. Added postmaster_mailfrom=xxx to callout verification option.
3321 10. Added mailfrom=xxxx to the callout verification option, for verify=
3324 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3325 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3327 12. Inserted some casts to stop certain compilers warning when using pointer
3328 differences as field lengths or precisions in printf-type calls (mostly
3329 affecting debugging statements).
3331 13. Added optional readline() support for -be (dynamically loaded).
3333 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3334 same clock tick as a message's arrival, so that its received time was the
3335 same as the "first fail" time on the retry record, and that message
3336 remained on the queue past the ultimate address timeout, every queue runner
3337 would try a delivery (because it was past the ultimate address timeout) but
3338 after another failure, the ultimate address timeout, which should have then
3339 bounced the address, did not kick in. This was a "< instead of <=" error;
3340 in most cases the first failure would have been in the next clock tick
3341 after the received time, and all would be well.
3343 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3344 being recognized when the domain list was tested by the match_domain
3345 condition in an expansion string.
3347 16. Added the ${str2b64: operator.
3349 17. Exim was always calling setrlimit() to set a large limit for the number of
3350 processes, without checking whether the existing limit was already
3351 adequate. (It did check for the limit on file descriptors.) Furthermore,
3352 errors from getrlimit() and setrlimit() were being ignored. Now they are
3353 logged to the main and panic logs, but Exim does carry on, to try to do its
3354 job under whatever limits there are.
3356 18. Imported PCRE 5.0.
3358 19. Trivial typo in log message " temporarily refused connection" (the leading
3361 20. If the log selector return_path_on_delivery was set and an address was
3362 redirected to /dev/null, the delivery process crashed because it assumed
3363 that a return path would always be set for a "successful" delivery. In this
3364 case, the whole delivery is bypassed as an optimization, and therefore no
3367 21. Internal re-arrangement: the function for sending a challenge and reading
3368 a response while authentication was assuming a zero-terminated challenge
3369 string. It's now changed to take a pointer and a length, to allow for
3370 binary data in such strings.
3372 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3374 23. Exim was not respecting finduser_retries when seeking the login of the
3375 uid under which it was called; it was always trying 10 times. (The default
3376 setting of finduser_retries is zero.) Also, it was sleeping after the final
3377 failure, which is pointless.
3379 24. Implemented tls_on_connect_ports.
3381 25. Implemented acl_smtp_predata.
3383 26. If the domain in control=submission is set empty, Exim assumes that the
3384 authenticated id is a complete email address when it generates From: or
3385 Sender: header lines.
3387 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3388 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3389 chown and chgrp in /bin and hostname in /usr/bin.
3391 28. Exim was keeping the "process log" file open after each use, just as it
3392 does for the main log. This opens the possibility of it remaining open for
3393 long periods when the USR1 signal hits a daemon. Occasional processlog
3394 errors were reported, that could have been caused by this. Anyway, it seems
3395 much more sensible not to leave this file open at all, so that is what now
3398 29. The long-running daemon process does not normally write to the log once it
3399 has entered its main loop, and it closes the log before doing so. This is
3400 so that log files can straightforwardly be renamed and moved. However,
3401 there are a couple of unusual error situations where the daemon does write
3402 log entries, and I had neglected to close the log afterwards.
3404 30. The text of an SMTP error response that was received during a remote
3405 delivery was being truncated at 512 bytes. This is too short for some of
3406 the long messages that one sometimes sees. I've increased the limit to
3409 31. It is now possible to make retry rules that apply only when a message has a
3410 specific sender, in particular, an empty sender.
3412 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3413 it possible to be selective about when SMTP synchronization is enforced.
3415 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3417 32. Implemented hosts_connection_nolog.
3419 33. Added an ACL for QUIT.
3421 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3424 35. Added mailbox_size and mailbox_filecount to appendfile.
3426 36. Added control = no_multiline_responses to ACLs.
3428 37. There was a bug in the logic of the code that waits for the clock to tick
3429 in the case where the clock went backwards by a substantial amount such
3430 that the microsecond fraction of "now" was more than the microsecond
3431 fraction of "then" (but the whole seconds number was less).
3433 38. Added support for the libradius Radius client library this is found on
3434 FreeBSD (previously only the radiusclient library was supported).
3440 1. When certain lookups returned multiple values in the form name=value, the
3441 quoting of the values was not always being done properly. Specifically:
3442 (a) If the value started with a double quote, but contained no whitespace,
3444 (b) If the value contained whitespace other than a space character (i.e.
3445 tabs or newlines or carriage returns) it was not quoted.
3446 This fix has been applied to the mysql and pgsql lookups by writing a
3447 separate quoting function and calling it from the lookup code. The fix
3448 should probably also be applied to nisplus, ibase and oracle lookups, but
3449 since I cannot test any of those, I have not disturbed their existing code.
3451 2. A hit in the callout cache for a specific address caused a log line with no
3452 reason for rejecting RCPT. Now it says "Previous (cached) callout
3453 verification failure".
3455 3. There was an off-by-one bug in the queryprogram router. An over-long
3456 return line was truncated at 256 instead of 255 characters, thereby
3457 overflowing its buffer with the terminating zero. As well as fixing this, I
3458 have increased the buffer size to 1024 (and made a note to document this).
3460 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3461 when Exim is waiting for an SMTP response from a remote server, Exim
3462 restarts its select() call on the socket, thereby resetting its timeout.
3463 This is not a problem when such interrupts are rare. Somebody set up a cron
3464 job to run exiwhat every 2 minutes, which is less than the normal select()
3465 timeout (5 or 10 minutes). This meant that the select() timeout never
3466 kicked in because it was always reset. I have fixed this by comparing the
3467 time when an interrupt arrives with the time at the start of the first call
3468 to select(). If more time than the timeout has elapsed, the interrupt is
3469 treated as a timeout.
3471 5. Some internal re-factoring in preparation for the addition of Sieve
3472 extensions (by MH). In particular, the "personal" test is moved to a
3473 separate function, and given an option for scanning Cc: and Bcc: (which is
3474 not set for Exim filters).
3476 6. When Exim created an email address using the login of the caller as the
3477 local part (e.g. when creating a From: or Sender: header line), it was not
3478 quoting the local part when it contained special characters such as @.
3480 7. Installed new OpenBSD configuration files.
3482 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3483 try to make them clearer.
3485 9. Callout options, other than the timeout value, were being ignored when
3486 verifying sender addresses in header lines. For example, when using
3488 verify = header_sender/callout=no_cache
3490 the cache was (incorrectly) being used.
3492 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3493 only the Cygwin environment.
3495 11. When return_path_on_delivery was set as a log selector, if different remote
3496 addresses in the same message used different return paths and parallel
3497 remote delivery occurred, the wrong values would sometimes be logged.
3498 (Whenever a remote delivery process finished, the return path value from
3499 the most recently started remote delivery process was logged.)
3501 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3502 lines when AUTH and/or TLS are in use. This is the "received protocol"
3503 field. Exim used to use "asmtp" for authenticated SMTP, without any
3504 indication (in the protocol name) for TLS use. Now it follows the RFC and
3505 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3506 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3507 names appear in log lines as well as in Received: header lines.
3509 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3510 extensions, and comparison tests, and to fix some bugs.
3512 14. Changes to the "personal" filter test:
3514 (1) The test was buggy in that it was just doing the equivalent of
3515 "contains" tests on header lines. For example, if a user's address was
3516 anne@some.where, the "personal" test would incorrectly be true for
3518 To: susanne@some.where
3520 This test is now done by extracting each address from the header in turn,
3521 and checking the entire address. Other tests that are part of "personal"
3522 are now done using regular expressions (for example, to check local parts
3523 of addresses in From: header lines).
3525 (2) The list of non-personal local parts in From: addresses has been
3526 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3527 taken from the Sieve specification recommendations.
3529 (3) If the message contains any header line starting with "List-" it is
3530 treated as non-personal.
3532 (4) The test for "circular" in the Subject: header line has been removed
3533 because it now seems ill-conceived.
3535 15. Minor typos in src/EDITME comments corrected.
3537 16. Installed latest exipick from John Jetmore.
3539 17. If headers_add on a router specified a text string that was too long for
3540 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3541 of string_sprintf() is now avoided.
3543 18. $message_body_size was not set (it was always zero) when running the DATA
3544 ACL and the local_scan() function.
3546 19. For the "mail" command in an Exim filter, no default was being set for
3547 the once_repeat time, causing a random time value to be used if "once" was
3548 specified. (If the value happened to be <= 0, no repeat happened.) The
3549 default is now 0s, meaning "never repeat". The "vacation" command was OK
3550 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3551 (I found it when inspecting the code).
3553 20. There is now an overall timeout for performing a callout verification. It
3554 defaults to 4 times the callout timeout, which applies to individual SMTP
3555 commands during the callout. The overall timeout applies when there is more
3556 than one host that can be tried. The timeout is checked before trying the
3557 next host. This prevents very long delays if there are a large number of
3558 hosts and all are timing out (e.g. when the network connections are timing
3559 out). The value of the overall timeout can be changed by specifying an
3560 additional sub-option for "callout", called "maxwait". For example:
3562 verify = sender/callout=5s,maxwait=20s
3564 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3565 to the end before writing, but this should make it even safer).
3567 22. Exim was forgetting that it had advertised PIPELINING for the second and
3568 subsequent messages on an SMTP connection. It was also not resetting its
3569 memory on STARTTLS and an internal HELO.
3571 23. When Exim logs an SMTP synchronization error within a session, it now
3572 records whether PIPELINING has been advertised or not.
3574 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3575 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3576 rather than long int.
3578 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3580 26. Added the never_mail option to autoreply.
3586 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3587 crash if the getsockname() call failed; this can happen if a connection is
3588 closed very soon after it is established. The problem was simply in the
3589 order in which certain operations were done, causing Exim to try to write
3590 to the SMTP stream before it had set up the file descriptor. The bug has
3591 been fixed by making things happen in the correct order.
3597 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3598 before the connection was closed, thus losing the rejection response.
3600 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3601 some early Solaris releases, but causes trouble in current releases where
3602 socklen_t is defined.
3604 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3607 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3610 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3611 that releasing the top of it at the end releases what was used for sub-
3612 expansions (unless the block got too big). However, discard this block if
3613 the first thing is a variable or header, so that we can use its block when
3614 it is dynamic (useful for very large $message_headers, for example).
3616 6. Lookups now cache *every* query, not just the most recent. A new, separate
3617 store pool is used for this. It can be recovered when all lookup caches are
3618 flushed. Lookups now release memory at the end of their result strings.
3619 This has involved some general refactoring of the lookup sources.
3621 7. Some code has been added to the store_xxx() functions to reduce the amount
3622 of flapping under certain conditions.
3624 8. log_incoming_interface used to affect only the <= reception log lines. Now
3625 it causes the local interface and port to be added to several more SMTP log
3626 lines, for example "SMTP connection from", and rejection lines.
3628 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3630 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3632 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3633 could be overwritten at the end of the current message (or the start of a
3634 new message if it was set in a HELO ACL). The value is now preserved for
3635 the duration of the SMTP connection.
3637 12. If a transport had a headers_rewrite setting, and a matching header line
3638 contained an unqualified address, that address was qualified, even if it
3639 did not match any rewriting rules. The underlying bug was that the values
3640 of the flags that permit the existence of unqualified sender and recipient
3641 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3642 non-local messages, and by -bnq for local messages) were not being
3643 preserved with the message after it was received.
3645 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3646 "next input=" as part of the text comprising the host identity instead of
3647 the correct text. The code was using the same buffer for two different
3648 strings. However, depending on which order the printing function evaluated
3649 its arguments, the bug did not always show up. Under Linux, for example, my
3650 test suite worked just fine.
3652 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3653 doesn't work with some older versions of Perl. It has been changed to "my",
3654 which in any case is probably the better facility to use.
3656 15. A really picky compiler found some instances of statements for creating
3657 error messages that either had too many or two few arguments for the format
3660 16. The size of the buffer for calls to the DNS resolver has been increased
3661 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3662 for addresses that have a lot of PTR records. This alleviates a problem; it
3663 does not fully solve it.
3665 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3666 buffer now truncates the list and logs the incident, which is the same
3667 action as happens when Exim is looking up a host name and its aliases.
3668 Previously in this situation something unpredictable would happen;
3669 sometimes it was "internal error: store_reset failed".
3671 18. If a server dropped the connection unexpectedly when an Exim client was
3672 using GnuTLS and trying to read a response, the client delivery process
3673 crashed while trying to generate an error log message.
3675 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3676 string, for example:
3678 warn message = H1: something\nH2: something
3680 the text was added as a single header line from Exim's point of view
3681 though it ended up OK in the delivered message. However, searching for the
3682 second and subsequent header lines using $h_h2: did not work. This has been
3683 fixed. Similarly, if a system filter added multiple headers in this way,
3684 the routers could not see them.
3686 20. Expanded the error message when iplsearch is called with an invalid key to
3687 suggest using net-iplsearch in a host list.
3689 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3690 ACLs are no longer actually imposed (and a message to that effect is
3693 22. If a "gecos" field in a passwd entry contained escaped characters, in
3694 particular, if it contained a \" sequence, Exim got it wrong when building
3695 a From: or a Sender: header from that name. A second bug also caused
3696 incorrect handling when an unquoted " was present following a character
3697 that needed quoting.
3699 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3700 was not being matched caselessly.
3702 24. Arranged for all hyphens in the exim.8 source to be escaped with
3705 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3706 properly. Recipient callout cache records were still being keyed to include
3707 the sender, even when use_sender was set false. This led to far more
3708 callouts that were necessary. The sender is no longer included in the key
3709 when use_sender is false.
3711 26. Added "control = submission" modifier to ACLs.
3713 27. Added the ${base62d: operator to decode base 62 numbers.
3715 28. dnsdb lookups can now access SRV records.
3717 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3718 the configuration file.
3720 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3721 (-v) mode. This makes the output for a verbose queue run more intelligible.
3723 31. Added a use_postmaster feature to recipient callouts.
3725 32. Added the $body_zerocount variable, containing the number of binary zero
3726 bytes in the message body.
3728 33. The time of last modification of the "new" subdirectory is now used as the
3729 "mailbox time last read" when there is a quota error for a maildir
3732 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3734 35. Added +ignore_unknown as a special item in host lists.
3736 36. Code for decoding IPv6 addresses in host lists is now included, even if
3737 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3738 address was recognized as an IP address, but was then not correctly decoded
3739 into binary, causing unexpected and incorrect effects when compared with
3746 1. Very minor rewording of debugging text in manualroute to say "list of
3747 hosts" instead of "hostlist".
3749 2. If verify=header_syntax was set, and a header line with an unqualified
3750 address (no domain) and a large number of spaces between the end of the
3751 name and the colon was received, the reception process suffered a buffer
3752 overflow, and (when I tested it) crashed. This was caused by some obsolete
3753 code that should have been removed. The fix is to remove it!
3755 3. When running in the test harness, delay a bit after writing a bounce
3756 message to get a bit more predictability in the log output.
3758 4. Added a call to search_tidyup() just before forking a reception process. In
3759 theory, someone could use a lookup in the expansion of smtp_accept_max_
3760 per_host which, without the tidyup, could leave open a database connection.
3762 5. Added the variables $recipient_data and $sender_data which get set from a
3763 lookup success in an ACL "recipients" or "senders" condition, or a router
3764 "senders" option, similar to $domain_data and $local_part_data.
3766 6. Moved the writing of debug_print from before to after the "senders" test
3769 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3770 problems for message scanning, either using a data ACL, or using
3771 local_scan() because the Received: header was not generated till after they
3772 were called (in order to set the time as the time of reception completion).
3773 I have revised the way this works. The header is now generated after the
3774 body is received, but before the ACL or local_scan() are called. After they
3775 are run, the timestamp in the header is updated.
3781 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3782 before starting a queue runner without re-exec. This happened only when
3783 deliver_drop_privilege was set or when the Exim user was set to root. The
3784 effect of the bug was that timeouts during subsequent deliveries caused
3785 crashes instead of being properly handled. The handler is now left at its
3786 default (and expected) setting.
3788 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3789 message, again when deliver_drop_privilege is set or Exim is run as root.
3790 The bug described in (1) was not present in this case, but the tidying up
3791 of the other signals was missing. I have made the two cases consistent.
3793 3. The ignore_target_hosts setting on a manualroute router was being ignored
3794 for hosts that were looked up using the /MX notation.
3796 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3799 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3800 operated on the sender address. After changing the $sender_address to <>
3801 for the sender address verify, Exim was re-instated it as the original
3802 (before rewriting) address, but remembering that it had rewritten it, so it
3803 wasn't rewriting it again. This bug also had the effect of breaking the
3804 sender address verification caching when the sender address was rewritten.
3806 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3807 This has been changed so that if the ip literal address matches
3808 ignore_target_hosts, the router declines.
3810 7. Added expansion conditions match_domain, match_address, and match_local_
3811 part (NOT match_host).
3813 8. The placeholder for the Received: header didn't have a length field set.
3815 9. Added code to Exim itself and to exim_lock to test for a specific race
3816 condition that could lead to file corruption when using MBX delivery. The
3817 issue is with the lockfile that is created in /tmp. If this file is removed
3818 after a process has opened it but before that process has acquired a lock,
3819 there is the potential for a second process to recreate the file and also
3820 acquire a lock. This could lead to two Exim processes writing to the file
3821 at the same time. The added code performs the same test as UW imapd; it
3822 checks after acquiring the lock that its file descriptor still refers to
3823 the same named file.
3825 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3826 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3829 11. Added the smtp_active_hostname option. If used, this will typically be made
3830 to depend on the incoming interface address. Because $interface_address is
3831 not set up until the daemon has forked a reception process, error responses
3832 that can happen earlier (such as "too many connections") no longer contain
3835 12. If an expansion in a condition on a "warn" statement fails because a lookup
3836 defers, the "warn" statement is abandoned, and the next ACL statement is
3837 processed. Previously this caused the whole ACL to be aborted.
3839 13. Added the iplsearch lookup type.
3841 14. Added ident_timeout as a log selector.
3843 15. Added tls_certificate_verified as a log selector.
3845 16. Added a global option tls_require_ciphers (compare the smtp transport
3846 option of the same name). This controls incoming TLS connections.
3848 17. I finally figured out how to make tls_require_ciphers do a similar thing
3849 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3850 before starting the TLS session.
3852 18. Tabs are now shown as \t in -bP output.
3854 19. If the log selector return_path_on_delivery was set, Exim crashed when
3855 bouncing a message because it had too many Received: header lines.
3857 20. If two routers both had headers_remove settings, and the first one included
3858 a superfluous trailing colon, the final name in the first list and the
3859 first name in the second list were incorrectly joined into one item (with a
3860 colon in the middle).
3866 1. Added -C and -D options to the exinext utility, mainly to make it easier
3867 to include in the automated testing, but these could be helpful when
3868 multiple configurations are in use.
3870 2. The exinext utility was not formatting the output nicely when there was
3871 an alternate port involved in the retry record key, nor when there was a
3872 message id as well (for retries that were specific to a specific message
3873 and a specific host). It was also confused by IPv6 addresses, because of
3874 the additional colons they contain. I have fixed the IPv4 problem, and
3875 patched it up to do a reasonable job for IPv6.
3877 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3878 delivery, the log line now contains "pipelined" if PIPELINING was used.
3880 4. An SMTP transport process used to panic and die if the bind() call to set
3881 an explicit outgoing interface failed. This has been changed; it is now
3882 treated in the same way as a connect() failure.
3884 5. A reference to $sender_host_name in the part of a conditional expansion
3885 that was being skipped was still causing a DNS lookup. This no longer
3888 6. The def: expansion condition was not recognizing references to header lines
3889 that used bh_ and bheader_.
3891 7. Added the _cache feature to named lists.
3893 8. The code for checking quota_filecount in the appendfile transport was
3894 allowing one more file than it should have been.
3896 9. For compatibility with Sendmail, the command line option
3904 and sets the incoming protocol and host name (for trusted callers). The
3905 host name and its colon can be omitted when only the protocol is to be set.
3906 Note the Exim already has two private options, -pd and -ps, that refer to
3907 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3908 "s", but I don't think that's a major issue.
3910 10. A number of refactoring changes to the code, none of which should affect
3913 (a) The number of logging options was getting close to filling up the
3914 32-bit word that was used as a bit map. I have split them into two classes:
3915 those that are passed in the argument to log_write(), and those that are
3916 only ever tested independently outside of that function. These are now in
3917 separate 32-bit words, so there is plenty of room for expansion again.
3918 There is no change in the user interface or the logging behaviour.
3920 (b) When building, for example, log lines, the code previously used a
3921 macro that called string_cat() twice, in order to add two strings. This is
3922 not really sufficiently general. Furthermore, there was one instance where
3923 it was actually wrong because one of the argument was used twice, and in
3924 one call a function was used. (As it happened, calling the function twice
3925 did not affect the overall behaviour.) The macro has been replaced by a
3926 function that can join an arbitrary number of extra strings onto a growing
3929 (c) The code for expansion conditions now uses a table and a binary chop
3930 instead of a serial search (which was left over from when there were very
3931 few conditions). Also, it now recognizes conditions like "pam" even when
3932 the relevant support is not compiled in: a suitably worded error message is
3933 given if an attempt is made to use such a condition.
3935 11. Added ${time_interval:xxxxx}.
3937 12. A bug was causing one of the ddress fields not to be passed back correctly
3938 from remote delivery subprocesses. The field in question was not being
3939 subsequently used, so this caused to problems in practice.
3941 13. Added new log selectors queue_time and deliver_time.
3943 14. Might have fixed a bug in maildirsizefile handling that threw up
3944 "unexpected character" debug warnings, and recalculated the data
3945 unnecessarily. In any case, I expanded the warning message to give more
3948 15. Added the message "Restricted characters in address" to the statements in
3949 the default ACL that block characters like @ and % in local parts.
3951 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3952 Three changes have been made:
3954 (a) There was a serious bug; a negative response to MAIL caused the whole
3955 recipient domain to be cached as invalid, thereby blocking all messages
3956 to all local parts at the same domain, from all senders. This bug has
3957 been fixed. The domain is no longer cached after a negative response to
3958 MAIL if the sender used is not empty.
3960 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3963 (c) A new callout option, "use_sender" has been added for people who want
3964 the modified behaviour.
3970 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3973 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3974 indeed breaks things for older releases.
3976 3. Added additional logging to the case where there is a problem reading data
3977 from a filter that is running in a subprocess using a pipe, in order to
3978 try to track down a specific problem.
3980 4. Testing facility fudge: when running in the test harness and attempting
3981 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3982 getting "No route to host". Convert this to a timeout.
3984 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3987 6. Some OS don't have socklen_t but use size_t instead. This affects the
3988 fifth argument of getsockopt() amongst other things. This is now
3989 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3990 can be set for individual OS. I have set it for SunOS5, OSF1, and
3991 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3992 some earlier ones do not.
3994 7. Change 4.30/15 was not doing the test caselessly.
3996 8. The standard form for an IPv6 address literal was being rejected by address
3997 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3998 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3999 this, as well as the form without the "IPv6" on the front (but only when
4000 address literals are enabled, of course).
4002 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4004 10. Exim crashed if a message with an empty sender address specified by -f
4005 encountered a router with an errors_to setting. This could be provoked only
4006 by a command such as
4010 where an empty string was supplied; "<>" did not hit this bug.
4012 11. Installed PCRE release 4.5.
4014 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4015 remained set. It is now erased.
4017 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4018 times from message ids (which are base 36 rather than the normal 62).
4020 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4021 were being counted as actual protocol errors, and logged if the log
4022 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4023 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4024 and DATA following a set of rejected RCPTs do not count as protocol errors.
4025 In other words, Exim assumes they were pipelined, though this may not
4026 actually be the case. Of course, in all cases the client gets an
4027 appropriate error code.
4029 15. If a lookup fails in an ACL condition, a message about the failure may
4030 be available; it is used if testing the ACL cannot continue, because most
4031 such messages specify what the cause of the deferral is. However, some
4032 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4033 that caused an old message to be retained and used if a later statement
4034 caused a defer, replacing the real cause of the deferral.
4036 16. If an IP address had so many PTR records that the DNS lookup buffer
4037 was not large enough to hold them, Exim could crash while trying to process
4038 the truncated data. It now detects and logs this case.
4040 17. Further to 4.21/58, another change has been made: if (and only if) the
4041 first line of a message (the first header line) ends with CRLF, a bare LF
4042 in a subsequent header line has a space inserted after it, so as not to
4043 terminate the header.
4045 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4046 unnecessarily, used atoi() instead of strtol(), and didn't check the
4047 termination when getting file sizes from file names by regex.
4049 19. Completely re-implemented the support for maildirsize files, in the light
4050 of a number of problems with the previous contributed implementation
4051 (4.30/29). In particular:
4053 . If the quota is zero, the maildirsize file is maintained, but no quota is
4056 . If the maildir directory does not exist, it is created before any attempt
4057 to write a maildirsize file.
4059 . The quota value in the file is just a cache; if the quota is changed in
4060 the transport, the new value overrides.
4062 . A regular expression is available for excluding directories from the
4065 20. The autoreply transport checks the characters in options that define the
4066 message's headers; it allows continued headers, but it was checking with
4067 isspace() after an embedded newline instead of explicitly looking for a
4070 21. If all the "regular" hosts to which an address was routed had passed their
4071 expiry times, and had not reached their retry times, the address was
4072 bounced, even if fallback hosts were defined. Now Exim should go on to try
4075 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4076 equivalent code in the SMTP transport. Some hosts send humungous responses
4077 to HELO/EHLO, more than 1024 it seems.
4079 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4080 gives compiler warnings in some environments. I've now done it "properly",
4083 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4084 (because of problems with the built-in one) was declared to return uschar *
4085 instead of char *, causing compiler failure.
4087 25. Fixed a file descriptor leak when processing alias/forward files.
4089 26. Fixed a minor format string issue in dbfn.c.
4091 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4093 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4094 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4097 29. When a sender address is verified, it is cached, to save repeating the test
4098 when there is more than one recipient in a message. However, when the
4099 verification involves a callout, it is possible for different callout
4100 options to be set for different recipients. It is too complicated to keep
4101 track of this in the cache, so now Exim always runs a verification when a
4102 callout is required, relying on the callout cache for the optimization.
4103 The overhead is duplication of the address routing, but this should not be
4106 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4107 to be verified with callout=postmaster, and the main callout worked but the
4108 postmaster check failed, the verification correctly failed. However, if a
4109 subsequent RCPT command asked for sender verification *without* the
4110 postmaster check, incorrect caching caused this verification also to fail,
4113 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4114 it was not caching the DNS options (qualify_single, search_parents) that
4115 were used when the lookup failed. A subsequent lookup with different
4116 options therefore always gave the same answer, though there were cases
4117 where it should not have. (Example: a "domains = !$mx_any" option on a
4118 dnslookup router: the "domains" option is always processed without any
4119 widening, but the router might have qualify_single set.) Now Exim uses the
4120 cached value only when the same options are set.
4122 32. Added John Jetmore's "exipick" utility to the distribution.
4124 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4125 than a timeout (e.g. a certificate is required, and is not provided), an
4126 Exim server now closes the connection immediately. Previously it waited for
4127 the client to close - but if the client is SSL, it seems that they each
4128 wait for each other, leading to a delay before one of them times out.
4130 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4131 maintained 0.8.x compatibility because I don't think many are using it, and
4132 it is clearly obsolete.
4134 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4137 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4138 client certificate was expired. A simple patch fixes this, though I don't
4139 understand the full logic of why the verify callback is called multiple
4142 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4143 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4144 which causes problems with some clients (such as the Certicom SSL Plus
4145 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4146 disables the coutermeasure allowing Eudora to connect."
4148 38. Exim was not checking that a write() to a log file succeeded. This could
4149 lead to Bad Things if a log got too big, in particular if it hit a file
4150 size limit. Exim now panics and dies if it cannot write to a log file, just
4151 as it does if it cannot open a log file.
4153 39. Modified OS/Makefile-Linux so that it now contains
4155 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4157 The two -D definitions ensure that Exim is compiled with large file
4158 support, which makes it possible to handle log files that are bigger than
4161 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4162 instance) a domain was checked against a named list that involved a lookup,
4163 causing $domain_data to be set, then another domain was checked against the
4164 same list, then the first domain was re-checked, the value of $domain_data
4165 after the final check could be wrong. In particular, if the second check
4166 failed, it could be set empty. This bug probably also applied to
4169 41. The strip_trailing_dot option was not being applied to the address given
4170 with the -f command-line option.
4172 42. The code for reading a message's header from the spool was incrementing
4173 $received_count, but never initializing it. This meant that the value was
4174 incorrect (doubled) while delivering a message in the same process in which
4175 it was received. In the most common configuration of Exim, this never
4176 happens - a fresh exec is done - but it can happen when
4177 deliver_drop_privilege is set.
4179 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4180 it now includes up to 150 characters of the unexpected data in the log
4183 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4184 and building data strings. The size of both of these buffers was 10 000
4185 bytes - far larger than anybody would *ever* want, thought I. Needless to
4186 say, somebody hit the limit. I have increased the maximum line length to
4187 20 000 and the maximum data length of concatenated lines to 100 000. I have
4188 also fixed two bugs, because there was no checking on these buffers. Tsk,
4189 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4190 buffer is too small.
4192 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4193 lsearch lookups. Now it does.
4195 46. When parsing a route_list item in a manualroute router, a fixed-length
4196 buffer was used for the list of hosts. I made this 1024 bytes long,
4197 thinking that nobody would ever have a list of hosts that long. Wrong.
4198 Somebody had a whole pile of complicated expansion conditions, and the
4199 string was silently truncated, leading to an expansion error. It turns out
4200 that it is easier to change to an unlimited length (owing to other changes
4201 that have happened since this code was originally written) than to build
4202 structure for giving a limitation error. The length of the item that
4203 expands into the list of hosts is now unlimited.
4205 47. The lsearch lookup could not handle data where the length of text line was
4206 more than 4095 characters. Such lines were truncated, leading to shortened
4207 data being returned. It should now handle lines of any length.
4209 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4210 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4213 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4214 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4215 no longer happens in comment lines. A list of replacements is now placed
4216 at the head of all of the source files, except those whose only change is
4217 to replace PERL_COMMAND in the very first #! line.
4219 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4220 messages on the queue, with a bottom-up merge sort, using code contributed
4221 by Michael Haardt. This should make operations like -bp somewhat faster on
4222 large queues. It won't affect queue runners, except when queue_run_in_order
4225 51. Installed eximstats 1.31 in the distribution.
4227 52. Added support for SRV lookups to the dnslookup router.
4229 53. If an ACL referred to $message_body or $message_body_end, the value was not
4230 reset for any messages that followed in the same SMTP session.
4232 54. The store-handling optimization for building very long strings was not
4233 differentiating between the different store pools. I don't think this
4234 actually made any difference in practice, but I've tidied it.
4236 55. While running the routers to verify a sender address, $sender_address
4237 was still set to the sender address. This is wrong, because when routing to
4238 send a bounce to the sender, it would be empty. Therefore, I have changed
4239 it so that, while verifying a sender address, $sender_address is set to <>.
4240 (There is no change to what happens when verifying a recipient address.)
4242 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4243 target A or AAAA records (if not already returned) without resetting the
4244 qualify_single or search_parents options of the DNS resolver. These are
4245 inappropriate in this case because the targets of MX and SRV records must
4246 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4247 target that, when qualified, matched something in the local domain. These
4248 two options are now turned off when doing these lookups.
4250 57. It seems that at least some releases of Reiserfs (which does not have the
4251 concept of a fixed number of inodes) returns zero and not -1 for the
4252 number of available inodes. This interacted badly with check_spool_inodes,
4253 which assumed that -1 was the "no such thing" setting. What I have done is
4254 to check that the total number of inodes is greater than zero before doing
4255 the test of how many are available.
4257 58. When a "warn" ACL statement has a log_message modifier, the message is
4258 remembered, and not repeated. This is to avoid a lot of repetition when a
4259 message has many recipients that cause the same warning to be written.
4260 Howewer, Exim was preserving the list of already written lines for an
4261 entire SMTP session, which doesn't seem right. The memory is now reset if a
4262 new message is started.
4264 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4265 debugging output unless log_rewrite was also set.
4267 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4268 of (int)(handle) when we know that handle contains (void *)(-1).
4270 61. The Exim daemon panic-logs an error return when it closes the incoming
4271 connection. However "connection reset by peer" seems to be common, and
4272 isn't really an error worthy of noting specially, so that particular error
4275 62. When Exim is trying to find all the local interfaces, it used to panic and
4276 die if the ioctl to get the interface flags failed. However, it seems that
4277 on at least one OS (Solaris 9) it is possible to have an interface that is
4278 included in the list of interfaces, but for which you get a failure error
4279 for this call. This happens when the interface is not "plumbed" into a
4280 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4281 failure of the "get flags" call assumes that the interface is down.
4283 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4284 makes life easier for people who are doing arithmetic on fields extracted
4285 from dates, where you often get leading zeros that should not be
4286 interpreted as octal.
4288 64. Added qualify_domain to the redirect router, to override the global
4291 65. If a pathologically long header line contained very many addresses (the
4292 report of this problem mentioned 10 000) and each of them was rewritten,
4293 Exim could use up a very large amount of memory. (It kept on making new
4294 copies of the header line as it rewrote, and never released the old ones.)
4295 At the expense of a bit more processing, the header rewriting function has
4296 been changed so that it no longer eats memory in this way.
4298 66. The generation of the Received: header has been moved from the time that a
4299 message starts to be received, to the time that it finishes. The timestamp
4300 in the Received: header should now be very close to that of the <= log
4301 line. There are two side-effects of this change:
4303 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4304 logged header lines no longer include the local Received: line, because
4305 it has not yet been created. The same applies to a copy of the message
4306 that is returned to a non-SMTP sender when a message is rejected.
4308 (b) When a filter file is tested using -bf, no additional Received: header
4309 is added to the test message. After some thought, I decided that this
4312 This change does not affect the value of $received_for. It is still set
4313 after address rewriting, but before local_scan() is called.
4315 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4317 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4318 gave an unhelpful panic error message, and a defer error. I have managed to
4319 change this behaviour so that it now rejects any supplied certificate,
4320 which seems right, as the list of acceptable certificates is empty.
4322 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4323 gave an unhelpful defer error. I have not managed to make this reject any
4324 supplied certificates, but the error message it gives is "no certificate
4325 supplied", which is not helpful.
4327 70. exigrep's output now also includes lines that are not associated with any
4328 message, but which match the given pattern. Implemented by a patch from
4329 Martin Sluka, which also tidied up the Perl a bit.
4331 71. Recipient callout verification, like sender verification, was using <> in
4332 the MAIL FROM command. This isn't really the right thing, since the actual
4333 sender may affect whether the remote host accepts the recipient or not. I
4334 have changed it to use the actual sender in the callout; this means that
4335 the cache record is now keyed on a recipient/sender pair, not just the
4336 recipient address. There doesn't seem to be a real danger of callout loops,
4337 since a callout by the remote host to check the sender would use <>.
4338 [SEE ABOVE: changed after hitting problems.]
4340 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4341 temporary errors. However, in the case of such a code being given after
4342 the end of a data transmission (i.e. after ".") Exim was failing to write
4343 a retry record for the message. (Yes, there was some broken host that was
4344 actually sending 8xx at this point.)
4346 73. An unknown lookup type in a host list could cause Exim to panic-die when
4347 the list was checked. (An example that provoked this was putting <; in the
4348 middle of a list instead of at the start.) If this happened during a DATA
4349 ACL check, a -D file could be left lying around. This kind of configuration
4350 error no longer causes Exim to die; instead it causes a defer errror. The
4351 incident is still logged to the main and panic logs.
4353 74. Buglet left over from Exim 3 conversion. The message "too many messages
4354 in one connection" was written to the rejectlog but not the mainlog, except
4355 when address rewriting (yes!) was being logged.
4357 75. Added write_rejectlog option.
4359 76. When a system filter was run not as root (that is, when system_filter_user
4360 was set), the values of the $n variables were not being returned to the
4361 main process; thus, they were not subsequently available in the $sn
4364 77. Added +return_path_on_delivery log selector.
4366 78. A connection timeout was being treated differently from recipients deferred
4367 when testing hosts_max_try with a message that was older than the host's
4368 retry timeout. (The host should not be counted, thus allowing all hosts to
4369 be tried at least once before bouncing.) This may have been the cause of an
4370 occasionally reported bug whereby a message would remain on the queue
4371 longer than the retry timeout, but would be bounced if a delivery was
4372 forced. I say "may" because I never totally pinned down the problem;
4373 setting up timeout/retry tests is difficult. See also the next item.
4375 79. The ultimate address timeout was not being applied to errors that involved
4376 a combination of host plus message (for example, a timeout on a MAIL
4377 command). When an address resolved to a number of possible hosts, and they
4378 were not all tried for each delivery (e.g. because of hosts_max_try), a
4379 message could remain on the queue longer than the retry timeout.
4381 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4384 81. Fixed an obscure SMTP outgoing bug which required at least the following
4385 conditions: (a) there was another message waiting for the same server;
4386 (b) the server returned 5xx to all RCPT commands in the first message so
4387 that the message was not completed; (c) the server dropped the connection
4388 or gave a negative response to the RSET that Exim sends to abort the
4389 transaction. The observed case was a dropped connection after DATA that had
4390 been sent in pipelining mode. That is, the server had advertised PIPELINING
4391 but was not implementing it correctly. The effect of the bug was incorrect
4392 behaviour, such as trying another host, and this could lead to a crash.
4398 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4399 and daemon.c were passed as pointers to ints; they should have been
4400 pointers to socklen_t variables (which are typically unsigned ints).
4402 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4405 3. Fixed a really odd bug that affected only the testing scheme; patching a
4406 certain fixed string in the binary changed the value of another string that
4407 happened to be identical to the end of the original first string.
4409 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4410 name", it returns that address as the IP address. On some operating
4411 systems (e.g. Solaris), it also passes back the IP address string as the
4412 "host name". However, on others (e.g. Linux), it passes back an empty
4413 string. Exim wasn't checking for this, and was changing the host name to an
4414 empty string, assuming it had been canonicized.
4416 5. Although rare, it is permitted to have more than one PTR record for a given
4417 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4418 all the names associated with an address, because they do in Solaris.
4419 However, it seems that they do not in Linux for data that comes from the
4420 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4421 I found this out when I moved to a new Linux workstation and tried to run
4422 the Exim test suite.
4424 To get round this problem I have changed the code so that it now does its
4425 own call to the DNS to look up PTR records when searching for a host name.
4426 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4427 addresses that are only in /etc/hosts are still found.
4429 This behaviour is, however, controlled by an option called host_lookup_
4430 order, which defaults to "bydns:byaddr". If people want to use the other
4431 order, or indeed, just use one or the other means of lookup, they can
4432 specify it in this variable.
4434 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4435 some operating systems, this comes back from gethostbyaddr() as an empty
4436 string, and this is what Exim used to test for. However, it seems that in
4437 other systems, "." is yielded. Exim now tests for this case too.
4439 7. The values of check_spool_space and check_log_space are now held internally
4440 as a number of kilobytes instead of an absolute number of bytes. If a
4441 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4442 kilobyte. This means that much larger values can be stored.
4444 8. Exim monitor: an attempt to get the action menu when not actually pointing
4445 at a message produces an empty menu entitled "No message selected". This
4446 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4447 no entries in it ("Shell widget menu has zero width and/or height"). So I
4448 have added a single, blank menu entry in this case.
4450 9. Added ${quote_local_part.
4452 10. MIME decoding is now applied to the contents of Subject: header lines when
4455 11. Now that a reference to $sender_host_address automatically causes a reverse
4456 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4457 host lookup before query-style lookups in lists that might use this
4458 variable. This has therefore been abolished, and the "net-" prefix is no
4459 longer necessary for query-style lookups.
4461 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4462 have been a typo for LFLAGS, so it has been changed.
4464 13. The install script calls Exim with "-C /dev/null" in order to find the
4465 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4466 to be output. Howeve, since Exim outputs its version number before the
4467 error, it didn't break the script. It just looked ugly. I fixed this by
4468 always allowing "-C /dev/null" if the caller is root.
4470 14. Ignore overlarge ACL variable number when reading spool file - insurance
4471 against a later release with more variables having written the file.
4473 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4474 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4475 this, as well as the form without the "IPv6" on the front.
4477 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4478 OS/Makefile-Darwin file.
4480 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4481 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4483 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4484 variable. However, some OS use ints for inodes. I've added cast to long int
4485 to get rid of the compiler warning.
4487 19. I had forgotten to lock out "/../" in configuration file names when
4488 ALT_CONFIG_PREFIX was set.
4490 20. Routers used for verification do not need to specify transports. However,
4491 if such a router generated a host list, and callout was configured, Exim
4492 crashed, because it could not find a port number from the (non-existent)
4493 transport. It now assumes port 25 in this circumstance.
4495 21. Added the -t option to exigrep.
4497 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4498 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4499 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4500 them is set, LOOKUP_LSEARCH is forced.
4502 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4504 24. Added sender and host information to the "rejected by local_scan()" log
4505 line; previously there was no indication of these.
4507 25. Added .include_if_exists.
4509 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4510 Linux. It turns out that not all file systems support this. Apparently some
4511 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4512 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4513 not supported on the file descriptor, is now ignored when Exim is trying to
4514 sync a directory. This applies only to Linux.
4516 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4518 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4519 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4522 29. Added support for maildirsize files from supplied patch (modified a bit).
4524 30. The use of :fail: followed by an empty string could lead Exim to respond to
4525 sender verification failures with (e.g.):
4527 550 Verification failed for <xxx>
4528 550 Sender verify failed
4530 where the first response line was missing the '-' that indicates it is not
4531 the final line of the response.
4533 31. The loop for finding the name of the user that called Exim had a hardwired
4534 limit of 10; it now uses the value of finduser_retries, which is used for
4535 all other user lookups.
4537 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4540 33. Exim was neglecting to zero errno before one call of strtol() when
4541 expanding a string and expecting an integer value. On some systems this
4542 resulted in spurious "integer overflow" errors. Also, it was casting the
4543 result into an int without checking.
4545 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4546 did not work. The code looks as if it has *never* worked, though it appears
4547 to have been documented since at least releast 1.62. I have made it work.
4549 35. The "timeout_DNS" error in retry rules, also documented since at least
4550 1.62, also never worked. As it isn't clear exactly what this means, and
4551 clearly it isn't a major issue, I have abolished the feature by treating it
4552 as "timeout", and writing a warning to the main and panic logs.
4554 36. The display of retry rules for -brt wasn't always showing the error code
4557 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4558 timeout_connect_A, timeout_connect_MX.
4560 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4561 to the empty sender.
4563 39. The daemon was not analysing the content of -oX till after it had closed
4564 stderr and disconnected from the controlling terminal. This meant that any
4565 syntax errors were only noted on the panic log, and the return code from
4566 the command was 0. By re-arranging the code a little, I've made the
4567 decoding happen first, so such errors now appear on stderr, and the return
4568 code is 1. However, the actual setting up of the sockets still happens in
4569 the disconnected process, so errors there are still only recorded on the
4572 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4573 connections (as happens on some IP stacks) was logged at start up time as
4574 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4575 it from "IPv6 and IPv4", which means that two separate sockets are being
4578 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4579 says whether AF_INET or AF_INET6 was passed as an argument.
4581 42. Exiwhat output was messed up when time zones were included in log
4584 43. Exiwhat now gives more information about the daemon's listening ports,
4585 and whether -tls-on-connect was used.
4587 44. The "port" option of the smtp transport is now expanded.
4589 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4590 silently ignored. Now an error message is written to the main and panic
4593 46. There's a new ACL modifier called "logwrite" which writes to a log file
4594 as soon as it is encountered.
4596 47. Added $local_user_uid and $local_user_gid at routing time.
4598 48. Exim crashed when trying to verify a sender address that was being
4601 49. Exim was recognizing only a space character after ".include". It now also
4602 recognizes a tab character.
4604 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4605 extracting the relevant information from the specification. The man page no
4606 longer contains scrambled data for the -d option, and I've added a section
4607 at the front about calling Exim under different names.
4609 51. Added "extra_headers" argument to the "mail" command in filter files.
4611 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4614 53. Installed eximstats 1.29.
4616 54. Added transport_filter_timeout as a generic transport option.
4618 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4619 Cc: header lines. This was required by RFC 822, but it not required by RFC
4622 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4623 incoming messages that did not have them. Now it does so only if the
4624 message originates locally, that is, if there is no associated remote host
4625 address. When Resent- header lines are present, this applies to the Resent-
4626 lines rather than the non-Resent- lines.
4628 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4629 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4631 58. Messages for configuration errors now include the name of the main
4632 configuration file - useful now that there may be more than one file in a
4633 list (.included file names were always shown).
4635 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4636 for those rare installations that do not start the daemon as root or run it
4637 setuid root. I've cut out the call to initgroups() if the daemon is not
4640 60. The Exim user and group can now be bound into the binary as text strings
4641 that are looked up at the start of Exim's processing.
4643 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4645 62. Added $mailstore_basename variable.
4647 63. Installed patch to sieve.c from Michael Haardt.
4649 64. When Exim failed to open the panic log after failing to open the main log,
4650 the original message it was trying to log was written to stderr and debug
4651 output, but if they were not available (the usual case in production), it
4652 was lost. Now it is written to syslog before the two lines that record the
4653 failures to open the logs.
4655 65. Users' Exim filters run in subprocesses under the user's uid. It is
4656 possible for a "deliver" command or an alias in a "personal" command to
4657 provoke an address rewrite. If logging of address rewriting is configured,
4658 this fails because the process is not running as root or exim. There may be
4659 a better way of dealing with this, but for the moment (because 4.30 needs
4660 to be released), I have disabled address rewrite logging when running a
4661 filter in a non-root, non-exim process.
4667 1. The buildconfig auxiliary program wasn't quoting the value set for
4668 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4669 not defined. This bug was masked in 4.22 by the effect that was fixed in
4672 2. Some messages that were rejected after a message id was allocated were
4673 shown as "incomplete" by exigrep. It no longer does this for messages that
4674 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4676 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4677 have allow_domain_literals set, the ID did not get logged in the <= line.
4678 Domain literals are now always recognized in Message-ID: header lines.
4680 4. The first argument for a ${extract expansion item is the key name or field
4681 number. Leading and trailing spaces in this item were not being ignored,
4682 causing some misleading effects.
4684 5. When deliver_drop_privilege was set, single queue runner processes started
4685 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4686 same command in the process it spins off) were not dropping privilege.
4688 6. When the daemon running as "exim" started a queue runner, it always
4689 re-executed Exim in the spun-off process. This is a waste of effort when
4690 deliver_drop_privilege is set. The new process now just calls the
4691 queue-runner function directly.
4697 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4700 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4701 ignored. Though the use of -oP was forcing the writing of a pid file, it
4702 was always written to the default place.
4704 3. If the message "no IP address found for host xxxx" is generated during
4705 incoming verification, it is now followed by identification of the incoming
4706 connection (so you can more easily find what provoked it).
4708 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4710 5. Added some features to "harden" Exim a bit more against certain attacks:
4712 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4713 be put in Local/Makefile. This is like the never_users runtime option,
4714 but it cannot be overridden. The default setting is "root".
4716 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4717 prefix string with which any file named in a -C command line option
4720 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4721 is retained for -C and -D only if the caller of Exim is root. Without
4722 it, the exim user may also use -C and -D and retain privilege.
4724 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4725 command line option is disabled.
4727 6. Macro names set by the -D option must start with an upper case letter, just
4728 like macro names defined in the configuration file.
4730 7. Added "dereference=" facility to LDAP.
4732 8. Two instances of the typo "uknown" in the source files are fixed.
4734 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4735 the Configure-Makefile script screwed up while processing it.
4737 10. Incorporated PCRE 4.4.
4739 11. The SMTP synchronization check was not operating right at the start of an
4740 SMTP session. For example, it could not catch a HELO sent before the client
4741 waited for the greeting. There is now a check for outstanding input at the
4742 point when the greeting is written. Because of the duplex, asynchronous
4743 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4744 way, but not yet received, when the check is performed.
4746 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4747 on TCP/IP sockets, because this apparently causes some broken clients to
4750 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4751 unchanged) from the Cygwin maintainer.
4753 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4754 was supported instead of testing for mbx. Effectively a typo.
4756 15. The spa authenticator server code was not checking that the input it
4757 received was valid base64.
4759 16. The debug output line for the "set" modifier in ACLs was not showing the
4760 name of the variable that was being set.
4762 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4764 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4765 lookup on its own. Something else was needed to trigger the lookup. For
4766 example, a match in host_lookup or the need for a host name in a host list.
4767 Now, if $sender_host_name is referenced and the host name has not yet been
4768 looked up, a lookup is performed. If the lookup fails, the variable remains
4769 empty, and $host_lookup_failed is set to "1".
4771 19. Added "eqi" as a case-independent comparison operator.
4773 20. The saslauthd authentication condition could segfault if neither service
4774 nor realm was specified.
4776 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4777 error message that was logged was misleading, and incoming SMTP
4778 connections were dropped. The message is now more accurate, and temporary
4779 errors are given to SMTP connections.
4781 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4782 (except RSET) with a 421 error, until QUIT is received. However, it was
4783 failing to send a response to QUIT.
4785 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4786 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4787 ACL is now run after the helo_try_verify_hosts code.
4789 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4790 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4791 case-independent, so other case variants are also recognized). Apparently
4792 some systems use these upper case variants.
4794 25. If more than two messages were waiting for the same host, and a transport
4795 filter was specified for the transport, Exim sent two messages over the
4796 same TCP/IP connection, and then failed with "socket operation on non-
4797 socket" when it tried to send the third.
4799 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4801 27. The extern definition of crypt16() in expand.c was not being excluded when
4802 the OS had its own crypt16() function.
4804 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4805 as a preferred synonym for return_size_limit, both as an option and as an
4808 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4810 30. Changed the default configuration ACL to relax the local part checking rule
4811 for addresses that are not in any local domains. For these addresses,
4812 slashes and pipe symbols are allowed within local parts, but the sequence
4813 /../ is explicitly forbidden.
4815 31. SPA server authentication was not clearing the challenge buffer before
4818 32. log_message in a "warn" ACL statement was writing to the reject log as
4819 well as to the main log, which contradicts the documentation and doesn't
4820 seem right (because no rejection is happening). So I have stopped it.
4822 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4823 However, I am unable to do any testing of this.
4825 34. Fixed an infelicity in the appendfile transport. When checking directories
4826 for a mailbox, to see if any needed to be created, it was accidentally
4827 using path names with one or more superfluous leading slashes; tracing
4828 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4830 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4831 contents are added to the log line that is written for every discarded
4832 recipient. (Previously a log_message setting was ignored.)
4834 36. The ${quote: operator now quotes the string if it is empty.
4836 37. The install script runs exim in order to find its version number. If for
4837 some reason other than non-existence or emptiness, which it checks, it
4838 could not run './exim', it was installing it with an empty version number,
4839 i.e. as "exim-". This error state is now caught, and the installation is
4842 38. An argument was missing from the function that creates an error message
4843 when Exim fails to connect to the socket for saslauthd authentication.
4844 This could cause Exim to crash, or give a corrupted message.
4846 39. Added isip, isip4, and isip6 to ${if conditions.
4848 40. The ACL variables $acl_xx are now saved with the message, and can be
4849 accessed later in routers, transports, and filters.
4851 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4852 strings in the file are not string-expanded.
4854 42. If a MAIL command specified a SIZE value that was too large to fit into an
4855 int variable, the check against message_size_limit failed. Such values are
4856 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4857 day this will have to be increased, but I don't think I want to be around
4858 when emails are that large.
4865 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4866 iconv() is not standard in FreeBSD.
4868 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4869 IPv6 enabled. The observed symptom was a segmentation fault on return from
4870 the function os_common_find_running_interfaces() in src/os.c.
4872 3. In the check_special_case() function in daemon.c I had used "errno" as an
4873 argument name, which causes warnings on some systems. This was basically a
4874 typo, since it was named "eno" in the comments!
4876 4. The code that waits for the clock to tick (at a resolution of some fraction
4877 of a second) so as to ensure message-id uniqueness was always waiting for
4878 at least one whole tick, when it could have waited for less. [This is
4879 almost certainly not relevant at current processor speeds, where it is
4880 unlikely to ever wait at all. But we try to future-proof.]
4882 5. The function that sleeps for a time interval that includes fractions of a
4883 second contained a race. It did not block SIGALRM between setting the
4884 timer, and suspending (a couple of lines later). If the interval was short
4885 and the sigsuspend() was delayed until after it had expired, the suspension
4886 never ended. On busy systems this could lead to processes getting stuck for
4889 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4890 process, before it forks any delivery processes. The open lookup caching
4891 mechanism meant that the open file or database connection was passed into
4892 the delivery process. The problem was that delivery processes always tidy
4893 up cached lookup data. This could cause a problem for the next delivery
4894 process started by the queue runner, because the external queue runner
4895 process does not know about the closure. So the next delivery process
4896 still has data in the lookup cache. In the case of a file lookup, there was
4897 no problem because closing a file descriptor in a subprocess doesn't affect
4898 the parent. However, if the lookup was caching a connection to a database,
4899 the connection was closed, and the second delivery process was likely to
4900 see errors such as "PGSQL: query failed: server closed the connection
4901 unexpectedly". The problem has been fixed by closing all cached lookups
4902 in a queue runner before running a delivery process.
4904 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4905 seem to have the "const" qualifier which it has on other OS. I've
4908 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4909 *of the same type* (client or server) with the same public name that an
4910 error should be diagnosed.
4912 9. When Exim looked up a host name for an IP address, but failed to find the
4913 original IP address when looking up the host name (a safety check), it
4914 output the message "<ip address> does not match any IP for NULL", which was
4915 confusing, to say the least. The bug was that the host name should have
4916 appeared instead of "NULL".
4918 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4919 user that is built into the binary, and the -C or -D options is used, root
4920 privilege is dropped before the configuration file is read. In addition,
4921 logging is switched to stderr instead of the normal log files. If the
4922 configuration then re-defines the Exim user, the unprivileged environment
4923 is probably not what is expected, so Exim logs a panic warning message (but
4926 However, if deliver_drop_privilege is set, the unprivileged state may well
4927 be exactly what is intended, so the warning has been cut out in that case,
4928 and Exim is allowed to try to write to its normal log files.
4934 1. smtp_return_error_details was not giving details for temporary sender
4935 or receiver verification errors.
4937 2. Diagnose a configuration error if two authenticators have the same public
4940 3. Exim used not to create the message log file for a message until the first
4941 delivery attempt. This could be confusing when incoming messages were held
4942 for policy or load reasons. The message log file is now created at the time
4943 the message is received, and an initial "Received" line is written to it.
4945 4. The automatically generated man page for command line options had a minor
4946 bug that caused no ill effects; however, a more serious problem was that
4947 the procedure for building the man page automatically didn't always
4948 operate. Consequently, release 4.20 contains an out-of-date version. This
4949 shouldn't happen again.
4951 5. When building Exim with embedded Perl support, the script that builds the
4952 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4953 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4955 6. The freeze_tell option was not being used for messages that were frozen on
4956 arrival, either by an ACL or by local_scan().
4958 7. Added the smtp_incomplete_transaction log selector.
4960 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4961 was accepting AUTH without a new EHLO.
4963 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4964 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4967 10. Logging of TCP/IP connections (when configured) now happens in the main
4968 daemon process instead of the child process, so that the TCP/IP connection
4969 count is more accurate (but it can never be perfect).
4971 11. The use of "drop" in a nested ACL was not being handled correctly in the
4972 outer ACL. Now, if condition failure induced by the nested "drop" causes
4973 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4974 or "require"), the connection is dropped.
4976 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4977 that yield "discard" can now be used with an "accept" or a "discard" verb,
4978 but an error is generated for any others (because I can't see a useful way
4979 to define what should happen).
4981 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4982 are now processed in the same way as lines in the Exim configuration file.
4983 In particular, continuation lines are supported.
4985 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4987 15. Added -ti meaning -t -i.
4989 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4990 domains, and warn by logging if others are found.
4992 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4993 returned from GIFCONF when Exim is trying to find the list of interfaces on
4994 a host. The code in os.c has been modified to copy each ifreq to an aligned
4995 structure in all cases.
4997 Also, in some cases, the returned ifreq's were being copied to a 'struct
4998 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4999 means the last couple of bytes of an IPv6 address could be chopped if the
5000 ifreq contained only a normal sockaddr (14 bytes storage).
5002 18. Named domain lists were not supported in the hosts_treat_as_local option.
5003 An entry such as +xxxx was not recognized, and was treated as a literal
5006 19. Ensure that header lines added by a DATA ACL are included in the reject log
5007 if the ACL subsequently rejects the message.
5009 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5010 MD5 (which is deprecated).
5012 21. When testing a filter file using -bf, Exim was writing a message when it
5013 took the sender from a "From " line in the message, but it was not doing so
5014 when it took $return_path from a Return-Path: header line. It now does.
5016 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5017 with a valid header line field name (a series of printing characters
5018 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5020 23. Changed "disc" in the source to "disk" to conform to the documentation and
5021 the book and for uniformity.
5023 24. Ignore Sendmail's -Ooption=value command line item.
5025 25. When execve() failed while trying to run a command in a pipe transport,
5026 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5027 could be confused with a return value of 69 from the command itself. This
5028 has been changed to 127, the value the shell returns if it is asked to run
5029 a non-existent command. The wording for the related log line suggests a
5030 non-existent command as the problem.
5032 26. If received_header_text expands to an empty string, do not add a Received:
5033 header line to the message. (Well, it adds a token one on the spool, but
5034 marks it "old" so that it doesn't get used or transmitted.)
5036 27. Installed eximstats 1.28 (addition of -nt option).
5038 28. There was no check for failure on the call to getsockname() in the daemon
5039 code. This can fail if there is a shortage of resources on the system, with
5040 ENOMEM, for example. A temporary error is now given on failure.
5042 29. Contrary to the C standard, it seems that in some environments, the
5043 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5044 program. Exim now does this explicitly; it affects the formatting of
5045 timestamps using strftime().
5047 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5048 complaints. I've now initialized all the variables, to avoid this.
5050 32. Header lines added by a system filter were not being "seen" during
5051 transport-time rewrites.
5053 33. The info_callback() function passed to OpenSSL is set up with type void
5054 (*)(SSL *, int, int), as described somewhere. However, when calling the
5055 function (actually a macro) that sets it up, the type void(*)() is
5056 expected. I've put in a cast to prevent warnings from picky compilers.
5058 34. If a DNS black list lookup found a CNAME record, but there were no A
5059 records associated with the domain it pointed at, Exim crashed.
5061 35. If a DNS black list lookup returned more than one A record, Exim ignored
5062 all but the first. It now scans all returned addresses if a particular IP
5063 value is being sought. In this situation, the contents of the
5064 $dnslist_value variable are a list of all the addresses, separated by a
5067 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5068 to accept a host name and all its aliases if the forward lookup for any of
5069 them yielded the IP address of the incoming connection. Now it accepts only
5070 those names whose forward lookup yields the correct IP address. Any other
5071 names are discarded. This closes a loophole whereby a rogue DNS
5072 administrator could create reverse DNS records to break through a
5073 wildcarded host restriction in an ACL.
5075 37. If a user filter or a system filter that ran in a subprocess used any of
5076 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5077 the wrong values were passed to the pipe command ($thisaddress had the
5078 value of $0, $0 had the value of $1, etc). This bug was introduced by
5079 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5081 38. Improved the line breaking for long SMTP error messages from ACLs.
5082 Previously, if there was no break point between 40 and 75 characters, Exim
5083 left the rest of the message alone. Two changes have been made: (a) I've
5084 reduced the minimum length to 35 characters; (b) if it can't find a break
5085 point between 35 and 75 characters, it looks ahead and uses the first one
5086 that it finds. This may give the occasional overlong line, but at least the
5087 remaining text gets split now.
5089 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5090 file descriptors might be low, and that setting 1000 would always raise it.
5091 It turns out that in some environments, the limit is already over 1000 and
5092 that lowering it causes trouble. So now Exim takes care not to decrease it.
5094 40. When delivering a message, the value of $return_path is set to $sender_
5095 address at the start of routing (routers may change the value). By an
5096 oversight, this default was not being set up when an address was tested by
5097 -bt or -bv, which affected the outcome if any router or filter referred to
5100 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5101 log only when "message" or "log_message" are set. However, if one of the
5102 conditions was an address verification, or a call to a nested ACL, the
5103 messages generated by the underlying test were being passed through. This
5104 no longer happens. The underlying message is available in $acl_verify_
5105 message for both "message" and "log_message" expansions, so it can be
5106 passed through if needed.
5108 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5109 new expansion $bh_ to give the encoded byte string without charset
5110 translation. Translation happens only if iconv() is available; HAVE_ICONV
5111 indicates this at build time. HEADERS_CHARSET gives the charset to
5112 translate to; headers_charset can change it in the configuration, and
5113 "headers charset" can change it in an individual filter file.
5115 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5116 that creates RFC 2047 encoded "words" labels them as that charset instead
5117 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5118 expansion operator; (ii) when Exim creates a From: line for a local
5119 message; (iii) when a header line is rewritten to include a "phrase" part.
5121 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5122 buggy, causing it to skip the first lines of messages whose message ID
5123 ended in 'D'. This would not have bitten before Exim release 4.14, because
5124 message IDs were unlikely to end in 'D' before then. The effect was to have
5125 incorrect size information for certain domains.
5127 45. #include "config.h" was missing at the start of the crypt16.c module. This
5128 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5131 46. If there was a timeout during a "random" callout check, Exim treated it as
5132 a failure of the random address, and carried on sending RSET and the real
5133 address. If the delay was just some slowness somewhere, the response to the
5134 original RCPT would be taken as a response to RSET and so on, causing
5135 mayhem of various kinds.
5137 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5138 when I implemented it. It didn't allow for the fact that some option values
5139 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5140 the right test for positive values.
5142 48. Domain names in DNS records are case-independent. Exim always looks them up
5143 in lower case. Some resolvers return domain names in exactly the case they
5144 appear in the zone file, that is, they may contain uppercase letters. Not
5145 all resolvers do this - some return always lower case. Exim was treating a
5146 change of case by a resolver as a change of domain, similar to a widening
5147 of a domain abbreviation. This triggered its re-routing code and so it was
5148 trying to route what was effectively the same domain again. This normally
5149 caused routing to fail (because the router wouldn't handle the domain
5150 twice). Now Exim checks for this case specially, and just changes the
5151 casing of the domain that it ultimately uses when it transmits the message
5154 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5157 50. If a filter generated a file delivery with a non-absolute name (possible if
5158 no home directory exists for the router), the forbid_file option was not
5161 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5162 the existing equality matching.
5164 52. Exim was using ints instead of ino_t variables in some places where it was
5165 dealing with inode numbers.
5167 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5168 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5169 called TMPDIR, and if it finds it is different, it changes its value.
5171 54. The smtp_printf() function is now made available to local_scan() so
5172 additional output lines can be written before returning. There is also an
5173 smtp_fflush() function to enable the detection of a dropped connection.
5174 The variables smtp_input and smtp_batched_input are exported to
5177 55. Changed the default runtime configuration: the message "Unknown user"
5178 has been removed from the ACL, and instead placed on the localuser router,
5179 using the cannot_route_message feature. This means that any verification
5180 failures that generate their own messages won't get overridden. Similarly,
5181 the "Unrouteable address" message that was in the ACL for unverifiable
5182 relay addresses has also been removed.
5184 56. Added hosts_avoid_esmtp to the smtp transport.
5186 57. The exicyclog script was not checking for the esoteric option
5187 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5188 will work only if exicyclog is run under the appropriate euid.
5190 58. Following a discussion on the list, the rules by which Exim recognises line
5191 endings on incoming messages have been changed. The -dropcr and drop_cr
5192 options are now no-ops, retained only for backwards compatibility. The
5193 following line terminators are recognized: LF CRLF CR. However, special
5194 processing applies to CR:
5196 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5197 nor a local message in the state where . is a terminator.
5199 (ii) If a bare CR is encountered in a header line, an extra space is added
5200 after the line terminator so as not to end the header. The reasoning
5201 behind this is that bare CRs in header lines are most likely either
5202 to be mistakes, or people trying to play silly games.
5204 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5205 was being incorrectly given as 18 bytes larger than it should have been.
5206 This is a VOB (very old bug).
5208 60. This may never have affected anything current, but just in case it has:
5209 When the local host is found other than at the start of a list of hosts,
5210 the local host, those with the same MX, and any that follow, are discarded.
5211 When the list in question was part of a longer list of hosts, the following
5212 hosts (not currently being processed) were also being discarded. This no
5213 longer happens. I'm not sure if this situation could ever has previously
5216 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5219 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5220 header. This is something that is recommended in a new Internet Draft, and
5221 is something that is documented as being done by Sendmail. There are two
5222 possible values. For messages generated by the autoreply transport, Exim
5225 Auto-Submitted: auto-replied
5227 whereas for all other generated messages (e.g. bounces) it adds
5229 Auto-Submitted: auto-generated
5231 63. The "personal" condition in filters now includes a test for the
5232 Auto-Submitted: header. If it contains the string "auto-" the message it
5233 not considered personal.
5235 64. Added rcpt_include_affixes as a generic transport option.
5237 65. Added queue_only_override (default true).
5239 66. Added the syslog_duplication option.
5241 67. If what should have been the first header line of a message consisted of
5242 a space followed by a colon, Exim was mis-interpreting it as a header line.
5243 It isn't of course - it is syntactically invalid and should therefore be
5244 treated as the start of the message body. The misbehaviour could have
5245 caused a number of strange effects, including loss of data in subsequent
5246 header lines, and spool format errors.
5248 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5249 client host had authenticated. This control can now be exercised by an ACL
5250 for more flexibility.
5252 69. By default, callouts do not happen when testing with -bh. There is now a
5253 variant, -bhc, which does actually run the callout code, including
5254 consulting and updating the callout cache.
5256 70. Added support for saslauthd authentication, courtesy of Alexander
5259 71. If statvfs() failed on the spool or log directories while checking their
5260 size for availability, Exim confusingly gave the error "space shortage".
5261 Furthermore, in debugging mode it crashed with a floating point exception.
5262 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5263 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5264 serious problem, Exim now writes to the main and panic logs when this
5265 happens, with details of the failure. It then refuses to accept the
5266 incoming message, giving the message "spool directory problem" or "log
5267 directory problem" with a 421 code for SMTP messages.
5269 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5270 0, 1, and 2 exist, because some OS complain for execs without them (see
5271 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5272 descriptors. However, the code omitted to check that the open succeeded,
5273 causing mysterious errors if for some reason the permissions on /dev/null
5274 got screwed. Now Exim writes a message to the main and panic logs, and
5275 bombs out if it can't open /dev/null.
5277 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5278 interact so that it is all more flexible. It is supposed to remain
5279 backwards compatible. Also added extra_local_interfaces.
5281 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5282 to bomb out with an assertion failure - to the client this appears as a
5283 connection drop. This problem occurs in the part of the code that was taken
5284 from the Samba project. Fortunately, the assertion is in a very simple
5285 function, so I have fixed this by reproducing the function inline in the
5286 one place where it is called, and arranging for authentication to fail
5287 instead of killing the process with assert().
5289 75. The SPA client code was not working when the server requested OEM rather
5290 than Unicode encoding.
5292 76. Added code to make require_files with a specific uid setting more usable in
5293 the case where statting the file as root fails - usually a non-root-mounted
5294 NFS file system. When this happens and the failure is EACCES, Exim now
5295 forks a subprocess and does the per-uid checking as the relevant uid.
5297 77. Added process_log_path.
5299 78. If log_file_path was not explicitly set, a setting of check_log_space or
5300 check_log_inodes was ignored.
5302 79. If a space check for the spool or log partitions fails, the incident is now
5303 logged. Of course, in the latter case the data may get lost...
5305 80. Added the %p formatting code to string_format() so that it can be used to
5306 print addresses in debug_print(). Adjusted all the address printing in the
5307 debugging in store.c to use %p rather than %d.
5309 81. There was a concern that a line of code in smtp_in.c could overflow a
5310 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5311 initially expressed, the concern was not well-founded, because trailing
5312 spaces are removed early. However, if the trailing spaces were followed by
5313 a NULL, they did not get removed, so the overflow was possible. Two fixes
5316 (a) I re-wrote the offending code in a cleaner fashion.
5317 (b) If an incoming SMTP command contains a NULL character, it is rejected
5320 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5321 runs initgroups(), so that if the Exim user is in any additional groups,
5322 they will be used during message reception.
5328 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob